BUILDINGS AND WIND

A Software-Based Design Methodology

by

Stephanie Fleming

A thesis
presented to the University of Waterloo
in fulfilment of the
thesis requirement for the degree of
Master of Architecture
in
Engineering

Waterloo, Ontario, Canada, 2015

© Stephanie Fleming 2015
AUTHOR’S DECLARATION

AUTHOR’S DECLARATION
I hereby declare that I am the sole author of this thesis. This is a
true copy of the thesis, including any required final revisions, as
accepted by my examiners.

I understand that my thesis may be made electronically available

to the public.

ii
 ABSTRACT

ABSTRACT
There is a reciprocal relationship between wind and buildings, as the speeds and patterns of wind flow around the initial building
they each affect the other. Building form affects wind by altering form design. This tests the appropriateness of the wind conditions
its speed and flow patterns, and can be used to create desirable for the exterior programs that must be accommodated around the
wind conditions around the building. Wind, in turn, exerts load on building. The speed with which these results are provided allows
the building, which can be reduced with aerodynamic forms and the architect to refine and re-test many iterations of their design
resisted with structural systems. This establishes a relationship until the building form creates the desired wind conditions. The
between wind conditions, the building form that creates these wind CFD software is then used to evaluate the aerodynamics of the
conditions, and the structure that stabilizes the form against these building form by providing information about the wind pressure
wind conditions. that is exerted on each building face. The architect can change the
building form to reduce the wind pressure acting on it, and then
This relationship is investigated through the development of a re-test the form with the CFD software to ensure that improved
design methodology that allows architects to consider, in the early aerodynamics have been achieved without compromising the
design stages, how wind and buildings affect each other. The thesis surrounding wind conditions. Then, the wind pressure information
does not serve to propose a building; rather, it will use a building that is provided by the CFD software is input into the FEA software
as a means for developing this method. The method consists to predict how the building will react to combined wind and gravity
of a pairing of computational fluid dynamics (CFD) software and loading. This information informs the schematic design of the
finite element analysis (FEA) software. While this pairing has not building’s structural system, which is developed through another
been widely explored within the context of architectural design, iterative process using the FEA software.
the combined use of these software programs allows architects
to integrate wind engineering considerations into their current The production of accurate wind and structural data is not the
architectural practices, without having to acquire extensive goal of this thesis, since accurate results are not currently available
engineering knowledge. Software also provides architects with a due to software limitations. Instead, this thesis seeks to develop
means of quickly testing multiple design iterations in relation to these a design method that will increase in accuracy as CFD and FEA
engineering considerations, because the software can perform software programs continue to be improved. In the future, CFD
engineering calculations or simulations much faster than if the and FEA software programmers could potentially draw from this
architect were to learn and perform these calculations themselves. method to create programs that can be used together, to allow
architects to consider wind as a generator of architectural form
For each building design iteration, CFD software is used to simulate within a streamlined, software-based workflow.

iii
ACKNOWLEDGEMENTS

ACKNOWLEDGEMENTS
I would like to extend my sincerest thanks to Elizabeth English for
her confidence in this thesis. I could not have finished it without her
constant guidance, encouragement and patience. Thanks for believing
in it, and in me.

Thank you to Matthew Spremulli, without whom I may not have

stumbled upon the possibilities that wind has to offer. His extraordinary
enthusiasm and dedication were essential to the beginnings of this
thesis.

Thanks to Lloyd Hunt for his insightful suggestions, and for adding his
usual dose of poeticism to an otherwise technical subject.

Thank you to Dr. Tom Mara for providing an engineering perspective

with his thoughtful feedback at the defence.

Thanks also to Maya Przybylski for her wisdom and calming influence
that kept me on track in M2.

Thank you to my family for always being there for me. I am so grateful
for your unconditional love and support. Special thanks to my mom for
her meticulous editing!

Thanks to all of my friends for their encouragements and distractions;

both were greatly appreciated!

And to Nashin and Kate: you have been there since the beginning.
I could not have made it through this thesis, or the past six years,
without your unwavering support, humour and friendship. Thanks for
everything.
iv
 TABLE OF CONTENTS

TABLE OF CONTENTS
ii AUTHOR’S DECLARATION 105 WIND AFFECTS FORM
106 How Wind Affects Form
iii ABSTRACT 107 Aerodynamic Forms Library
113 CFD Iterations to Improve Form Aerodynamics
iv ACKNOWLEDGEMENTS
123 WIND AFFECTS STRUCTURE
v TABLE OF CONTENTS 124 How Wind Affects Structure
127 Structural Systems Library
vi LIST OF FIGURES 131 Structural System Selection
132 FEA Iterations to Develop Structural System
1 WORKFLOW
2 Integration of Architectural and Engineering Processes 145 PROGRAM AFFECTS FORM AND STRUCTURE
6 Designing with Wind 146 How Program Affects Form and Structure
14 CFD Software 147 Interior Program
17 FEA Software 156 Exterior Program Elevations
24 Software Selection 164 Refinement of Structure for Form and Program
32 Selected Programs
34 Accessibility of Method 169 CONCLUSION
36 Software Sequence 170 Design Method Applications
180 Summary
39 PROGRAM AND SITE 184 Discussion and Further Development
41 Program and Site Conditions
42 Program 186 ENDNOTES
54 Site
193 BIBLIOGRAPHY
59 FORM AFFECTS WIND
60 How Form Affects Wind 196 GLOSSARY
61 Wind Effects Library
85 CFD Iterations to Create Wind Conditions 198 APPENDIX A | PAPER FOR ICWE14

v
LIST OF FIGURES

LIST OF FIGURES
Figure Caption Page Figure Caption Page
Source Source

WORKFLOW Fig. 1.7. Screenshot from computational fluid dynamics software.

Video still. From “Autodesk Simulation CFD: Success Stories”.
5

Microdesk. https://www.youtube.com/watch?v=2KfAI_zCdqo.
Fig. 1.1. Trees are naturally structurally efficient. 2
James, Linda. “Tree Branches.” Photograph. Fig. 1.8. Wind can bring comfort and prosperity. 6
PublicDomainPictures.net. http://www.publicdomainpictures. Left: “Bessie Ellen - 1904 Original West Country Trading Ship.”
net/view-image.php?image=2385. Photograph. Classic Sailing. http://www.classic-sailing.co.uk/
vessels/bessie-ellen.
Fig. 1.2. This rock formation maintains a structurally efficient form. 2
“Rainbow Bridge 1.” Photograph. June 2010. Trip Advisor. Middle: Video still. From “Wind Turbine Technology Program
http://www.tripadvisor.ca/Attraction_Review-g60911- Student Testimonial.” Mitchell Technical Institute. https://www.
d145217-Reviews-Rainbow_Bridge-Lake_Powell_Utah. mitchelltech.edu/programs/on-campus/energy-production-
html#photos. transmission/wind-turbine-technology.
Fig. 1.3. Stacking blocks can develop structural intuition. 3 Right: Photograph. Flickr Hive Mind. http://fiveprime.org/flickr_
Photograph. LaughMoreAbundant. https://laughmoreabundant. hvmnd.cgi?method=GET&page=11&photo_number=50&tag_
wordpress.com/2013/09/. mode=all&search_type=Tags&sorting=Interestingness&pho
to_type=75&noform=t&search_domain=Tags&sort=Interestingn
Fig. 1.4. Feeling a diving board’s deflection can develop structural 3 ess&textinput=wind,windy.
intuition.
“Maggie-Ward-standing-on-the-end-of-a-diving-board-at- Fig. 1.9. Wind can also act destructively. 6
Coney-Island-Brooklyn-July-24-1888-520×415.” Photograph. Left: “This home was mainly damaged by falling trees and flying
Big Island Chronicle. http://www.bigislandchronicle. debris during Hurricane Katrina.” Photograph. World-Weather-
com/2013/03/03/commentary-no-i-am-not-going-to-serve- Travellers-Guide.com. http://www.world-weather-travellers-
on-the-planning-commission/. guide.com/hurricane-proof-home.html.
Fig. 1.5. Small-scale physical models demonstrate structural 4 Middle: Snider, Bryan. “A dust storm approaches Phoenix,
behaviour. Arizona on July 5, 2011.” Photograph. July 5, 2011.
Wagner, Rosemarie. “Students in Prof. Wagner’s courses Meteorology News.com. http://meteorologynews.com/offbeat/
explore the behavior of materials along with relationships dust-storm-shrouds-phoenix-in-zero-visibility/.
between curvature and stress in those materials.” Photograph.
Right: “Record-holder: A storm in 2006 which dropped almost
Fabric Architecture. http://fabricarchitecturemag.com/
27inches of snow in New York City is the current record-holder.
articles/0708_f2_networks.html.
Above, a man walks his two golden retrievers through the
Fig. 1.6. Screenshot from finite element analysis software. 4 blizzard.” Photograph. Mail Online. http://www.dailymail.co.uk/
“RSTAB 8.” Screenshot. Dlubal. https://www.dlubal.com/en/. news/article-2925546/Major-snowstorm-snarl-commute-
BLIZZARD-like-conditions-foot-snow-Northeast.html.

vi
 LIST OF FIGURES

Figure Caption Page Figure Caption Page

Source Source

Fig. 1.10. The sun’s rays strike the Earth’s surface at different angles. 6 Fig. 1.21. A tornado is a rotating column of air. 12
By author. Persoff, Jason. Photograph. Huffington Post. http://
www.huffingtonpost.com/2014/10/16/tornado-swarms-
Fig. 1.11. Molecules move from high to low pressure areas. 7 study_n_5998328.html.
By author.
Fig. 1.22. Cladding is often damaged by tornadoes. 12
Fig. 1.12. Global wind patterns. 7 “Here’s one of the facade outside.” Photograph. Dallas
By author. Metropolis. http://forum.dallasmetropolis.com/showthread.
Data from: php/148-Amazing-tornado-damage-pics-from-inside-the-Bank-
De Villiers, Marq. Windswept: The Story of Wind and Weather. One-Tower!.
New York: Walker, 2006.
Fig. 1.23. When building facades are damaged, the interior is  13
Fig. 1.13. Latitudinal bands. 8 exposed.
By author. Smith, Bryan. “Firefighters respond to a building facade collapse
on 8th Avenue after high winds hit NYC.” Photograph. 2012.
Data from: Daily News. http://www.nydailynews.com/new-york/hurricane-
De Villiers, Marq. Windswept: The Story of Wind and Weather. sandy-300-miles-nyc-article-1.1194527.
New York: Walker, 2006.
Fig. 1.24. Digital building model. 14
Fig. 1.14. Buildings shield leeward spaces from the wind. 9 By author.
By author.
Fig. 1.25. Remaining space around the digital building model. 14
Fig. 1.15. Wind is channeled and accelerated between buildings. 9 By author.
By author.
Fig. 1.26. Remaining space divided into small boxes. 14
Fig. 1.16. Wind accelerates around corners. 9 By author.
By author.
Fig. 1.27. Wind speed data slice. 15
Fig. 1.17. Wind accelerates through openings between buildings. 10 By author. Vasari screenshot.
By author.
Fig. 1.28. Flow line animation. 15
Fig. 1.18. Tall buildings direct high-speed wind down to street level. 10 By author. Flow Design screenshot.
By author.
Fig. 1.29. Vector field. 15
Fig. 1.19. Pressure differentials create wind at coastlines. 11 By author. Flow Design screenshot.
By author.
Fig. 1.30. Wind pressure data slice. 16
Fig. 1.20. Pressure differentials cause wind to blow into cities from 11 By author. Flow Design screenshot.
surrounding rural areas.
By author. Fig. 1.31. Surface wind pressure gradient. 16
By author. Flow Design screenshot.

vii
LIST OF FIGURES

Figure Caption Page Figure Caption Page

Source Source

Fig. 1.32. A digital model is broken down into finite elements. 17 Fig. 1.42. Block with left face restrained and 1 kN load on right face 19
By author. (top); resulting tensile strain and exaggerated deformation
from FEA software (bottom).
Fig. 1.33. The particles of a material under tensile stress get pulled 18 By author. Scan&Solve screenshots.
apart.
By author. Fig. 1.43. Compressive strain is the shortening of a unit length of 19
material.
Fig. 1.34. Block under tensile stress with 1 kN loads on end faces 18 By author.
(top); stress distribution and exaggerated deformation from
FEA software (bottom). Fig. 1.44. Block with left face restrained and 1 kN load on right 19
By author. Scan&Solve screenshots. face (top); resulting compressive strain and exaggerated
deformation from FEA software (bottom).
Fig. 1.35. The particles of a material under compressive stress get 18 By author. Scan&Solve screenshots.
pushed together.
By author. Fig. 1.45. A body under tensile stress elongates in the direction of 20
the force and shortens in the direction perpendicular to the
Fig. 1.36. Block under compressive stress with 1 kN loads on 18 force.
end faces (top); stress distribution and exaggerated By author.
deformation from FEA software (bottom).
By author. Scan&Solve screenshots. Fig. 1.46. Block under tensile stress with 1 kN loads on end faces 20
(top); displacement of each point and exaggerated
Fig. 1.37. The particles of a material under shear stress slide relative 19 deformation from FEA software (bottom).
to each other. By author. Scan&Solve screenshots.
By author.
Fig. 1.47. A body under compressive stress shortens in the direction 20
Fig. 1.38. Block under shear stress with 1 kN loads on edges (top); 19 of the force and elongates in the direction perpendicular to
stress distribution and exaggerated deformation from FEA the force.
software (bottom). By author.
By author. Scan&Solve screenshots.
Fig. 1.48. Block under compressive stress with 1 kN loads on end 20
Fig. 1.39. The particles of a material under bending stress shorten on 19 faces (top); displacement of each point and exaggerated
one side and elongate on the other side. deformation from FEA software (bottom).
By author. By author. Scan&Solve screenshots.
Fig. 1.40. Block under bending stress restrained on bottom edges 19 Fig. 1.49. A body under shear stress lengthens along one diagonal 21
of end faces with 1 kN load on top face (top); stress and shortens along the other diagonal.
distribution and exaggerated deformation from FEA By author.
software (bottom).
By author. Scan&Solve screenshots. Fig. 1.50. Block under shear stress with 1 kN loads on edges (top); 21
displacement of each point and exaggerated deformation
Fig. 1.41. Tensile strain is the elongation of a unit length of material. 19 from FEA software (bottom).
By author. By author. Scan&Solve screenshots.

viii
 LIST OF FIGURES

Figure Caption Page Figure Caption Page

Source Source

Fig. 1.51. A body subject to bending elongates on one side and 21 Fig. 1.56. Simulation CFD software ranking diagram. 28
shortens on the other side. By author.
By author.
Data from:
Fig. 1.52. Block under bending stress restrained on bottom edges of 21 “Features: Autodesk CFD software benefits.” Autodesk.
end faces with 1 kN load on top face (top); displacement Accessed January 28, 2015. http://www.autodesk.com/
of each point and exaggerated deformation from FEA products/cfd/features/key-differentiators/gallery-view.
software (bottom).
Jamie Ware, AEC Account Executive, IMAGINiT Technologies,
By author. Scan&Solve screenshots.
e-mail message, January 30, 2015.
Fig. 1.53. Smaller finite elements (right) provide a better 22 “Overview: Computational fluid dynamics software.” Autodesk.
approximation of a form than larger finite elements (left). Accessed January 28, 2015. http://www.autodesk.com/
By author. products/cfd/overview.
Fig. 1.54. Vasari software ranking diagram. 26 Fig. 1.57. Scan&Solve software ranking diagram. 29
By author. By author.
Data from: Data from:
“Autodesk Vasari.” Autodesk. Accessed January 28, 2015. “Scan&Solve Full License (Commercial).” Intact Solutions.
http://autodeskvasari.com. Accessed January 28, 2015. http://intact-solutions-inc.
Part 7: Running Wind Tunnel Simulations. Autodesk Building myshopify.com/collections/scan-solve-full-license-commercial.
Solutions. September 6, 2011. Video. http://www.youtube. “Scan-and-Solve for Rhino.” Intact Solutions. Accessed
com/watch?v=PhuYa8aXYjU&index=31&list=PLF8ACB9BED5 January 28, 2015. http://www.scan-and-solve.com.
6DE576.
Yares, Evan. “Meshless FEA: A new way to solve linear static
Fig. 1.55. Flow Design software ranking diagram. 27 FEA problems.” Design World. Last modified May 11, 2012.
By author. http://www.designworldonline.com/meshless-fea-a-new-way-
Data from: to-solve-linear-static-fea-problems/#_????.
“Autodesk Flow Design.” Autodesk. Accessed January 28, Fig. 1.58. SOFiSTiK Rhino Interface software ranking diagram. 30
2015. http://www.autodesk.com/store/flow-design?licenseTyp By author.
e=cloudSub&support=basic&term=annual.
Data from:
“Features: Flow simulation tools.” Autodesk. Accessed January Laura Baumgärtner Dipl.-Ing., Sales and Consulting, SOFiSTiK,
28, 2015. http://www.autodesk.com/products/flow-design/ e-mail message, February 10, 2015.
features/all/list-view.
“SOFiSTiK Rhinoceros Interface.” McNeel. Last modified
The Flow Design Story. Autodesk. February 12, 2015. Video. September 20, 2011. http://blog.rhino3d.com/2011/09/
http://knowledge.autodesk.com/support/flow-design/learn- sofistik-rhinoceros-interface.html.
explore/caas/CloudHelp/cloudhelp/ENU/FlowDesign/files/
GUID-6D24A84A-A4ED-4B91-9720-46C6DF6CFC87-htm.
html.

ix
LIST OF FIGURES

Figure Caption Page Figure Caption Page

Source Source

“SOFiSTiK Rhinoceros Interface.” SOFiSTiK. Accessed Fig. 2.2. Exterior program diagram. 43
February 4, 2015. http://www.sofistik.com/en/solutions/ By author.
structural-fea/rhinoceros-interface/.
Data from:
Fig. 1.59. Simulation Mechanical software ranking diagram. 31 Adair, Josh, Nels Estlund, Braden Etchison, Drew Gelderd, and
By author. Joe Yoder. “Compact Wind Acceleration Turbine.” Prezi. Last
modified March 18, 2013. https://prezi.com/ozzjc94bhsjx/
Data from: compact-wind-acceleration-turbine/.
Jamie Ware, AEC Account Executive, IMAGINiT Technologies,
e-mail message, January 30, 2015. Brown, Lance. “Wind and Snow Fences.” Factsheet,
British Columbia Ministry of Agriculture, Food and
“Overview: Finite element analysis and modeling software.” Fisheries, April 1997. http://www.agf.gov.bc.ca/resmgmt/
Autodesk. Accessed January 28, 2015. http://www.autodesk. publist/300Series/307230-1.pdf.
com/products/simulation-mechanical/overview.
“Course 102 - Wind Knowledge.” A Wind of Change.
Fig. 1.60. Flow line animation from Flow Design. 32 Accessed November 12, 2014. http://www.awindofchange.
By author. Flow Design screenshot. com/lessons/lesson1-2.html.
Fig. 1.61. Wind pressure colour gradient from Flow Design. 32 Dabiri, John O. “Potential order-of-magnitude enhancement of
By author. Flow Design screenshot. wind farm power density via counter-rotating vertical-axis wind
turbine arrays.” Journal of Renewable and Sustainable Energy
Fig. 1.62. Wind speed data slice from Vasari. 33
3, 043104 (2011): 1-12.
By author. Vasari screenshot.
Feng, Guoying, Tim De Troyer, and Mark C. Runacres.
Fig. 1.63. Input loading (blue) and reactions (red) in Scan&Solve. 33 “Optimizing the Land Use for Wind Farms Using Vertical Axis
By author. Screenshot from Scan&Solve. Wind Turbines.” Conference poster, EWEA 2014, Barcelona,
Fig. 1.64. Displacement colour gradient overlaid on deflection  33 Spain. http://proceedings.ewea.org/annual2014/conference/
animation from Scan&Solve. posters/PO_192_EWEApresentation2014.pdf.
By author. Screenshot from Scan&Solve. “How to fly a kite.” National Kite Month. Accessed November
12, 2014. http://www.nationalkitemonth.org/how-to-fly-a-kite/.
Fig. 1.65. Software sequence diagram. 37
By author. “How to Snowkite.” Snowkiting.com. Accessed November 12,
2014. http://www.snowkiting.com/howto.
PROGRAM AND SITE “Introducing Snowfer.” Snowfer. Accessed November 12,
2014. http://www.snowfer.com/intro.html.
Fig. 2.1. Exterior program list. 42
By author. “Snowfer Frequently Asked Questions.” Snowfer. Accessed
November 12, 2014. http://www.snowfer.com/snowferfaq.
html.
“Snowkiting FAQs.” MACkite. Accessed November 12, 2014.
http://www.mackiteboarding.com/snowkitefaq.htm#wind.

x
 LIST OF FIGURES

Figure Caption Page Figure Caption Page

Source Source

Svitil, Kathy. “Wind-turbine Placement Produces Tenfold Wesoff, Eric. “Optiwind’s Innovation in Medium Wind.”
Power Increase, Caltech Researchers Say.” Caltech. Last GreenTechMedia. Last modified January 26, 2011. http://
modified July 13, 2011. http://www.caltech.edu/news/wind- www.greentechmedia.com/articles/read/optiwinds-medium-
turbine-placement-produces-tenfold-power-increase-caltech- wind.
researchers-say-1700.
Fig. 2.5. Vertical axis wind turbine wind and spatial requirements. 46
Wesoff, Eric. “Optiwind’s Innovation in Medium Wind.” By author.
GreenTechMedia. Last modified January 26, 2011. http://
www.greentechmedia.com/articles/read/optiwinds-medium- Data from:
wind. Dabiri, John O. “Potential order-of-magnitude enhancement of
wind farm power density via counter-rotating vertical-axis wind
“Wind Turbine Site Selection.” Solacity Inc. Accessed turbine arrays.” Journal of Renewable and Sustainable Energy
November 12, 2014. http://www.solacity.com/siteselection. 3, 043104 (2011): 1-12.
htm.
Feng, Guoying, Tim De Troyer, and Mark C. Runacres.
“Windbreaks.” Penn State College of Agricultural Sciences. “Optimizing the Land Use for Wind Farms Using Vertical Axis
Accessed November 12, 2014. http://extension.psu.edu/ Wind Turbines.” Conference poster, EWEA 2014, Barcelona,
plants/plasticulture/production-details/windbreaks. Spain. http://proceedings.ewea.org/annual2014/conference/
posters/PO_192_EWEApresentation2014.pdf.
WindTronics. “BTPS 6500 Wind Turbine - Blade Tip Power
System.” Product guide, November 2010. http://www. Svitil, Kathy. “Wind-turbine Placement Produces Tenfold
windtronics.eu.com/wp-content/uploads/International-Catalog. Power Increase, Caltech Researchers Say.” Caltech. Last
pdf. modified July 13, 2011. http://www.caltech.edu/news/wind-
turbine-placement-produces-tenfold-power-increase-caltech-
Fig. 2.3. Horizontal axis wind turbine wind and spatial requirements. 44
researchers-say-1700.
By author.
Fig. 2.6. Blade tip power system wind and spatial requirements. 47
Data from:
By author.
“Wind Turbine Site Selection.” Solacity Inc. Accessed
November 12, 2014. http://www.solacity.com/siteselection. Data from:
htm. WindTronics. “BTPS 6500 Wind Turbine - Blade Tip Power
System.” Product guide, November 2010. http://www.
Fig. 2.4. Compact wind acceleration turbine wind and spatial 45 windtronics.eu.com/wp-content/uploads/International-Catalog.
requirements.
pdf.
By author.
Fig. 2.7. Kite flying wind and spatial requirements. 48
Data from:
By author.
Adair, Josh, Nels Estlund, Braden Etchison, Drew Gelderd, and
Joe Yoder. “Compact Wind Acceleration Turbine.” Prezi. Last Data from:
modified March 18, 2013. https://prezi.com/ozzjc94bhsjx/ “Course 102 - Wind Knowledge.” A Wind of Change.
compact-wind-acceleration-turbine/. Accessed November 12, 2014. http://www.awindofchange.
com/lessons/lesson1-2.html.

xi
LIST OF FIGURES

Figure Caption Page Figure Caption Page

Source Source

“How to fly a kite.” National Kite Month. Accessed November Fig. 2.12. Snowkiting wind and spatial requirements. 53
12, 2014. http://www.nationalkitemonth.org/how-to-fly-a-kite/. By author.
Fig. 2.8. Field sports wind and spatial requirements. 49 Data from:
By author. “How to Snowkite.” Snowkiting.com. Accessed November 12,
2014. http://www.snowkiting.com/howto.
Data from:
Brown, Lance. “Wind and Snow Fences.” Factsheet, “Snowkiting FAQs.” MACkite. Accessed November 12, 2014.
British Columbia Ministry of Agriculture, Food and http://www.mackiteboarding.com/snowkitefaq.htm#wind.
Fisheries, April 1997. http://www.agf.gov.bc.ca/resmgmt/
publist/300Series/307230-1.pdf. Fig. 2.13. Regina’s average annual wind speed. 54
By author.
Fig. 2.9. Tennis and badminton wind and spatial requirements. 50
Data from:
By author.
“Wind & weather statistics Regina Airport.” Windfinder. Last
Data from: modified July 2015. http://www.windfinder.com/windstatistics/
“Windbreaks.” Penn State College of Agricultural Sciences. regina_airport.
Accessed November 12, 2014. http://extension.psu.edu/
plants/plasticulture/production-details/windbreaks. Fig. 2.14. A flat, open site was chosen. 54
By author.
Fig. 2.10. Snow build-up for sledding wind and spatial requirements. 51
By author. Fig. 2.15. Regina’s population. 54
By author.
Data from:
Brown, Lance. “Wind and Snow Fences.” Factsheet, Fig. 2.16. The site is near the University of Regina. 54
British Columbia Ministry of Agriculture, Food and “Black.” Logo. University of Regina. http://www.uregina.ca/
Fisheries, April 1997. http://www.agf.gov.bc.ca/resmgmt/ external/communications/visual-id/logo.html.
publist/300Series/307230-1.pdf. Fig. 2.17. Site Plan. 55
Fig. 2.11. Snow windsurfing wind and spatial requirements. 52 By author.
By author. Data from:
Data from: Google Maps. Accessed June 6, 2015. https://www.google.
“Introducing Snowfer.” Snowfer. Accessed November 12, ca/maps.
2014. http://www.snowfer.com/intro.html. Fig. 2.18. Annual wind rose for the chosen site. 56
“Snowfer Frequently Asked Questions.” Snowfer. Accessed By author.
November 12, 2014. http://www.snowfer.com/snowferfaq. Data from:
html. Autodesk Vasari Beta 3.

xii
 LIST OF FIGURES

Figure Caption Page Figure Caption Page

Source Source

Fig. 2.19. Monthly wind roses for the chosen site. 57 Fig. 3.12. Flow Design simulation - effect not shown. 65
By author. By author. Flow Design screenshot.
Data from: Fig. 3.13. Corner effect. 66
“Wind & weather statistics Regina Airport.” Windfinder. Last By author.
modified July 2015. http://www.windfinder.com/windstatistics/
regina_airport. Fig. 3.14. Vasari simulation. 66
By author. Vasari screenshot.

FORM AFFECTS WIND Fig. 3.15. Flow Design simulation. 66

By author. Flow Design screenshot.
Fig. 3.1. Breezeway effect. 62 Fig. 3.16. Cumulative effect. 67
By author. By author.
Fig. 3.2. Vasari simulation - effect not shown. 62 Fig. 3.17. Vasari simulation - effect not shown. 67
By author. Vasari screenshot. By author. Vasari screenshot.
Fig. 3.3. Flow Design simulation - effect not shown. 62 Fig. 3.18. Flow Design simulation. 67
By author. Flow Design screenshot. By author. Flow Design screenshot.
Fig. 3.4. Channel effect. 63 Fig. 3.19. Diverting effect. 68
By author. By author.
Fig. 3.5. Vasari simulation. 63 Fig. 3.20. Vasari simulation - effect not shown. 68
By author. Vasari screenshot. By author. Vasari screenshot.
Fig. 3.6. Flow Design simulation - effect not shown. 63 Fig. 3.21. Flow Design simulation - effect not shown. 68
By author. Flow Design screenshot. By author. Flow Design screenshot.
Fig. 3.7. Channelling effect. 64 Fig. 3.22. Downwash effect. 69
By author. By author.
Fig. 3.8. Vasari simulation. 64 Fig. 3.23. Vasari simulation - effect not shown. 69
By author. Vasari screenshot. By author. Vasari screenshot.
Fig. 3.9. Flow Design simulation. 64 Fig. 3.24. Flow Design simulation - effect not shown. 69
By author. Flow Design screenshot. By author. Flow Design screenshot.
Fig. 3.10. Combined row and downwash effect. 65 Fig. 3.25. Funneling effect. 70
By author. By author.
Fig. 3.11. Vasari simulation - effect not shown. 65 Fig. 3.26. Vasari simulation. 70
By author. Vasari screenshot. By author. Vasari screenshot.

xiii
LIST OF FIGURES

Figure Caption Page Figure Caption Page

Source Source

Fig. 3.27. Flow Design simulation. 70 Fig. 3.42. Reduction of downwash effect - canopy. 75
By author. Flow Design screenshot. By author.
Fig. 3.28. Staggering effect. 71 Fig. 3.43. Reduction of downwash effect - setback. 75
By author. By author.
Fig. 3.29. Vasari simulation. 71 Fig. 3.44. Vasari simulation - effect not shown. 75
By author. Vasari screenshot. By author. Vasari screenshot.
Fig. 3.30. Flow Design simulation - effect not shown. 71 Fig. 3.45. Flow Design simulation - effect not shown. 75
By author. Flow Design screenshot. By author. Flow Design screenshot.
Fig. 3.31. Venturi effect. 72 Fig. 3.46. Solid windbreak effect. 76
By author. By author.
Fig. 3.32. Venturi effect induces suction. 72 Fig. 3.47. Vasari simulation. 76
By author. By author. Vasari screenshot.
Fig. 3.33. Vasari simulation. 72 Fig. 3.48. Flow Design simulation. 76
By author. Vasari screenshot. By author. Flow Design screenshot.
Fig. 3.34. Flow Design simulation. 72 Fig. 3.49. Combined row and downwash effect. 77
By author. Flow Design screenshot. By author.
Fig. 3.35. Cumulative effect. 73 Fig. 3.50. Vasari simulation. 77
By author. By author. Vasari screenshot.
Fig. 3.36. Vasari simulation - effect not shown. 73 Fig. 3.51. Flow Design simulation. 77
By author. Vasari screenshot. By author. Flow Design screenshot.
Fig. 3.37. Flow Design simulation - effect not shown. 73 Fig. 3.52. Courtyard effect. 78
By author. Flow Design screenshot. By author.
Fig. 3.38. Porous windbreak effect. 74 Fig. 3.53. Flow Design simulation (effect not shown in Vasari). 78
By author. By author. Flow Design screenshots.
Fig. 3.39. Vasari simulation. 74 Fig. 3.54. Downwash effect. 79
By author. Vasari screenshot. By author.
Fig. 3.40. Flow Design simulation. 74 Fig. 3.55. Vasari simulation - effect not shown. 79
By author. Flow Design screenshot. By author. Vasari screenshot.
Fig. 3.41. Reduction of downwash effect - podium. 75 Fig. 3.56. Flow Design simulation - effect not shown. 79
By author. By author. Flow Design screenshot.

xiv
 LIST OF FIGURES

Figure Caption Page Figure Caption Page

Source Source

Fig. 3.57. Row effect. 80 Fig. 3.72. Iteration 3 Vasari screenshot. 88

By author. By author. Vasari screenshot.
Fig. 3.58. Vasari simulation - effect not shown. 80 Fig. 3.73. Iteration 4 Vasari screenshot. 88
By author. Vasari screenshot. By author. Vasari screenshot.
Fig. 3.59. Flow Design simulation. 80 Fig. 3.74. Iteration 5 Vasari screenshot. 89
By author. Flow Design screenshot. By author. Vasari screenshot.
Fig. 3.60. Vortex shedding. 81 Fig. 3.75. Iteration 6 Vasari screenshot. 89
By author. By author. Vasari screenshot.
Fig. 3.61. Flow Design simulation (flow lines not shown in Vasari). 81 Fig. 3.76. Iteration 7 Vasari screenshot. 91
By author. Flow Design screenshot. By author. Vasari screenshot.
Fig. 3.62. Wake effect. 82 Fig. 3.77. Iteration 7 Vasari screenshot. 91
By author. By author. Vasari screenshot.
Fig. 3.63. Vasari simulation. 82 Fig. 3.78. Iteration 8 Vasari screenshot. 92
By author. Vasari screenshot. By author. Vasari screenshot.
Fig. 3.64. Flow Design simulation. 82 Fig. 3.79. Iteration 8 Vasari screenshot. 92
By author. Flow Design screenshot. By author. Vasari screenshot.
Fig. 3.65. Courtyard effect. 83 Fig. 3.80. Iteration 9 Vasari screenshot. 93
By author. By author. Vasari screenshot.
Fig. 3.66. Flow Design simulation (effect not shown in Vasari). 83 Fig. 3.81. Iteration 9 Vasari screenshot. 93
By author. Flow Design screenshots. By author. Vasari screenshot.
Fig. 3.67. Distance from obstruction. 84 Fig. 3.82. Iteration 10 Vasari screenshot. 94
By author. By author. Vasari screenshot.
Fig. 3.68. Vasari simulation. 84 Fig. 3.83. Iteration 10 Vasari screenshot. 94
By author. Vasari screenshot. By author. Vasari screenshot.
Fig. 3.69. Flow Design simulation. 84 Fig. 3.84. Iteration 11 Vasari screenshot. 95
By author. Flow Design screenshot. By author. Vasari screenshot.
Fig. 3.70. Iteration 1 Vasari screenshot. 87 Fig. 3.85. Iteration 11 Vasari screenshot. 95
By author. Vasari screenshot. By author. Vasari screenshot.
Fig. 3.71. Iteration 2 Vasari screenshot. 87 Fig. 3.86. Form with exterior programs. 96
By author. Vasari screenshot. By author.

xv
LIST OF FIGURES

Figure Caption Page Figure Caption Page

Source Source

Fig. 3.87. Wind speed from Vasari. 97 Fig. 4.5. Flow Design simulation. 109
By author. Vasari screenshot. By author. Flow Design screenshots.
Fig. 3.88. Flow lines from Flow Design. 97 Fig. 4.6. Effect of form on wind force. 110
By author. Flow Design screenshot. By author.
Fig. 3.89. Exterior programs when wind blows from the SE in 98 Fig. 4.7. Flow Design simulation - effect not shown. 110
summer. By author. Flow Design screenshots.
By author.
Fig. 4.8. Effect of exposure on wind force. 111
Fig. 3.90. Exterior programs when wind blows from the SE in winter. 99 By author.
By author.
Fig. 4.9. Flow Design simulation. 111
Fig. 3.91. Form with exterior programs. 100 By author. Flow Design screenshots.
By author.
Fig. 4.10. Effect of porosity on wind force. 112
Fig. 3.92. Wind speed from Vasari. 101 By author.
By author. Vasari screenshot.
Fig. 4.11. Flow Design simulation. 112
Fig. 3.93. Flow lines from Flow Design. 101 By author. Flow Design screenshots.
By author. Flow Design screenshot.
Fig. 4.12. Iteration 1 Flow Design screenshots. 114
Fig. 3.94. Exterior programs when wind blows from the NW in 102 By author. Flow Design screenshots.
summer.
By author. Fig. 4.13. Iteration 2 Flow Design screenshots. 114
By author. Flow Design screenshots.
Fig. 3.95. Exterior programs when wind blows from the NW in winter. 103
By author. Fig. 4.14. Iteration 3 Flow Design screenshots. 115
By author. Flow Design screenshots.

WIND AFFECTS FORM Fig. 4.15. Iteration 4 Flow Design screenshots. 115
By author. Flow Design screenshots.
Fig. 4.1. Wind pressure distribution over a building. 106 Fig. 4.16. Iteration 5 Flow Design screenshots. 116
By author. By author. Flow Design screenshots.
Fig. 4.2. Effect of orientation on wind pressure. 108 Fig. 4.17. Iteration 6 Flow Design screenshots. 116
By author. By author. Flow Design screenshots.
Fig. 4.3. Flow Design simulation. 108 Fig. 4.18. Iteration 7 Flow Design screenshots. 117
By author. Flow Design screenshots. By author. Flow Design screenshots.
Fig. 4.4. Effect of wind speed on wind pressure. 109 Fig. 4.19. Iteration 7 Flow Design screenshots. 117
By author. By author. Flow Design screenshots.

xvi
 LIST OF FIGURES

Figure Caption Page Figure Caption Page

Source Source

Fig. 4.20. Iteration 8 Flow Design screenshots. 118 Fig. 5.7. Diaphragm. 127
By author. Flow Design screenshots. By author.

Fig. 4.21. Iteration 8 Flow Design screenshots. 118 Fig. 5.8. Bracing. 127
By author. Flow Design screenshots. By author.

Fig. 4.22. Iteration 9 Flow Design screenshots. 119 Fig. 5.9. Moment-resisting connections. 127
By author. Flow Design screenshots. By author.

Fig. 4.23. Iteration 9 Flow Design screenshots. 119 Fig. 5.10. Plywood on wood framing. 128
By author. Flow Design screenshots. “A wall is raised.” Photograph. You Can Build Your Own House.
http://boardsandbricks.com/tag/radiant/.
Fig. 4.24. Iteration 10 Flow Design screenshots. 120
By author. Flow Design screenshots. Fig. 5.11. Solid concrete wall. 128
Photograph. PCG. http://www.pcg-group.com.my/index.
Fig. 4.25. Iteration 10 Flow Design screenshots. 120 asp?p=/static/the-advantage-of-using-shear-wall-method-of-
By author. Flow Design screenshots. construction.html.
Fig. 4.26. Iteration 11 Flow Design screenshots. 121 Fig. 5.12. Masonry wall. 128
By author. Flow Design screenshots. Schumin, Ben. “Looking inside, elevator shaft openings around
Fig. 4.27. Iteration 11 Flow Design screenshots. 121 the building’s concrete core have been boarded up, and the
By author. Flow Design screenshots. floor slabs are completely bare.” Photograph. February 9,
2005. The Schumin Web. http://www.schuminweb.com/
photography/photo-2005/urban-demolition/.
WIND AFFECTS STRUCTURE Fig. 5.13. A frame without surfacing moves to the side under lateral 128
loads.
Fig. 5.1. Alongwind load response. 124
By author.
By author.
Fig. 5.14. A diaphragm subjected to lateral loads. 128
Fig. 5.2. Crosswind load response. 124
By author.
By author.
Fig. 5.15. Steel framing. 129
Fig. 5.3. Torsional load response. 124
“Typical bracing in a multi-storey frame.” Photograph.
By author.
SteelConstruction.info. http://www.steelconstruction.info/
Fig. 5.4. Uplift. 125 Design.
By author.

Fig. 5.5. Gravity loads. 127

By author.

Fig. 5.6. Wind loads. 127

By author.

xvii
LIST OF FIGURES

Figure Caption Page Figure Caption Page

Source Source

Fig. 5.16. Wood framing. 129 Fig. 5.25. The frame deforms as a cohesive unit to absorb wind and 130
“This internal wall frame has to be strong enough to carry the gravity loads.
weight of the floor above it. Notice that it still has temporary By author.
diagonal bracing in place, which was used to help keep
the frame square while it was being installed.” Photograph. Fig. 5.26. Wind pressure gradients from Flow Design. 132
Workspace Training. http://www.workspacetraining.com.au/ By author. Flow Design screenshots.
timberplustoolbox/toolbox13_05/unit7_assembling_wall_ Fig. 5.27. The building divided into pressure zones. 133
frames/section1_wall_frame_components/lesson1_your_job. By author.
htm.
Fig. 5.28. Pressure zone matrix. 133
Fig. 5.17. Bracing configurations. 129 By author.
By author.
Fig. 5.29. Loading applied to zone 1 in Scan&Solve. 134
Fig. 5.18. Wind load on a frame. 129 By author. Scan&Solve screenshot.
By author.
Fig. 5.30. Deflection animation of zone 1 from Scan&Solve. 134
Fig. 5.19. One diagonal member. 129 By author. Scan&Solve screenshot.
By author.
Fig. 5.31. Zone 1 divided into sub-zones. 134
Fig. 5.20. Two diagonal members. 129 By author. Scan&Solve screenshot.
By author.
Fig. 5.32. Zone 1A iteration 1 Scan&Solve screenshot. 136
Fig. 5.21. Steel plates and fasteners. 130 By author. Scan&Solve screenshot.
Coleman, Jason. “Heavy Timer Connection.” Photograph.
January 14, 2007. flickr. https://www.flickr.com/photos/jason_ Fig. 5.33. Zone 1A iteration 2 Scan&Solve screenshot. 136
coleman/360316812. By author. Scan&Solve screenshot.
Fig. 5.22. Bolted connection. 130 Fig. 5.34. Zone 1A iteration 3 Scan&Solve screenshot. 137
Meyer Boake, Terri. Photograph. tboake.com. http://www. By author. Scan&Solve screenshot.
tboake.com/SSEF1/bolt.shtml.
Fig. 5.35. Zone 1A iteration 4 Scan&Solve screenshot. 137
Fig. 5.23. Welded connection. 130 By author. Scan&Solve screenshot.
Meyer Boake, Terri. “Detail of welded connection for steel
truss.” Photograph. tboake.com. http://www.tboake.com/ Fig. 5.36. Zone 1B iteration 1 Scan&Solve screenshot. 138
steel/waiward.html. By author. Scan&Solve screenshot.

Fig. 5.24. Steel-reinforced concrete joint. 130 Fig. 5.37. Zone 1B iteration 2 Scan&Solve screenshot. 138
“Rebar connection system / reinforced concrete.” Rendering. By author. Scan&Solve screenshot.
ArchiExpo. http://www.archiexpo.com/prod/ancon-building- Fig. 5.38. Zone 1B iteration 3 Scan&Solve screenshot. 139
products/product-69643-1358449.html. By author. Scan&Solve screenshot.

xviii
 LIST OF FIGURES

Figure Caption Page Figure Caption Page

Source Source

Fig. 5.39. Zone 1B iteration 4 Scan&Solve screenshot. 139 Fig. 6.6. Building and exterior programs when wind blows from the 153
By author. Scan&Solve screenshot. NW in winter.
By author.
Fig. 5.40. Schematic structural drawing (left) and pressure zones 141
overlaid onto structural drawing (right). Fig. 6.7. Wind speed from Vasari. 154
By author. By author. Vasari screenshot.

Fig. 5.41. Building form. 142 Fig. 6.8. Wind speed from Vasari. 155
By author. By author. Vasari screenshot.

Fig. 5.42. Adjusted building form to reflect the inclusion of the steel 142 Fig. 6.9. Vasari screenshot at 2m. 157
structure. By author. Vasari screenshot.
By author.
Fig. 6.10. Vasari screenshot at 2m. 157
Fig. 5.43. Wind speed from Vasari. 143 By author. Vasari screenshot.
By author. Vasari screenshots.
Fig. 6.11. Vasari screenshot at 4m. 158
Fig. 5.44. Wind speed from Vasari. 143 By author. Vasari screenshot.
By author. Vasari screenshots.
Fig. 6.12. Vasari screenshot at 4m. 158
By author. Vasari screenshot.
PROGRAM AFFECTS FORM AND STRUCTURE Fig. 6.13. Vasari screenshot at 6m. 159
By author. Vasari screenshot.
Fig. 6.1. Building form. 149
By author. Fig. 6.14. Vasari screenshot at 6m. 159
By author. Vasari screenshot.
Fig. 6.2. Building with program. 149
By author. Fig. 6.15. Vasari screenshot at 8m. 160
By author. Vasari screenshot.
Fig. 6.3. Building and exterior programs when wind blows from the 150
SE in summer. Fig. 6.16. Vasari screenshot at 8m. 160
By author. By author. Vasari screenshot.
Fig. 6.4. Building and exterior programs when wind blows from the 151 Fig. 6.17. Vasari screenshot at 10m. 161
SE in winter. By author. Vasari screenshot.
By author.
Fig. 6.18. Vasari screenshot at 10m. 161
Fig. 6.5. Building and exterior programs when wind blows from the 152 By author. Vasari screenshot.
NW in summer.
By author. Fig. 6.19. Vasari screenshot at 20m. 162
By author. Vasari screenshot.

xix
LIST OF FIGURES

Figure Caption Page Figure Caption Page

Source Source

Fig. 6.20. Vasari screenshot at 20m. 162 Fig. 7.6. Canopies deflect wind. 171
By author. Vasari screenshot. By author.

Fig. 6.21. Vasari screenshot at 60m. 163 Fig. 7.7. Podiums deflect wind. 171
By author. Vasari screenshot. By author.

Fig. 6.22. Vasari screenshot at 60m. 163 Fig. 7.8. Setbacks shelter pedestrians from downwash. 171
By author. Vasari screenshot. By author.

Fig. 6.23. Structural layout for building form and interior program. 165 Fig. 7.9. Breezeways deflect wind before it reaches ground level. 171
By author. By author.

Fig. 6.24. Key drawing (left) and building cross-section (spread). 166-7 Fig. 7.10. Porous screens decrease wind speed. 171
By author. By author.

Fig. 7.11. Small parks in Toronto’s downtown. 172

CONCLUSION By author. Base image from Google Maps. https://www.
google.ca/maps.
Fig. 7.1. Buildings creating fast or turbulent winds can make it 170
Fig. 7.12. Central Park in New York City. 172
uncomfortable for pedestrians to walk at street level.
By author. Base image from Google Maps. https://www.
Left: “Students have a laugh in Blackpool as strong winds blow
google.ca/maps.
umbrellas inside out.  Coastal areas will be hit by winds of up to
80mph.” Photograph. 2014. Mail Online. http://www.dailymail. Fig. 7.13. Ventilation corridors in Masdar City. 172
co.uk/news/article-2800523/storm-clouds-brewing-britain- By author. Base image: “Aerial View of Proposed Master plan
braced-hurricane-gonzalo-hit-uk-midnight.html. of Masdar City (Eastern Orientation).” Rendering. Bright Ideas.
Right: “BATTEN THE HATCHES: Severe northwest gales http://blog.invention.smithsonian.org/category/notes-from-the-
were expected to hit Wairarapa and Wellington later today.” director/.
Photograph. 2012. The Dominion Post. http://www.stuff. Fig. 7.14. Pressure differentials induce an intake of fresh rural air into 173
co.nz/dominion-post/news/wellington-weather/7623120/ cities.
Severe-wind-warning-for-afternoon. By author.
Fig. 7.2. Tall buildings direct high-speed wind down to street level. 170 Fig. 7.15. Korčula, Croatia blocks cold winds but is permeable to 173
By author. warm winds.
Fig. 7.3. Wind is channeled and accelerated between buildings. 170 By author.
By author. Fig. 7.16. Masdar City day wind rose. 174
Fig. 7.4. Wind accelerates around corners. 171 By author.
By author.
Fig. 7.17. Masdar City night wind rose. 174
Fig. 7.5. Wind accelerates through openings between buildings. 171 By author.
By author.

xx
 LIST OF FIGURES

Figure Caption Page Figure Caption Page

Source Source

Fig. 7.18. Hot winds from the northwest enter the corridors during 174 Fig. 7.28. Snow can block doors. 178
the day. “Entrance to Bartels blocked by a mountain snow after
By author. Storm Nemo.” Photograph. The Charger Bulletin. http://
www.chargerbulletin.com/2013/02/14/36-inches-of-winter-
Fig. 7.19. Cool winds from the east enter the corridors at night. 174 wonderland/.
By author.
Fig. 7.29. Snow can block vents. 178
Fig. 7.20. Winds disperse pollutants in urban areas. 175 Fuller, Steve. “These shots illustrate an external furnace vent
“Beijing, China.” Photograph. Trend Police. http://www. before and after it has been cleared of snow. Clearing such
trendpolice.com/3165/delhi-worlds-most-polluted-city/. vents helps prevent carbon monoxide from building up inside
Fig. 7.21. Global winds carry air pollution around the world. 175 a home.” Photograph. The Ellsworth American. http://www.
Putman, William. “A lot of the airborne particles in the Earth’s ellsworthamerican.com/featured/clear-vents-doors-pathways-
atmosphere come from natural sources, such as desert dust vehicles.
(red-orange) and sea salt (blue). But there’s also soot from fires Fig. 7.30. Buildings must be able to withstand the load exerted by  178
(green and yellow) and sulfur emissions (white) from burning snow.
fossil fuel.” Still from animation. 2007. NPR. http://www.npr. “Snow.” Photograph. February 2006. Crawford Family Website.
org/sections/goatsandsoda/2015/03/08/391056439/why- http://crawfordfamilywebsite.com/steamboat_2006_p1.htm.
chinas-pollution-could-be-behind-our-cold-snowy-winters.
Fig. 7.31. Simulated snow on a building model in a wind tunnel. 178
Fig. 7.22. Xeritown day wind rose. 176 “Scale model test of snow accumulation.” Photograph. Cermak
By author. Peterka Petersen. http://www.cppwind.com/blogs/snow-
Fig. 7.23. During the day, cool winds are channeled through the city 176 loading-drifting.
for cooling and ventilation. Fig. 7.32. Svalbard Science Centre. 179
By author. Dale, Nils-Petter. Photograph. Phaidon Atlas. http://
Fig. 7.24. Xeritown night wind rose. 176 phaidonatlas.com/building/svalbard-science-centre/303.
By author. Fig. 7.33. Physical model. 179
Fig. 7.25. At night, hot winds are diverted over the tops of the 176 “One of the models used to develop the complex design,
buildings. shaped like a lobster.” Photograph. RCI Online. http://www.
By author. rci-online.org/interface/2009-02-robinson.pdf.

Fig. 7.26. Xeritown. 177 Fig. 7.34. CFD simulation. 179

“Masterplan of Xeritown.” Rendering. ConstructionWeekOnline. Dale, Nils-Petter. Photograph. ArchDaily. http://www.archdaily.
com. http://www.constructionweekonline.com/article-4552- com/3506/svalbard-science-centre-jva.
dubais_xeritown/1/. Fig. 7.35. Design method sequence. 181
Fig. 7.27. Varied building heights induce turbulence at the skyline. 177 By author.
By author.

xxi
 1 “Wind is moving air. The air has a particular mass (density or
weight) and moves in a particular direction . . . When the moving
fluid air encounters a stationary object, there are several effects
that combine to exert a force on the object.”
James Ambrose and Dimitry Vergun, Simplified Building Design for Wind
and Earthquake Forces1

2 “What a character the wind is . . . An all powerful ruler, sometimes

terrible, sometimes charitable . . . he who can be used, avoided,
or fled from, but who can never be tamed.”
Guy de Maupassant, Sur l’eau2

xxiii
 WORKFLOW

1
WORKFLOW | INTEGRATION OF ARCHITECTURAL AND ENGINEERING PROCESSES

INTEGRATION OF ARCHITECTURAL AND ENGINEERING PROCESSES

Within current architectural practice there is a separation between Qualitative structural understanding
architectural and engineering design processes, as engineering The ability to know how a structure behaves without referring
analysis is not often integrated into the early architectural design to measurements or calculations.
phases. Instead, it is performed by an engineer after the initial building
design has already been developed by the architect, at a point when
it is often difficult, costly, and time-consuming to change the design to
accommodate the engineering requirements. This separation is widely
acknowledged in the fields of both architecture and engineering, and
members of both professions have worked on developing methods of
integrating the two design processes. This thesis explores one such
method of integration, through the creation of a design methodology
that architects can use to consider wind loading and effects as design
informants that are integrated into the initial architectural design phases.
This allows both structural and wind engineering considerations to be
included in the architectural design process.

In antiquity, the architect was the “master builder”1 who understood Fig. 1.1. Trees are naturally structurally efficient.
and was able to execute both building design and construction.2
However, when the Industrial Revolution spurred the rapid creation
of many new building materials and technologies, it became difficult
for a single person to master them all.3 This fostered the creation of
the structural engineering profession.4 Structural engineers became
experts in building construction technologies,5 while architects
specialized in the spatial and aesthetic design of buildings.6 This
has resulted in a divergence between the roles and priorities of the
architect and the structural engineer.

One method of integrating structural engineering considerations

into the architectural design process is for the architect to develop
a qualitative structural understanding. There are several ways in
which a structural form can be generated without scientific structural
analysis, including observation of natural forms, structural intuition, Fig. 1.2. This rock formation maintains a structurally efficient form.

2
 WORKFLOW | INTEGRATION OF ARCHITECTURAL AND ENGINEERING PROCESSES

Structural intuition and physical models. These tools can help architects develop a
The ability to immediately understand how an object or qualitative structural understanding, which is necessary for them to
material will act under load, without necessarily knowing why. consider space and structure simultaneously in the design process.

For example, natural forms (Fig. 1.1, Fig. 1.2) are made structurally
efficient to withstand the environment, and have always provided
humans with forms that we can mimic and know that they will
be structurally stable.7 As we grow up, structural intuition is
subconsciously developed through activities such as stacking blocks
as a child (Fig. 1.3) or feeling a diving board deflect under our weight
(Fig. 1.4), which allow everyone to develop an idea of how structural
forms work without necessarily knowing why they work.8 Small-
scale physical models (Fig. 1.5) can also be used to demonstrate a
form’s structural behaviour.9 Because these methods of structural
form generation do not require scientific structural analysis, they allow
architects to design structures with a qualitative idea of whether or not
they will work. This understanding needs to be developed by architects
to be able to carry out a design method that integrates architectural
and engineering considerations.

Fig. 1.3. Stacking blocks can develop structural intuition. This qualitative understanding of engineering principles may also
be developed through the use of software, such as finite element
analysis (FEA) software (Fig. 1.6). Gary Black and Stephen Duff,
who are architecture and engineering professors at the University
of California, Berkeley, developed a structural education model for
architecture students that emphasizes the importance of learning
structural design, rather than just structural analysis, and facilitates
this through the use of finite element analysis software as a learning
tool.10 They taught this model with great success for six years at
the University of California, Berkeley.11 This educational model
was developed because many structures courses for architecture
students, both at the time and today, only teach a condensed version
of the structural analysis courses for engineering students, which do
not cover structural design.12 As a result, architectural graduates are
Fig. 1.4. Feeling a diving board’s deflection can develop structural intuition. often unable to apply structural principles during the architectural

3
WORKFLOW | INTEGRATION OF ARCHITECTURAL AND ENGINEERING PROCESSES

design process.13 To combat this, the use of FEA software as a

learning tool allows users of this educational model to quickly see
and understand how forces act within the structures they design
and input into the program without their having to learn engineering
calculations.14 The quick feedback that is provided allows students
to develop a qualitative structural understanding, and teaches them
how to design while constantly considering structural feasibility.15
As they build up this qualitative structural understanding, architects
will eventually develop structural intuition.16 This intuition allows a
designer to think about structural behaviour early in the design
process, without needing calculations to understand how their
design will react under load.17 This intuition is developed with
experience and experimentation that is facilitated by the use of Fig. 1.5. Small-scale physical models demonstrate structural behaviour.
the FEA software that simulates structural behaviour.18 Through
repeated use of the software, architects can learn to predict what
the outcome of the structural analysis will be.19
Finite element analysis (FEA) software
A similar intuition may be developed with the use of computational Software that evaluates the structural behaviour of an input model
fluid dynamics (CFD) software (Fig. 1.7), which in this thesis is under input loading conditions and produces numbers, graphics
used to describe the movement of wind. While it simulates fluid and animations to convey this behaviour.
flow rather than structural behaviour, it provides similar advantages
to the FEA software as it performs the engineering calculations to
determine this fluid behaviour so that the user does not have to
learn and perform these complicated calculations manually. The
simulations that are generated by the software and visually represent
the wind, rather than quantify it, are easier for users who are not
trained in wind engineering to understand and interpret. Repeated
use of the software allows the user to develop an understanding
of wind behaviour, so that they are eventually able to predict the
results of the simulation even without using the software. This
allows them to make more informed design decisions earlier in the
design process that consider the effects of wind on and around Fig. 1.6. Screenshot from finite element analysis software.
buildings.

4
 WORKFLOW | INTEGRATION OF ARCHITECTURAL AND ENGINEERING PROCESSES

Structural design The design method that is developed in this thesis integrates

The process of determining the form of a structure that will allow both wind and structural engineering considerations into early
it to withstand subjected loads. architectural design stages by pairing CFD software and FEA
software. Software can be a valuable design tool for architects
Structural analysis because it allows them to integrate engineering considerations
The process of calculating the types and magnitudes of stresses into their current architectural practices, without having to acquire
and deformations in a structure subjected to loads. extensive engineering knowledge. It provides visualizations of wind
patterns and structural behaviour, which are easier than numerical
data for architects to understand and interpret, since they are
Structural behaviour trained in visual media. Software also provides architects with a
The manner in which a structure acts or functions under loading means of quickly testing multiple design iterations in relation to these
conditions. engineering considerations, because the software can perform
engineering calculations and simulations much faster than if the
Computational fluid dynamics (CFD) software architect were to learn and perform these calculations themselves.
Software that simulates the flow of fluids, including wind, around Repeated use of the software allows the user to gradually develop
an input model and produces numbers, graphics and animations an intuitive understanding that allows them to predict what the
to convey this flow. results of the software analysis will likely be, which will eventually
influence their design decisions even before the software analysis
is run.

This design method does not intend for the architect to replace
the structural or wind engineer. Instead, it equips the architect
with the necessary knowledge and tools to design a building that
considers wind effects and building aerodynamics in its initial form
generation, which may be refined by the wind engineer in a later
design phase. The method also allows the architect to design a
building that accommodates a feasible structural system that the
structural engineer can then easily adjust and detail. This approach
allows the architect and engineers to work towards shared goals,
thereby streamlining co-ordination between them. It also eliminates
costly and time-consuming design revisions that can occur when
wind effects and loading are only considered in the later stages of
the design process. This method serves to re-integrate engineering
and architectural design processes within the current working
Fig. 1.7. Screenshot from computational fluid dynamics software. practices of architects and engineers.

5
WORKFLOW | DESIGNING WITH WIND

DESIGNING WITH WIND

Throughout history, humans have tried to understand the seemingly
arbitrary nature of the wind, as it both creates life and takes it
away.20 Wind makes the Earth habitable by cooling the equator
and warming the poles,21 it brings prosperity by carrying our ships
and spinning our turbines, and it provides the welcome relief of a
forgiving breeze on a hot day (Fig. 1.8). It can also damage our Fig. 1.8. Wind can bring comfort and prosperity.
buildings and landscapes, carry dust storms and harsh winter
chills, drive people mad with its relentless sound, and even kill
people with its force (Fig. 1.9). Whether it comes as a blessing or
a danger, we are at wind’s mercy, so humans study how and why
wind works the way it does so that we may decipher its arbitrary
behaviour and be prepared to face it.22 Fig. 1.9. Wind can also act destructively.

HOW WIND IS CREATED

The creation of wind begins with the sun, as the curve of the earth
causes the sun’s rays to strike the earth’s surface directly at the
equator and more obliquely at the poles (Fig. 1.10).23 As a result,
the sun’s rays are distributed over a smaller area at the equator and
a larger area near the poles, making the air hotter at the equator
and cooler at the poles.24 Since the molecules comprising hotter
air are less dense than those that make up the colder air, this
temperature difference creates air pressure differentials throughout
the atmosphere.25 These atmospheric pressure differentials cause
wind, weather, and climate.26
Whenever there is a differential in the environment, nature has
a tendency to want to equalize it.27 Because the mass of the
atmosphere is mostly constant, this equalization is achieved by
re-distributing the atmospheric gases, and moving some of the
molecules away from high pressure areas and into low pressure
areas to balance the atmospheric pressure around the globe (Fig.
1.11).28 This movement of atmospheric gases is the wind.29 Fig. 1.10. The sun’s rays strike the Earth’s surface at different angles.

6
 WORKFLOW | DESIGNING WITH WIND

HIGH LOW EVEN Wind speeds are determined by the magnitude of the pressure
differential, as the larger the difference in pressure between the two
areas, the faster the gases move between them.30 Wind direction is
determined by the locations of the high and low pressure areas in
relation to each other, as air moves directly from the high pressure
area to the low pressure area.31 However, this straight line of
movement is deflected by the Earth’s rotation.32 This deflection,
coupled with the temperature differentials between the poles and
the equator, create the global wind patterns.
Fig. 1.11. Molecules move from high to low pressure areas.
GLOBAL WIND PATTERNS
At a global scale, wind and weather systems generally follow and
repeat the same patterns.33 At the equator, where the solar radiation
POLAR EASTERLIES is the most concentrated, the warmed air rises and creates an area
60° of low atmospheric pressure closer to the Earth’s surface.34 This
WESTERLIES draws in air from the semi-tropical latitudes, but as this air moves
towards the equator, the Earth’s rotation from west to east causes
30° the moving air to turn to the right in the northern hemisphere and
to the left in the southern hemisphere until it is moving parallel to
TRADE WINDS the equator.35 This moving air is called the trade winds (Fig. 1.12),36
whose consistency has been known and depended upon for sailing
for hundreds of years.37
0°
The trade winds are warmed at the equator, causing them to rise
TRADE WINDS
and drift towards the poles.38 At about 30° latitude, they are cooled
and sink back down towards the Earth’s surface.39 Some of this
cooled air moves back to the equator again to become part of
30° the trade winds, while some of it moves towards the low-pressure
areas in the mid-latitudes.40 As this air moves polewards, the
WESTERLIES Earth’s rotation causes it to turn right in the northern hemisphere
and left in the southern hemisphere.41 These are the westerlies (Fig.
60° 1.12),42 which have also been exploited by sailors for hundreds of
POLAR EASTERLIES years for the colonization of, or trade with, foreign lands.43 These
regions have the most turbulent winds, as warm equatorial air and
Fig. 1.12. Global wind patterns. cool polar air meet here to cause gales and storms.44

7
WORKFLOW | DESIGNING WITH WIND

While some of the air in the mid-latitudes forms the westerlies, some
60°
of the air higher up in the atmosphere moves back to the equator POLAR EASTERLIES
to join the trade winds, and some of the air near the Earth’s surface
moves further towards the poles.45 At the poles, the cold air sinks to 38°
the ground to create areas of high pressure at the Earth’s surface.46
WESTERLIES 30°
The air in these high pressure areas then moves back towards the
equator, to the low-pressure areas next to the westerlies.47 As this
air moves it is deflected to the right in the northern hemisphere and
HORSE LATITUDES 5°
to the left in the southern hemisphere to create the polar easterlies
(Fig. 1.12).48 0°
In between these wind patterns lie latitudinal bands of windless 5°
regions. The doldrums sit over the equator, usually extending about TRADE WINDS
5° to each side (Fig. 1.13).49 In between the trade winds and the
westerlies lies another band of calm air called the horse latitudes
(Fig. 1.13).50 These are still, low-pressure regions that are as well- DOLDRUMS
30°
known to sailors as the trade winds and the westerlies, but are DOLDRUMS
avoided rather than exploited.51 38°
TRADE WINDS

HOW WIND CREATES CLIMATE HORSE LATITUDES

These global wind patterns move large air masses around the WESTERLIES 60°
world.52 Contained within these air masses is solar energy, so POLAR EASTERLIES
through the movement of the wind, much of the solar energy at the
equator is transported towards the poles.53 If wind did not occur Fig. 1.13. Latitudinal bands.
to balance the distribution of the globe’s solar energy, the equator
would be too hot and the poles would be too cold for the survival
of many species that inhabit the earth today.54 The only habitable
place on earth would be a small band in between the poles and the
equator.55 Through this distribution of air masses and solar energy,
wind patterns create regions of weather and climate around the
world.56

8
 WORKFLOW | DESIGNING WITH WIND

BUILDINGS AFFECT WIND

ZONE OF 
Wind patterns are usually fairly consistent and predictable at a
DECREASED SPEED global scale, and if the Earth’s surface was completely even, wind
MAX 30h
would flow smoothly and predictably at smaller scales as well.57
However, this level of surface consistency is uncommon, except
over oceans.58 Small-scale surface variations, such as buildings,
vegetation, and topography, alter local wind speeds and patterns
and make local winds difficult to predict.59
h
When designing a building, it is important to understand how global
Fig. 1.14. Buildings shield leeward spaces from the wind. winds affect the climate of the region, as well as how conditions
around the building will alter local winds. Surface variations and
temperature differentials are two factors that complicate wind at a
local scale.
slow

fast SURFACE VARIATIONS
low
SPEED SOLID WINDBREAK 
high 

EFFECT As air flows over smooth surfaces, it is altered less than when it
flows over surfaces with much variation.60 Features that prevent the
PRESSURE

surface from being flat get in the way of the wind and complicate
Fig. 1.15. Wind is channeled and accelerated between buildings.
its flow by speeding it up, slowing it down, or creating turbulence.
Such surface features include continents, topography, vegetation,
buildings, and cities.61
slow

fast
- PRESSSURE
Wind tends to slow down in cities because the surface roughness
of the buildings slows down the wind flow, and because buildings
low
SPEED

CHANNELLING EFFECT
high 

shield adjacent leeward spaces from the wind (Fig. 1.14).62 However,
+ PRESSSURE
PRESSURE

the wind can interact with buildings in ways that causes it to speed
w
up in local areas. Buildings that are parallel to each other can create
ZONE OF INCREASED 
channels that increase the speed of the wind flowing between them
WIND SPEED
(Fig. 1.15);63 wind accelerates around building corners (Fig. 1.16)64
w
and through building openings (Fig. 1.17);65 and tall building faces
slow
Fig. 1.16. Wind accelerates around corners. that are exposed to oncoming wind can direct the higher-speed
winds at higher elevations down the building face to the street level
fast (Fig. 1.18).66 In these ways, buildings and cities alter local winds
SPEED and make them difficult to predict.
slow
9
fast
low
SPEED

CORNER EFFECT
high 

PRESSURE
WORKFLOW | DESIGNING WITH WIND

TEMPERATURE DIFFERENTIALS
Different landscapes radiate heat at different rates, and these local
temperature differentials between two landscapes create pressure
differentials that induce or alter wind.67 Because land heats up
faster than water, the resulting pressure differentials create wind
at every coastline (Fig. 1.19).68 Landscapes that radiate heat faster
than others also cause wind at the boundary between them, such
as the boundary between deserts, which radiate heat more quickly,
MAX 2h

and grasslands, which radiate heat more slowly.69

The temperature difference between urban and rural areas also h

influences local wind patterns to create an influx of wind into the

city.70 The heat conductivity of buildings and paved surfaces is Fig. 1.17. Wind accelerates through openings between buildings.
much higher than that of soil and vegetation, so cities become
warmer as they absorb more heat.71 Underground drainage in the slow

city also contributes to warmer city temperatures, as the energy

that would be used to evaporate this water if it was above-ground, low
fast

remains in the city in the form of heat.72 These factors contribute to SPEED

VENTURI EFFECT
high  STAGNATION 
the urban heat island effect, which in turn affects the wind patterns PRESSURE POINT

between the city and the adjacent countryside.73 Hot city air rises,
creating an area of low pressure within the city closer to the ground
level.74 As a result, air from the surrounding rural, high-pressure h

areas blows into the city (Fig. 1.20).75 (2/3)h

TORNADOES
In addition to the surface variations and temperature differentials
that complicate wind, the tendency of air to form vortices also
makes wind patterns less predictable.76 One type of vortex is a
tornado.77 With their sudden formations,78 unpredictable paths,79 Fig. 1.18. Tall buildings direct high-speed wind down to street level.
and blistering wind speeds,80 they are a type of extreme wind that
should be considered when designing buildings,81 especially in
slow

slow
tornado-prone areas. low
fast
SPEED
fast
A tornado is a column of air that rotates around a central point, and high 

PRESSURE DOWNWASH EFFECT
as the air gets closer to the centre, it rotates faster (Fig. 1.21).82 The SPEED

10
 WORKFLOW | DESIGNING WITH WIND

LOW PRESSURE

HIGH PRESSURE

Fig. 1.19. Pressure differentials create wind at coastlines.

HIGH PRESSURE LOW PRESSURE

Fig. 1.20. Pressure differentials cause wind to blow into cities from surrounding rural areas.

LABELS:
LEGEND: COOL AIR | WARM AIR
UNDER TREES: HIGH PRESSURE
UNDER CITY: LOW PRESSURE
COOL AIR WARM AIR
LABELS:
LEGEND: COOL AIR | WARM AIR
11
UNDER TREES: HIGH PRESSURE
UNDER CITY: LOW PRESSURE
WORKFLOW | DESIGNING WITH WIND

wind speeds within these columns of air can exceed 450 km/h,83
and as they move across the Earth’s surface they can quickly and
unpredictably change both speed and direction, even turning 180
degrees.84 They are only made visible by the condensation that
they contain, as well as the dust and debris such as soil, rocks,
vegetation, and parts of buildings and infrastructure that they pick
up along their way.85 While the sequence of tornado formation is
not clear, tornadoes occur when there is warm, humid air at ground
level and cooler air at a higher elevation.86 The air just above the
Earth’s surface must be warm enough to prevent the warm, ground-
level air from rising into the cooler air above it.87 If a sudden cold
front comes in and moves this layer of warm air, the warm, humid Fig. 1.21. A tornado is a rotating column of air.
air at ground level swiftly rises up into the colder air and causes a
tornado to form.88 This can happen suddenly and without warning,
making tornadoes very difficult to predict.89 A tornado warning
usually comes only minutes before it strikes, which is usually
insufficient time for the residents of the area to find shelter.90 As a
result, tornadoes can cause injury and death, as well as financial
losses and substantial destruction of buildings and infrastructure.91
While some locations are more prone to tornadoes than others,
tornadoes can occur all over the world.92 Buildings located in
areas that are prone to tornadoes should ideally be designed to
withstand their extreme winds.93 However, because tornadoes are
low-probability events, most buildings are not specifically designed
to resist the damage that they can cause, although these buildings’
structures can usually withstand the damage even if their cladding
cannot.94 There are three main ways in which tornadoes may inflict
damage upon buildings. First, the extreme winds exert high wind
forces on buildings that can destroy their cladding and strain their
structural systems.95 These forces may be resisted by stiffening
the building’s structural system and using resilient construction for
the roof and facades.96 Second, tornadoes cause rapid drops in
the exterior air pressure.97 Because these pressure drops occur
in a matter of seconds, buildings do not have time to adjust to
the pressure difference between the interior and exterior.98 This Fig. 1.22. Cladding is often damaged by tornadoes.

12
 WORKFLOW | DESIGNING WITH WIND

causes the roof and walls to be blown outwards.99 This could be

prevented with resilient construction that allows roofs and walls to
resist increased pressure loads, or by venting between the building
interior and exterior.100 Buildings may also be damaged during
tornadoes by debris that is picked up by the tornado and hits the
buildings.101 The damage caused by debris may be reduced with
resilient building construction to allow buildings to better withstand
the impact.102 Building roofs and facades are often damaged by
tornadoes in these three ways, as the cladding is not designed
to as high a factor of safety as the structure (Fig. 1.22).103 When
the cladding is damaged, the building interior is exposed to the
outside and is also damaged by the high winds and flying debris
(Fig. 1.23).104 Because of the ways in which tornadoes can affect
buildings, these extreme winds should be a consideration during
building design.

DESIGNING WITH WIND

Fig. 1.23. When building facades are damaged, the interior is exposed.
Wind has a large impact on the built environment.105 It is also the
most dynamic climate parameter that architects must work with.106
The prominence and unpredictability of wind, especially at the
scale of a building,107 provides an extreme context within which to
test the integration of architecture, wind engineering, and structural
engineering through the development of a design methodology.
Wind effects on buildings are constantly changing, so buildings
may be designed to resist these fluctuating wind loads, as well
as alter its patterns to create desirable wind conditions around
the building. As such, the methodology that is developed in this
thesis harnesses the productive nature of the wind, while resisting
its destructive nature. We may be at wind’s mercy, but if we can
understand the wind, we may manipulate it to serve our purposes.

13
WORKFLOW | CFD SOFTWARE

CFD SOFTWARE
HOW IT WORKS
Computational fluid dynamics (CFD) uses complex equations to describe
the movement of heat or fluids (such as air and water) in and around solid
objects108 over time.109 For the purpose of this thesis, CFD is used to
simulate wind. Complex differential equations are used to describe this fluid
movement, but their complexity requires that they be solved with a CFD
code as it is impossible to solve them by hand.110 The code first subtracts Fig. 1.24. Digital building model.
the building model (Fig. 1.24) from the total volume of the digital model
space.111 It then analyzes the remaining volume, which is the empty space
around the building model (Fig. 1.25).112 This space is then divided into
small boxes (Fig. 1.26), and the motion of the fluid within each box, as well
as how it interacts with the fluid in the neighbouring boxes, is calculated
using the differential equations.113 These calculations are then repeated for
each box and for subsequent moments in time, to simulate the movement
of the fluid within the entire space around the building model.114 The
simulation parameters, which for this thesis are wind speed and direction,
are input by the user before the simulation is run.

ACCURACY Fig. 1.25. Remaining space around the digital building model.

The accuracy of a CFD simulation depends on several factors. Some

codes are designed to be more accurate than others, but because they are
inherently more complex, the hardware requirements are more intensive
and it is also often necessary for a CFD specialist to run them.115 The
accuracy also depends on the grid resolution that is set either manually by
the user, or automatically by the software.116 A higher grid resolution will yield
more accurate simulations as it divides the volume around the model into
smaller boxes, but the trade-off for this increase in accuracy is an increase
in computing time.117 The accuracy of the simulation is also affected by the
extent to which the surrounding surface features are modeled.118 These
surface features can have such a substantial influence on the behaviour
of the fluid surrounding the model that without considering their impact, Fig. 1.26. Remaining space divided into small boxes.

14
 WORKFLOW | CFD SOFTWARE

the simulation results could be inaccurate. However, larger, more

complex models that include these surrounding features also
increase the computing time required to run the CFD simulation.119

WIND SPEED OUTPUT

INCREASED
For the purpose of this thesis, CFD software is used to simulate and
represent the flow of wind around a building. The software provides
AVERAGE
information about wind speed, patterns, and pressure, through a
variety of visual output styles from which the user may choose to
DECREASED
display the required information. The following visual output styles
Fig. 1.27. Wind speed data slice. may be obtained from the CFD software that is used in this thesis:
WIND SPEED DATA SLICE
Depicts the variation in wind speed around the building model
within a single horizontal or vertical plane, represented by the colour
variation of the data slice (Fig. 1.27).
WIND SPEED
INCREASED FLOW LINE ANIMATION
Animates the wind flow, conveying wind patterns with the movement
AVERAGE
of the flow lines and representing wind speed with the colours of
the flow lines (Fig. 1.28). Flow lines may be animated in a single
horizontal or vertical plane, or throughout the 3D volume around
DECREASED
the building model.
Fig. 1.28. Flow line animation.
VECTOR FIELD
Conveys wind patterns with the direction of each arrow, and
represents wind speed with the colours of the arrows, within a
single horizontal or vertical plane (Fig. 1.29)

WIND SPEED WIND PRESSURE DATA SLICE

INCREASED Depicts the variation in air pressure around the building model
within a single horizontal or vertical plane, represented by the
AVERAGE colour variation of the data slice (Fig. 1.30). It should be noted that
wind suction is represented as negative wind pressure, so larger
DECREASED magnitudes of wind suction are represented by colours at the
bottom of the legend in Fig. 1.30, and larger magnitudes of positive
Fig. 1.29. Vector field.

15
WORKFLOW | CFD SOFTWARE

wind pressure are represented by colours at the top of the legend.

Wind pressure of minimum magnitude is therefore represented by
colours in the middle of the legend.
SURFACE WIND PRESSURE GRADIENT
WIND PRESSURE
Depicts the variation in wind pressure acting over the surface of the +
building model, represented by the variation in the colours applied
to the model’s surface (Fig. 1.31). The legend in Fig. 1.31 conveys
MINIMUM
the same information as the legend in Fig. 1.30, which is described
above.
-

ARCHITECTURAL APPLICATIONS Fig. 1.30. Wind pressure data slice.

CFD simulations of wind flow are useful in architectural

applications, as they can determine pressure loads on facades120
and structures, simulate the flow of air within buildings for HVAC
studies, and simulate wind conditions around buildings for studies WIND PRESSURE
of pedestrian comfort and pollutant dispersion.121 However, within +
current architectural practice, these simulations are mainly used for
verification of the design at a later stage, and do not usually inform MINIMUM
the initial design of the building.122 They are also often performed
by CFD specialists who work with a consulting engineer,123 who -
then makes design recommendations to the architect based on the
Fig. 1.31. Surface wind pressure gradient.
results of the simulation.124 Rarely is the architect directly involved in
the CFD simulation process.125 However, engineer Ziad Boutanios,
who specializes in CFD, believes that there is potential for architects
to be more directly involved in this process, through the use of
simpler codes that they may run themselves to obtain visualizations
of general wind trends.126 This thesis seeks to do this by selecting
and appropriating existing CFD software, so that architects can run
the simulations themselves and obtain qualitative wind information
early in the design process that informs the schematic design of
their buildings. The visual representation of the wind that may be
obtained from the CFD software makes it easier for architects who
are not trained in wind engineering to understand and interpret the
results, and allow the results to inform their building designs.

16
 WORKFLOW | FEA SOFTWARE

FEA SOFTWARE
HOW IT WORKS
Finite elements Finite element analysis (FEA) is a computerized method that
Small pieces into which a digital model is divided to be analyzed predicts and simulates how a digital model will react to applied
by FEA software. forces.127 The software breaks down the digital model into small
pieces called finite elements and analyzes the behaviour of each
Meshing of these elements under the input loading, restraints, and material
The process of representing a physical entity with finite elements, properties.128 The assembly of the behaviour of all of the finite
by breaking it down into smaller pieces to re-build it as a set of elements conveys the global structural behaviour of the entire digital
points, edges, and faces that approximate the original model. model.129 For the purpose of this thesis, FEA software is used to
simulate and visualize the effects of forces on building structural
systems.

In order to run finite element analysis on a digital model, steps must

be taken to make the model into a finite element (FE) model that is
defined by its geometry, its constraints, its material properties, and
the applied loads.130 To create an FE model, a digital model is first
made of the geometry to be analyzed.131 Within the FEA software,
the user then specifies the types, magnitudes and locations of the
applied loads, the locations of the restraints, and the properties of
the material out of which the full-scale model would be made.132
Finally, the model must be meshed to turn it into a FE model.133
Meshing is the process of representing a physical entity with finite
Fig. 1.32. A digital model is broken down into finite elements. elements,134 by breaking down the model into smaller pieces135 to
re-build it as a set of points, edges, and faces that approximate
Mesh the original model (Fig. 1.32).136 This set of finite elements is called
A set of finite elements made up of points, edges, and faces that a mesh.137 While there are theoretically no limits on the shapes
approximate the original model. that these finite elements may be, in practice only simple shapes
are used.138 They may be one-dimensional lines, two-dimensional
triangles or quadrilaterals, or three-dimensional tetrahedrons or
hexahedrons.139 Because the finite elements are simply-shaped
pieces, they are only able to approximate the more complexly-

17
WORKFLOW | FEA SOFTWARE

shaped original model to a certain degree of accuracy that Stress

depends on the size of the finite elements,140 unless the original A quantity that describes all the internal forces acting within a
model was already built as a mesh.141 Splitting the model into body of material.
these small, simply-shaped elements allows them to be analyzed
with simple equations, rather than using complex equations on the
entire, more complex digital model.142 The meshing process can be
time-consuming and complicated, and while most FEA programs
require the user to carry out this process manually, this thesis uses Fig. 1.33. The particles of a material under tensile stress get pulled apart.
a program that performs this step automatically. Once the mesh is
created, the analysis is run to determine how the mesh will behave Fig. 1.34. Block under tensile
stress with 1 kN loads on end
under the input loads, constraints, and material properties. faces (top); stress distribution
and exaggerated deformation
After the analysis has been run, the designer looks at the results from FEA software (bottom).
to find any areas of the model that are structurally inadequate,
STRESS
and then alters the digital model to improve its performance under 1 450 Pa Tension
load.143 The FEA process is then repeated with the altered digital
model to ensure that its structural performance has adequately
improved with the changes.144
470 Pa Tension

OUTPUT
When a force is applied to a material, the force produces stress
within that material that causes it to deform.145 This deformation is
Fig. 1.35. The particles of a material under compressive stress get pushed
seen in strain, which measures change of length per unit length,146 together.
and displacement, which describes the movement of each point of
the model under the force.147 Stress, strain, and displacement are Fig. 1.36. Block under
the three types of output that are available from the FEA software compressive stress with 1 kN
loads on end faces (top); stress
that is used in this thesis. distribution and exaggerated
deformation from FEA software
STRESS (bottom).
Stress is a quantity that describes all the internal forces acting within STRESS
a body of material, and is measured in units of force per area.148 470 Pa Compression

FEA software represents stress as colour gradients over the model

surface, with the colour representing the amount of stress that
the material experiences at that point under load. The two basic 1 450 Pa Compression

18
 WORKFLOW | FEA SOFTWARE

states of stress are tension and compression.149 When a material

experiences tensile stress, its particles tend to be pulled apart (Fig.
1.33, Fig. 1.34), whereas under compressive stress, the particles
tend to be pushed together (Fig. 1.35, Fig. 1.36).150 All complex
Fig. 1.37. The particles of a material under shear stress slide relative to each states of stress are combinations of tension and compression,151
other. such as shear stress, in which the particles of the material slide
Fig. 1.38. Block under shear relative to each other152 as one cross-section diagonal lengthens
stress with 1 kN loads on edges
(top); stress distribution and
in tension and the other shortens in compression (Fig. 1.37, Fig.
exaggerated deformation from 1.38),153 or bending stress, in which the material curves under load
FEA software (bottom). and the fibers shorten in compression on one side and elongate in
tension on the other side (Fig. 1.39, Fig. 1.40).154
STRESS
2 600 Pa Tension
0 Pa STRAIN
Strain describes the deformation in a direction that is caused by
stress, and is a unitless quantity that is measured in change in
11 170 Pa Compression
length per unit length.155 FEA software represents strain as colour
Due to software limitations that
restrict how restraints are specified, gradients over the model surface, with the colour representing the
the deformation shown in Fig. 1.38 amount of strain that each part of the model would experience
is not entirely accurate.
under load. Tensile strain is the elongation of a unit length of material
(Fig. 1.41, Fig. 1.42), whereas compressive strain is the shortening
of a unit length of material (Fig. 1.43, Fig. 1.44).156

DISPLACEMENT
Fig. 1.39. The particles of a material under bending stress shorten on one side
and elongate on the other side. Displacement describes where and by how much a point of a
Fig. 1.40. Block under bending body moves while the body experiences stress, and is measured
stress restrained on bottom in units of length.157 FEA software represents the displacement
edges of end faces with 1 kN of each point of a model as a colour gradient over the model’s
load on top face (top); stress
distribution and exaggerated
surface, with the colour at any given location representing the
deformation from FEA software amount of displacement of that point. This displacement of each
(bottom). point causes the body as a whole to undergo deformation,158
STRESS unless all of the points displace the same amount.159 In this case,
3 550 Pa Tension
0 Pa the body translates, or moves, rather than deforms.160 The FEA
software can create animations of the model deforming under load
as each of its points is displaced, to assist the user in visualizing the
39 700 Pa Compression model’s structural behaviour. A body under tensile stress deforms

19
WORKFLOW | FEA SOFTWARE

Strain Displacement
A quantity that describes deformation in a direction. A measure of where and by how much a point of a body moves
while the body experiences stress.
Deformation
The action of a body changing shape as the points within the
body displace different amounts.
Fig. 1.41. Tensile strain is the elongation of a unit length of material.

Fig. 1.42. Block with left face

restrained and 1 kN load on
Fig. 1.45. A body under tensile stress elongates in the direction of the force
right face (top); resulting
and shortens in the direction perpendicular to the force.
tensile strain and exaggerated
deformation from FEA software Fig. 1.46. Block under tensile
(bottom). stress with 1 kN loads on end
faces (top); displacement of
STRAIN
0.005
each point and exaggerated
deformation from FEA software
(bottom).
DISPLACEMENT
4.43 mm
0.002

0 mm

Fig. 1.43. Compressive strain is the shortening of a unit length of material. Fig. 1.47. A body under compressive stress shortens in the direction of the
force and elongates in the direction perpendicular to the force.
Fig. 1.44. Block with left face Fig. 1.48. Block under
restrained and 1 kN load on compressive stress with 1
right face (top); resulting kN loads on end faces (top);
compressive strain and displacement of each point
exaggerated deformation from and exaggerated deformation
FEA software (bottom). from FEA software (bottom).
STRAIN DISPLACEMENT
0.003 4.43 mm

0.001 0 mm

20
 WORKFLOW | FEA SOFTWARE

as it elongates in the direction of the force and shortens in the

direction perpendicular to the force (Fig. 1.45, Fig. 1.46), while a
body under compressive stress shortens in the direction of the force
Fig. 1.49. A body under shear stress lengthens along one diagonal and and elongates in the direction perpendicular to the force (Fig. 1.47,
shortens along the other diagonal. Fig. 1.48).161 While in shear, the cross-section of a body lengthens
Fig. 1.50. Block under shear along the diagonal that is in tension and shortens along the diagonal
stress with 1 kN loads on in compression,162 causing a rectangular cross-section to deform
edges (top); displacement of into a parallelogram (Fig. 1.49, Fig. 1.50).163 When a body is subject
each point and exaggerated
deformation from FEA software
to bending, the material on one side elongates in tension while the
(bottom). other side shortens in compression (Fig. 1.51, Fig. 1.52).164
DISPLACEMENT
8.68 mm
ACCURACY
The accuracy of FEA depends on the size of the finite elements.165
0 mm The smaller they are, the more accurate the analysis will be.166
Due to software limitations that Finite elements usually have straight sides, so when they are used
restrict how restraints are specified,
the deformation shown in Fig. 1.50 to approximate a curved or more complex form, a finer mesh
is not entirely accurate. with more, smaller finite elements is a better approximation of
the original form than one with fewer, larger finite elements (Fig.
1.53).167 As the number of finite elements increases and their size
consequently decreases, the finite element solution converges to
the exact solution, which is approached asymptotically.168 However,
Fig. 1.51. A body subject to bending elongates on one side and shortens on while a finer mesh provides more accurate analysis results, it also
the other side. uses more of the computer’s memory and requires longer solving
Fig. 1.52. Block under bending times.169
stress restrained on bottom
edges of end faces with The size that the finite elements should be in any given analysis
1 kN load on top face (top);
displacement of each point is decided by the person running the analysis, and depends on
and exaggerated deformation both the model being analyzed as well as the type of information
from FEA software (bottom). that the user wants to obtain from the analysis.170 There is no
DISPLACEMENT rule or method that may be used to determine element size;171
18.73 mm
rather, it is determined with the user’s engineering knowledge and
experience.172 It is best practice in engineering to use the simplest
model and run the simplest analysis that will still provide the required
0 mm information and level of accuracy.173 The user will often run several

21
WORKFLOW | FEA SOFTWARE

finite element analyses on the same model, specifying different

element sizes each time, to determine which size provides an
adequate level of accuracy without taking too long to compute.174
Engineering knowledge can also inform element size, as the user
could specify a coarse mesh over most of a model with a finer
mesh around a small area of the model that they know will have a
large impact on the structural behaviour of the model as a whole
(such as an opening around which there would be a high stress
concentration).175 This provides a more accurate analysis at the
location where a high level of accuracy is the most important,
without taking a lot of time to compute the results over the rest of
the model where the same level of accuracy is not needed.

While architects who use FEA software should follow the

engineering practice of using the simplest model and the largest
element size that will still provide the required level of accuracy, they
may not posses the same knowledge and experience as engineers
to be able to determine what that size would be. The FEA software
that is used in this thesis automatically meshes the model, so that
the architect does not have to execute this process themselves. Fig. 1.53. Smaller finite elements (right) provide a better approximation of a
form than larger finite elements (left).
However, it is still beneficial for the architect to understand how to
determine an appropriate element size, especially if they would like
to use different FEA software in which they would need to mesh
their models manually. To find the appropriate element size for a
given problem, architects could run two finite element analyses on
the same model, but specify a mesh for the second analysis that
contains twice the number of finite elements as the mesh for the first
analysis.176 If the two analyses have similar results, the architect can
be confident that the coarser mesh is adequate to obtain accurate

22
 WORKFLOW | FEA SOFTWARE

information.177 If, however, the two analyses produce different structural engineering. However, because architects are not trained
results, this indicates that the finer mesh is necessary to obtain as extensively in structural concepts as engineers, they need to
accurate information.178 This allows the architect to determine the develop a structural understanding to ensure that they are able to
appropriate element size by comparing the visual output from the interpret the FEA results correctly.
analyses, rather than relying on engineering knowledge to make
this decision. Wind speed and turbulence are relatively easy concepts for
architects to understand, as everyone has felt the wind. This
Specifically addressing the use of FEA in wind loading applications, means that once the wind is visualized by CFD software, the
the software used in this thesis assumes that the forces applied architect running the simulation can interpret the results and
to the model are static, and that they increase slowly and then understand what they mean by drawing on their own experience
remain consistent.179 This is contrary to the dynamic nature of wind of the wind. FEA software, however, simulates structural behaviour
loading, which can change suddenly and intensely. The software and produces results of stress, strain and displacement, which
therefore disregards the dynamic effects of wind and instead are more abstract concepts in which architects are not extensively
simulates average wind loads. trained and with which they may not have direct experience. While
it can be simple for architects to run the analysis, they must have
ARCHITECTURAL APPLICATIONS an understanding of these concepts to be able to interpret the
results of the analysis, and know how they should adjust the model
Most FEA software is intended to be used by engineers to analyze accordingly to improve its structural performance. This may be
and improve the structural properties of an existing model.180 It is achieved through repeated use of the FEA software.182 Software
used as a design tool, as FEA is run continuously throughout the has a lot of potential to allow architects to work within the realm of
design process so that the engineer can understand the structural structural engineering, as it means that they do not need to have
implications of each design iteration.181 In the design method extensive structural knowledge to be able to test many design
developed in this thesis, FEA is also used as a design tool to test iterations. Through these tests, they can learn what the analysis
and refine iterations of building structure, much in the same way as results mean and how changes to the model affect the results. With
it is used by engineers. The FEA software is used to simulate and experience and experimentation facilitated by the FEA software,
represent the effects of combined wind and gravity loading on the the architect will eventually develop structural intuition that will allow
building’s form and structural system. Graphics and animations of them to consider structural behaviour earlier in the architectural
structural behaviour are used instead of numerical data, as they design process.
are more easily understood by architects who are not trained in

23
WORKFLOW | FEA SOFTWARE

SOFTWARE SELECTION
Most CFD and FEA programs are intended for use in engineering, for architects to learn new 3D-modeling software in addition to
rather than architectural, applications. This design method selects the CFD and FEA software, making the design method more
software that can be adapted for use in architectural applications, accessible to architects.
and uses them in a way that is appropriate for the initial architectural
design stages. To ensure that this design method is accessible to 2. FREE EDUCATIONAL LICENSE
architects, many types of CFD and FEA programs were researched The software must be free for students, to encourage its use
to determine which are most appropriate for use in this design in a studio setting. This would allow students to learn these
method. programs while in school, so that they would be proficient in
them when seeking jobs after graduation. This criteria would
REQUIRED CRITERIA make it more feasible for architectural firms to adopt these
programs into their current working methods, since it would
In order for a CFD or FEA software to be considered for use in this be easier to find employees who know how to use them. It
method, it has to meet the following required criteria: is advantageous if the program is also free for commercial
use; however, since architectural firms have greater financial
1. 3D-MODELING SOFTWARE COMPATIBILITY resources than students, and since very few programs are free
The software must either be a plug-in for, or run files from, for commercial use, it is not a requirement.
3D-modeling programs that are commonly used by architects
or geared towards architectural, rather than engineering, EVALUATION FACTORS
applications. Rhinoceros and Revit are two such programs,
so the chosen CFD and FEA programs must be compatible After the elimination of software that did not meet the above criteria,
with at least one of these two programs. Rhinoceros can be the remaining programs were evaluated based on the following
used to create initial building massing models on which wind factors to determine which should be used in this design method:
studies may be performed. The building can then be modeled
in Revit to continue to develop the project and its wind studies 1. EASE
in more detail through the later project phases. Executing the It is advantageous if the software is easy to learn, as architects
design method with these two programs eliminates the need will be more inclined to learn it. This also allows them to
24
 WORKFLOW | SOFTWARE SELECTION

implement the design method sooner, instead of spending develop a methodology that will become more accurate as
more time learning a complex software. CFD and FEA software is improved, so accuracy is not as
important at this stage of the method’s development.
2. SPEED
It is advantageous if the software is able to quickly provide 4. 3D-MODELING FUNCTION
the user with results. This includes considerations of how fast It is advantageous if the software includes 3D-modeling
a model can be set up for evaluation, any geometry clean- functions, because once the CFD or FEA results are obtained,
up that the software might require, and how long it takes the the design can be adjusted accordingly within the same
software to process a result. The faster the program can work, program. This eliminates the need to switch programs to make
the more design iterations can be tested and refined. the changes to the digital model, and then re-export the model
to run the evaluation. This speeds up the design method.
3. ACCURACY
It is advantageous if the software provides accurate analysis 5. PRICE
results. However, this factor is the least valued, as for the It is advantageous if the software is inexpensive, as it is more
purposes of initial design development, it is more advantageous feasible that architectural firms would integrate a design method
to use a program that is easy to learn and that quickly tests that uses inexpensive software into their working practices.
multiple design iterations, rather than use one that is complex
but provides completely accurate data. For example, it is SOFTWARE SELECTION
better for this method to employ a CFD program that is fast
but less accurate, since no CFD software is as accurate as The following CFD and FEA programs were researched and
a physical wind tunnel. The chosen CFD programs therefore considered in relation to these five factors. The diagrams rank
simulate general patterns of wind flow, but are not relied upon each software in terms of the factors, to determine which should
for quantitative results. It is also preferable to select an FEA be used in this design method. The larger the area of the shaded
software that is easy to use, rather than one that provides polygon, the more advantageous the program is for the purpose of
accurate results but is too complex to be feasibly integrated this thesis.
into the initial design stages. The purpose of this thesis is to

25
WORKFLOW | SOFTWARE SELECTION

VASARI VASARI | CFD

Software Type EASE

CFD easy to learn with 
Made By knowledge of Revit
Autodesk
Software Compatibility
Rhino files | Revit files
Intended Use
This program is much like a simplified version of Revit, with
a similar interface and similar tools. It was designed to be PRICE SPEED
used for initial massing, environment, and energy studies of free quick to set up 
wind settings 
architectural projects.183 After the massing has been refined
and process 
according to these studies, the model can be easily exported simulation
to Revit for continued project development. The environmental
tools include a wind tunnel simulator that allows users to
visualize the air flow around digital 3D models. It simulates
only major qualitative wind trends, and is intended to provide
insight into wind patterns at early stages of building massing.184
Although the speed with which results can be obtained is useful MODELING ACCURACY
during early project stages, the designer should be aware that included shows qualitative 
the CFD analysis is not always especially accurate. However, wind trends
this does not negate the usefulness of these results to depict
general wind trends. Vasari can run files from Rhinoceros and Fig. 1.54. Vasari software ranking diagram.
Revit, and also includes a 3D-modeling function within the
program.185

26
 WORKFLOW | SOFTWARE SELECTION

FLOW DESIGN | CFD FLOW DESIGN

EASE Software Type

easy to learn CFD
Made By
Autodesk
Software Compatibility
Rhino files | Revit plug-in
Intended Use
PRICE SPEED Autodesk Flow Design simulates a wind tunnel to allow users
$210/ quick to set up  to visualize the air flow around digital 3D models.186 The
year wind settings 
simplicity of the program’s set-up and operation makes it
and process 
simulation ideal for designers who don’t have the time or the need to
learn a comprehensive CFD program. It simulates only major
qualitative wind trends, and is intended to provide insight into
wind patterns at early stages of building massing.187 Like
Vasari, the speed with which results can be obtained is useful
during early project stages, but the designer should be aware
MODELING ACCURACY that the CFD analysis is not always accurate. The program,
only included as  shows qualitative  however, is still useful in depicting general wind trends. Flow
Revit plug-in wind trends Design can run files from Rhinoceros, or can be used as a
plug-in for Revit.188
Fig. 1.55. Flow Design software ranking diagram.

27
WORKFLOW | SOFTWARE SELECTION

SIMULATION CFD SIMULATION CFD | CFD

Software Type EASE

CFD hard to learn
Made By
Autodesk
Software Compatibility
Revit files
Intended Use
This program provides relatively accurate fluid dynamics PRICE SPEED
$12 999 slow to set up 
analysis tools.189 It is intended for use by professionals who
wind settings 
need to regularly perform complex CFD simulations and and process 
obtain accurate results.190 However, it is still not as accurate simulation
as a physical wind tunnel simulation, so other programs that
are more accessible to architects, even if not as accurate as
Simulation CFD, are preferred for the purposes of this thesis.
Also, it is recommended that users of this program attend a
2-3 day fundamentals course, as well as hire a CFD consultant
for additional help with the program,191 which would likely MODELING ACCURACY
make architectural firms less inclined to integrate this software not included provides quantitative 
into their working methods. Due to the steep learning curve, wind data
as well as intensive hardware requirements and high cost,192
this program has been deemed unrealistic to incorporate Fig. 1.56. Simulation CFD software ranking diagram.
into architects’ design practices and has been removed from
consideration for use in this thesis.

28
 WORKFLOW | SOFTWARE SELECTION

SCAN & SOLVE | FEA SCAN&SOLVE
EASE Software Type
easy to learn FEA
with knowledge of  Made By
Rhino
Intact Solutions
Software Compatibility
Rhino plug-in
Intended Use
Scan&Solve is a plug-in for Rhinoceros that allows the
designer to apply materials, restraints, and loads to the
PRICE SPEED
$995 quick to set  Rhinoceros model, and then evaluates the model’s reaction
up model and  to the simulated forces.193 The program works with native
process  Rhinoceros geometry, and unlike many other FEA programs,
simulation it does not require a separate meshed model in order to
perform the analysis.194 The analysis can be quite accurate,
depending on the grid resolution that is set by the user.195 It is
intended to provide designers with finite element analysis of
both conceptual and detailed models.196

MODELING ACCURACY
included as  can be very good with 
Rhino plug-in high grid resolution

Fig. 1.57. Scan&Solve software ranking diagram.

29
WORKFLOW | SOFTWARE SELECTION

SOFiSTiK RHINO INTERFACE SOFiSTiK RHINO INTERFACE | FEA

Software Type EASE

FEA hard to learn
Made By
SOFiSTiK
Software Compatibility
Rhino plug-in
Intended Use
The SOFiSTiK Rhino Interface integrates Rhinoceros into the PRICE SPEED
$10 800 text-based 
SOFiSTiK environment.197 It creates a mesh from a Rhinoceros
input takes
model on which it then performs finite element analysis.198 This time to
would provide the advantage of allowing architects to use the execute
Rhinoceros interface that they are familiar with to carry out
this analysis. However, although the analysis can be carried
out within Rhinoceros, all loading must be applied via text-
based input, making the program too complex to be feasibly
integrated into architects’ design practices.199 It is intended to
be used by structural engineers,200 so it has been removed MODELING ACCURACY
from consideration for use in this thesis. included as  good
Rhino plug-in

Fig. 1.58. SOFiSTiK Rhino Interface software ranking diagram.

30
 WORKFLOW | SOFTWARE SELECTION

SIMULATION MECHANICAL | FEA SIMULATION MECHANICAL

EASE Software Type

hard to learn FEA
Made By
Autodesk
Software Compatibility
Rhino files
Intended Use
PRICE SPEED Simulation Mechanical allows designers to both model
$24 999 slow to set up  geometry and run detailed FEA simulations on that geometry
model and 
within the same program.201 Due to the steep learning curve,
process 
simulation large hardware requirements, and high cost,202 this program
has been deemed unrealistic to incorporate into architects’
design practices and has been removed from consideration
for use in this thesis.

MODELING ACCURACY
included good

Fig. 1.59. Simulation Mechanical software ranking diagram.

31
WORKFLOW | SOFTWARE SELECTION

SELECTED PROGRAMS
Based on the evaluation of the five factors in relation to the
considered software programs, Autodesk Vasari and Autodesk
Flow Design are the selected CFD programs to be used in the
design method, and Scan&Solve is the selected FEA program.

FLOW DESIGN
In this design method, Flow Design’s flow line animations (Fig. 1.60) Fig. 1.60. Flow line animation from Flow Design.
and colour gradients representing the wind pressure acting on the
model surface (Fig. 1.61) are used to visualize wind turbulence and
pressure on and around the building.

Fig. 1.61. Wind pressure colour gradient from Flow Design.

32
 WORKFLOW | SELECTED PROGRAMS

VASARI
Vasari’s horizontal data slices depicting wind speed (Fig. 1.62) are
used to evaluate the surrounding wind conditions that are created
by each building form iteration.

Fig. 1.62. Wind speed data slice from Vasari.

SCAN&SOLVE
Colour gradients representing the displacement of each point of the
model under load overlaid onto deflection animations of the model
under load (Fig. 1.64), is the most useful output from Scan&Solve
for this design method. Scan&Solve also provides visualizations
of the input loading, represented by the blue arrows in Fig. 1.63,
with red arrows added by the author to show the reactions to the
Fig. 1.63. Input loading (blue) and Fig. 1.64. Displacement colour loading.
reactions (red) in Scan&Solve. gradient overlaid on deflection
animation from Scan&Solve.

33
WORKFLOW | ACCESSIBILITY OF METHOD

ACCESSIBILITY OF METHOD
Most CFD and FEA software is designed for use by engineers, rather training than engineers that would allow them to use complex
than architects. However, software is a valuable tool for architects engineering software. It is therefore critical to select simple
as it allows them to integrate engineering considerations into their programs that can be easily learned and integrated into current
design methods. This may only be made possible if software is architectural working practices.
selected that could be feasibly integrated into current architectural
working practices, so that architects do not have to substantially 2. QUICK TO RETURN RESULTS
alter their current methods to be able to incorporate the use of The programs used in this thesis can all quickly provide the
the software. The design method that is developed in this thesis is user with results. This allows more design iterations to be
therefore made accessible to architects by selecting software that tested and refined because the software can perform the
is normally intended for engineering use, and appropriating it for calculations and simulations much faster than if the architect
the initial architectural design stages. The selected programs have were to have to learn to do them manually. For the purpose of
the following qualities that make them appropriate for architectural initial architectural design, it is more advantageous to be able
purposes, and make them potentially easy to integrate within to test these design iterations quickly and refine them based
architects’ current working methods: on general results provided by the software, instead of using
a software that provides unnecessarily detailed results for this
1. EASY TO LEARN design stage and takes much longer to do so. The speed with
The selected programs are easy to learn even without previous which the results may be obtained from each of the programs
experience with similar types of software. This will make also allows the architect to test enough iterations to eventually
architects more inclined to take the time to learn them, and develop an intuition of what these results will be, even before
allow them to spend less time learning the software and more the simulations are run.
time using it to test their design iterations. Within the field of
engineering, it is generally good practice to use the simplest 3. AUTOMATE PROCESSES
software that also fulfills the analysis requirements for which While many CFD and FEA programs require the user to manually
it is needed.203 This practice is especially advantageous when adjust many of the settings, the chosen programs automate a
engineering software is being used in the early architectural lot of these processes to simplify the operation of the software
design stages, as architects have less knowledge and and reduce the knowledge that is required from the user to
34
 WORKFLOW | ACCESSIBILITY OF METHOD

run the simulations. For example, the chosen CFD software 5. REQUIRE TYPICAL HARDWARE
automatically sets the grid resolution, so that the user only has None of the chosen programs have intense hardware
to input wind speed and direction. Similarly, the complicated requirements, and can even be run on a laptop if necessary.
and time-consuming meshing process that must be done This means that architectural firms wanting to integrate these
manually in many FEA programs is automated in the chosen programs into their working practices will be able to do so with
FEA software. This means that architects do not have to spend their current computers.
time learning how to mesh their models. This automation
of complicated processes allows architects to integrate 6. COMPATIBLE WITH ARCHITECTURAL 3D-MODELING
engineering considerations into their current architectural SOFTWARE
practices, without having to acquire extensive knowledge All of the programs used in this method are compatible with
about engineering principles and software operation. 3D-modeling programs that are already commonly used
by architects. This allows architects to use their current
4. PROVIDE VISUAL OUTPUT 3D-modeling software to design the building, and then test the
The selected programs provide visual output, rather than only digital model with the CFD and FEA software. This eliminates
numerical output, to convey the results of the simulations. the need for architects to learn new 3D-modeling software that
These outputs include graphics and animations that provide is compatible with the CFD and FEA software.
qualitative, rather than quantitative, wind and structural
information. Visual output is easier than numerical output for 7. FREE EDUCATIONAL LICENSES AVAILABLE
architects to understand and interpret because they are more All of the programs provide free educational licenses, so that
familiar with visual media. Additionally, qualitative information students may learn these programs while in school to be able
and general trends are more useful than exact results for to use them when they enter the workforce. This would allow
informing the early design stages. The design method architectural firms to hire employees who already know how
developed for this thesis informs architects of the visual output to use these programs, to be able to integrate them into their
that is best able to provide them with the necessary information current design methods.
to make design revisions, so that they may focus only on the
output that is relevant to their purposes.
35
WORKFLOW | SOFTWARE SEQUENCE

SOFTWARE SEQUENCE
The three programs that have been selected for use in this design that the surrounding wind patterns are altered desirably by the
method are all compatible with each other, as files from each of adjusted building form.
the programs can either be opened in, or exported to, any of the
other programs. This is necessary because many of the steps in 4. The building model is then exported from Vasari and tested
the design method need to be repeated as the building model is in Flow Design to evaluate the aerodynamics of the building
refined and re-tested, requiring the digital model to be opened and form. This may be done with the information provided by Flow
edited in all three programs throughout the design process. This Design about the wind pressure that is exerted on each building
sequence is shown in Fig. 1.65, along with the file formats and face. The architect can change the building form in Vasari to
conversion processes that should be used to maintain compatibility reduce the wind pressure acting on it, and then re-export
among the programs. and re-test the model in Flow Design to ensure that improved
aerodynamics have been achieved without compromising the
1. For each building design iteration, Vasari is first used to model surrounding wind conditions.
the building form.
5. Next, the Vasari model is exported to Rhino, in which a
2. Vasari is then used to simulate the speed of the wind flow structural bay is modeled that will be repeated throughout the
around the Vasari model. This tests the wind conditions that building form.
will be created around the building. The model may be adjusted
within Vasari and re-tested in the simulated wind conditions as 6. Finite element analysis is then run on the Rhino model of the
necessary. structural bay, as the wind pressure information that is provided
by Flow Design is input into Scan&Solve to predict how the bay
3. The building model is then exported to Flow Design, which is will react to combined wind and gravity loading. The model of
used to evaluate the turbulence around the building model with the bay may be adjusted as needed within Rhino, and then the
flow line animations. The architect can change the building finite element analysis may be re-run on the adjusted model
form in Vasari to reduce the wind turbulence as necessary, and until it is structurally adequate.
then re-export and re-test the model in Flow Design to ensure
36
 WORKFLOW | SOFTWARE SEQUENCE

MODEL 

1
BUILDING 

2
FORM
VASARI .rv WIND SPEED
t  
.rvt    
E CFD
STRUCTURAL ANALYSIS

XP
.rvt 
VASARI

OR
BAY FEA    E

6
t
.sa XP
        .rvt

T F
SCAN&SOLVE   OR .rvt
O

RO
IN E AS T
.3dm

M V
V

IT,  H

 F

.rv
EV  R

RO
SA

ASA
t   
O R OM

M 

  EX
.sat    IMPORT INT T FR

RI    
VA
L
DE

SAR

POR
OR

.sat
Y MO

I    .d
.3dm     EXP

T FROM 
wg    RUN MESH
L BA

ODEL

VASARI     .
A
TUR

G M
C

sat
TON
IN
U
MODEL 

5 TR

SS
A

U
STRUCTURAL S
M

RB 
WIND

3
BAY

AN
TURBULENCE

D
 C
RHINO RE
CFD
AT
.3dm
m ES
3d OLI
     .
ANALYSIS
IN RHIN AS D 
O, SAVE 
FLOW DESIGN
.sat
STEP WIND 

4
SOFTWARE FILE C PRESSURE
ONV
file extension ERS CFD 
ION ANALYSIS
FLOW DESIGN
.sat

LEGEND Fig. 1.65. Software sequence diagram.

37
 PROGRAM AND SITE

39
 PROGRAM AND SITE | PROGRAM AND SITE CONDITIONS

PROGRAM AND SITE CONDITIONS

The thesis does not serve to propose a specific building. Instead, primarily to the creation and resistance of wind conditions. The site
it uses the design of a small, low-rise building as a means of was chosen to provide an extreme context within which to develop
developing the proposed design method that may be applied to the design method.
a multitude of building, program and site requirements. As such,
the exterior programs are not a program proposal, but a means of Once the wind and spatial conditions are defined with the chosen
providing a variety of wind conditions to be created with the building program and site, the design method is developed within these
form. Similarly, the building form does not reflect interior program specified conditions, using a single building on a simple site to
placement or size requirements, but is instead a form that responds provide a basis for potentially more complex contexts.

41
PROGRAM AND SITE | PROGRAM

PROGRAM
The exterior program activities (Fig. 2.1) serve to provide a variety of FAST WIND
wind condition requirements to be created with the building form,
HORIZONTAL AXIS WIND TURBINE
and are not a program proposal in themselves. These programs
consist of pairings of wind energy generation technologies and
seasonal sports that require specific wind conditions. To support COMPACT WIND ACCELERATION TURBINE
the exterior programs, the building itself could accommodate
energy storage, space to record and compare energy generation KITE FLYING
data from the wind energy generation technologies to be tested on
the site, as well as equipment storage and change room facilities SNOWKITING
to support the sports programs. The building form does not reflect
interior building design, program sizes or program placement SNOW WINDSURFING
requirements, but is instead a form that creates the wind conditions
that are required for the exterior programs.
SLOW WIND
These exterior programs have been diagrammed and sorted BLADE TIP POWER SYSTEM
according to the area that they each require (Fig. 2.2). This diagram
serves as a tool to develop different combinations of programs FIELD SPORTS
for different design iterations, as it conveys which programs have
similar wind and spatial requirements, and which can occur within
SNOW BUILD-UP FOR SLEDDING
the same physical space in alternating seasons. Drawings have
also been made that document the detailed wind and spatial
requirements for each of the exterior programs (Fig. 2.3-Fig. MINIMAL WIND
2.12). These drawings may be referred to by the architect when TENNIS AND BADMINTON
developing building forms to create these required exterior spaces
and wind conditions.
TURBULENT WIND
VERTICAL AXIS WIND TURBINE Fig. 2.1. Exterior program list.

42 YEAR-ROUND WIND ENERGY GENERATION PROGRAM
SUMMER SPORTS PROGRAM
WINTER SPORTS PROGRAM
MINIMAL WIND
TENNIS AND BADMINTON

TURBULENT WIND PROGRAM AND SITE | PROGRAM

VERTICAL AXIS WIND TURBINE
Fig. 2.2. Exterior program diagram.

YEAR-ROUND WIND ENERGY GENERATION PROGRAM
230 000m2
SUMMER SPORTS PROGRAM
WINTER SPORTS PROGRAM

MIN  MAX
SPEED AREA SPEED 29 000m2

265m2

1 600m2

3 600m2
2 300m 2

260m2 1 500m2

3 500m2

7 200m2

WIND SPEED
km/h 0 5 10 15 20 25 30 35 40 45 50 55 60 65

43
PROGRAM AND SITE | PROGRAM

d 3d - 5d
MIN 20h
MIN 10m
h
2h

3d - 5d 20h

ZONE OF INCREASED 
TURBULENCE

6d - 10d

slow

fast
SPEED
IDEAL WIND SPEED km/h

15 120

HORIZONTAL AXIS WIND TURBINE
Fig. 2.3. Horizontal axis wind turbine wind and spatial requirements.

44
 PROGRAM AND SITE | PROGRAM

28m

MIN 20h

61m

slow

fast
SPEED
IDEAL WIND SPEED km/h

20+

COMPACT WIND ACCELERATION TURBINE
Fig. 2.4. Compact wind acceleration turbine wind and spatial requirements.

45
PROGRAM AND SITE | PROGRAM

10m

MIN 4d MIN 4d
slow
MIN 4d MIN 4d
fast
SPEED MIN 4d

IDEAL WIND SPEED km/h

15 40

VERTICAL AXIS WIND TURBINE
Fig. 2.5. Vertical axis wind turbine wind and spatial requirements.

46
 PROGRAM AND SITE | PROGRAM

>0m

1.8m
>0m
MIN 10m

MIN 10m

slow

fast
SPEED
IDEAL WIND SPEED km/h

10 60

BLADE TIP POWER SYSTEM
Fig. 2.6. Blade tip power system wind and spatial requirements.

47
PROGRAM AND SITE | PROGRAM

IDEAL WIND SPEED km/h
IDEAL WIND SPEED km/h
15 40
10 25

MIN 3h 5

1 45°

2 4

STRING
LENGTH
slow

30°
fast
SPEED

KITE FLYING 30°

Fig. 2.7. Kite flying wind and spatial requirements.

48
 PROGRAM AND SITE | PROGRAM

MIN 10h

8h

slow

fast
SPEED
IDEAL WIND SPEED km/h

05

FIELD SPORTS
Fig. 2.8. Field sports wind and spatial requirements.

49
PROGRAM AND SITE | PROGRAM

MIN 10h
8h

slow

fast
SPEED
IDEAL WIND SPEED km/h

TENNIS AND BADMINTON
Fig. 2.9. Tennis and badminton wind and spatial requirements.

50
 PROGRAM AND SITE | PROGRAM

50% FENCE 
DENSITY

70% FENCE 
DENSITY

85% FENCE 
DENSITY
10h
100% FENCE 
DENSITY
6h

h 4h

4h

4h

35h

slow

fast
SPEED
IDEAL WIND SPEED km/h

15+

SNOW BUILD-UP FOR SLEDDING
Fig. 2.10. Snow build-up for sledding wind and spatial requirements.

51
PROGRAM AND SITE | PROGRAM

MIN 3h

slow

fast
SPEED
IDEAL WIND SPEED km/h

10 50

SNOW WINDSURFING
Fig. 2.11. Snow windsurfing wind and spatial requirements.

52
 PROGRAM AND SITE | PROGRAM

35-60°

h
MIN 45m

MIN 3h

MIN 45m
MIN 45m

slow

fast
SPEED
IDEAL WIND SPEED km/h

10 50

SNOWKITING
Fig. 2.12. Snowkiting wind and spatial requirements.

53
PROGRAM AND SITE | SITE

SITE
The site on which the design method has been developed (Fig. 22
24 24

2.17) is an open field on the outskirts of southern Regina. 20 20 20 20 20 20 20 AVERAGE ANNUAL WIND SPEED

18 18
20 km/h

Regina is the fifth windiest city in Canada, with an average annual

wind speed of 20 km/h (Fig. 2.13).1 These high wind loads provide
an extreme context within which to develop the design method. J F M A M J J A S O N D Fig. 2.13. Regina’s average annual wind speed.

The flat, open site (Fig. 2.14) provides an opportunity to manipulate

the wind with only the building form, as there are minimal site
conditions to alter the wind immediately around the building. The
large site is adjacent to a residential neighbourhood so the public Fig. 2.14. A flat, open site was chosen.
can easily access the exterior sports programs.

Of the top five windiest cities, it has the largest population for use
of the exterior programs (Fig. 2.15).2 238 000 Fig. 2.15. Regina’s population.

The site is located near the University of Regina (Fig. 2.16), whose
existing wind turbine studies3 could potentially pair with the wind
energy generation technologies that surround the building. Fig. 2.16. The site is near the University of Regina.

54
 PROGRAM AND SITE | SITE

SITE PLAN
0 500m

PROPOSED SITE
0 500m

Y
UNIVERSITY 

KWA
OF REGINAPROPOSED SITE
SASKATCHEWAN 

Y
UNIVERSITY 

A PA

KWA
POLYTECHNIC
OF REGINA
RESIDENTIAL 

CAN
SASKATCHEWAN 

R
BUILDINGS

A PA
POLYTECHNIC
COMMERCIAL 

WAS
RESIDENTIAL 

CAN
BUILDINGS
BUILDINGS
COMMERCIAL 

WAS
FARM FIELDS
BUILDINGS
ALBERT STREET
FARM FIELDS

ALBERT STREET

Y
WA
GH

Y
I
 H

WA
DA

GH
NA

I
 H
CA

DA
S-

NA
AN

CA
TR

S-
AN
TR

Fig. 2.17. Site plan.

55
 N

PROGRAM AND SITE | SITE

ANNUAL WIND ROSE

AMOUNT OF TIME WIND 
WIND SPEED m/s BLOWS FROM DIRECTION
10+
8-9
6-8
5-6
4-5
3-4
1-3
0-1 LESS MORE

AMOUNT OF TIME WIND 
WIND SPEED m/s BLOWS FROM DIRECTION
10+
8-9
6-8
5-6
4-5
3-4
1-3
0-1 LESS MORE
Fig. 2.18. Annual wind rose for the chosen site.

AMOUNT OF TIME WIND 
WIND SPEED m/s BLOWS FROM DIRECTION 56
10+
8-9
6-8
5-6
 PROGRAM AND SITE | SITE

MONTHLY WIND ROSES

N
25%
20%
15%
10%
5%

MONTH JANUARY FEBRUARY MARCH APRIL

AVERAGE WIND SPEED 22 km/h 20 km/h 22 km/h 24 km/h

MAY JUNE JULY AUGUST

24 km/h 22 km/h 18 km/h 18 km/h

Fig. 2.19. Monthly wind roses SEPTEMBER OCTOBER NOVEMBER DECEMBER

for the chosen site. 20 km/h 20 km/h 20 km/h 20 km/h

57
 FORM AFFECTS WIND

59
FORM AFFECTS WIND | HOW FORM AFFECTS WIND

HOW FORM AFFECTS WIND

As air flows over surfaces with a lot of variation, it is altered more than manipulation of the wind should be done to create wind conditions
when it flows over smooth surfaces.1 Forms that prevent a surface around the building that are appropriate for the exterior programs.
from being smooth get in the way of the wind and complicate its For each building design iteration, CFD software is used to simulate
flow by speeding it up, slowing it down, or creating turbulence.2 the speeds and patterns of wind flow around the initial building form
Such surface features include buildings. design. The speed with which these simulation results are provided
allows the architect to refine and re-test many iterations of their
The first step in the design method is to design a building so that its design until the building form creates the desired wind conditions.
form interacts with the wind to alter its speed and flow patterns. This

60
 FORM AFFECTS WIND | WIND EFFECTS LIBRARY

WIND EFFECTS LIBRARY

Manipulations of building form to increase and decrease wind able to depict the downwash effect, as the programs assume a
speed, turbulence, and pressure were studied from a broad range uniform wind speed and do not consider the higher wind speeds
of published sources3 and compiled into a library of wind effects. that exist in reality at higher elevations. It was also learned through
For each technique, a wind effect that has been studied in real these simulations that Vasari’s horizontal data slices are more
wind environments is compared with the results of the simulation detailed than those from Flow Design, but Flow Design’s flow line
of the effect in both Vasari and Flow Design. Architects may refer animations and colour gradients representing the wind pressure
to this library of effects to alter the building geometry in order to acting over the model surface are more accurate than those from
create the specific wind conditions that are required for the exterior Vasari. While CFD software can be a useful tool for quick tests of
programs. The software simulations of each of the effects allow the design iterations, the architect has to know how to interpret the
architect to become familiar with what each of the CFD programs results and be able to tell if they are an accurate representation of
can accurately represent, and what they don’t consider when the simulated effects, by possessing some knowledge of studied
computing results. For example, neither Vasari nor Flow Design is wind effects.

61
 FORM AFFECTS WIND | WIND EFFECTS LIBRARY

BREEZEWAY EFFECT
INCREASE WIND SPEED

Fig. 3.2. Vasari simulation - effect not shown.

Fig. 3.1. Breezeway effect.

WIND SPEED
INCREASED
slow

AVERAGE

fast
DECREASED
SPEED Fig. 3.3. Flow Design simulation - effect not shown.

62
fast
SPEED
 FORM AFFECTS WIND | WIND EFFECTS LIBRARY

CHANNEL EFFECT
INCREASE WIND SPEED

Fig. 3.5. Vasari simulation.

Fig. 3.4. Channel effect.

slow

WIND SPEED
fast INCREASED
low slow SPEED

CHANNEL EFFECT
high 
AVERAGE
PRESSURE

fast
DECREASED
SPEED Fig. 3.6. Flow Design simulation - effect not shown.

63
FORM AFFECTS WIND | WIND EFFECTS LIBRARY

CHANNELLING EFFECT
INCREASE WIND SPEED

Fig. 3.8. Vasari simulation.

Fig. 3.7. Channelling effect.

slow

WIND SPEED
low
fast
INCREASED
SPEED
slow
CHANNELLING EFFECT
high 

PRESSURE
AVERAGE

fast
DECREASED
SPEED Fig. 3.9. Flow Design simulation.

64
 FORM AFFECTS WIND | WIND EFFECTS LIBRARY

COMBINED ROW AND DOWNWASH EFFECT

INCREASE WIND SPEED

4h

Fig. 3.11. Vasari simulation - effect not shown.

Fig. 3.10. Combined row and downwash effect.

WIND SPEED
INCREASED
slow
slow

AVERAGE
fast
low
SPEED COMBINED ROW AND 
fast
DECREASED
SPEED
DOWNWASH EFFECT
high  Fig. 3.12. Flow Design simulation - effect not shown.
PRESSURE
65
FORM AFFECTS WIND | WIND EFFECTS LIBRARY

CORNER EFFECT
INCREASE WIND SPEED

- PRESSSURE

+ PRESSSURE

w Fig. 3.14. Vasari simulation.

ZONE OF INCREASED 
WIND SPEED

Fig. 3.13. Corner effect.

slow

WIND SPEED
fast INCREASED
low
slow SPEED

CORNER EFFECT
high 

PRESSURE AVERAGE

fast
DECREASED
SPEED Fig. 3.15. Flow Design simulation.

66
 FORM AFFECTS WIND | WIND EFFECTS LIBRARY

CUMULATIVE EFFECT
INCREASE WIND SPEED

MAX (2/3)h

Fig. 3.17. Vasari simulation - effect not shown.

Fig. 3.16. Cumulative effect.

slow

WIND SPEED
fast INCREASED
low
slow SPEED

CUMULATIVE EFFECT
high 

PRESSURE AVERAGE

fast
DECREASED
SPEED Fig. 3.18. Flow Design simulation.

67
FORM AFFECTS WIND | WIND EFFECTS LIBRARY

DIVERTING EFFECT
INCREASE WIND SPEED

Fig. 3.20. Vasari simulation - effect not shown.

Fig. 3.19. Diverting effect.

slow

fast WIND SPEED

low
SPEED INCREASED

DIVERTING EFFECT
high 
slow
PRESSURE

AVERAGE

fast
DECREASED
SPEED Fig. 3.21. Flow Design simulation - effect not shown.

68
 FORM AFFECTS WIND | WIND EFFECTS LIBRARY

DOWNWASH EFFECT
INCREASE WIND SPEED

STAGNATION 
POINT

(2/3)h Fig. 3.23. Vasari simulation - effect not shown.

Fig. 3.22. Downwash effect.

slow WIND SPEED

INCREASED
slow fast
low
SPEED
AVERAGE
DOWNWASH EFFECT
high 

PRESSURE

fast
DECREASED
SPEED Fig. 3.24. Flow Design simulation - effect not shown.

69
FORM AFFECTS WIND | WIND EFFECTS LIBRARY

FUNNELING EFFECT
INCREASE WIND SPEED

Fig. 3.26. Vasari simulation.

Fig. 3.25. Funneling effect. WIND SPEED

INCREASED
slow
FUNNELING EFFECTAVERAGE
FUNNELING EFFECT
fast
DECREASED
SPEED Fig. 3.27. Flow Design simulation.

70
 FORM AFFECTS WIND | WIND EFFECTS LIBRARY

STAGGERING EFFECT
INCREASE WIND SPEED

Fig. 3.29. Vasari simulation.

Fig. 3.28. Staggering effect.

WIND SPEED
INCREASED
slow slow

AVERAGE
fast
low
SPEED
fast
DECREASED
high  SPEED
STAGGERING EFFECT
Fig. 3.30. Flow Design simulation - effect not shown.
PRESSURE
71
FORM AFFECTS WIND | WIND EFFECTS LIBRARY

VENTURI EFFECT
INCREASE WIND SPEED

MAX 2h

Fig. 3.31. Venturi effect.

h
Fig. 3.33. Vasari simulation.

slow

fast
low
SPEED

VENTURI EFFECT
high 

PRESSURE
Fig. 3.32. Venturi effect induces suction.
WIND SPEED
INCREASED
slow

AVERAGE

fast
DECREASED
SPEED Fig. 3.34. Flow Design simulation.

72
 FORM AFFECTS WIND | WIND EFFECTS LIBRARY

CUMULATIVE EFFECT
DECREASE WIND SPEED

MAX (2/3)h

Fig. 3.36. Vasari simulation - effect not shown.

Fig. 3.35. Cumulative effect.

slow

WIND SPEED
fast
low INCREASED
SPEED
slow
CUMULATIVE EFFECT
high 

PRESSURE AVERAGE

fast
DECREASED
SPEED Fig. 3.37. Flow Design simulation - effect not shown.

73
FORM AFFECTS WIND | WIND EFFECTS LIBRARY

POROUS WINDBREAK EFFECT

DECREASE WIND SPEED

D 
ASE
D E CRE
F  PEED
E O S Fig. 3.39. Vasari simulation.
ZON WIND 

h 6h - 8h

MAX w w

Fig. 3.38. Porous windbreak effect.

slow

WIND SPEED
fast INCREASED
low
slow SPEED POROUS WINDBREAK 
EFFECT
high 

PRESSURE AVERAGE

fast
DECREASED
SPEED Fig. 3.40. Flow Design simulation.

74
 FORM AFFECTS WIND | WIND EFFECTS LIBRARY

REDUCTION OF DOWNWASH EFFECT

DECREASE WIND SPEED

Fig. 3.41. Reduction of downwash effect TOWER AND PODIUM

- podium.
TOWER AND PODIUM

Fig. 3.44. Vasari simulation - effect not shown.

CANOPY
CANOPY SETBACK
SETBACK
Fig. 3.43. Reduction of downwash
Fig. 3.42. Reduction of downwash
effect
slow - canopy.
slow effect - setback.
WIND SPEED
fast INCREASED
REDUCTION OF 
fast
low
low
slow SPEED
SPEED

DOWNWASH EFFECT
high  
high

PRESSURE
PRESSURE
AVERAGE

fast
DECREASED
SPEED Fig. 3.45. Flow Design simulation - effect not shown.

75
FORM AFFECTS WIND | WIND EFFECTS LIBRARY

SOLID WINDBREAK EFFECT

DECREASE WIND SPEED

ZONE OF 
DECREASED SPEED

MAX 30h
Fig. 3.47. Vasari simulation.

Fig. 3.46. Solid windbreak effect.

WIND SPEED
slow INCREASED
slow
fast
low
SPEED SOLID WINDBREAK 
AVERAGE

EFFECT
high 

PRESSURE fast
DECREASED
SPEED Fig. 3.48. Flow Design simulation.

76
 FORM AFFECTS WIND | WIND EFFECTS LIBRARY

COMBINED ROW AND DOWNWASH EFFECT

INCREASE WIND TURBULENCE

4h

Fig. 3.50. Vasari simulation.

Fig. 3.49. Combined row and downwash effect.

WIND SPEED
INCREASED
slow

slow AVERAGE

fast
fast DECREASED
low
SPEED COMBINED ROW AND Fig. 3.51. Flow Design simulation.
SPEED

DOWNWASH EFFECT
high  77
PRESSURE
FORM AFFECTS WIND | WIND EFFECTS LIBRARY

COURTYARD EFFECT
INCREASE WIND TURBULENCE

h
>2.4h h
>2.4h
h
>2.4h

1.4h-2.4h
1.4h-2.4h h
h
1.4h-2.4h
h

<1.4h
<1.4h h
h
<1.4h
Fig. 3.52. Courtyard effect.
h WIND SPEED
slow
INCREASED
slow
slow

slow

low
fast
AVERAGE
fast
low SPEED
fast SPEED

COURTYARD EFFECT
high 
low fast
COURTYARD EFFECT
high  SPEED
PRESSURE DECREASED

COURTYARD EFFECT
high  PRESSURE
SPEED Fig. 3.53. Flow Design simulation (effect not shown in Vasari).
PRESSURE
78
 FORM AFFECTS WIND | WIND EFFECTS LIBRARY

DOWNWASH EFFECT
INCREASE WIND TURBULENCE

STAGNATION 
POINT

(2/3)h
Fig. 3.55. Vasari simulation - effect not shown.

Fig. 3.54. Downwash effect.

slow
WIND SPEED
INCREASED
fast
low slow SPEED

DOWNWASH EFFECT
high 
AVERAGE
PRESSURE

fast
DECREASED
SPEED Fig. 3.56. Flow Design simulation - effect not shown.

79
 FORM AFFECTS WIND | WIND EFFECTS LIBRARY

ROW EFFECT
INCREASE WIND TURBULENCE

Fig. 3.58. Vasari simulation - effect not shown.

MIN 8h

Fig. 3.57. Row effect.

WIND SPEED
INCREASED
slow
slow

fast
AVERAGE
low
SPEED
fast
ROW EFFECT
high  DECREASED
PRESSURE SPEED Fig. 3.59. Flow Design simulation.

80
 FORM AFFECTS WIND | WIND EFFECTS LIBRARY

VORTEX SHEDDING
INCREASE WIND TURBULENCE

WIND SPEED
INCREASED

AVERAGE

DECREASED
Fig. 3.60. Vortex shedding. Fig. 3.61. Flow Design simulation (flow lines not shown in Vasari).

slow

low
slow fast
SPEED

VORTEX SHEDDING
high 

PRESSURE

fast
SPEED

81
FORM AFFECTS WIND | WIND EFFECTS LIBRARY

WAKE EFFECT
INCREASE WIND TURBULENCE

ZONE OF 
Fig. 3.63. Vasari simulation.
INCREASED WIND 
w TURBULENCE

h
w

Fig. 3.62. Wake effect.

slow
WIND SPEED
INCREASED
low slow fast
SPEED

AVERAGE
WAKE EFFECT
high 

PRESSURE

fast
DECREASED
SPEED Fig. 3.64. Flow Design simulation.

82
 FORM AFFECTS WIND | WIND EFFECTS LIBRARY

COURTYARD EFFECT
DECREASE WIND TURBULENCE

h
>2.4h h
>2.4h
h
>2.4h

1.4h-2.4h
1.4h-2.4h h
h
1.4h-2.4h
h

<1.4h
<1.4h h
h
<1.4h
Fig. 3.65. Courtyard effect.
h WIND SPEED
slow
INCREASED
slow
slow

slow

low
fast
AVERAGE
fast
low SPEED
fast SPEED

COURTYARD EFFECT
high 
low fast
COURTYARD EFFECT
high  SPEED
PRESSURE DECREASED

COURTYARD EFFECT
high  PRESSURE
SPEED Fig. 3.66. Flow Design simulation (effect not shown in Vasari).
PRESSURE
83
FORM AFFECTS WIND | WIND EFFECTS LIBRARY

DISTANCE FROM OBSTRUCTION

DECREASE WIND TURBULENCE

ZONE OF INCREASED 
TURBULENCE
Fig. 3.68. Vasari simulation.

h
MIN 3h

Fig. 3.67. Distance from obstruction.

slow

WIND SPEED
fast

DISTANCE FROM 
low INCREASED
SPEED
slow
OBSTRUCTION
high 

PRESSURE
AVERAGE

fast
DECREASED
SPEED Fig. 3.69. Flow Design simulation.

84
 FORM AFFECTS WIND | CFD ITERATIONS TO CREATE WIND CONDITIONS

CFD ITERATIONS TO CREATE WIND CONDITIONS

For the first step in the design method, a building form is modeled to the wind effects library, which catalogues ways of manipulating
and tested with CFD software to visualize the wind speeds and form to increase or decrease the surrounding wind speed and
patterns that are created around the building by the building form, turbulence. Iterations may also be tested in Flow Design to look at
and evaluate the appropriateness of these wind conditions for the the flow lines, which depict the wind turbulence that is generated
exterior programs that are intended to be accommodated. This by the building. Although this adds an extra step to the process, it
can be represented by Vasari’s horizontal data slices of wind speed is beneficial to ensure that there is no undesired turbulence created
that depict general wind patterns. After the first building iteration is by the building form. As CFD programs are improved in the future,
tested with the CFD software, the architect makes observations and the accuracy of this process will increase. To test this first step in the
adjusts the form within Vasari for the next iteration to improve the method, a building form was developed through eleven iterations,
appropriateness of the surrounding wind conditions for the exterior improving the surrounding wind conditions with each iteration.
programs. These adjustments to the form may be made by referring

85
FORM AFFECTS WIND | CFD ITERATIONS TO CREATE WIND CONDITIONS

WIND FROM MOST PREDOMINANT DIRECTION

The first six iterations were developed while only considering wind
coming from the predominant southeast direction. The second-
most predominant wind direction on the site was disregarded for
these first six iterations to allow the architect to become familiar
with the iterative process and the ways in which building form can
influence the wind within a simple site condition. These first six
iterations are depicted on the following pages.

86
 FORM AFFECTS WIND | CFD ITERATIONS TO CREATE WIND CONDITIONS

ITERATION 1

begin with two volumes to create sheltered areas

of decreased wind speed, as well as a channel of
increased wind speed between the volumes
channel of increased speed curves towards the shorter
form; in next iteration, test forms of the same length to
see if it affects the direction of the channel

area of decreased wind speed

WIND FROM SE
Fig. 3.70. Iteration 1 Vasari screenshot.

ITERATION 2

forms have been made the same length

channel of increased speed has straightened
channel of increased speed is short; in next iteration,
see if increasing the space between the forms
lengthens the channel

WIND SPEED
INCREASED

WIND FROM SE AVERAGE

Fig. 3.71. Iteration 2 Vasari screenshot.
DECREASED

87
FORM AFFECTS WIND | CFD ITERATIONS TO CREATE WIND CONDITIONS

ITERATION 3

space between forms has been increased

channel of increased speed is longer and faster; in
next iteration, see if slightly smaller opening between
forms will still create the same type of channel
turbulent wind on the leeward side of the forms; in next
iteration, make forms longer to see if they create calm
leeward zones

WIND FROM SE
Fig. 3.72. Iteration 3 Vasari screenshot.

ITERATION 4

forms lengthened to create calm leeward zones

space between forms has been made smaller, but is

still adequate to create a channel of increased wind
speed
sheltered area on the leeward side; in next iteration, try
to create a sheltered area on the windward side of the
form for snow build-up

WIND SPEED
INCREASED

WIND FROM SE AVERAGE

Fig. 3.73. Iteration 4 Vasari screenshot.
DECREASED

88
 FORM AFFECTS WIND | CFD ITERATIONS TO CREATE WIND CONDITIONS

ITERATION 5
sheltered windward area isn’t very large; in next
iteration, create a porous wall for snow build-up on its
leeward side instead
channel of increased wind speed curves into the zone
of increased speed at the side of the forms
don’t need this much building area; in next iteration
replace some of the building form with walls

WIND FROM SE
Fig. 3.74. Iteration 5 Vasari screenshot.

ITERATION 6

porous wall reduces wind speed and lets snow through

for snow build-up on the leeward side
some building mass was changed to freestanding
walls which decrease wind speed on the leeward side
this leeward zone isn’t very sheltered; in next iteration,
curve wall around more to shelter the zone from the
wind
WIND SPEED
INCREASED

WIND FROM SE AVERAGE

Fig. 3.75. Iteration 6 Vasari screenshot.
DECREASED

89
FORM AFFECTS WIND | CFD ITERATIONS TO CREATE WIND CONDITIONS

WIND FROM ALL PREDOMINANT DIRECTIONS

The site on which this building form is developed has two The wind energy generation technologies, unlike the sports
predominant wind directions: northwest and southeast.4 Since the programs, must remain stationary on the site. Because they are
wind comes from the southeast more often5, the first six iterations more limited, their optimal wind conditions are created when the
of the building form have been designed to create certain wind wind comes from the southeast. When the wind comes from the
conditions when the wind blows from only this direction. However, northwest, the conditions that are created around each energy
the northwest winds are nearly as predominant, so the next step generation technology are either the same as the conditions from
in the design of the building form should be to adjust it so that the the southeast, or provide different conditions within which the
desired wind conditions for the exterior programs will be created by technology may be tested. This provides an opportunity to study
the form when the wind blows from either of these two directions. and compare the energy generation technologies within different
While this site has only two directions from which the wind blows wind conditions.
most of the time, other sites may have more than two predominant
wind directions that should be considered. To adjust the building model to be able to create the desired wind
conditions when the wind blows from either predominant direction,
The sports programs are able to occur in different areas around the model may be adjusted and tested in Vasari. For each iteration,
the building depending on which direction the wind is coming from, the wind should be simulated from both directions to ensure that
as most of the required equipment is moveable. Although there any changes to improve wind conditions when wind comes from
is not a seasonal variation in the predominant wind direction on the northwest do not negatively impact wind conditions when wind
this particular site, if this method were to be applied to a site on comes from the southeast. The wind speed that should be input is
which wind comes from one direction in the summer and a different the average or highest speed for the specific wind direction being
direction in the winter, the seasonal sports programs could be tested. The following five iterations were used to adjust the form of
arranged around the building based on the seasonality of the site’s the building so that it is able to create the desired wind conditions
wind directions. On this site, however, all of the sports need to be from either predominant wind direction. The eleventh iteration
accommodated when the wind blows from either direction, as the accommodates all of the exterior programs in a wide variety of
wind comes from the northwest and southeast in both the summer wind conditions. After this iteration, this step in the method was
and winter.6 stopped and the resulting form was used to develop the next steps
in the design methodology.

90
 FORM AFFECTS WIND | CFD ITERATIONS TO CREATE WIND CONDITIONS

ITERATION 7
WALL ADDED AT NORTHWEST SIDE TO SHELTER SPACE
FROM WIND FROM BOTH DIRECTIONS

wind is slowed by porous wall for snow build-up

wall curved around more to better shelter the leeward area for
tennis and badminton
channel of increased wind speed for launch of kites, snowkites,
and snow windsurfers
large space with decreased wind speed for field sports
zones of increased wind speed at the sides of the forms for
snowkiting, snow windsurfing, and kite flying

WIND FROM SE
Fig. 3.76. Iteration 7 Vasari screenshot.

zones of increased wind speed at the sides of the forms for

snowkiting, snow windsurfing, and kite flying
wind is slowed by porous wall for snow build-up
channel of increased wind speed for launch of kites, snowkites,
and snow windsurfers
large space with decreased wind speed for field sports
added wall shelters area for tennis and badminton from wind
from either direction; in next iteration, curve this wall more to
minimize wind in this area when wind comes from the NW

WIND SPEED
INCREASED

WIND FROM NW AVERAGE

Fig. 3.77. Iteration 7 Vasari screenshot.
DECREASED

91
FORM AFFECTS WIND | CFD ITERATIONS TO CREATE WIND CONDITIONS

ITERATION 8
FORMS ROTATED AND STAGGERED TO CREATE SIDEWAYS
HIGH-SPEED WIND CHANNEL

placement of forms uses the staggering effect to direct the

channel of increased wind speed into the high-speed area at
the side

WIND FROM SE
Fig. 3.78. Iteration 8 Vasari screenshot.

no channel of high-speed wind is created when wind comes

from the NW; in next iteration, add a wall to create a high-speed
wind stream with NW wind

wall was curved more to reduce wind in this area, but a high-
speed wind stream enters the courtyard when wind comes from
the NW; fix this in a subsequent iteration

WIND SPEED
INCREASED

WIND FROM NW AVERAGE

Fig. 3.79. Iteration 8 Vasari screenshot. DECREASED

92
 FORM AFFECTS WIND | CFD ITERATIONS TO CREATE WIND CONDITIONS

ITERATION 9
WALL ADDED AT NORTH END OF FORM TO CREATE HIGH-
SPEED WIND CHANNEL WITH WIND FROM THE NW

added wall does not negatively impact wind conditions when

wind comes from the SE

WIND FROM SE
Fig. 3.80. Iteration 9 Vasari screenshot.

stream of increased wind speed comes into sheltered area when

wind blows from the NW; in next iteration, adjust form so that
this does not happen
wall added to create high-speed wind channel when wind blows
from the NW, to mirror the high-speed wind channel on the other
side of the building when wind comes from the SE

WIND SPEED
INCREASED

WIND FROM NW AVERAGE

Fig. 3.81. Iteration 9 Vasari screenshot.
DECREASED

93
FORM AFFECTS WIND | CFD ITERATIONS TO CREATE WIND CONDITIONS

ITERATION 10
WALL CURVED TO SHELTER COURTYARD FROM NW WINDS

curved wall does not impact wind conditions when wind blows
from the SE

WIND FROM SE
Fig. 3.82. Iteration 10 Vasari screenshot.

curved wall doesn’t alter wind conditions when wind blows from
the NW; in next iteration, replace with a straight, porous wall to
see if the curved wall is needed
wall is curved more to shelter the courtyard from the stream of
high-speed wind that was entering the courtyard during winds
from the NW

WIND SPEED
INCREASED

WIND FROM NW AVERAGE

Fig. 3.83. Iteration 10 Vasari screenshot. DECREASED

94
 FORM AFFECTS WIND | CFD ITERATIONS TO CREATE WIND CONDITIONS

ITERATION 11
CURVED WALL AT SOUTH END REPLACED WITH STRAIGHT,
POROUS WALL

straight, porous wall does not negatively alter wind conditions

when wind blows from the SE

WIND FROM SE
Fig. 3.84. Iteration 11 Vasari screenshot.

straight, porous wall does not negatively alter wind conditions

when wind blows from the NW

WIND SPEED
INCREASED

WIND FROM NW AVERAGE

Fig. 3.85. Iteration 11 Vasari screenshot.
DECREASED

95
FORM AFFECTS WIND | CFD ITERATIONS TO CREATE WIND CONDITIONS

ITERATION 11 | WIND FROM SE

SE

FAST WIND SLOW WIND MINIMAL WIND

HORIZONTAL AXIS  BLADE TIP  TENNIS AND 
WIND TURBINE POWER SYSTEM BADMINTON

COMPACT WIND  FIELD SPORTS
ACCELERATION TURBINE TURBULENT WIND
SNOW BUILD-UP  VERTICAL AXIS 
KITE FLYING
FOR SLEDDING WIND TURBINE

SNOWKITING

SNOW WINDSURFING

Fig. 3.86. Form with exterior programs.

96
 FORM AFFECTS WIND | CFD ITERATIONS TO CREATE WIND CONDITIONS

LAUNCH TO SIDE OF BUILDING

SLOW WIND FOR SNOW BUILD-UP

LAUNCH TO SIDE OF BUILDING

LAUNCH TO SIDE OF BUILDING

WIND SPEED TEST IN TURBULENT WIND
INCREASED

AVERAGE SLOW WIND REQUIRED

DECREASED
TEST IN LARGE SPACE

TEST IN LARGE SPACE

Fig. 3.87. Wind speed from Vasari.

TEST IN MINIMAL WIND

MINIMAL WIND REQUIRED

LAUNCH TO SIDE OF BUILDING

LAUNCH TO SIDE OF BUILDING

LAUNCH TO SIDE OF BUILDING

TEST IN TURBULENT WIND

TEST IN MINIMAL WIND

MINIMAL WIND REQUIRED

Fig. 3.88. Flow lines from Flow Design.

97
FORM AFFECTS WIND | CFD ITERATIONS TO CREATE WIND CONDITIONS

ITERATION 11 | WIND FROM SE | SUMMER

SE

Fig. 3.89. Exterior programs when wind blows from the SE in summer.

98
 FORM AFFECTS WIND | CFD ITERATIONS TO CREATE WIND CONDITIONS

ITERATION 11 | WIND FROM SE | WINTER

SE

Fig. 3.90. Exterior programs when wind blows from the SE in winter.

99
FORM AFFECTS WIND | CFD ITERATIONS TO CREATE WIND CONDITIONS

ITERATION 11 | WIND FROM NW

NW

FAST WIND SLOW WIND MINIMAL WIND

HORIZONTAL AXIS  BLADE TIP  TENNIS AND 
WIND TURBINE POWER SYSTEM BADMINTON

COMPACT WIND  FIELD SPORTS
ACCELERATION TURBINE TURBULENT WIND
SNOW BUILD-UP  VERTICAL AXIS 
KITE FLYING
FOR SLEDDING WIND TURBINE

SNOWKITING

SNOW WINDSURFING

Fig. 3.91. Form with exterior programs.

100
 FORM AFFECTS WIND | CFD ITERATIONS TO CREATE WIND CONDITIONS

TEST IN LARGE SPACE

TEST IN LARGE SPACE

SLOW WIND FOR SNOW BUILD-UP

TEST IN TURBULENT WIND

WIND SPEED
INCREASED SLOW WIND REQUIRED
AVERAGE
TEST IN MINIMAL WIND
DECREASED

MINIMAL WIND REQUIRED

Fig. 3.92. Wind speed from Vasari.

LAUNCH TO SIDE OF BUILDING

LAUNCH TO SIDE OF BUILDING

LAUNCH TO SIDE OF BUILDING

TEST IN MINIMAL WIND

MINIMAL WIND REQUIRED

LAUNCH TO SIDE OF BUILDING

LAUNCH TO SIDE OF BUILDING

LAUNCH TO SIDE OF BUILDING

Fig. 3.93. Flow lines from Flow Design.

101
FORM AFFECTS WIND | CFD ITERATIONS TO CREATE WIND CONDITIONS

ITERATION 11 | WIND FROM NW | SUMMER

NW

Fig. 3.94. Exterior programs when wind blows from the NW in summer.

102
 FORM AFFECTS WIND | CFD ITERATIONS TO CREATE WIND CONDITIONS

ITERATION 11 | WIND FROM NW | WINTER

NW

Fig. 3.95. Exterior programs when wind blows from the NW in winter.

103
 WIND AFFECTS FORM

105
WIND AFFECTS FORM | HOW WIND AFFECTS FORM

HOW WIND AFFECTS FORM

When a building obstructs wind flow, the wind exerts pressure on Windward
the building.1 In general, wind pressure acts perpendicular to the (adj.) The side of an obstruction that is facing the wind.
surface on which it is exerted.2 Positive wind pressure pushes the (adv.) On the side of an obstruction that is facing the wind.
windward building walls inwards (Fig. 4.1).3 Along the leading edges
Leeward
of the building that are exposed to the oncoming wind, the wind
(adj.) The side of an obstruction that is sheltered from the wind.
flow increases in velocity and detaches from the building surface.4
(adv.) On the side of an obstruction that is sheltered from the
This creates zones of negative pressure along the leeward building
wind.
faces and the streamwise building faces, which include the roof
and walls that are parallel to the direction of the wind flow (Fig.
4.1).5 This negative pressure pulls these building faces outwards.6
Within a courtyard, positive pressure is exerted on the wall facing
the oncoming wind, and negative pressure is exerted on the wall
facing away from the direction of the wind (Fig. 4.1).7

The shape of a building can affect the amount of wind pressure,

and the resultant wind force, that acts on its faces.8 While this wind
pressure cannot be completely eliminated, there are ways that
buildings may be formed to reduce the pressure and overall wind
force that is exerted on them.

The second step in the design method is carried out concurrently +

with the step that was outlined in the previous chapter, in which WIND FLOW -
WIND PRESSURE
the building form is designed to create particular surrounding wind
Fig. 4.1. Wind pressure distribution over a building.
conditions. For each building design iteration, CFD software is used
to evaluate the aerodynamics of the building form by providing Aerodynamics
information about the wind pressure that is exerted on each building The study of the interaction between solid forms and air when at
face. The speed with which these results are provided allows the least one of them is in motion, not necessarily referring to solid
architect to refine and re-test many iterations of their design until forms that are shaped to reduce drag caused by surrounding
the building form has the desired aerodynamic properties. wind.
Aerodynamic
The quality of a form to have a shape that reduces drag caused
by surrounding wind.
106
 WIND AFFECTS FORM | AERODYNAMIC FORMS LIBRARY

AERODYNAMIC FORMS LIBRARY

The following library of aerodynamic forms shows how forms can how accurately the software represents surface pressure variations,
be designed to increase or decrease the wind pressure or overall so that they will be able to tell if the simulated wind pressure over
wind force that is exerted over them. These forms were studied their building design iterations is an accurate representation of the
and compiled from a variety of published sources.9 This library may ways in which the building form would affect the wind pressure
be referred to while designing and adjusting the initial building form acting on it.
to reduce the amount of wind pressure or force acting on it. For
each factor, the studied effect is compared with the Flow Design It should be noted that because this thesis focuses on the design
simulation of the wind pressure over the form. The Flow Design of a low-rise building, the ways in which building height and aspect
pressure simulations allow the architect to become familiar with ratio affect wind pressure are not considered.

107
WIND AFFECTS FORM | AERODYNAMIC FORMS LIBRARY

ORIENTATION
SURFACE ORIENTATION IN RELATION TO WIND DIRECTION AFFECTS PRESSURE DISTRIBUTION
It should be noted that the colour ORIGINAL ADJUSTED
gradients representing pressure from PRESSURE LEGEND PRESSURE LEGEND
Flow Design do not always match the + +
colours of the original legend from Flow
Design (left), so an adjusted legend MINIMUM MINIMUM
(right) may also be used to interpret the
pressure distribution screenshots. - -

WINDWARD

Fig. 4.2. Effect of orientation on wind pressure.

ADJUSTED
PRESSURE LEGEND
+
slow +
MINIMUM
minimum LEEWARD
fast -

WIND FLOW - WIND PRESSURE Fig. 4.3. Flow Design simulation.

108
 WIND AFFECTS FORM | AERODYNAMIC FORMS LIBRARY

WIND SPEED
SLOWER SPEED DECREASES PRESSURE

SLOW WIND

WINDWARD

ADJUSTED
PRESSURE LEGEND
0.45 Pa

MINIMUM

-0.60 Pa LEEWARD

FAST WIND

WINDWARD

ADJUSTED
Fig. 4.4. Effect of wind speed on wind pressure. PRESSURE LEGEND
99.10 Pa
slow +
MINIMUM
minimum LEEWARD
fast -111.20 Pa

WIND FLOW - WIND PRESSURE Fig. 4.5. Flow Design simulation.

109
WIND AFFECTS FORM | AERODYNAMIC FORMS LIBRARY

FORM
STREAMLINED FORM DECREASES OVERALL WIND FORCE
It should be noted that although these
form manipulations reduce the overall
wind force acting on it, they may not ADJUSTED
necessarily reduce the wind pressure PRESSURE LEGEND
on localized areas of the surface.10 10.80 Pa

MINIMUM

-15.35 Pa WINDWARD

ADJUSTED
PRESSURE LEGEND
11.95 Pa

MINIMUM

-20.80 Pa WINDWARD

ADJUSTED
PRESSURE LEGEND
12.25 Pa

MINIMUM

-15.65 Pa WINDWARD

ADJUSTED
Fig. 4.6. Effect of form on wind force. PRESSURE LEGEND
11.85 Pa
slow +
MINIMUM
minimum -17.90 Pa WINDWARD
fast
WIND FLOW - WIND PRESSURE Fig. 4.7. Flow Design simulation - effect not shown.

110
 WIND AFFECTS FORM | AERODYNAMIC FORMS LIBRARY

EXPOSURE
UPWIND OBSTRUCTIONS DECREASE WIND FORCE
It should be noted that although
these upwind obstructions generally
ADJUSTED
decrease the overall wind force that
PRESSURE LEGEND
is exerted on the downwind form,
they may increase the wind pressure +
on localized areas of the downwind MINIMUM
form’s windward surface.11
- WINDWARD

ADJUSTED
PRESSURE LEGEND
+

MINIMUM

- WINDWARD

ADJUSTED
PRESSURE LEGEND
+

MINIMUM

- WINDWARD

ADJUSTED
Fig. 4.8. Effect of exposure on wind force. PRESSURE LEGEND
+
slow +
MINIMUM
minimum WINDWARD
fast -

WIND FLOW - WIND PRESSURE Fig. 4.9. Flow Design simulation.

111
WIND AFFECTS FORM | AERODYNAMIC FORMS LIBRARY

POROSITY
HIGHER POROSITY DECREASES WIND FORCE ON LEEWARD FACE
It should be noted that although a higher porosity reduces the overall wind force
on the form by reducing the suction on the leeward face, it may not necessarily
reduce the wind pressure on some localized areas of the surface.12

ORIGINAL
PRESSURE LEGEND
+

MINIMUM

- LEEWARD

ORIGINAL
Fig. 4.10. Effect of porosity on wind force. PRESSURE LEGEND
+
slow +
MINIMUM
minimum LEEWARD
fast -

WIND FLOW - WIND PRESSURE Fig. 4.11. Flow Design simulation.

112
 WIND AFFECTS FORM | CFD ITERATIONS TO IMPROVE FORM AERODYNAMICS

CFD ITERATIONS TO IMPROVE FORM AERODYNAMICS

For the next step in the design method, CFD software is used to The eleven iterations were developed to improve the aerodynamics
evaluate the aerodynamics of the building form by depicting how of the building form, looking at the wind pressure on both the
much wind pressure is exerted over the model’s surface. The amount windward and leeward building faces. Like the previous chapter, the
of wind pressure that acts across the surface is represented by first six iterations evaluate the wind pressure acting on the building
colour gradients in Flow Design. After each iteration is tested within when the wind comes only from the most predominant southeast
the simulated wind conditions, the designer makes observations direction, and iterations 7-11 consider wind pressure when wind
and adjusts the form for the next iteration to improve the form’s comes from both of the site’s predominant wind directions. These
aerodynamics. These adjustments to the form may be designed simulations of wind pressure from multiple directions ensure that
by referring to the aerodynamic forms library, which catalogues the building form is aerodynamically designed to reduce the wind
ways of manipulating the form to decrease the wind pressure or pressure acting on it within all of the site’s frequently-occurring
force acting on it. These observations and adjustments are done wind conditions.
concurrently with the adjustments to improve the surrounding wind
conditions that were described in the previous chapter, and as It should be noted that Flow Design shades some faces of the
such, the forms of each iteration in this chapter are the same as forms, rendering them black. The user of the software may rotate
those in the previous chapter. the form to bring the shaded faces into the light to see the wind
pressure distribution over them.

113
WIND AFFECTS FORM | CFD ITERATIONS TO IMPROVE FORM AERODYNAMICS

ITERATION 1
WIND FROM SE

simple forms are tested for the first iteration only to manipulate
wind patterns, not considering aerodynamics at this point

WINDWARD

Fig. 4.12. Iteration 1 Flow Design screenshots.

LEEWARD

ITERATION 2
WIND FROM SE

blunt, simple forms are still used in the second iteration, which
was altered only to change the wind patterns around the forms

WINDWARD

PRESSURE
+

MINIMUM
Fig. 4.13. Iteration 2 Flow Design screenshots.
LEEWARD
-

114
 WIND AFFECTS FORM | CFD ITERATIONS TO IMPROVE FORM AERODYNAMICS

ITERATION 3
WIND FROM SE

blunt, simple forms are still used in the third iteration, which was
altered only to change the wind patterns around the forms

WINDWARD

Fig. 4.14. Iteration 3 Flow Design screenshots.

LEEWARD

ITERATION 4
WIND FROM SE

the blunt forms were changed into two different shapes to

compare the differences in the wind pressure acting on them;
while there is no visible difference in pressure between the two
forms from the Flow Design output, the curved face on the right
WINDWARD would experience less pressure build-up than the straight face
on the left
Flow Design shows no visible difference between the wind
pressures on the leeward sides of the two forms

PRESSURE
+

MINIMUM
Fig. 4.15. Iteration 4 Flow Design screenshots. LEEWARD
-

115
WIND AFFECTS FORM | CFD ITERATIONS TO IMPROVE FORM AERODYNAMICS

ITERATION 5
WIND FROM SE
curved form was kept in this iteration for its good aerodynamic
properties
in next iteration, curve this form to reduce the wind pressure
build-up on it
WINDWARD

Fig. 4.16. Iteration 5 Flow Design screenshots.

LEEWARD

ITERATION 6
WIND FROM SE

form was curved to reduce the wind pressure acting on it

form is kept curved in this iteration for its good aerodynamic
properties

WINDWARD

part of the wall was made porous to reduce the wind force on
its leeward side
PRESSURE
+

MINIMUM
Fig. 4.17. Iteration 6 Flow Design screenshots.
LEEWARD
-

116
 WIND AFFECTS FORM | CFD ITERATIONS TO IMPROVE FORM AERODYNAMICS

ITERATION 7
WIND FROM SE

WINDWARD

Fig. 4.18. Iteration 7 Flow Design screenshots.

LEEWARD

WIND FROM NW

added wall to create sheltered space is curved into the direction

WINDWARD of the wind to reduce pressure on it when wind blows from the
NW

PRESSURE
+

MINIMUM
Fig. 4.19. Iteration 7 Flow Design screenshots. LEEWARD
-

117
WIND AFFECTS FORM | CFD ITERATIONS TO IMPROVE FORM AERODYNAMICS

ITERATION 8
WIND FROM SE
rotating and staggering the forms shelters this part
of the building that is leeward of the smaller form,
reducing wind pressure on that area

the form of the smaller building is curved into the

WINDWARD
direction of the wind to reduce pressure on it when
wind comes from the SE

Fig. 4.20. Iteration 8 Flow Design screenshots.

LEEWARD

WIND FROM NW

all edges of forms are curved to reduce wind pressure

build-up on them

WINDWARD

PRESSURE
+

MINIMUM
Fig. 4.21. Iteration 8 Flow Design screenshots.
LEEWARD
-

118
 WIND AFFECTS FORM | CFD ITERATIONS TO IMPROVE FORM AERODYNAMICS

ITERATION 9
WIND FROM SE

WINDWARD

Fig. 4.22. Iteration 9 Flow Design screenshots. LEEWARD

WIND FROM NW

the added wall to create a high-speed wind stream shelters this

area of the wall when wind comes from the NW, reducing wind
pressure acting on it
added wall is made curved into the direction of the wind to
WINDWARD reduce pressure on it when wind blows from the NW

PRESSURE
+

MINIMUM
Fig. 4.23. Iteration 9 Flow Design screenshots.
LEEWARD
-

119
WIND AFFECTS FORM | CFD ITERATIONS TO IMPROVE FORM AERODYNAMICS

ITERATION 10
WIND FROM SE

WINDWARD

Fig. 4.24. Iteration 10 Flow Design screenshots.

LEEWARD

WIND FROM NW

the extension of the wall, which is the only alteration in

this iteration, has little impact on wind pressure

WINDWARD

PRESSURE
+

MINIMUM
Fig. 4.25. Iteration 10 Flow Design screenshots.
LEEWARD
-

120
 WIND AFFECTS FORM | CFD ITERATIONS TO IMPROVE FORM AERODYNAMICS

ITERATION 11
WIND FROM SE

the removal of the curved wall, which is the only alteration in this
WINDWARD
iteration, has little impact on wind pressure

Fig. 4.26. Iteration 11 Flow Design screenshots.

LEEWARD

WIND FROM NW

WINDWARD

PRESSURE
+

MINIMUM
Fig. 4.27. Iteration 11 Flow Design screenshots.
LEEWARD
-

121
WIND AFFECTS STRUCTURE

123
WIND AFFECTS STRUCTURE | HOW WIND AFFECTS STRUCTURE

HOW WIND AFFECTS STRUCTURE

When wind pressure acts on a building, it exerts load on the building Alongwind load
which must be resisted by the building’s structural system.1 Because Wind load acting in the direction parallel to the mean wind
these wind loads usually act laterally, their resistance often requires velocity.
different structural systems than those used to resist vertical gravity
loads.2 The ways in which wind loads act on a building therefore Crosswind load
should be considered in the selection and design of the building’s Wind load acting in the direction perpendicular to the mean
structural system. wind velocity.
There are several types of wind loads, and each type induces a
different building response. Alongwind load is the wind load that
acts in the direction parallel to that of the mean wind velocity, and
causes the building to move in that same direction (Fig. 5.1).3 Similarly,
crosswind load acts on the building perpendicular to the direction of
the mean wind velocity, and the crosswind response of the building
is movement in the same direction (Fig. 5.2).4 Crosswind response of
slender buildings may be induced by vortex shedding.5 As wind blows
around a building, vortices form on the sides of the building and then
detach from the building.6 They tend to do this in a pattern, shedding
Fig. 5.1. Alongwind load response. Fig. 5.2. Crosswind load response.
alternately from each side (see Fig. 3.60 on page 81).7 As they shed,
they generate suction,8 which causes forces to act alternately on
the sides of the building in the crosswind direction.9 This makes the Torsional load
building oscillate in the crosswind direction.10 Finally, torsional load Wind load that induces twisting about the vertical axis.
causes the building to twist about its vertical axis (Fig. 5.3).11

While alongwind, crosswind and torsional loads act laterally, uplift (Fig.
5.4) is wind force that acts upwards, mainly on large roofs.12 As wind
flow detaches from the building surface at its windward edges, an area
of negative pressure is created over the roof13 that physically sucks
the roof upwards.14 Building structures should be designed to resist
both lateral and uplift forces, although the type of force that tends to
govern the design of the structural system depends on the building’s
shape and aspect ratio.15 In general, the resistance of lateral loads
and overturning governs the design of slender buildings, while uplift Fig. 5.3. Torsional load response.

124
 WIND AFFECTS STRUCTURE | HOW WIND AFFECTS STRUCTURE

Uplift forces are of more concern in the design of broad, low buildings.16 To
Wind force acting upwards in the direction perpendicular resist uplift forces, all structural members should be connected to create
to the mean wind velocity. a continuous load path that anchors the roof to the foundation, to transfer
the uplift forces down to the foundation through the structural members.17
This may be achieved with reinforcing bars in concrete or masonry walls,
anchoring steel columns at their bases with anchor bolts, or using strong
metal connectors between members of wood frame systems.18 While
uplift is an important consideration in the design of structural systems to
resist wind loads, it is not considered in the design method of this thesis
because the CFD software that is used does not simulate wind uplift
effects. This design method therefore only considers the resistance of
lateral wind loads. However, a consideration of uplift could be integrated
flow detaches from edge and
into the method if the software were to be improved in the future to be
creates area of negative pressure
Fig. 5.4. Upift. able to simulate uplift forces.

Alongwind, crosswind and torsional loads all have both static and dynamic
Aspect ratio components.19 Static wind load is caused by the mean wind pressure
Ratio of height to width. acting on a building over a period of time, and maintains a consistent
magnitude and location.20 As described in the previous chapter, the
wind pressure, and therefore the static wind load, acting on a building
Static wind load can be affected by the building’s form.21 Dynamic wind loads rapidly
Wind load that maintains a consistent magnitude and change in magnitude or location over much shorter periods of time.22
location over a period of time. They are caused by unsteady wind pressures, which are the result of
wind turbulence or the separation of the wind flow off the building surface
at the windward edges of the building.23 Whether a certain load acts
Dynamic wind load
statically or dynamically depends on the stiffness of the structure on
Wind load that rapidly changes in magnitude or location.
which it is exerted.24 A load that is applied to the structure for a longer
length of time than the structure’s fundamental period25 acts as a static
Fundamental period load, whereas if it is applied for a shorter length of time, it acts as a
The length of time required to complete one oscillation. dynamic load.26 Stiff structures have a shorter fundamental period than
more flexible structures, so a certain load, applied for a certain length
of time, could be static for a stiffer structure and dynamic for a more
Quasi-static wind load flexible structure, putting more strain on the flexible structure.27 For the
Static wind load with an increased magnitude to account purposes of simplifying wind loads, a quasi-static wind load assumes
for the wind load’s dynamic nature. that the wind acts as a static load, but the magnitude of that load is

125
WIND AFFECTS STRUCTURE | HOW WIND AFFECTS STRUCTURE

increased to account for the dynamic nature of the wind.28 All of moves in the wind. Some damping is provided by the stiffness of
these wind loads, and their static and dynamic components, need the building’s structural system, but supplementary dampers such
to be accounted for when designing a building’s structural system. as tuned mass dampers, tuned liquid dampers, viscous dampers,
and active damping systems, can increase the level of damping
There are several factors that affect how much a building responds that is provided to the building.37 The higher the level of structural
to wind loads:29 damping, the less the building responds to the wind.38
WIND VELOCITY
In addition to affecting the design of a building’s structural system,
Higher-speed winds generally exert more wind pressure on
wind also affects cladding design as it exerts pressure across the
buildings and induce more building response.30
surface of the cladding.39 The amount of wind pressure that acts
GEOMETRY on small areas of the cladding can vary greatly across the building’s
Streamlined shapes generally experience lower wind loads surface, as the magnitude of the wind pressure at any given
because the wind cannot build up enough pressure on a single, location is dependent on the building’s shape and orientation.40 For
blunt side.31 This causes streamlined buildings to experience less this reason, some small elements of the cladding such as mullions,
wind-induced response. glazing panels, sheathing, and shingles must be designed to
withstand increased wind pressures over their small surface areas.41
EXPOSURE These increased design wind pressures are often mandated by
The more exposed a building is to the oncoming wind, the more building codes.42 The various pressures that are exerted across
wind load will be exerted on the building, causing the building to the building’s surface are averaged to obtain the overall wind load
move more in the wind.32 Generally, surface features surrounding values that are used to design the building’s structural system.43
the building shelter it from the wind and reduce the wind load
that acts on it.33 However, surrounding features can sometimes For the third step in the design method, the wind pressure information
accelerate the wind flow around the building, causing the building’s that is provided by the CFD software is input into the FEA software
response to the wind to be greater.34 to predict how the building will react to combined wind and gravity
ORIENTATION loading. This allows the architect to design the building’s structural
If a building is oriented so that its most sensitive direction to the system, by developing models of single structural bays. Based on
wind is facing away from the oncoming wind flow, the building will the deflection animation and the colour gradient that represents
experience less wind-induced response.35 the displacement of each point of the model that is provided by the
FEA software, the architect can see where and how much the bay
MASS AND STIFFNESS will deflect in the wind and under gravity. The architect can then
Heavier, stiffer structures respond less to the wind than lighter stiffen the bay against this deflection, and run the finite element
or more flexible structures because of their shorter fundamental analysis on multiple iterations of bays, with the goal of increasing
periods.36 the stiffness of the assembly each time. This process is repeated
STRUCTURAL DAMPING LEVEL until the building’s structural system can adequately resist large
Structural damping decreases the amount by which the building movements in the wind.

126
 WIND AFFECTS STRUCTURE | STRUCTURAL SYSTEMS LIBRARY

STRUCTURAL SYSTEMS LIBRARY

A structural framing system can resist gravity loads (Fig. 5.5), but
must be braced, or stiffened, to resist movement under lateral wind
loads (Fig. 5.6).44 This can be done with diaphragms that make
the entire wall or floor plane act as a unified whole to resist lateral
loads (Fig. 5.7),45 bracing that stabilizes the frame under lateral
loads (Fig. 5.8),46 or moment-resisting connections that don’t allow
the structural members to rotate under lateral loads (Fig. 5.9).47
Earthquake loads are also a type of lateral load,48 and may be
resisted with these same structural systems that are used to resist
Fig. 5.5. Gravity loads. Fig. 5.6. Wind loads. wind loads.49

Fig. 5.7. Diaphragm. Fig. 5.8. Bracing. Fig. 5.9. Moment-

resisting connections.

127
WIND AFFECTS STRUCTURE | STRUCTURAL SYSTEMS LIBRARY

Fig. 5.10. Plywood on Fig. 5.11. Solid concrete Fig. 5.12. Masonry wall.
wood framing. wall.
DIAPHRAGMS
Diaphragms are rigid walls that are continuous through the height
of the building.50 They resist the tendency of the building floors to
slide relative to one another when subjected to lateral wind loads51
by acting in bending as stiff vertical cantilevers.52 They are made
of surfacing that is attached to the structural framing to make
the structural members act as a unified whole in resisting lateral
wind loads,53 and can take the form of plywood attached to wood
framing (Fig. 5.10), solid concrete walls (Fig. 5.11), or masonry
walls (Fig. 5.12).54 Without surfacing, individual structural members Fig. 5.13. A frame without surfacing moves to the side under lateral loads.
in a frame would move to the side under lateral loads (Fig. 5.13).55
The surfacing makes the entire wall act as a cohesive unit that
resists rotation under wind loads.56 This tendency to want to rotate
induces tension in the windward side of the wall and compression in
the leeward side, so the connections at the base of the diaphragm
must resist these forces and carry them down into the foundation
(Fig. 5.14).57 There should be minimal openings in these walls
to ensure that they are effective at resisting wind loads.58 They
may either be placed symmetrically within the building’s exterior
walls, or be used as interior walls, typically wrapping around the
RESIST COMPRESSION
building’s central core.59 Regardless of their placement, a minimum
of two perpendicular diaphragms are required within the building
to provide resistance against the wind that could come from any RESIST TENSION
direction.60
Fig. 5.14. A diaphragm subjected to lateral loads.

128
 WIND AFFECTS STRUCTURE | STRUCTURAL SYSTEMS LIBRARY

Fig. 5.15. Steel framing. Fig. 5.16. Wood framing.

BRACING
Bracing may also be used to stabilize a structural frame against
lateral wind loads. The diagonal members work to resist shear61
as they increase they rigidity of the frame and reduce the bending
stresses in the beams and columns.62 They may be made out of
steel (Fig. 5.15) or wood (Fig. 5.16) framing.63 There are many
configurations of diagonal bracing members that may be inserted
within a frame (Fig. 5.17).64 If only one diagonal member is used, it
must be able to resist both tension and compression as the wind
load could act on either side of the frame (Fig. 5.18) to elongate
or shorten the diagonal member (Fig. 5.19).65 If there are at least
two diagonal members, however, it is not necessary for both of
them to be good in both tension and compression, as one can
FILL WITH: work in tension to stiffen the frame when the wind comes from one
direction, and the other can work in tension to stiffen the frame
against wind from the other direction (Fig. 5.20).66 While bracing
is efficient at stiffening frames against wind loads, the insertion
of the diagonal members into the building’s framing is not always
Fig. 5.17. Bracing configurations. conducive to the interior building layout.67

Fig. 5.18. Wind load on Fig. 5.19. One Fig. 5.20. Two diagonal
a frame. diagonal member. members.

129
WIND AFFECTS STRUCTURE | STRUCTURAL SYSTEMS LIBRARY

Fig. 5.21. Steel plates and fasteners. Fig. 5.22. Bolted connection.

Fig. 5.23. Welded connection. Fig. 5.24. Steel-reinforced concrete

joint.
MOMENT-RESISTING CONNECTIONS
Moment-resisting connections, which occur at the joints between the
beams and columns in a frame and not within the frame’s opening,
prevent the frame’s structural members from rotating relative to
one another under lateral loads.68 They may be created with steel
plates and fasteners connecting wooden members (Fig. 5.21),69
bolted (Fig. 5.22) or welded (Fig. 5.23) connections between steel
members, or steel-reinforced joints between concrete members
(Fig. 5.24).70 These connections stiffen the entire frame so that it
deforms as a cohesive unit to absorb wind and gravity loads (Fig.
5.25).71 While they are less efficient at resisting shear than diagonal
bracing members, the lack of obstructions within the frame makes
it easier to plan interior building layouts and exterior window and
door placements.72 Fig. 5.25. The frame deforms as a cohesive unit to absorb wind and gravity
loads.

130
 WIND AFFECTS STRUCTURE | STRUCTURAL SYSTEM SELECTION

STRUCTURAL SYSTEM SELECTION

In the two previous steps of the design method, the wind effects system allows the architect to design small units of structure that
library and the aerodynamic forms library could be referred to for are repeated throughout the building. This creation of a single unit
design strategies to manipulate the wind conditions around the may be applied to buildings of various sizes, as the units may be
building and reduce the wind pressure acting on the building. repeated as many times as necessary. Its modularity also allows
Similarly, the structural systems library in this chapter described the architect to avoid modeling and testing the entire building’s
several systems that can be used to resist wind loads. However, structure, instead just designing a few different structural units. The
the structural systems library is not intended to be a reference simplicity of this system provides an appropriate base with which
during the design process, as for the purposes of the development to develop a methodology that could then potentially be applied to
of the design methodology, a steel-frame system has been chosen the design of one of the other structural systems described in the
and this step has been developed specifically for the planning of structural systems library.
this type of structural system. The modular nature of a steel-frame

131
WIND AFFECTS STRUCTURE | FEA ITERATIONS TO DEVELOP STRUCTURAL SYSTEM

FEA ITERATIONS TO DEVELOP STRUCTURAL SYSTEM

1. PRESSURE ZONES
WIND FROM SE
The first step in the structural system development is to divide the
building form into different pressure zones, based on the colour
gradients representing the wind pressure on the model’s surface
that were obtained from Flow Design in the previous chapter (Fig.
5.26). The building is divided into seven pressure zones (Fig. 5.27) WINDWARD
based on the areas that are subject to the same amount of wind
pressure according to the colour gradients. The reasons for these
variations in wind pressure across the model’s surface, such as
form, exposure, orientation, and porosity, are described in the
previous chapter.

Once the form has been divided into these pressure zones, the
average positive and negative wind pressure values for each zone LEEWARD
are obtained from the wind pressure colour gradients for each of
the site’s two predominant wind directions. These values are shown
in the pressure zone matrix (Fig. 5.28). However, the accuracy of WIND FROM NW
these values depends on the level of accuracy that may currently
be obtained from the CFD software, and as such, they may not
necessarily reflect the wind pressure that would be exerted on the
building in reality. For the purposes of this method, each zone is
assumed to be subjected to the highest combined positive and WINDWARD
negative wind pressure, out of the values obtained from both wind
directions. This ensures that the building’s structure will be able
to withstand the wind pressure within all of the site’s predominant
wind directions.
PRESSURE
+

MINIMUM
LEEWARD
-
Fig. 5.26. Wind pressure gradients from Flow Design.

132
 WIND AFFECTS STRUCTURE | FEA ITERATIONS TO DEVELOP STRUCTURAL SYSTEM

5 4

Fig. 5.27. The building divided into pressure zones.

WIND DIRECTION PRESSURE ZONE 1 ZONE 2 ZONE 3 ZONE 4 ZONE 5 ZONE 6 ZONE 7

WINDWARD 32 Pa 32 Pa   5 Pa 32 Pa 32 Pa   5 Pa   8 Pa

SE
LEEWARD - 28 Pa - 35 Pa - 35 Pa - 20 Pa - 10 Pa - 20 Pa - 20 Pa

WINDWARD   1 Pa 15 Pa 15 Pa 15 Pa   1 Pa 10 Pa 10 Pa

NW
LEEWARD - 15 Pa - 23 Pa - 36 Pa - 36 Pa - 12 Pa - 10 Pa - 22 Pa
Fig. 5.28. Pressure zone matrix.

133
WIND AFFECTS STRUCTURE | FEA ITERATIONS TO DEVELOP STRUCTURAL SYSTEM

2. FEA OF BUILDING MASSING

Fig. 5.29. Loading applied to zone 1 in Scan&Solve.
The wind pressure values for the highest combined positive
and negative pressure that were obtained from Flow Design in
the previous step, as well as gravity loading, are then input into
Scan&Solve and applied to the massing model of each pressure
zone, as shown for zone 1 (Fig. 5.29). The colour gradient that
represents the displacement of each point of the model, as well
as the deflection animation (Fig. 5.30), reveal areas within the zone
that may deflect more while under wind load than other areas within
the zone. In the case of zone 1, the form deflects the most at the
right side, where the massing is the thinnest. As such, zone 1 has
been divided into two sub-zones: 1A and 1B (Fig. 5.31). These
sub-zones are subjected to the same amount of wind pressure, Fig. 5.30. Deflection animation of zone 1 from Scan&Solve.
but require different structural configurations due to the difference
in geometry thickness. This process is repeated with each of the
other zones to determine if more sub-zone divisions are required.

1A

1B

Fig. 5.31. Zone 1 divided into sub-zones.

134
 WIND AFFECTS STRUCTURE | FEA ITERATIONS TO DEVELOP STRUCTURAL SYSTEM

3. STRUCTURAL BAYS
A structural bay is then modeled for zone 1A, and tested with 7 to develop a structural bay for each pressure zone within the
Scan&Solve by inputting gravity loading and the wind pressure for zone’s specified wind pressure conditions. Reducing the scope of
zone 1. The deflection of the bay in the wind is observed with the the structural design to a single structural bay allows the architect
displacement colour gradient and deflection animation. The colour to avoid the time-consuming task of modeling and testing the
gradient that represents displacement allows the architect to easily entire building’s structure, while still being able to understand the
read where and by how much the model will move under load, column spacing, member sizes, and bracing that will be required
although the accuracy of this quantitative data depends on the throughout the building.
accuracy of the FEA software. The deflection animation allows the
architect to actually see how the structural bay would deflect to This step in the design method does not intend for the architect to
know how and where to stiffen the model in the next iteration. It replace the structural engineer. In a later design phase, the structural
should be noted that the deflection animations are exaggerated engineer would perform a more thorough structural design and
to more obviously show where the deflection would occur. If analysis and would likely make adjustments to the structural bay that
there is too much deflection under the input loading, the architect is developed with this step. The importance of the method is that
adjusts the model of the bay to increase its stiffness. This may it provides the architect with a sense of the approximate structural
be accomplished by reducing column spacing, increasing member spacing, as well as an understanding that some form of bracing
sizes, or adding bracing. This process is repeated until a bay is will be required. The architect can account for this while designing
developed that has appropriate column spacing, member sizes, the building interior, so that no design decisions compromise the
and bracing to resist large movements in the wind. approximate structure that must be accommodated. This method
provides a way of integrating structural considerations into the
The first four iterations, for which the displacement colour gradients initial building design, which improves collaboration between the
and deflection animations are shown, were used to develop a architect and the structural engineer throughout many design
structural bay for zone 1A, and the following four iterations develop phases.
a bay for zone 1B. The process is then repeated for zones 2 through

135
WIND AFFECTS STRUCTURE | FEA ITERATIONS TO DEVELOP STRUCTURAL SYSTEM

ZONE 1A | ITERATION 1
BAY DIMENSIONS 15m x 15m, to fit one bay within 15m
wide building
BEAM SECTIONS 200mm x 400mm
COLUMN SECTIONS 200mm x 200mm
OBSERVATIONS The bay will deflect 45mm in some
places, which is excessive.
CHANGES TO MAKE In the next iteration, reduce the column
spacing to stiffen the bay and reduce its
Δd
45 mm deflection.

WIND DIRECTION
Fig. 5.32. Zone 1A iteration 1 Scan&Solve screenshot. 0 mm

ZONE 1A | ITERATION 2
BAY DIMENSIONS 7.5m x 7.5m, to fit two bays within 15m
wide building
BEAM SECTIONS 200mm x 400mm
COLUMN SECTIONS 200mm x 200mm
COMMENTS The bay will only deflect 8mm at most,
so maintain this column spacing so the
bays can fit evenly within the building
width.
Δd
8 mm CHANGES TO MAKE In the next iteration, add bracing to
further stiffen the structure.

WIND DIRECTION
Fig. 5.33. Zone 1A iteration 2 Scan&Solve screenshot. 0 mm

136
 WIND AFFECTS STRUCTURE | FEA ITERATIONS TO DEVELOP STRUCTURAL SYSTEM

ZONE 1A | ITERATION 3
BAY DIMENSIONS 7.5m x 7.5m
BEAM SECTIONS 200mm x 400mm
COLUMN SECTIONS 200mm x 200mm
COMMENTS Different types of bracing have been
added on each side of the bay to
compare their deflections. Two members
deflect less than one member, so use
two members in each frame.
Δd
3 mm CHANGES TO MAKE In the next iteration, try cross-bracing
to compare the deflection, although the
deflection of this upside-down V-bracing
WIND DIRECTION is adequately low so that it could be used
Fig. 5.34. Zone 1A iteration 3 Scan&Solve screenshot. 0 mm in frames where openings are needed.

ZONE 1A | ITERATION 4
BAY DIMENSIONS 7.5m x 7.5m
BEAM SECTIONS 200mm x 400mm
COLUMN SECTIONS 200mm x 200mm
COMMENTS Cross-bracing has been tested on
one side of the bay, and deflects less
than the upside-down V-bracing of the
previous iteration.
CONCLUSIONS Therefore, use this column spacing and
Δd
5 mm these member sizes, and insert cross-
bracing or V-bracing within the frames
that are parallel to the wind direction in
WIND DIRECTION zone 1A.
Fig. 5.35. Zone 1A iteration 4 Scan&Solve screenshot. 0 mm

137
WIND AFFECTS STRUCTURE | FEA ITERATIONS TO DEVELOP STRUCTURAL SYSTEM

ZONE 1B | ITERATION 1
BAY DIMENSIONS 10m x 1m, because zone 1B is a wall
and not occupiable building space
BEAM SECTIONS 200mm x 400mm
COLUMN SECTIONS 200mm x 200mm
COMMENTS No bracing is used in this iteration. The
bay will deflect 11mm.
CHANGES TO MAKE In the next iteration, decrease the
Δd
11 mm lengthwise column spacing, increase
the wall thickness, and add bracing to
increase the bay’s stiffness.

0 mm WIND DIRECTION
Fig. 5.36. Zone 1B iteration 1 Scan&Solve screenshot.

ZONE 1B | ITERATION 2
BAY DIMENSIONS 7.5m x 2m, to try a shorter and wider
bay
BEAM SECTIONS 200mm x 400mm
COLUMN SECTIONS 200mm x 200mm
COMMENTS A single bracing member is inserted in
the frames that are parallel to the wind
direction. The deflection is low, at 2mm.
Δd
2 mm CHANGES TO MAKE In the next iteration, see if a shallower
wall will still experience an adequately
small amount of deflection under load.

0 mm WIND DIRECTION
Fig. 5.37. Zone 1B iteration 2 Scan&Solve screenshot.

138
 WIND AFFECTS STRUCTURE | FEA ITERATIONS TO DEVELOP STRUCTURAL SYSTEM

ZONE 1B | ITERATION 3
BAY DIMENSIONS 7.5m x 1.5m, to try a shallower bay
BEAM SECTIONS 200mm x 400mm
COLUMN SECTIONS 200mm x 200mm
COMMENTS The deflection is adequately low at 2mm.
CHANGES TO MAKE In the next iteration, see if a shallower
wall will have still have an adequately
small amount of deflection under load.
Δd
2 mm

0 mm WIND DIRECTION
Fig. 5.38. Zone 1B iteration 3 Scan&Solve screenshot.

ZONE 1B | ITERATION 4
BAY DIMENSIONS 7.5m x 1m
BEAM SECTIONS 200mm x 400mm
COLUMN SECTIONS 200mm x 200mm
COMMENTS The deflection is adequately low at
2.5mm, and the wall is thinner than the
previous iterations.
CONCLUSIONS Therefore use this column spacing and
Δd
2.5 mm these member sizes, and insert bracing
within the frames that are parallel to the
wind direction in zone 1B.

0 mm WIND DIRECTION
Fig. 5.39. Zone 1B iteration 4 Scan&Solve screenshot.

139
WIND AFFECTS STRUCTURE | FEA ITERATIONS TO DEVELOP STRUCTURAL SYSTEM

4. STRUCTURAL SYSTEM
After a structural bay has been developed for each zone and compromises the other. This consideration of the building interior
sub-zone, the bays are inserted within the massing to create the will be elaborated upon in the following chapter.
building’s structural system. This schematic structural drawing
(Fig. 5.40) provides the architect with a comprehensive visual The intention of this part of the design method is for the architect
of the required structural density throughout the building. The to develop the structure to a point where they can see how it will
architect can use this visual to design the building’s interior and the impact the building design. By doing this structural layout in the
exterior facade. Knowing the approximate spacing of columns and initial design phases, the architect gets a realistic idea of what the
placement of bracing at this early design stage can inform program structure needs to be, to make sure that no future design moves will
layout according to available space and potential for glazing. compromise the approximate structure that they know has to be
If larger member spacing would be beneficial in certain areas to included. This method does not intend for the architect to replace
create larger open spaces, the architect can test new bays with the structural engineer. Instead, it equips the architect to design
the FEA software, using the same iterative process to increase the a building that can accommodate a feasible structural system,
column spacing, while also increasing member sizing or adding and develop a schematic structural design that the engineer can
bracing to compensate for the longer spans. Once such a bay has then easily size and detail. It provides a way for the architect to
been developed so that its deflection in the wind is acceptably low, develop a qualitative understanding of wind loads and the structure
these larger bays can be inserted where they are needed within the required to resist them, to integrate into the initial building design
building. The architect can repeat this process as necessary until and improve collaboration between the architect and the structural
the structural and architectural designs work together, and neither and wind engineers.

140
 WIND AFFECTS STRUCTURE | FEA ITERATIONS TO DEVELOP STRUCTURAL SYSTEM

4
5
3
2
1

Fig. 5.40. Schematic structural drawing (left) and pressure zones overlaid onto structural drawing (right).

141
WIND AFFECTS STRUCTURE | FEA ITERATIONS TO DEVELOP STRUCTURAL SYSTEM

5. MODIFIED FORM
The form of the building (Fig. 5.41) can then be adjusted to reflect
the inclusion of the steel structure (Fig. 5.42). For example, curved
sections could be faceted, with each side length equal to the length
of the column spacing. This modified form is then re-tested with
Vasari to ensure that the change to the form did not undesirably
alter the wind conditions around the building (Fig. 5.43, Fig. 5.44).
If the wind conditions are found to have changed in certain areas,
this would affect the wind pressure that is applied to that area of the
building. A new pressure zone would be created and a new structural Fig. 5.41. Building form.
bay for that zone would be developed with the same iterative FEA
process described in this chapter. If the wind conditions are altered
so much as to become undesirable, the form itself would need to
be adjusted so that it fosters the intended wind conditions, yet is
also an appropriate shape to be able to accommodate a steel-
frame structure. This step completes the feedback loop between
the CFD and FEA programs. At this point, the design method may
be repeated as many times as necessary to develop a building that
both shapes and resists the wind, before the form is refined at a
finer scale in the next and final step in the design method.

Fig. 5.42. Adjusted building form to reflect the inclusion of the steel structure.

142
 WIND AFFECTS STRUCTURE | FEA ITERATIONS TO DEVELOP STRUCTURAL SYSTEM

LAUNCH TO SIDE OF BUILDING

SLOW WIND FOR SNOW BUILD-UP

LAUNCH TO SIDE OF BUILDING

LAUNCH TO SIDE OF BUILDING

TEST IN TURBULENT WIND

WIND SPEED SLOW WIND REQUIRED

INCREASED
TEST IN LARGE SPACE
AVERAGE
TEST IN LARGE SPACE
DECREASED

WIND FROM SE TEST IN MINIMAL WIND

Fig. 5.43. Wind speed from Vasari. MINIMAL WIND REQUIRED

SLOW WIND REQUIRED

SLOW WIND FOR SNOW BUILD-UP

TEST IN LARGE SPACE

TEST IN LARGE SPACE

TEST IN TURBULENT WIND

WIND SPEED MINIMAL WIND REQUIRED

INCREASED
TEST IN MINIMAL WIND
AVERAGE
LAUNCH TO SIDE OF BUILDING
DECREASED
LAUNCH TO SIDE OF BUILDING
WIND FROM NW
Fig. 5.44. Wind speed from Vasari.
LAUNCH TO SIDE OF BUILDING

143
PROGRAM AFFECTS FORM AND STRUCTURE

145
PROGRAM AFFECTS FORM AND STRUCTURE | HOW PROGRAM AFFECTS FORM AND STRUCTURE

HOW PROGRAM AFFECTS FORM AND STRUCTURE

The steps of the methodology that were described in the previous be accommodated. Once the model is changed accordingly, it
chapters allow the architect to develop a building form that creates should be re-tested in Vasari to ensure that the adjustments to the
desirable wind conditions for a set of exterior programs while form did not undesirably alter the surrounding wind conditions.
reducing and resisting the wind loads that are generated by these
wind conditions. This may be achieved through one or multiple EXTERIOR PROGRAM ELEVATIONS
iterations of the methodology thus far. Now that the form has been
established, it should be refined in this fourth and final step of In previous chapters, the building form was developed based on
the design method, so that it could feasibly be developed from a the wind conditions at a single elevation close to the ground level.
form into a building during a later design phase. The intent of this However, some of the exterior programs would occur at other
refinement is not to substantially alter the design, but to make small vertical elevations. The wind conditions around the building form
adjustments to the form and the structure so that it could be made should therefore be tested in Vasari at all of the elevations at which
into a building with both interior and exterior programs that would these programs would occur, to ensure that the wind conditions
function within the site’s wind conditions. are appropriate for the exterior programs.

The following considerations should be made to refine the building

form that was developed in the previous chapters:
REFINEMENT OF STRUCTURE FOR FORM AND
PROGRAM
INTERIOR PROGRAM
The layout of the structural bays should be adjusted to fit within the
new building form and to accommodate the interior program. The
The form that was developed through the previous steps should be
structure may be made into a prominent feature of the architecture.
adjusted at a finer scale so that the interior program could feasibly

146
 PROGRAM AFFECTS FORM AND STRUCTURE | INTERIOR PROGRAM

INTERIOR PROGRAM
Although detailed interior building layouts are not a part of this serve to support the wind energy generation technologies and
design method, the building form that has been developed should exterior sporting activities that exist within the wind conditions that
be able to feasibly accommodate the required interior programs. it is the building’s primary purpose to create. For the next step in
This may be achieved through minor modifications to the building the refinement of the building, modifications to the building form
form. The purpose of the building that is developed through this were made to accommodate these interior programs. Then, the
methodology is to alter the wind patterns around the building to adjusted form was tested in Vasari, simulating wind coming from
create appropriate wind conditions for the exterior programs, and both of the site’s predominant wind directions to ensure that the
to provide interior space to support these exterior programs. As small adjustments to the form did not alter the appropriateness of
such, the interior program requirements are minimal, and only the wind conditions for the exterior programs.

147
PROGRAM AFFECTS FORM AND STRUCTURE | INTERIOR PROGRAM

The following interior programs were accommodated through slight

modifications to the building’s form (Fig. 6.1) that allow the form to
feasibly be made into a building (Fig. 6.2) that accommodates the
exterior programs within both predominant wind directions (Fig.
6.3-Fig. 6.6):

INTERIOR SPECTATOR AREA to use this space to accommodate back-of-house

This area used to be a wall with the sole purpose programs such as change rooms.
of sheltering the courtyard space, but its width
was increased to 5m to accommodate an interior EQUIPMENT STORAGE
spectator area. This space has sliding glass doors The equipment storage is located next to the
on both sides so that visitors to the building can change rooms, at the south end of the building. Like
easily access the exterior, or remain inside and look the change rooms, only a single storey is required
out towards the activities happening on either side. for this program, so it may be included in this area.
It needed to be 9m high to adequately shelter the
adjacent spaces from the wind, so this area may be ENTRANCE
a double-height feature space within the building. The entrance to the building is located between
the change rooms and interior spectator area, for
EXTERIOR SPECTATOR STANDS easy access by visitors who are either players or
Stands were added to the two freestanding walls spectators of the exterior sports activities.
that were initially made part of the form to create
a sheltered courtyard and a channel of high-speed ENERGY STORAGE
wind. These walls that were previously only included The energy that is generated by the wind energy
to shape the wind around the building now also generation technologies is stored in the smaller
serve a programmatic purpose. Stands were also building volume. The volume height was increased
added to one side of the smaller building form. to two storeys to accommodate the large energy
These exterior stands provide space for spectators storage equipment.
to observe the sporting areas that cannot be viewed
from the interior spectator area. OFFICE
Office space is also located inside the smaller
CHANGE ROOMS building volume. This space would be used to study
Change rooms have been added towards the south and compare data from the energy generation
end of the building. This part of the building only technologies, so that the site may be used as an
needs to be a single storey high to generate the energy testing and experimentation facility as well
desired wind conditions in that area, so it is logical as a community sports facility.

148
 PROGRAM AFFECTS FORM AND STRUCTURE | INTERIOR PROGRAM

Fig. 6.2. Building with program. Fig. 6.1. Building form.

149
PROGRAM AFFECTS FORM AND STRUCTURE | INTERIOR PROGRAM

WIND FROM SE | SUMMER

SE

Fig. 6.3. Building and exterior programs when wind blows from the SE in summer.

150
 PROGRAM AFFECTS FORM AND STRUCTURE | INTERIOR PROGRAM

WIND FROM SE | WINTER

SE

Fig. 6.4. Building and exterior programs when wind blows from the SE in winter.

151
PROGRAM AFFECTS FORM AND STRUCTURE | INTERIOR PROGRAM

WIND FROM NW | SUMMER

NW

Fig. 6.5. Building and exterior programs when wind blows from the NW in summer.

152
 PROGRAM AFFECTS FORM AND STRUCTURE | INTERIOR PROGRAM

WIND FROM NW | WINTER

NW

Fig. 6.6. Building and exterior programs when wind blows from the NW in winter.

153
PROGRAM AFFECTS FORM AND STRUCTURE | INTERIOR PROGRAM

CFD SIMULATION
TEST IN LARGE SPACE
The modified form was tested in Vasari with wind
simulated from both predominant directions (Fig.
6.7, Fig. 6.8), to ensure that the changes to the LAUNCH TO SIDE OF BUILDING
form did not undesirably alter the wind conditions
for the exterior programs.

LAUNCH TO SIDE OF BUILDING

LAUNCH TO SIDE OF BUILDING

SLOW WIND FOR SNOW BUILD-UP

TEST IN TURBULENT WIND

WIND SPEED
INCREASED

AVERAGE SLOW WIND REQUIRED

DECREASED

WIND FROM SE TEST IN LARGE SPACE

Fig. 6.7. Wind speed from Vasari.

TEST IN MINIMAL WIND

MINIMAL WIND REQUIRED

154
 PROGRAM AFFECTS FORM AND STRUCTURE | INTERIOR PROGRAM

TEST IN LARGE SPACE

TEST IN LARGE SPACE

SLOW WIND FOR SNOW BUILD-UP

SLOW WIND REQUIRED

TEST IN TURBULENT WIND

TEST IN MINIMAL WIND

WIND SPEED
INCREASED

AVERAGE MINIMAL WIND REQUIRED

DECREASED

WIND FROM NW LAUNCH TO SIDE OF BUILDING

Fig. 6.8. Wind speed from Vasari.

LAUNCH TO SIDE OF BUILDING

LAUNCH TO SIDE OF BUILDING

155
PROGRAM AFFECTS FORM AND STRUCTURE | EXTERIOR PROGRAM ELEVATIONS

EXTERIOR PROGRAM ELEVATIONS

Although the initial building form was developed based on the wind elevations. The following Vasari data slices show that appropriate
conditions that would be created around the building at a single wind conditions are created for all of the exterior programs at their
vertical elevation, the exterior programs would actually occur at a respective elevations. It should be noted that Vasari does not
variety of vertical elevations. For example, kites would be launched simulate the increase in wind speed as elevation increases, so any
at ground level and therefore would require high-speed wind just data slice that is completely above the building only displays the
above the ground in the kite-launching area, but because kites are uniform input wind speed. However, as CFD programs improve,
flown at higher elevations, the open space for kite-flying would need they may be able to depict accurate wind speeds at elevations
to have fast winds higher up. To account for the wind conditions at above the building. This accuracy would make these results more
different elevations in the design of the building form, wind speed useful in determining whether or not there are appropriate wind
data slices were taken in Vasari at each of the elevations at which conditions for all of the exterior programs at all of the required
the programs would occur, to ensure that the sports and energy elevations.
generation technologies would be able to function properly at these

156
 PROGRAM AFFECTS FORM AND STRUCTURE | EXTERIOR PROGRAM ELEVATIONS

2m ELEVATION
LAUNCH

LAUNCH

SNOW BUILD-UP AT GROUND LEVEL

LAUNCH

KICKS AND THROWS

WIND FROM SE
Fig. 6.9. Vasari screenshot at 2m. TENNIS AND BADMINTON

SNOW BUILD-UP AT GROUND LEVEL

KICKS AND THROWS

TENNIS AND BADMINTON

LAUNCH

LAUNCH WIND SPEED

INCREASED
WIND FROM NW AVERAGE
LAUNCH
Fig. 6.10. Vasari screenshot at 2m.
DECREASED

157
PROGRAM AFFECTS FORM AND STRUCTURE | EXTERIOR PROGRAM ELEVATIONS

4m ELEVATION

LAUNCH

LAUNCH

SNOW BUILD-UP AT HIGHER LEVEL

LAUNCH

KICKS AND THROWS

WIND FROM SE
Fig. 6.11. Vasari screenshot at 4m.

SNOW BUILD-UP AT HIGHER LEVEL

KICKS AND THROWS

LAUNCH

LAUNCH
WIND SPEED
INCREASED
LAUNCH
WIND FROM NW AVERAGE
Fig. 6.12. Vasari screenshot at 4m.
DECREASED

158
 PROGRAM AFFECTS FORM AND STRUCTURE | EXTERIOR PROGRAM ELEVATIONS

6m ELEVATION

LAUNCH

LAUNCH

KICKS AND THROWS

WIND FROM SE
Fig. 6.13. Vasari screenshot at 6m.

KICKS AND THROWS

LAUNCH

LAUNCH

WIND SPEED
INCREASED
WIND FROM NW AVERAGE
Fig. 6.14. Vasari screenshot at 6m.
DECREASED

159
PROGRAM AFFECTS FORM AND STRUCTURE | EXTERIOR PROGRAM ELEVATIONS

8m ELEVATION

LAUNCH

LAUNCH

KICKS AND THROWS

WIND FROM SE
Fig. 6.15. Vasari screenshot at 8m.

KICKS AND THROWS

LAUNCH

LAUNCH

WIND SPEED
INCREASED

WIND FROM NW AVERAGE

Fig. 6.16. Vasari screenshot at 8m.
DECREASED

160
 PROGRAM AFFECTS FORM AND STRUCTURE | EXTERIOR PROGRAM ELEVATIONS

10m ELEVATION
WIND ABOVE BUILDING NOT SHOWN IN VASARI

LAUNCH

LAUNCH

KICKS AND THROWS

HEIGHT OF TURBINE

WIND FROM SE HEIGHT OF TURBINE

Fig. 6.17. Vasari screenshot at 10m.

KICKS AND THROWS

HEIGHT OF TURBINE

HEIGHT OF TURBINE

LAUNCH

LAUNCH
WIND FROM NW
Fig. 6.18. Vasari screenshot at 10m.

161
PROGRAM AFFECTS FORM AND STRUCTURE | EXTERIOR PROGRAM ELEVATIONS

20m ELEVATION
WIND ABOVE BUILDING NOT SHOWN IN VASARI

FLYING

FLYING

WIND FROM SE
Fig. 6.19. Vasari screenshot at 20m.

FLYING

FLYING

WIND FROM NW
Fig. 6.20. Vasari screenshot at 20m.

162
 PROGRAM AFFECTS FORM AND STRUCTURE | EXTERIOR PROGRAM ELEVATIONS

60m ELEVATION
WIND ABOVE BUILDING NOT SHOWN IN VASARI

HEIGHT OF TURBINE

HEIGHT OF TURBINE

WIND FROM SE
Fig. 6.21. Vasari screenshot at 60m.

HEIGHT OF TURBINE

HEIGHT OF TURBINE

WIND FROM NW
Fig. 6.22. Vasari screenshot at 60m.

163
PROGRAM AFFECTS FORM AND STRUCTURE | REFINEMENT OF STRUCTURE FOR FORM AND PROGRAM

REFINEMENT OF STRUCTURE FOR FORM AND PROGRAM

The last step in the refinement of the building is to co-ordinate the informed the decision to clad the freestanding wall with perforated
form, program and structure. The structural bays may be adjusted panels. The perforations alleviate some of the suction within the
to fit within the refined form and to accommodate the interior courtyard, while still slowing down the wind when it comes from the
program. Required alterations to the structure to accommodate other direction to continue to shelter the courtyard.1 The sizes of
these considerations could include changes to floor-to-floor the perforations transition from large at the ends of the courtyard,
heights, number of floors, or desired column spacing for the interior to small in the middle of the wall. Without the perforations, a
spaces. At this point, the structure may be made into a prominent considerable amount of wind could enter the courtyard through
feature of the architecture, as it may impact the use or aesthetic the narrow openings at either end when the wind comes from
of the space. The final structural layout that fits within the building the northwest.2 The gradual transition in perforation size means
form and accommodates the interior program is shown in Fig. 6.23. that there are no longer only two small openings through which
substantial winds could blow. This reduces the amount of wind that
The interaction between the building form, the program, and the enters the courtyard.3
structure is shown in the section in Fig. 6.24. This sectional study

164
 PROGRAM AFFECTS FORM AND STRUCTURE | REFINEMENT OF STRUCTURE FOR FORM AND PROGRAM

bracing in bays was removed to accommodate office

space; could use moment-resisting connections
instead to keep frames open

structure increased to two full storeys

with single-storey bays for office space
and energy storage equipment

these bays were widened to 5m and increased

to two storeys high to accommodate interior
spectator area column spacing
adjusted to fit large
garage door into the
equipment storage area

3m-wide, double-height, open

bays inserted along this side of the
building for interior spectator area

bracing removed in every other bay to

accommodate change rooms

bays increased to two storeys high and bracing removed to

create an open entrance space; could use moment-resisting
connections instead to keep frames open

Fig. 6.23. Structural layout for building form and interior program.

165
PROGRAM AFFECTS FORM AND STRUCTURE | REFINEMENT OF STRUCTURE FOR FORM AND PROGRAM

SE

Fig. 6.24. Key drawing (left) and building cross-section (spread).

166
PROGRAM AFFECTS FORM AND STRUCTURE | REFINEMENT OF STRUCTURE FOR FORM AND PROGRAM

167
 CONCLUSION

169
CONCLUSION | DESIGN METHOD APPLICATIONS

DESIGN METHOD APPLICATIONS

The design method that is developed in this thesis considers the
relationship between wind, building form, and building structure,
to use the wind in a productive manner instead of simply viewing it
as something to be resisted. The use of CFD software allows the
architect to form the building to harness the wind for both leisure
and economy, through the creation of appropriate wind conditions
for the exterior sports programs and the wind energy generation
technologies. The combined use of the CFD and FEA software also Fig. 7.1. Buildings creating fast or turbulent winds can make it uncomfortable
allows the architect to form the building in a way that reduces the for pedestrians to walk at street level.
wind loads that are exerted on it, and design the structure to resist
the wind loads that cannot be mitigated by the building form.
There are other applications to which this method may be applied
STAGNATION 
POINT

in order to harness or manipulate the wind for a specific purpose.

The following wind issues should be considered in all building and h

city design, and they emphasize the importance of considering the

(2/3)h

relationship between wind and buildings that is the focus of the

design method. The method may be applied to building design to
address these issues early in the architectural design process.
Fig. 7.2. Tall buildings direct high-speed wind down to street level.

PEDESTRIAN COMFORT
slow

fast
low
SPEED

Although wind tends to slow down in cities, it can interact with high 

PRESSURE DOWNWASH EFFECT
buildings in several ways that cause it to speed up or become
turbulent at ground level, making it uncomfortable or even
dangerous for pedestrians (Fig. 7.1).1 Tall buildings can direct high-
speed wind down their windward faces and to the street level with
the downwash effect (Fig. 7.2).2 Buildings running parallel to each
other can channel the wind and accelerate it (Fig. 7.3).3 Wind also Fig. 7.3. Wind is channeled and accelerated between buildings.
accelerates around building corners (Fig. 7.4)4 and through small slow
openings in or between buildings (Fig. 7.5).5 slow

fast
Building forms may be designed to alter the wind around them
fast
low
SPEED

CHANNELLING EFFECT
high 

PRESSURE SPEED

170
 CONCLUSION | DESIGN METHOD APPLICATIONS

- PRESSSURE

+ PRESSSURE

ZONE OF INCREASED 
WIND SPEED

Fig. 7.4. Wind accelerates around corners.

slow
TOWER AND PODIUM
fast
Fig. 7.7. Podiums deflect wind.
TOWER AND PODIUM
low
SPEED

CORNER EFFECT
high 

PRESSURE

MAX 2h

Fig. 7.5. Wind accelerates through openings between buildings.

slow
TOWER AND PODIUM CANOPY Fig. 7.8. Setbacks shelter pedestrians
SETBACK from downwash.
slow CANOPY SETBACK
fast
low
SPEED
slow

VENTURI EFFECT
high 
fast

REDUCTION OF 
PRESSURE
low
SPEED fast
low
REDUCTION OF 
DOWNWASH EFFECT
high  SPEED

DOWNWASH EFFECT
PRESSURE high 

PRESSURE

Fig. 7.9. Breezeways deflect wind before it reaches ground level.

slow

Fig. 7.6. Canopies deflect wind.

CANOPY SETBACK low
fast
SPEED

BREEZEWAY EFFECT
high 

PRESSURE
slow
ED 
EAS
slow E O
ECR
F D PEED
ZON WIND S
fast
low
SPEED
fast
REDUCTION OF  h 6h - 8h

DOWNWASH EFFECT
high  MAX w w

PRESSURE
SPEED Fig. 7.10. Porous screens decrease wind speed.

171

slow

fast
low
SPEED POROUS WINDBREAK 
high 
CONCLUSION | DESIGN METHOD APPLICATIONS

so that it is not uncomfortable for pedestrians. For example,

canopies (Fig. 7.6) and podiums (Fig. 7.7) reduce the downwash
effect by directing the wind away from the ground, while setbacks
provide sheltered spaces for pedestrians to walk, out of the flow
of the downwash (Fig. 7.8).6 A breezeway, which is an open space
through a building partway up its height, allows the high-speed
wind that travels down the windward building face to be deflected
horizontally through the opening before it reaches ground level (Fig.
7.9).7 To reduce wind speed in areas where wind is accelerated
horizontally through, around, or between buildings, porous screens
may be installed (Fig. 7.10).8
Fig. 7.11. Small parks in Toronto’s downtown.
The design method that is developed in this thesis may be used to
form buildings to create comfortable ground-level wind conditions
for pedestrians, rather than wind conditions specifically for sports
and energy generation technologies. The same iterative CFD
process could be run, but with the goal of reducing wind speed
and turbulence at the building’s base. After this first step in which
the wind conditions around the building are refined, the rest of the
steps to develop the building could be carried out the same way as
described in the previous chapters.

COOLING AND VENTILATION Fig. 7.12. Central Park in New York City.
For hundreds of years, the wind has been used by architects for
natural cooling and ventilation of buildings and cities.9 If buildings
and cities are designed to use the wind advantageously in this
manner, they could create more comfortable interior and urban
environments while reducing the cooling and ventilation loads on
the buildings’ mechanical systems. The wind environment therefore
should be considered by building designers and urban planners,
so that building and city layout, orientation, and shape may be
designed to use the winds advantageously.10
Building and city form can be designed to alter the wind patterns
around and through the buildings, to induce natural cooling and Fig. 7.13. Ventilation corridors in Masdar City.

172
 CONCLUSION | DESIGN METHOD APPLICATIONS

ventilation. Open spaces are necessary within cities to let the air
flow through them to use the wind for cooling and ventilation.11
This may be accomplished with small open spaces, such as the
many small parks across Toronto’s downtown (Fig. 7.11), with one
large open space, as in the case of New York City’s Central Park
(Fig. 7.12),12 or with ventilation corridors running through the city,
such as those in the proposed Masdar City in Abu Dhabi, UAE
(Fig. 7.13).13 Cities should also be permeable to the surrounding
HIGH PRESSURE LOW PRESSURE rural areas so that an intake of fresh rural air is achieved (Fig. 7.14),
rather than recycling polluted city air throughout these ventilation
LABELS: COOL AIR WARM AIR spaces.14 This fresh air moving through the city can then be taken
LABELS:
LEGEND: COOL AIR | WARM AIR in by buildings for the cooling and ventilation of their interiors.
Fig. 7.14.
LEGEND: COOLPressure differentials
AIR | WARM AIR induce an intake of fresh rural air into cities.
UNDER TREES:
UNDER
UNDER CITY:
HIGH PRESSURE
LOWTREES: HIGH PRESSURE
PRESSURE Cities may be oriented according to prevailing wind directions to
UNDER CITY: LOW PRESSURE encourage this flow of air into and throughout the city. For example,
the street patterns of Korčula, Croatia, close the city against the
cold winter winds from the north, but orient the streets so that the
east-west summer winds can enter through these open corridors
and ventilate the city when it is needed during the hot months
(Fig. 7.15).15 Similarly, the ventilation corridors of Masdar City are
oriented along the axes of the site’s two prevailing wind directions,
which alternate on a diurnal cycle (Fig. 7.16, Fig. 7.17).16 During the
day, hot winds coming from the northwest flow into the corridors
and are cooled as water evaporates into the air from the vegetation
(Fig. 7.18).17 The cooled air then gets diverted down side streets.18
At night, cool winds from the east enter the city from the other side
and are also diverted along side streets (Fig. 7.19).19
The design method of this thesis may be applied to the development
of buildings and cities that are conducive to natural cooling and
ventilation. The CFD simulations may be run for a model of a city
or neighbourhood, instead of a single building, to predict the wind
flows between the buildings. The city form may be adjusted with
the same iterative process as the one described in this thesis to
improve the wind patterns for cooling and ventilation of the city.
CFD simulations may also be run for a building interior to visualize
Fig. 7.15. Korčula, Croatia blocks cold winds but is permeable to warm winds. the air flow within, rather than around, the building. This would allow
173
CONCLUSION | DESIGN METHOD APPLICATIONS

Fig. 7.16. Masdar City day wind rose. Fig. 7.17. Masdar City night wind rose.

Fig. 7.18. Hot winds from the northwest enter the corridors during the day. Fig. 7.19. Cool winds from the east enter the corridors at night.

174
 CONCLUSION | DESIGN METHOD APPLICATIONS

the architect to make adjustments to the interior layout to improve

the natural cooling and ventilation within the building.

POLLUTANT DISPERSION
All winds, from strong to weak, disperse pollutants in urban areas
(Fig. 7.20).20 This is because cities emit their own pollutants,21 but
also because global winds carry pollutants around the world (Fig.
7.21) to contaminate cities that may not produce as much pollution
themselves.22 Winds can either spread these pollutants throughout
the city, or carry them away from the city.23 Pollutant dispersion
should therefore be a consideration in architectural and urban
design, so that buildings and cities may be designed to remove
pollutants from the air and make the environment healthier for its
residents.24
Building forms can alter the wind patterns around them, so that
Fig. 7.20. Winds disperse pollutants in urban areas. pollutants are carried by the wind away from or out of the city. For
example, the open spaces that may be used for city ventilation
should distribute clean air, rather than contaminated air, throughout
the city. These ventilation spaces should therefore not be near
industrial zones, major highways, or other areas that produce a
lot of pollution.25 It is especially advantageous if these ventilation
spaces are planted with vegetation, as the temperature difference
between the warmer city and the cooler green space induces wind
flow from the green space into the city, in the same way wind flows
into the city from surrounding rural areas.26 The clean air from the
green space moves into the city to cool, ventilate, and remove
pollutants rather than distribute them.27 Cities may also be designed
to direct winds from one direction through the city for cooling and
ventilation, while directing winds from another direction overtop of
the city so that they do not mix with the air at ground level.28 If
winds from a certain direction tend to carry pollutants, either from
Fig. 7.21. Global winds carry air pollution around the world. international sources carried by global wind patterns, or from local
pollution sources, this strategy can direct the polluted winds up
and over the city. Such a strategy is employed in the design of

175
 CONCLUSION | DESIGN METHOD APPLICATIONS

Fig. 7.22. Xeritown day wind rose.

>2.4x
Fig. 7.23. During the day, cool winds are channeled through the city for cooling and ventilation.
LABELS:
TOP: h=x and w= <0.7 x
BOTTOM: h=x and w= >2.4x

Fig. 7.24. Xeritown night wind rose.

<0.7x
Fig. 7.25. At night, hot winds are diverted over the tops of the buildings.

176
 CONCLUSION | DESIGN METHOD APPLICATIONS

Xeritown (Fig. 7.26), a proposed extension of Dubai, UAE, although it

is used to force hot, rather than polluted, winds over the city.29 During
the day, the cool winds from the sea (Fig. 7.22) are channeled through
the city for cooling and ventilation (Fig. 7.23).30 Long, open spaces are
aligned in the direction of these winds so that the wind will dip down
into them and cool the ground level.31 This may be achieved with open
spaces that have a width-to-height ratio of at least 2.4.32 At night,
the hot wind that comes from the desert (Fig. 7.24) is diverted over
the tops of the buildings (Fig. 7.25).33 In this direction, the buildings
are stepped up towards the centre of the urban area, with the rising
edges facing the oncoming desert winds.34 The spaces between the
buildings in this direction have a width-to-height ratio of less than 0.7,
Fig. 7.26. Xeritown. so that the wind will skim overtop of the openings instead of moving
down into them.35 These two strategies ensure that the unwanted
hot winds blow up and over the city instead of heating up the street
level. These strategies could also be employed to ensure that polluted
winds from one direction are kept out of the city, while the city is kept
permeable to clean air from another direction. Another strategy is to
vary building heights along a city skyline to induce turbulence at the
tops of the buildings (Fig. 7.27).36 This turbulence exchanges polluted
air closer to ground level with fresher air from the top of the city, to
continually recycle the air at street level.37 The skyline of Xeritown was
designed to have building heights with enough variation to induce this
air cycling and remove pollutants.38
The design method developed in this thesis may be used to create
urban forms that direct pollutants out of, rather than throughout, the
city. With CFD simulations of wind patterns, the user may visualize
where pollutants will be distributed by these winds. This allows them
to develop iterations of the urban area, so that its form creates wind
Fig. 7.27. Varied building heights induce turbulence at the skyline. patterns that will minimize pollutant dispersion within the city.

SNOW AND ICE ACCUMULATION

When wind blows snow and ice, it affects where snow and ice
accumulates on and around buildings. This can cause particular

177
CONCLUSION | DESIGN METHOD APPLICATIONS

issues that are addressed in building codes. The Ontario Building

Code, for example, stipulates that buildings must be designed so
that exit doors are never blocked by the accumulation of snow or
ice for consistently accessible egress (Fig. 7.28),39 and that snow
and ice may not accumulate over vents for air intake or exhaust
so that the building’s ventilation is not compromised (Fig. 7.29).40
Without paying particular attention to a site’s wind patterns during
building design, the ways in which snow and ice accumulation will
happen can be difficult to predict, and can cause these problems
to occur after construction is complete. The Ontario Building
Fig. 7.28. Snow can block doors. Fig. 7.29. Snow can block vents.
Code also specifies that the building structure and cladding must
be designed to withstand snow loads (Fig. 7.30),41 and provides
formulas that may be used to calculate the design snow load for
a particular building on a particular site.42 However, these formulas
only provide snow load values for a small number of building forms
and environmental situations.43 For example, the role that the wind
plays in the accumulation of snow, and the resultant snow load
that is exerted on the building, is only considered with the wind
exposure factor in the provided formula.44 This factor is determined
based on only the number of sides of the building that are exposed
to the wind, and the nature of the building’s roof obstructions that
would block the wind flow.45 The more the building is exposed Fig. 7.30. Buildings must be able to withstand the load exerted by snow.
to the wind, the more the wind exposure factor may be reduced,
which reduces the snow load that the building must be designed to
resist.46 Similarly, the building form is considered in the formula with
only the slope factor, which is determined by the slope of the roof
and the roofing material,47 and the shape factor, which depends on
the shape of the roof, the presence of roof valleys in which snow
could accumulate, and the number of roof levels between which
snow could drift.48 These factors can sometimes be too generic
to allow the designer to ensure that wind-induced snow and ice
accumulation will not overload the building,49 so it is beneficial to
study in more detail the ways in which wind affects snow and ice
accumulation on a particular building and site.
Buildings may be formed so that snow and ice is not accumulated Fig. 7.31. Simulated snow on a building model in a wind tunnel.

178
 CONCLUSION | DESIGN METHOD APPLICATIONS

in undesirable areas. The accumulation patterns may be tested

with a building model in a wind tunnel, using simulated snow made
out of, for example, crushed walnut shells, sawdust, or sand (Fig.
7.31).50 This allows the designer to visualize how snow will fall and
accumulate on a building while the wind is blowing, or how the
wind will cause snow that is already on the ground to drift and build
up on and around the building.51 The results of the wind tunnel test
may also help to determine the design snow loads that should be
used when designing the building’s structure,52 rather than relying
on the more generic formulas provided by the building code. CFD
Fig. 7.32. Svalbard Science Centre. software may also be used to simulate how the wind will impact
snow and ice accumulation. The Svalbard Science Centre (Fig.
7.32) by Jarmund/Vigsnaes AS Architects was designed using
CFD software for this purpose.53 Consideration of the strong wind
flows across the site was a primary driver in the generation of the
building’s geometry, as its irregular plan, pitched roofs and sloped
walls were formed to avoid unwanted snow drifts around doors
and windows.54 This is an especially important consideration
because of the building’s northern location in Longyearbyen,
Norway.55 Through several iterations, the building form was tested
with both physical models (Fig. 7.33) as well as more detailed CFD
simulations (Fig. 7.34) to visualize where the wind would blow the
Fig. 7.33. Physical model. snow around the building.56 Both of these kinds of tests allow the
architect to design the building form to control the accumulation of
snow and ice. This avoids the accessibility and loading issues that
are associated with unwanted accumulation.
The design method of this thesis may be applied in a similar
manner. The user may run the CFD simulations particularly to
observe the ways in which the site’s wind patterns would cause
snow and ice to accumulate on and around the building. Through
the iterative process, the building form could be refined so that
snow and ice accumulates only in desired areas, to keep entrances
and exits accessible, to keep windows clear to let natural light into
the building, and to reduce the snow loads acting on the building.
Fig. 7.34. CFD simulation.

179
CONCLUSION | SUMMARY

SUMMARY
A design method (Fig. 7.35) in which architects use wind effects and STEP 3 | WIND AFFECTS STRUCTURE
loading as a design generator may be carried out with a pairing of
Next, the wind pressure information that is provided by Flow Design
computational fluid dynamics software and finite element analysis
is used to divide the building form into pressure zones based on
software.
the areas that are subjected to approximately the same amount
of wind pressure. A structural bay is then modeled for one of the
STEP 1 | FORM AFFECTS WIND
pressure zones and tested with Scan&Solve by inputting gravity
First, a building form is modeled in Vasari, and a CFD simulation loading and the average positive and negative wind pressure
is run in Vasari to visualize the wind speeds and patterns that are values for that zone that were obtained from Flow Design. If there is
created around the building by the building form. The architect then too much deflection, the architect adjusts the model of the bay to
makes observations and adjusts the form within Vasari for the next increase its stiffness and reduce its deflection under wind loading,
iteration to improve the appropriateness of the surrounding wind and then tests the adjusted model with Scan&Solve until the bay
conditions for the exterior programs. They may repeat this step to is able to resist large movements in the wind. This step is repeated
refine and re-test many iterations of their design until the building to develop a bay for each pressure zone and the bays are laid out
form creates the desired wind conditions. within the building form to create the building’s structural system.

STEP 2 | WIND AFFECTS FORM STEP 4 | PROGRAM AFFECTS FORM AND STRUCTURE
In the second step, which is carried out concurrently with the Finally, the form of the building is adjusted so that it could feasibly
first step, a CFD simulation is run in Flow Design to evaluate the accommodate both exterior and interior programs. This adjusted
aerodynamics of the building form by providing information about model should be tested with a CFD simulation in Vasari to ensure
the wind pressure that is exerted on each building face. The that the changes to the form did not negatively alter the wind
architect then makes observations and adjusts the form for the conditions around the building. The layout of the structural bays is
next iteration to improve the form’s aerodynamics. This step may then adjusted to fit within the refined form. This step completes the
be repeated until the building form has the desired aerodynamic feedback loop between the CFD and FEA software, and the design
properties. method may be repeated as many times as necessary to develop
a building that both shapes and resists the wind.

180
 CONCLUSION | SUMMARY

STEP 1 | FORM AFFECTS WIND STEP 2 | WIND AFFECTS FORM
VASARI FLOW DESIGN
MODEL BUILDING FORM

RUN CFD FOR WIND CONDITIONS RUN CFD FOR WIND PRESSURE

EVALUATE WIND CONDITIONS EVALUATE FORM AERODYNAMICS

ADJUST BUILDING FORM

STEP 3 | WIND AFFECTS STRUCTURE
SCAN&SOLVE FOR RHINO
DIVIDE BUILDING FORM INTO PRESSURE ZONES

MODEL STRUCTURAL BAY

RUN FEA FOR STRUCTURAL BEHAVIOUR

EVALUATE STRUCTURAL BEHAVIOUR

ADJUST STRUCTURAL BAY

LAY OUT BAYS IN BUILDING FORM TO CREATE STRUCTURAL SYSTEM

STEP 4 | PROGRAM AFFECTS FORM AND STRUCTURE
RHINO
ADJUST BUILDING FORM TO ACCOMMODATE PROGRAM

Fig. 7.35. Design ADJUST STRUCTURAL BAYS TO FIT WITHIN BUILDING FORM

method sequence.

181
CONCLUSION | SUMMARY

Constant feedback is an important part of a designer’s workflow. by these tests, they can learn what the analysis results mean and
It allows the designer to test a design iteration, receive feedback how changes to their design affects these results. This feedback
about how their design works, evaluate how well it works for its informs their subsequent design iterations, and eventually the
required purposes, and then learn from this feedback and apply architect will develop an intuition of how to form their buildings
it to subsequent design iterations. Throughout this process, they to control how they interact with the wind. This is beneficial for
begin to understand which aspects of their designs work for their practicing architects who work with disciplines that are tangential
purposes and which do not. Feedback is especially important to architectural design, as well as architecture students who may
in an architectural workflow when the architect is working with learn through this feedback while they are in school, and then apply
disciplines in which they are not extensively trained, such as wind this learned intuition in the workplace after graduation.
and structural engineering. Feedback is essential for them to learn
how their designs interact with the wind, since it is unlikely that The feedback loop that is developed in this thesis may easily
the architect would already possess an intuitive understanding of be incorporated into existing architectural working practices, to
this relationship. The design methodology that is developed in this enhance them rather than change them. The method uses software
thesis provides this essential feedback through the CFD and FEA that is easy for architects to learn without previous experience with
software, as it produces visual output to show the architect how similar software, and the software automates complex processes
their design influences wind speeds and patterns, moderates wind to make it easier to use. All of the programs are also compatible
pressure, and reacts to wind loading. Software has a lot of potential with each other, expediting the process of exporting and importing
to allow architects to work within the realms of wind and structural model files between programs. The selected software is compatible
engineering, as it means that they do not need to have extensive with 3D-modeling programs that are already commonly used in
engineering knowledge to be able to test many design iterations in architectural offices, and may be run on typical computer hardware
relation to these considerations. Through the feedback provided that is already owned by most firms. The feedback that is provided

182
 CONCLUSION | SUMMARY

by the software is also in the form of visual output, conveying process to develop a schematic structural system for the project,
wind patterns and forces through a medium that architects are inputting into the FEA software the loads that are applicable. The
trained to understand. All of these considerations make the design ease with which the steps may be separated makes them feasible
method feasible to integrate into current architectural practices, to integrate into current architectural practices, because only the
both within their current hardware and software use, as well as steps that are relevant to a design stage need be carried out. The
their interpretation and production of graphic output. method may therefore be used to consider wind, structure, or both
in the architectural design process, depending on the designer’s
The design method is also easy to integrate into architectural specific needs.
workflows because designers may implement only one step
of the method at a time to develop specific aspects of their Using CFD and FEA software in a way that is appropriate for
designs. The fact that each of the steps within the method may architectural applications allows architects to integrate wind and
be repeated as many times as necessary means that they may structural engineering considerations into the early design stages of
also stand alone as iterative, repeatable processes in themselves. their current working practices. As CFD and FEA programs continue
For example, step one may be used to run CFD simulations to to be improved, the accuracy and usefulness of this method will
evaluate and adjust building form iterations for considerations of increase. It is the hope of the author that software programmers
pedestrian comfort, cooling and ventilation, pollutant dispersion, or may draw from this method to create more accurate CFD and FEA
snow and ice accumulation, as described in the previous section programs that are conducive to architectural applications, so that
of this chapter. The second step could be carried out on its own in the near future, architects may use this design methodology to
for detailed aerodynamic studies of building forms. Even if wind integrate wind, structural and architectural design processes.
was not a major consideration in a building design, the third step
of the method could be integrated into the architectural design

183
CONCLUSION | DISCUSSION AND FURTHER DEVELOPMENT

DISCUSSION AND FURTHER DEVELOPMENT

The defence of this thesis generated discussion about the further The process by which the design method is carried out would
development of this design methodology. Considerations of more be slightly different, for example, on a site with more than two
complex wind conditions, as well as the aesthetic design of the predominant wind directions. In the first step, CFD simulations
building, should be incorporated into the method if it is to be would have to be run for each of the wind directions. This would
developed in the future. take more time, and would require the building to be designed
so that appropriate wind conditions for the exterior programs
The design methodology was carried out in this thesis within a were created within all of these conditions, which could be more
simple wind condition of two predominant wind directions that are challenging to accommodate. However, the process of the first
opposite to each other. This provided simple parameters within step would essentially be carried out the same way. The second
which to develop and test the method. However, it was brought step would also become more challenging, as the building would
up in the discussion that many sites have different, more complex ideally need to be formed in a way that optimizes its aerodynamics
wind conditions, and the question was raised as to whether or not within many wind directions, instead of just two directions along
the design method would be able to be carried out within these the same axis, as considered in this thesis. Finally, the third step
other conditions. As such, the first step in the further development could be carried out much in the same way as in this thesis, but the
of the design method would be to test the method within different architect would have to find the worst-case loading condition for
wind conditions. Repeating the method with different parameters each pressure zone out of all of the predominant wind directions,
would inevitably reveal steps of the method that were designed to instead of just looking at two. All of these considerations would
work best within the original site conditions, and may not work as make the design method take longer to execute, although testing
well within the new conditions. Working through these issues as the method within multiple site conditions could potentially reveal
they are discovered, and developing the method within multiple ways of streamlining the method. Also, these tests would inevitably
wind conditions, would allow the method to be refined until it could reveal more complications that are not predicted here. However,
be applied within the conditions of any given site or project. repeating this method within several different wind conditions would

184
 CONCLUSION | DISCUSSION AND FURTHER DEVELOPMENT

be a necessary step in the development of the design methodology, designed with the methodology to fit within this form that is the
to ensure that it may be used for any site or project. product of both architectural and engineering considerations.

It was also brought up in the discussion following the defence Even if this design methodology is not developed further after this
that because the building was formed almost entirely based thesis, it has still provided me with insights that I may apply in the
on considerations of surrounding wind conditions and form architectural field. It has allowed me to gain an understanding of
aerodynamics, the aesthetic and functional considerations that engineering topics that are often tangential to architectural design.
normally drive the architectural design process do not influence the Having an understanding of these considerations, even if not a
resulting building form. The method could therefore be developed mastery of them, will help with the collaboration between architects
to provide a step in which these factors would influence the design, and engineers that is a necessity in architectural practice. This
so that the building could be created based on considerations of would provide the opportunity to work with architects to make
both engineering and architecture. This could be accommodated by them aware of the implications that wind and structural engineering
expanding the fourth step to allow the building form to be adjusted considerations may have on the architecture of a building. It would
more substantially based on programmatic considerations or also allow for the possibility of involvement in wind or structural
building aesthetic. The resulting building would, however, need to engineering design processes, as the knowledge gained in
be re-tested with the design method to see how the new form and this thesis would allow me to collaborate more efficiently with
structure functioned within the site’s wind conditions. The building engineers, while offering an architect’s perspective on these wind
may then need to be adjusted to fix any wind-related issues that and structural considerations.
may result. A potentially more streamlined approach would be to
create a schematic building design first, based only on architectural Whether the design methodology itself is developed further, or
considerations, and then run the design methodology to adjust the whether the insights gained from it are applied to architectural
form to create more desirable surrounding wind conditions and to practice, it has been a valuable investigation into the integration of
improve its aerodynamics. The building structure could then be architectural and engineering design.

185
ENDNOTES

ENDNOTES
FRONT MATTER 15 Ibid.
16 Black and Duff, 48.
1 James Ambrose and Dimitry Vergun, Simplified Building Design for Wind and 17 Black and Duff, 40.
Earthquake Forces (New York: John Wiley & Sons, Inc., 1990), 5. 18 Black and Duff, 46.
2 Guy de Maupassant, Sur l’eau (1888), quoted in Marq De Villiers, Windswept: 19 Ibid.
The Story of Wind and Weather (New York: Walker, 2006), 16. 20 Marq De Villiers, Windswept: The Story of Wind and Weather (New York:
Walker, 2006), 2.

WORKFLOW 21 Mareike Krautheim, Ralf Pasel, Sven Pfeiffer, and Joachim Schultz-Granberg,
City and Wind: Climate as an Architectural Instrument (Berlin: DOM Publishers,
2014), 43.
1 “Part One: Schism,” in Bridging the Gap: Rethinking the Relationship
22 De Villiers, 24-5.
of Architect and Engineer, ed. Deborah Gans (New York: Van Nostrand
Reinhold, 1991), 1. 23 De Villiers, 73.
2 William L. MacDonald, The Architecture of the Roman Empire Volume 1: An 24 Ibid.
Introductory Study (United States of America: Yale University, 1965), 138. 25 Krautheim, Pasel, Pfeiffer, and Schultz-Granberg, 40.
3 David Billington, “The New Art of Engineering,” in Bridging the Gap: Rethinking 26 De Villiers, 73.
the Relationship of Architect and Engineer, ed. Deborah Gans (New York: Van 27 Ibid.
Nostrand Reinhold, 1991), 4.
28 Krautheim, Pasel, Pfeiffer, and Schultz-Granberg, 41.
4 Ibid.
29 De Villiers, 74.
5 Ibid.
30 Krautheim, Pasel, Pfeiffer, and Schultz-Granberg, 42.
6 Peter McCleary, “The Differing Projects of Life of the Architect and the
31 Ibid.
Engineer,” in Bridging the Gap: Rethinking the Relationship of Architect and
Engineer, ed. Deborah Gans (New York: Van Nostrand Reinhold, 1991), 40. 32 De Villiers, 74.
7 Olga Popovic Larsen, Conceptual Structural Design: Bridging the gap 33 “How Earth Made Us: Wind,” directed by Nigel Walk (2010; BBC, 2010),
between architects and engineers, ed. by Andy Tyas (London: Thomas television.
Telford, 2003), 6. 34 De Villiers, 74.
8 Larsen, 17. 35 Ibid.
9 Larsen, 61. 36 Ibid.
10 R. Gary Black and Stephen Duff, “A Model for Teaching Structures: Finite 37 “How Earth Made Us: Wind.”
Element Analysis in Architectural Education,” Journal of Architectural 38 De Villiers, 74-5.
Education 48, no. 1 (1994): 38.
39 Ibid.
11 Ibid.
40 De Villiers, 75.
12 Black and Duff, 39.
41 Ibid.
13 Ibid.
42 Ibid.
14 Black and Duff, 45.
43 “How Earth Made Us: Wind.”

186
 ENDNOTES

44 De Villiers, 82-3. 76 De Villiers, 99.

45 De Villiers, 75. 77 De Villiers, 100.
46 “Polar Winds,” Kids Earth Science, accessed July 15, 2015, http://www.kids- 78 De Villiers, 104.
earth-science.com/polar-winds.html. 79 Michele Melaragno, Wind In Architectural And Environmental Design (New
47 Ibid. York: Van Nostrand Reinhold Company, 1982), 93.
48 Ibid. 80 De Villiers, 101.
49 De Villiers, 81. 81 Leighton Cochran, Wind Issues in the Design of Buildings (Reston, Virginia:
50 De Villiers, 82. American Society of Civil Engineers, 2012), 17.
51 De Villiers, 81. 82 De Villiers, 99.
52 Krautheim, Pasel, Pfeiffer, and Schultz-Granberg, 43. 83 De Villiers, 101.
53 Ibid. 84 Melaragno, 93.
54 De Villiers, 74. 85 Melaragno, 123.
55 Ibid. 86 De Villiers, 104-5.
56 Krautheim, Pasel, Pfeiffer, and Schultz-Granberg, 53. 87 Ibid.
57 De Villiers, 91. 88 De Villiers, 105.
58 Ibid. 89 De Villiers, 104.
59 De Villiers, 90. 90 De Villiers, 106.
60 De Villiers, 84. 91 Melaragno, 62.
61 De Villiers, 90. 92 Melaragno, 64.
62 Krautheim, Pasel, Pfeiffer, and Schultz-Granberg, 64. 93 Cochran, 17.
63 Ibid. 94 Cochran, 36.
64 Jessica Bennett, “Wind Design Guide” (architectural aerodynamics course 95 Melaragno, 150.
reading, Victoria University of Wellington, http://www.victoria.ac.nz/ 96 Ibid.
architecture/centres/cbpr/publications/architectural-aerodynamics/pdfs/ 97 De Villiers, 101.
BBSC_433_Jessica-Bennett_Wind-Design-Guide.pdf), 13.
98 Ibid.
65 Bennett, 14.
99 Ibid.
66 Krautheim, Pasel, Pfeiffer, and Schultz-Granberg, 64.
100 Melaragno, 150.
67 De Villiers, 90.
101 Cochran, 35.
68 Ibid.
102 Melaragno, 150.
69 Ibid.
103 Cochran, 36.
70 Krautheim, Pasel, Pfeiffer, and Schultz-Granberg, 63.
104 Ibid.
71 Krautheim, Pasel, Pfeiffer, and Schultz-Granberg, 124.
105 Krautheim, Pasel, Pfeiffer, and Schultz-Granberg, 56.
72 Ibid.
106 Krautheim, Pasel, Pfeiffer, and Schultz-Granberg, 12.
73 Krautheim, Pasel, Pfeiffer, and Schultz-Granberg, 64.
107 Krautheim, Pasel, Pfeiffer, and Schultz-Granberg, 56.
74 Ibid.
108 “Overview: Computational fluid dynamics software,” Autodesk, accessed
75 Ibid. January 28, 2015, http://www.autodesk.com/products/cfd/overview.

187
ENDNOTES

109 Krautheim, Pasel, Pfeiffer, and Schultz-Granberg, 86. 140 Hutton, 4.

110 “Overview: Computational fluid dynamics software.” 141 Spremulli, September 18, 2015.
111 Krautheim, Pasel, Pfeiffer, and Schultz-Granberg, 116-7. 142 Kurowski, 10.
112 Ibid. 143 Kurowski, 132.
113 Krautheim, Pasel, Pfeiffer, and Schultz-Granberg, 86. 144 Ibid.
114 Ibid. 145 Mario Salvadori and Robert Heller, Structure in Architecture: The Building of
115 Ziad Boutanios, PEng., BinkZ, e-mail message, July 3, 2015. Buildings (New Jersey: Prentice-Hall, 1986), 86.
116 Krautheim, Pasel, Pfeiffer, and Schultz-Granberg, 86. 146 Kim and Sankar, 17.
117 Ibid. 147 “Displacement,” Encyclopædia Britannica, accessed March 3, 2015, http://
www.britannica.com/science/displacement-mechanics.
118 Ibid.
148 “Stress,” Encyclopædia Britannica, accessed March 3, 2015, http://www.
119 Ibid.
britannica.com/science/stress-physics.
120 Ibid.
149 Salvadori and Heller, 86.
121 Boutanios, July 3, 2015.
150 Ibid.
122 Krautheim, Pasel, Pfeiffer, and Schultz-Granberg, 86.
151 Ibid.
123 Boutanios, July 3, 2015.
152 Salvadori and Heller, 94.
124 Krautheim, Pasel, Pfeiffer, and Schultz-Granberg, 86.
153 Salvadori and Heller, 97.
125 Boutanios, July 3, 2015.
154 Salvadori and Heller, 100.
126 Ibid.
155 “Mechanics of Materials: Strain,” eFunda, accessed March 3, 2015, http://
127 “Finite element analysis,” Autodesk, accessed January 28, 2015, http://www. www.efunda.com/formulae/solid_mechanics/mat_mechanics/strain.cfm.
autodesk.com/solutions/finite-element-analysis.
156 Salvadori and Heller, 86.
128 Ibid.
157 “Displacement.”
129 Black and Duff, 43.
158 Dr. Elizabeth English, University of Waterloo, meeting, September 19, 2015.
130 Paul M. Kurowski, Finite Element Analysis Design for Engineers (Warrendale:
159 English, September 21, 2015.
SAE International, 2004), 6.
160 Ibid.
131 Kurowski, 5-6.
161 Salvadori and Heller, 89.
132 Ibid.
162 Salvadori and Heller, 96.
133 Kurowski, 61.
163 Salvadori and Heller, 94.
134 David V. Hutton, Fundamentals of Finite Element Analysis (New York: The
McGraw-Hill Companies, Inc., 2004), 4. 164 Salvadori and Heller, 100.
135 Kurowski, 47. 165 Hutton, 4.
136 Matthew Spremulli, University of Waterloo, document revisions, September 166 Ibid.
18, 2015. 167 Ibid.
137 Hutton, 4. 168 Ibid.
138 Kurowski, 21. 169 “Effective Uses of Scan&Solve: Resolution and Convergence,” Intact
139 Nam-Ho Kim and Bhavani V. Sankar, Introduction to Finite Element Analysis Solutions, accessed January 28, 2015, http://www.intact-solutions.com/
and Design (Hoboken: John Wiley & Sons, Inc., 2009), 271. sns_documents/sns_2014/resolution.htm.

188
 ENDNOTES

170 Kim and Sankar, 267. Design World, last modified May 11, 2012, http://www.designworldonline.
171 Kim and Sankar, 271. com/meshless-fea-a-new-way-to-solve-linear-static-fea-problems/#_????.
172 Kim and Sankar, 267. 197 “SOFiSTiK Rhinoceros Interface,” SOFiSTiK, accessed February 4, 2015,
http://www.sofistik.com/en/solutions/structural-fea/rhinoceros-interface/.
173 Kurowski, 137.
198 “SOFiSTiK Rhinoceros Interface,” McNeel, last modified September 20,
174 Kim and Sankar, 271.
2011, http://blog.rhino3d.com/2011/09/sofistik-rhinoceros-interface.html.
175 Kim and Sankar, 267.
199 Laura Baumgärtner Dipl.-Ing., Sales and Consulting, SOFiSTiK, e-mail
176 Kim and Sankar, 272. message, February 10, 2015.
177 Ibid. 200 Ibid.
178 Ibid. 201 “Overview: Finite element analysis and modeling software,” Autodesk,
179 “Linear Static Analysis: Mathematical Model: Linear Elasticity,” Intact accessed January 28, 2015, http://www.autodesk.com/products/simulation-
Solutions, accessed January 28, 2015, http://www.intact-solutions.com/ mechanical/overview.
sns_documents/sns_2014/mathematical.htm. 202 Ware, January 30, 2015.
180 Kurowski, 1. 203 Kurowski, 137.
181 Kurowski, 2.
182 Black and Duff, 46.
183 “Autodesk Vasari,” Autodesk, accessed January 28, 2015, http:// PROGRAM AND SITE
autodeskvasari.com.
184 “Part 7: Running Wind Tunnel Simulations,” Autodesk Building Solutions 1 “Wind & weather statistics Regina Airport,” Windfinder, last modified July
(September 6, 2011), video, http://www.youtube.com/watch?v=PhuYa8aXYj 2015, http://www.windfinder.com/windstatistics/regina_airport.
U&index=31&list=PLF8ACB9BED56DE576. 2 “Population of census metropolitan areas,” Statistics Canada, last modified
185 “Autodesk Vasari.” February 11, 2015, http://www.statcan.gc.ca/tables-tableaux/sum-som/l01/
cst01/demo05a-eng.htm.
186 “Features: Flow simulation tools,” Autodesk, accessed January 28, 2015,
http://www.autodesk.com/products/flow-design/features/all/list-view. 3 Esam Hussein, “Dean’s Message,” University of Regina, accessed January
13, 2015, http://www.uregina.ca/engineering/about-us/deans-message.
187 Ibid. html.
188 Ibid.
189 “Features: Autodesk CFD software benefits,” Autodesk, accessed January 28,
2015, http://www.autodesk.com/products/cfd/features/key-differentiators/ FORM AFFECTS WIND
gallery-view.
190 Ibid. 1 Marq De Villiers, Windswept: The Story of Wind and Weather (New York:
191 Jamie Ware, AEC Account Executive, IMAGINiT Technologies, e-mail Walker, 2006), 84.
message, January 30, 2015. 2 De Villiers, 90.
192 Ibid. 3 Jessica Bennett, “Wind Design Guide” (architectural aerodynamics course
193 “Scan-and-Solve for Rhino,” Intact Solutions, accessed January 28, 2015, reading, Victoria University of Wellington, http://www.victoria.ac.nz/
http://www.scan-and-solve.com. architecture/centres/cbpr/publications/architectural-aerodynamics/pdfs/
BBSC_433_Jessica-Bennett_Wind-Design-Guide.pdf).
194 Ibid.
195 “Effective Uses of Scan&Solve: Resolution and Convergence.” Leighton Cochran, Wind Issues in the Design of Buildings (Reston, Virginia:
American Society of Civil Engineers, 2012).
196 Evan Yares, “Meshless FEA: A new way to solve linear static FEA problems,”

189
ENDNOTES

Mareike Krautheim, Ralf Pasel, Sven Pfeiffer, and Joachim Schultz-Granberg, 11 Ibid.
City and Wind: Climate as an Architectural Instrument (Berlin: DOM Publishers, 12 Ibid.
2014).
Theodore Stathopoulos, “Introduction to Environmental Aerodynamics,” in
Environmental Wind Engineering and Design of Wind Energy Structures, ed. WIND AFFECTS STRUCTURE
Charalambos C. Baniotopoulos, Claudio Borri and Theodore Stathopoulos
(New York: SpringerWienNewYork, 2011). 1 Leighton Cochran, Wind Issues in the Design of Buildings (Reston, Virginia:
American Society of Civil Engineers, 2012), 1.
4 “Wind & weather statistics Regina Airport,” Windfinder, last modified July
2 Mario Salvadori and Robert Heller, Structure in Architecture: The Building of
2015, http://www.windfinder.com/windstatistics/regina_airport.
Buildings (New Jersey: Prentice-Hall, 1986), 20.
5 Ibid.
3 Cochran, 47.
6 Ibid.
4 Ibid.
5 Cochran, 49.
WIND AFFECTS FORM 6 Peter A. Irwin, “Vortices and tall buildings: A recipe for resonance,” Physics
Today, September 2010, http://www.math.lsa.umich.edu/~krasny/math654_
1 Leighton Cochran, Wind Issues in the Design of Buildings (Reston, Virginia: irwin.pdf, 68.
American Society of Civil Engineers, 2012), 47. 7 Ibid.
2 Ibid. 8 Ibid.
3 Cochran, 61. 9 Cochran, 49.
4 Cochran, 58. 10 Ibid.
5 Cochran, 61. 11 Cochran, 47.
6 Ibid. 12 Ibid.
7 Mareike Krautheim, Ralf Pasel, Sven Pfeiffer, and Joachim Schultz-Granberg, 13 Cochran, 61.
City and Wind: Climate as an Architectural Instrument (Berlin: DOM Publishers, 14 James Ambrose and Dimitry Vergun, Simplified Building Design for Wind and
2014), 71. Earthquake Forces (New York: John Wiley & Sons, Inc., 1990), 93.
8 Mario Salvadori and Robert Heller, Structure in Architecture: The Building of 15 Cochran, 54.
Buildings (New Jersey: Prentice-Hall, 1986), 20.
16 Ibid.
9 J.D. Holmes, C.J. Baker, E.C. English, and E.C.C. Choi, “Wind structure and
17 Ibid.
codification,” Wind and Structures: An International Journal (2005).
18 Ambrose and Vergun, 93.
Leighton Cochran, Wind Issues in the Design of Buildings (Reston, Virginia: 19 Cochran, 47.
American Society of Civil Engineers, 2012).
20 Cochran, 49.
Mareike Krautheim, Ralf Pasel, Sven Pfeiffer, and Joachim Schultz-Granberg, 21 Ibid.
City and Wind: Climate as an Architectural Instrument (Berlin: DOM Publishers,
22 Salvadori and Heller, 30.
2014).
23 Cochran, 49.
Mario Salvadori and Robert Heller, Structure in Architecture: The Building of 24 Salvadori and Heller, 33.
Buildings (New Jersey: Prentice-Hall, 1986).
25 Salvadori and Heller, 32.
10 Dr. Elizabeth English, University of Waterloo, meeting, September 21, 2015. 26 Salvadori and Heller, 33.

190
 ENDNOTES

27 Ibid. 60 Ibid.
28 Dr. Elizabeth English, University of Waterloo, meeting, July 29, 2015. 61 Ibid.
29 Cochran, 51. 62 Salvadori and Heller, 186.
30 Salvadori and Heller, 20. 63 Ambrose and Vergun, 78.
31 Cochran, 51. 64 Ambrose and Vergun, 73.
32 Ibid. 65 Ambrose and Vergun, 71.
33 Ibid. 66 Ibid.
34 Ibid. 67 Salvadori and Heller, 186.
35 Ibid. 68 Ambrose and Vergun, 82.
36 Ibid. 69 Maik Gehloff, Dipl.-Ing. (FH), M.A.Sc, “Specifying Modern Timber
37 Ibid. Connections,” Construction Canada, last modified May 21, 2014, http://
www.constructioncanada.net/specifying-modern-timber-connections/.
38 Ibid.
70 Ambrose and Vergun, 83.
39 Cochran, 40.
71 Ambrose and Vergun, 82.
40 Cochran, 40-1.
72 Cochran, 54.
41 Cochran, 41.
42 Cochran, 41-2.
43 Cochran, 41. PROGRAM AFFECTS FORM AND STRUCTURE
44 Salvadori and Heller, 20.
45 “Shearwalls 101: Why You Can’t Have a Window There,” BuildLLC, accessed 1 Dr. Elizabeth English, University of Waterloo, meeting, September 19, 2015.
September 3, 2015, http://blog.buildllc.com/2014/05/shearwalls-101-why- 2 Ibid.
you-cant-have-a-window-there/. 3 Ibid.
46 Cochran, 54.
47 Ambrose and Vergun, 82.
48 Cochran, 5. CONCLUSION
49 Mario Salvadori, Why Buildings Stand Up (New York: W. W. Norton &
Company, Inc., 1980), 53. 1 Mareike Krautheim, Ralf Pasel, Sven Pfeiffer, and Joachim Schultz-Granberg,
City and Wind: Climate as an Architectural Instrument (Berlin: DOM Publishers,
50 Cochran, 54. 2014), 64.
51 Ibid. 2 Theodore Stathopoulos, “Introduction to Environmental Aerodynamics,” in
52 Salvadori and Heller, 196. Environmental Wind Engineering and Design of Wind Energy Structures, ed.
53 “Shearwalls 101: Why You Can’t Have a Window There.” Charalambos C. Baniotopoulos, Claudio Borri and Theodore Stathopoulos
54 Ambrose and Vergun, 64. (New York: SpringerWienNewYork, 2011), 4.
55 “Shearwalls 101: Why You Can’t Have a Window There.” 3 Stathopoulos, 7.
56 Ibid. 4 Jessica Bennett, “Wind Design Guide” (architectural aerodynamics course
reading, Victoria University of Wellington, http://www.victoria.ac.nz/
57 Ibid.
architecture/centres/cbpr/publications/architectural-aerodynamics/pdfs/
58 Cochran, 54. BBSC_433_Jessica-Bennett_Wind-Design-Guide.pdf), 13.
59 Ibid.

191
ENDNOTES

5 Bennett, 14. 35 Krautheim, Pasel, Pfeiffer, and Schultz-Granberg, 140.

6 Stathopoulos, 7. 36 Ibid.
7 Bennett, 24. 37 Ibid.
8 Stathopoulos, 7. 38 Ibid.
9 Krautheim, Pasel, Pfeiffer, and Schultz-Granberg, 6. 39 “Ontario Regulation 350/06 Building Code” (building code, Building and
10 Yang Li, “Computational Fluid Dynamics Technology and Its Application in Development Branch of the Ministry of Municipal Affairs and Housing, Ontario,
Wind Environment Analysis,” Journal of Urban Technology 17, no. 3 (2010): May 31, 2006) 3.4.6.10.(2).
80. 40 “Ontario Regulation 350/06 Building Code,” 6.2.3.12.(3).
11 Krautheim, Pasel, Pfeiffer, and Schultz-Granberg, 127. 41 “Ontario Regulation 350/06 Building Code,” 4.1.6.1.(1).
12 Krautheim, Pasel, Pfeiffer, and Schultz-Granberg, 65. 42 “Ontario Regulation 350/06 Building Code,” 4.1.6.2.(1).
13 Krautheim, Pasel, Pfeiffer, and Schultz-Granberg, 136. 43 Leighton Cochran, Wind Issues in the Design of Buildings (Reston, Virginia:
14 Krautheim, Pasel, Pfeiffer, and Schultz-Granberg, 126. American Society of Civil Engineers, 2012), 82.
15 Krautheim, Pasel, Pfeiffer, and Schultz-Granberg, 65. 44 “Ontario Regulation 350/06 Building Code,” 4.1.6.2.(3)-(4).
16 Krautheim, Pasel, Pfeiffer, and Schultz-Granberg, 136. 45 Ibid.
17 Ibid. 46 Ibid.
18 Ibid. 47 “Ontario Regulation 350/06 Building Code,” 4.1.6.2.(5)-(7).
19 Ibid. 48 “Ontario Regulation 350/06 Building Code,” 4.1.6.2.(8).
20 Chris J. Baker, “Wind effects in the urban environment - considerations for 49 Cochran, 82.
human health, comfort and safety,” in Wind Effects on Buildings and Urban 50 “Snow loading and drifting,” Cermak Peterka Petersen, last modified
Environment, ed. Y. Tamura, (Proceedings of ISWE1 - The First International September 7, 2014, http://www.cppwind.com/blogs/snow-loading-drifting.
Symposium on Wind Effects on Buildings and Urban Environment, 2004), 97. 51 Cochran, 82.
21 Krautheim, Pasel, Pfeiffer, and Schultz-Granberg, 124. 52 Ibid.
22 Marq De Villiers, Windswept: The Story of Wind and Weather (New York: 53 “Svalbard Science Centre / JVA,” ArchDaily, last modified July 10, 2008,
Walker, 2006), 208. http://www.archdaily.com/3506/svalbard-science-centre-jva.
23 Krautheim, Pasel, Pfeiffer, and Schultz-Granberg, 56. 54 Ibid.
24 Baker, 96. 55 Ibid.
25 Krautheim, Pasel, Pfeiffer, and Schultz-Granberg, 65. 56 Ibid.
26 Krautheim, Pasel, Pfeiffer, and Schultz-Granberg, 64.
27 Krautheim, Pasel, Pfeiffer, and Schultz-Granberg, 126.
28 Krautheim, Pasel, Pfeiffer, and Schultz-Granberg, 138.
29 Ibid.
30 Ibid.
31 Krautheim, Pasel, Pfeiffer, and Schultz-Granberg, 140.
32 Ibid.
33 Krautheim, Pasel, Pfeiffer, and Schultz-Granberg, 138.
34 Krautheim, Pasel, Pfeiffer, and Schultz-Granberg, 141.

192
 BIBLIOGRAPHY

BIBLIOGRAPHY
Adair, Josh, Nels Estlund, Braden Etchison, Drew Gelderd, and Joe Yoder. Dabiri, John O. “Potential order-of-magnitude enhancement of wind farm power
“Compact Wind Acceleration Turbine.” Prezi. Last modified March 18, 2013. density via counter-rotating vertical-axis wind turbine arrays.” Journal of Renewable
https://prezi.com/ozzjc94bhsjx/compact-wind-acceleration-turbine/. and Sustainable Energy 3, 043104 (2011): 1-12.

Ambrose, James, and Dimitry Vergun. Simplified Building Design for Wind and de Maupassant, Guy. Sur l’eau. 1888. Quoted in De Villiers, Marq. Windswept:
Earthquake Forces. New York: John Wiley & Sons, Inc., 1990. The Story of Wind and Weather. New York: Walker, 2006, 16.

“Autodesk Flow Design.” Autodesk. Accessed January 28, 2015. http://www. De Villiers, Marq. Windswept: The Story of Wind and Weather. New York: Walker,
autodesk.com/store/flow-design?licenseType=cloudSub&support=basic&term=a 2006.
nnual.
“Displacement.” Encyclopædia Britannica. Accessed March 3, 2015. http://www.
“Autodesk Vasari.” Autodesk. Accessed January 28, 2015. http://autodeskvasari. britannica.com/science/displacement-mechanics.
com.
Dr. Elizabeth English, University of Waterloo, meetings, July 29, 2015 - September
Baker, Chris J. “Wind effects in the urban environment - considerations for human 21, 2015.
health, comfort and safety.” In Wind Effects on Buildings and Urban Environment,
edited by Y. Tamura, 89-101. Proceedings of ISWE1 - The First International “Effective Uses of Scan&Solve: Resolution and Convergence.” Intact Solutions.
Symposium on Wind Effects on Buildings and Urban Environment, 2004. Accessed January 28, 2015. http://www.intact-solutions.com/sns_documents/
sns_2014/resolution.htm.
Bennett, Jessica. “Wind Design Guide.” Architectural aerodynamics course
reading, Victoria University of Wellington, http://www.victoria.ac.nz/architecture/ “Features: Autodesk CFD software benefits.” Autodesk. Accessed January 28,
centres/cbpr/publications/architectural-aerodynamics/pdfs/BBSC_433_Jessica- 2015. http://www.autodesk.com/products/cfd/features/key-differentiators/
Bennett_Wind-Design-Guide.pdf. gallery-view.

Billington, David. “The New Art of Engineering.” In Bridging the Gap: Rethinking “Features: Flow simulation tools.” Autodesk. Accessed January 28, 2015. http://
the Relationship of Architect and Engineer, edited by Deborah Gans, 3-21. New www.autodesk.com/products/flow-design/features/all/list-view.
York: Van Nostrand Reinhold, 1991.
Feng, Guoying, Tim De Troyer, and Mark C. Runacres. “Optimizing the Land Use
Black, R. Gary, and Stephen Duff. “A Model for Teaching Structures: Finite Element for Wind Farms Using Vertical Axis Wind Turbines.” Conference poster, EWEA
Analysis in Architectural Education.” Journal of Architectural Education 48, no. 1 2014, Barcelona, Spain. http://proceedings.ewea.org/annual2014/conference/
(1994): 38-55. posters/PO_192_EWEApresentation2014.pdf.

Brown, Lance. “Wind and Snow Fences.” Factsheet, British Columbia Ministry of “Finite element analysis.” Autodesk. Accessed January 28, 2015. http://www.
Agriculture, Food and Fisheries, April 1997. http://www.agf.gov.bc.ca/resmgmt/ autodesk.com/solutions/finite-element-analysis.
publist/300Series/307230-1.pdf.
Gehloff, Maik, Dipl.-Ing. (FH), M.A.Sc. “Specifying Modern Timber
Cochran, Leighton. Wind Issues in the Design of Buildings. Reston, Virginia: Connections.” Construction Canada. Last modified May 21, 2014. http://www.
American Society of Civil Engineers, 2012. constructioncanada.net/specifying-modern-timber-connections/.

“Course 102 - Wind Knowledge.” A Wind of Change. Accessed November 12, Holmes, J.D., C.J. Baker, E.C. English, and E.C.C. Choi. “Wind structure and
2014. http://www.awindofchange.com/lessons/lesson1-2.html. codification.” Wind and Structures: An International Journal (2005).

193
BIBLIOGRAPHY

How Earth Made Us: Wind. Directed by Nigel Walk. 2010. BBC, 2010. Television. MacDonald, William L. The Architecture of the Roman Empire Volume 1: An
Introductory Study. United States of America: Yale University, 1965.
“How to fly a kite.” National Kite Month. Accessed November 12, 2014. http://
www.nationalkitemonth.org/how-to-fly-a-kite/. Matthew Spremulli, University of Waterloo, document revisions, September 18,
2015.
“How to Snowkite.” Snowkiting.com. Accessed November 12, 2014. http://www.
snowkiting.com/howto. McCleary, Peter. “The Differing Projects of Life of the Architect and the Engineer.”
In Bridging the Gap: Rethinking the Relationship of Architect and Engineer, edited
Hussein, Esam. “Dean’s Message.” University of Regina. Accessed January 13, by Deborah Gans, 3-21. New York: Van Nostrand Reinhold, 1991.
2015. http://www.uregina.ca/engineering/about-us/deans-message.html.
“Mechanics of Materials: Strain.” eFunda. Accessed March 3, 2015. http://www.
Hutton, David V. Fundamentals of Finite Element Analysis. New York: The McGraw- efunda.com/formulae/solid_mechanics/mat_mechanics/strain.cfm.
Hill Companies, Inc., 2004.
Melaragno, Michele. Wind In Architectural And Environmental Design. New York:
“Introducing Snowfer.” Snowfer. Accessed November 12, 2014. http://www. Van Nostrand Reinhold Company, 1982.
snowfer.com/intro.html.
“Ontario Regulation 350/06 Building Code.” Building code, Building and
Irwin, Peter A. “Vortices and tall buildings: A recipe for resonance.” Physics Today, Development Branch of the Ministry of Municipal Affairs and Housing, Ontario,
September 2010. http://www.math.lsa.umich.edu/~krasny/math654_irwin.pdf. May 31, 2006.

Jamie Ware, AEC Account Executive, IMAGINiT Technologies, e-mail message, “Overview: Computational fluid dynamics software.” Autodesk. Accessed January
January 30, 2015. 28, 2015. http://www.autodesk.com/products/cfd/overview.

Kim, Nam-Ho, and Bhavani V. Sankar. Introduction to Finite Element Analysis and “Overview: Finite element analysis and modeling software.” Autodesk. Accessed
Design. Hoboken: John Wiley & Sons, Inc., 2009. January 28, 2015. http://www.autodesk.com/products/simulation-mechanical/
overview.
Krautheim, Mareike, Ralf Pasel, Sven Pfeiffer, and Joachim Schultz-Granberg. City
and Wind: Climate as an Architectural Instrument. Berlin: DOM Publishers, 2014. Part 7: Running Wind Tunnel Simulations. Autodesk Building Solutions. September
6, 2011. Video. http://www.youtube.com/watch?v=PhuYa8aXYjU&index=31&list
Kurowski, Paul M. Finite Element Analysis Design for Engineers. Warrendale: SAE =PLF8ACB9BED56DE576.
International, 2004.
“Part One: Schism.” In Bridging the Gap: Rethinking the Relationship of Architect
Larsen, Olga Popovic. Conceptual Structural Design: Bridging the gap between and Engineer, edited by Deborah Gans, 1-2. New York: Van Nostrand Reinhold,
architects and engineers, Edited by Andy Tyas. London: Thomas Telford, 2003. 1991.
Laura Baumgärtner Dipl.-Ing., Sales and Consulting, SOFiSTiK, e-mail message, “Polar Winds.” Kids Earth Science. Accessed July 15, 2015. http://www.kids-
February 10, 2015. earth-science.com/polar-winds.html.
Li, Yang. “Computational Fluid Dynamics Technology and Its Application in Wind “Population of census metropolitan areas.” Statistics Canada. Last modified
Environment Analysis.” Journal of Urban Technology 17, no. 3 (2010): 67-81. February 11, 2015. http://www.statcan.gc.ca/tables-tableaux/sum-som/l01/
cst01/demo05a-eng.htm.
“Linear Static Analysis: Mathematical Model: Linear Elasticity.” Intact Solutions.
Accessed January 28, 2015. http://www.intact-solutions.com/sns_documents/ Salvadori, Mario. Why Buildings Stand Up. New York: W. W. Norton & Company,
sns_2014/mathematical.htm. Inc., 1980.

194
 BIBLIOGRAPHY

Salvadori, Mario, and Robert Heller. Structure in Architecture: The Building of The Flow Design Story. Autodesk. February 12, 2015. Video. http://knowledge.
Buildings. New Jersey: Prentice-Hall, 1986. autodesk.com/support/flow-design/learn-explore/caas/CloudHelp/cloudhelp/
ENU/FlowDesign/files/GUID-6D24A84A-A4ED-4B91-9720-46C6DF6CFC87-
“Scan&Solve Full License (Commercial).” Intact Solutions. Accessed January htm.html.
28, 2015. http://intact-solutions-inc.myshopify.com/collections/scan-solve-full-
license-commercial. Wesoff, Eric. “Optiwind’s Innovation in Medium Wind.” GreenTechMedia. Last
modified January 26, 2011. http://www.greentechmedia.com/articles/read/
“Scan-and-Solve for Rhino.” Intact Solutions. Accessed January 28, 2015. http:// optiwinds-medium-wind.
www.scan-and-solve.com.
“Wind & weather statistics Regina Airport.” Windfinder. Last modified July 2015.
“Shearwalls 101: Why You Can’t Have a Window There.” BuildLLC. Accessed http://www.windfinder.com/windstatistics/regina_airport.
September 3, 2015. http://blog.buildllc.com/2014/05/shearwalls-101-why-you-
cant-have-a-window-there/. “Wind Turbine Site Selection.” Solacity Inc. Accessed November 12, 2014.
http://www.solacity.com/siteselection.htm.
“Snow loading and drifting.” Cermak Peterka Petersen. Last modified September
7, 2014. http://www.cppwind.com/blogs/snow-loading-drifting. “Windbreaks.” Penn State College of Agricultural Sciences. Accessed November
12, 2014. http://extension.psu.edu/plants/plasticulture/production-details/
“Snowfer Frequently Asked Questions.” Snowfer. Accessed November 12, 2014. windbreaks.
http://www.snowfer.com/snowferfaq.html.
WindTronics. “BTPS 6500 Wind Turbine - Blade Tip Power System.” Product
“Snowkiting FAQs.” MACkite. Accessed November 12, 2014. http://www. guide, November 2010. http://www.windtronics.eu.com/wp-content/uploads/
mackiteboarding.com/snowkitefaq.htm#wind. International-Catalog.pdf.

“SOFiSTiK Rhinoceros Interface.” McNeel. Last modified September 20, 2011. Yares, Evan. “Meshless FEA: A new way to solve linear static FEA problems.”
http://blog.rhino3d.com/2011/09/sofistik-rhinoceros-interface.html. Design World. Last modified May 11, 2012. http://www.designworldonline.com/
meshless-fea-a-new-way-to-solve-linear-static-fea-problems/#_????.
“SOFiSTiK Rhinoceros Interface.” SOFiSTiK. Accessed February 4, 2015. http://
www.sofistik.com/en/solutions/structural-fea/rhinoceros-interface/. Ziad Boutanios, PEng., BinkZ, e-mail message, July 3, 2015.

Stathopoulos, Theodore. “Introduction to Environmental Aerodynamics.” In

Environmental Wind Engineering and Design of Wind Energy Structures, edited by
Charalambos C. Baniotopoulos, Claudio Borri and Theodore Stathopoulos, 3-30.
New York: SpringerWienNewYork, 2011.

“Stress.” Encyclopædia Britannica. Accessed March 3, 2015. http://www.

britannica.com/science/stress-physics.

Svitil, Kathy. “Wind-turbine Placement Produces Tenfold Power Increase, Caltech

Researchers Say.” Caltech. Last modified July 13, 2011. http://www.caltech.
edu/news/wind-turbine-placement-produces-tenfold-power-increase-caltech-
researchers-say-1700.

“Svalbard Science Centre / JVA.” ArchDaily. Last modified July 10, 2008. http://
www.archdaily.com/3506/svalbard-science-centre-jva.

195
GLOSSARY

GLOSSARY
AERODYNAMIC the quality of a form to have a shape that reduces drag caused by
surrounding wind
AERODYNAMICS the study of the interaction between solid forms and air when at
least one of them is in motion, not necessarily referring to solid
forms that are shaped to reduce drag caused by surrounding wind
ALONGWIND LOAD wind load acting in the direction parallel to the mean wind velocity
ASPECT RATIO ratio of height to width
COMPUTATIONAL FLUID DYNAMICS (CFD) SOFTWARE software that simulates the flow of fluids, including wind, around an
input model and produces numbers, graphics and animations to
convey this flow
CROSSWIND LOAD wind load acting in the direction perpendicular to the mean wind
velocity
DEFORMATION the action of a body changing shape as the points within the body
displace different amounts
DISPLACEMENT a measure of where and by how much a point of a body moves
while the body experiences stress
DYNAMIC WIND LOAD wind load that rapidly changes in magnitude or location
FINITE ELEMENT ANALYSIS (FEA) SOFTWARE software that evaluates the structural behaviour of an input model
under input loading conditions and produces numbers, graphics
and animations to convey this behaviour
FINITE ELEMENTS small pieces into which a digital model is divided to be analyzed by
FEA software
FUNDAMENTAL PERIOD the length of time required to complete one oscillation
LEEWARD (adj.) the side of an obstruction that is sheltered from the wind
(adv.) on the side of an obstruction that is sheltered from the wind

196
 GLOSSARY

MESH A set of finite elements made up of points, edges, and faces that
approximate the original model
MESHING the process of representing a physical entity with finite elements, by
breaking it down into smaller pieces to re-build it as a set of points,
edges, and faces that approximate the original model
QUALITATIVE STRUCTURAL UNDERSTANDING the ability to know how a structure behaves without referring to
measurements or calculations
QUASI-STATIC WIND LOAD static wind load with an increased magnitude to account for the
wind load’s dynamic nature
STATIC WIND LOAD wind load that maintains a consistent magnitude and location over
a period of time
STRAIN a quantity that describes deformation in a direction
STRESS a quantity that describes all the internal forces acting within a body
of material
STRUCTURAL ANALYSIS the process of calculating the types and magnitudes of stresses
and deformations in a structure subjected to loads
STRUCTURAL BEHAVIOUR the manner in which a structure acts or functions under loading
conditions
STRUCTURAL DESIGN the process of determining the form of a structure that will allow it
to withstand subjected loads
STRUCTURAL INTUITION the ability to immediately understand how an object or material will
act under load, without necessarily knowing why
TORSIONAL LOAD wind load that induces twisting about the vertical axis
UPLIFT wind force acting upwards in the direction perpendicular to the
mean wind velocity
WINDWARD (adj.) the side of an obstruction that is facing the wind
(adv.) on the side of an obstruction that is facing the wind

197
APPENDIX A | PAPER FOR ICWE14

APPENDIX A | PAPER FOR ICWE14

This appendix is a paper that was co-wirtten by the author of this
thesis for publication and presentation at the 14th International
Conference on Wind Engineering, which took place from June 21-
26, 2015 in Porto Alegre, Brazil. The paper is a summary of the
content of this thesis.

198
 APPENDIX A | PAPER FOR ICWE14

1 2

Due to current software limitations, this design method may not necessarily produce accurate wind and structural data. Instead,
this method will increase in accuracy as CFD and FEA programs continue to be improved. In the future, CFD and FEA software
programmers could potentially draw from this method to create programs that can be used together, to allow architects to
consider wind during the early design stages within a streamlined workflow. This design method does not intend for the
architect to replace the structural or wind engineer. Instead, it equips the architect with the necessary knowledge and tools to
design a building that considers wind effects and building aerodynamics in its initial form generation, which may be refined by
the wind engineer in a later design phase. The method also allows the architect to design a building that accommodates a
feasible structural system that the structural engineer can then easily adjust and detail. This approach allows the architect and
engineers to work towards shared goals, thereby streamlining co-ordination between them. It also eliminates costly and time-
CFD and FEA Software in Architectural Design Methods consuming design revisions that can occur when wind effects and loading are only considered in the later stages of the design
process.
Stephanie Fleming1, Elizabeth English2
1
School of Architecture, Faculty of Engineering, University of Waterloo, Waterloo, Ontario, Canada 2 SOFTWARE
2
School of Architecture, Faculty of Engineering, University of Waterloo, Waterloo, Ontario, Canada 2.1 CFD and FEA Software in Architectural Applications
email: skflemin@uwaterloo.ca, english@ecenglish.ca
In this design method, computational fluid dynamics (CFD) software is used to simulate and represent the flow of wind around
buildings, and provide information about the amount of wind pressure that acts on the building faces. Visual, rather than
numerical, output is used, as it is easier for architects who are not trained in wind engineering to understand and interpret results
ABSTRACT: There is a reciprocal relationship between wind and buildings, as they each affect the other. Building form affects
that visually represent the wind, rather than quantify it.
wind by altering its speed and flow patterns, and can be used to create desirable wind conditions for a set of exterior programs.
Finite element analysis (FEA) software is used to simulate how a digital model will react to applied forces. When loads,
Wind, in turn, exerts load on the building, which can be reduced with aerodynamic forms and resisted with appropriate
restraints, and material properties are applied within the software to the digital model, the software breaks down the digital
structural systems. This establishes a relationship between wind conditions, the building form that influences these wind
model into small pieces called "finite elements" and analyzes the behaviour of each of these elements under the input loading,
conditions, and the structure that stabilizes the form against these wind conditions. This relationship is investigated through the
restraints, and material properties. Splitting the model into small, simply-shaped elements allows them to be analyzed with
development of a design method that allows architects to consider, in the early architectural design stages, how wind and
simple equations, rather than using complex equations on the entire digital model. The assembly of the behaviour of all of the
buildings affect each other. The method consists of a pairing of computational fluid dynamics (CFD) software and finite element
finite elements conveys the global structural behaviour of the entire digital model. In this design method, FEA software is used
analysis (FEA) software. For each building design iteration, CFD software is used to simulate the speeds and patterns of wind
to simulate and represent the effects of combined wind and gravity loading on the building's form and structural system.
flow around the initial building form design, as well as evaluate the aerodynamics of the building form by providing information
Graphics and animations of structural behaviour are used instead of numerical data, as they are more easily understood by
about the wind pressure that is exerted on each building face. The speed with which these results are provided allows the
architects who are not trained in structural engineering. [2, 3]
architect to refine and re-test many iterations of their design until the building form creates the desired wind conditions. Then,
Most CFD and FEA software programs are intended for use in engineering, rather than architectural, applications. This design
the wind pressure information that is provided by the CFD software is input into the FEA software to predict how the building
method selects software programs that can be adapted for use in architectural applications, and uses them in a way that is
will react to combined wind and gravity loading. This information informs the schematic design of the building’s structural
appropriate for the initial architectural design stages.
system, which is developed through another iterative process using the FEA software. This method allows architects to consider
wind as a generator of architectural form within a streamlined, software-based workflow. 2.2 Software Selection
To ensure that this design method is accessible to architects, many types of CFD and FEA software programs were researched
KEY WORDS: Computational Fluid Dynamics; CFD; Finite Element Analysis; FEA; Architectural Design; Design Method; to determine which are most appropriate for architectural applications. In order for a CFD or FEA software to be considered for
Form Generation; Aerodynamics; Structure. use in this method, it has to meet the following required criteria:
1. 3D-modeling software compatibility: The software must either be a plug-in for, or run files from, 3D-modeling programs
1 INTRODUCTION that are commonly used by architects or geared towards architectural, rather than engineering, applications. Rhinoceros
In antiquity, the architect was the “master builder” who understood and was able to execute both building design and and Revit are two such programs, so the chosen CFD and FEA programs must be compatible with at least one of these
construction. However, when the Industrial Revolution spurred the rapid creation of many new building materials and two programs. Rhinoceros can be used to create initial building massing models on which wind studies may be performed.
technologies, it became difficult for a single person to master them all. This fostered the creation of the structural engineering The building can then be modeled in Revit to continue to develop the project and its wind studies in more detail through
profession. Structural engineers became experts in building construction technologies, while architects specialized in the spatial the later project phases. Executing the design method with these two programs eliminates the need for architects to learn
and aesthetic design of buildings. This has resulted in a divergence between the roles and priorities of the architect and the new 3D-modeling software in addition to the CFD and FEA software, making the design method more accessible to
structural engineer. [1] This divergence is commonly seen within current architectural practice, as structural analysis is not often architects.
integrated into early architectural design phases. Instead, it is performed by an engineer after the initial building design has 2. Open-source software: The software must be free for students, to encourage its use in a studio setting. This would allow
already been developed by the architect, at a point when it is difficult, costly, and time-consuming to change the design to students to learn these programs while in school, so that they would be proficient in them when seeking jobs after
accommodate the structure. This divergence between structural and architectural processes is widely acknowledged in the fields graduation. This criteria would make it more feasible for architectural firms to adopt these software programs into their
of both architecture and engineering, and members of both professions have worked on developing methods of integrating the current working methods, since it would be easier to find employees who know how to use them. It is advantageous if the
two design processes. This paper explores one such method of integration, through the creation of a design method that program is also free for commercial use; however, since architectural firms have greater financial resources than students,
architects can use to consider wind loading and effects as design informants that are integrated into the initial architectural and since very few programs are free for commercial use, it is not a requirement.
design phases. After the elimination of software that did not meet the above criteria, the remaining programs were evaluated based on the
The design method that is described in this paper integrates both structural and wind engineering considerations into early following factors:
architectural design stages by pairing CFD software and FEA software. Software can be a valuable design tool for architects 1. Ease: It is advantageous if the software is easy to learn, as architects will be more inclined to learn it. This also allows
because it allows them to integrate engineering considerations into their current architectural practices, without having to them to implement the design method sooner, instead of spending more time learning a complex software.
acquire extensive engineering knowledge. It provides visualizations of wind patterns and structural behaviour, which are easier 2. Speed: It is advantageous if the software is able to quickly provide the user with results. This includes considerations of
than numerical data for architects to understand and interpret, since they are trained in visual media. Software also provides how fast a model can be set up for evaluation, any geometry clean-up that the software might require, and how long it
architects with a means of quickly testing multiple design iterations in relation to these engineering considerations, because the takes the software to process a result. The faster the program can work, the more design iterations can be tested and
software can perform engineering calculations and simulations much faster than if the architect were to learn and perform these refined.
calculations themselves. Repeated use of the software allows the user to gradually develop an intuitive understanding that allows 3. Accuracy: It is advantageous if the software provides accurate analysis results. However, this factor is the least valued, as
them to predict what the results of the software analysis will likely be, which will eventually influence their design decisions for the purposes of initial design development, it is more advantageous to use a program that is easy to learn and that
even before the software analysis is run [2]. quickly tests multiple design iterations, rather than use one that is complex but provides completely accurate data. For

14th International Conference on Wind Engineering – Porto Alegre, Brazil – June 21-26, 2015 14th International Conference on Wind Engineering – Porto Alegre, Brazil – June 21-26, 2015

199
APPENDIX A | PAPER FOR ICWE14

3 4

example, it is better for this method to employ a CFD program that is fast but less accurate, since no CFD software is as 2.5 Scan&Solve
accurate as a physical wind tunnel. The chosen CFD programs therefore simulate general patterns of wind flow, but are Scan&Solve is a plug-in for Rhinoceros that allows the designer to apply materials, restraints, and loads to the Rhinoceros
not relied upon for quantitative results. It is also preferable to select an FEA software that is easy to use, rather than one model, and then evaluates the model’s reaction to the simulated forces [6]. The program works with native Rhinoceros
that provides accurate results but is too complex to be feasibly integrated into the initial design stages. The purpose of this geometry, and unlike many other FEA programs, it does not require a separate meshed model in order to perform the analysis
paper is to develop a methodology that will become more accurate as CFD and FEA software is improved, so accuracy is [6]. The analysis can be quite accurate, depending on the grid resolution that is set by the user [7]. Colour gradients representing
not as important at this stage of the method’s development. the displacement of the model under load (Figure 5), as well as deflection animations of the model under load (Figure 6), are the
4. 3D-modeling function: It is advantageous if the software includes 3D-modeling functions, because once the CFD or FEA most useful output for this design method.
results are obtained, the design can be adjusted accordingly within the same program. This eliminates the need to switch
programs to make the changes to the digital model, and then re-export the model to run the evaluation. This speeds up the
design method.
5. Price: It is advantageous if the software is inexpensive, as it is more feasible that architectural firms would implement a
design method that uses inexpensive software into their working practices.
Based on the evaluation of these five factors in relation to the considered software programs, Autodesk Vasari and Autodesk
Flow Design are the selected CFD programs to be used in the design method, and Scan&Solve is the selected FEA program.
2.3 Autodesk Vasari
Autodesk Vasari was designed to be used for initial massing, environment, and energy studies of architectural projects [4]. The
environmental tools include a wind tunnel simulator that allows users to visualize the air flow around digital 3D models. It
simulates only major qualitative wind trends, and is intended to provide insight into wind patterns at early stages of building
massing. Although the speed with which results can be obtained is useful during early project stages, the designer should be
aware that the CFD analysis is not always especially accurate. However, this does not negate the usefulness of these results to
depict general wind trends. Vasari can run files from Rhinoceros and Revit, and also includes a 3D-modeling function within the Figure 5. Colour gradient representing displacement from Figure 6. Deflection animation from Scan&Solve.
program [4]. In this design method, Vasari's horizontal data slices depicting wind speed (Figure 1) are used to evaluate the Scan&Solve.
surrounding wind conditions that are created by each building form iteration.
3 DESIGN METHOD CONDITIONS
For this paper, the design of a small, low-rise building is used as a means of developing the proposed design method that may
be applied to many projects with varying building, program and site requirements. The paper does not serve to propose a
specific building. However, to define the conditions within which this method is developed, a program and site have been
chosen.
3.1 Program
The exterior program activities serve to provide a variety of wind condition requirements to be created with the building form,
and are not a program proposal in themselves. These programs consist of pairings of wind energy generation technologies and
seasonal sports that require specific wind conditions (Figure 7). To support the exterior programs, the building itself could
accommodate energy storage, space to record and compare energy generation data from the wind energy generation
Figure 1. Horizontal data slice from Vasari, depicting wind speed at 2m above the ground. technologies to be tested on the site, as well as equipment storage and change room facilities to support the sports programs. The
building form does not reflect interior building design, program sizes or program placement requirements, but is instead a form
that creates the wind conditions that are required for the exterior programs.
2.4 Autodesk Flow Design
Autodesk Flow Design simulates a wind tunnel to allow users to visualize the air flow around digital 3D models [5]. The
simplicity of the program's set-up and operation makes it ideal for designers who don’t have the time or the need to learn a
comprehensive CFD program. It simulates only major qualitative wind trends, and is intended to provide insight into wind
patterns at early stages of building massing. Like Vasari, the speed with which results can be obtained is useful during early
project stages, but the designer should be aware that the CFD analysis is not always accurate. The program, however, is still
useful in depicting general wind trends. Flow Design can run files from Rhinoceros, or can be used as a plug-in for Revit [5].
While Flow Design's horizontal data slices are not as detailed as those from Vasari, its vertical data slices depicting wind speed
(Figure 2), flow line animations (Figure 3), and colour gradients representing the wind pressure acting on the model surface
(Figure 4) are used in this method to visualize wind speed, turbulence, and pressure on and around the building.

Figure 7. Wind conditions for exterior programs.

Figure 2. Vertical data Figure 3. Flow line Figure 4. Colour gradient

slice depicting wind animation from Flow representing wind pressure on
speed from Flow Design. Design. model surface from Flow Design.

14th International Conference on Wind Engineering – Porto Alegre, Brazil – June 21-26, 2015 14th International Conference on Wind Engineering – Porto Alegre, Brazil – June 21-26, 2015

200
 APPENDIX A | PAPER FOR ICWE14

5 6

A set of drawings has been made that conveys the wind and spatial requirements for each exterior program, such as 4.2 Wind Effects Library
snowkiting, in which someone on skis or a snowboard is pulled across the snow by a large kite (Figure 8). These drawings may Manipulations of building form to increase and decrease wind speed, turbulence, and pressure were studied from a broad range
be referred to by the architect when developing building forms to create these required exterior spaces and wind conditions. of published sources [10, 11, 12, 13, 14] and compiled by the first author into a library of wind effects. For each technique, a
wind effect that has been studied in real wind environments is compared with the results of the simulation of the effect in both
Vasari and Flow Design, such as the drawing and simulations of the Venturi effect (Figure 10) [10]. While the wind effects
library is too extensive to include in this paper, it is a part of the first author’s Master of Architecture thesis. Architects may refer
to this library of effects to alter the building geometry in order to create the specific wind conditions that are required for the
exterior programs. The software simulations of each of the effects allow the architect to become familiar with what each of the
CFD programs can accurately represent, and what they don't consider when computing results. For example, neither Vasari nor
Flow Design is able to depict the downwash effect, as the programs assume a uniform wind speed and do not consider the higher
wind speeds that exist in reality at higher elevations. It was also learned through these simulations that Vasari’s horizontal data
slices are more detailed than those from Flow Design, but Flow Design’s vertical data slices, flow line animations, and colour
gradients representing the wind pressure acting over the model surface are more accurate than those from Vasari. While CFD
software can be a useful tool for quick tests of design iterations, the architect has to know how to interpret the results and be able
to tell if they are an accurate representation of the simulated effects, by possessing some knowledge of studied wind effects.

Figure 8. Wind and spatial requirements for snowkiting.

3.2 Site
The site on which the method will be developed is an open field on the outskirts of southern Regina, in Saskatchewan, Canada.
Regina is one of the windiest cities in Canada, with an average annual wind speed of 20 km/h [8]. These high wind loads
provide an extreme context within which to develop the design method, and the flat, open site provides an opportunity to Figure 10. Studied effect, Vasari simulation, and Flow Design simulation of the Venturi effect.
manipulate the wind with only the building form, as there are minimal site conditions to alter the wind immediately around the
building. The wind on the site comes predominantly from the south-east, but can also come from the north-west (Figure 9) [8].
The site is located near the University of Regina, whose existing wind turbine studies [9] could potentially pair with the wind 4.3 CFD Process
energy generation technologies that surround the building.
For the first step in the process, a building form is modeled and tested with CFD software to visualize the wind speeds and
patterns that are created around the building by the building form, and evaluate the appropriateness of these wind conditions for
the exterior programs that are intended to be accommodated. This can be represented by Vasari's horizontal data slices of wind
speed that depict general wind patterns. After the first building iteration is tested with the CFD software, the architect makes
observations and adjusts the form for the next iteration to improve the appropriateness of the surrounding wind conditions for
the exterior programs. These adjustments to the form may be made by referring to the wind effects library that was mentioned in
section 4.2, which catalogues ways of manipulating form to increase or decrease the surrounding wind speed and turbulence.
Iterations may also be tested in Flow Design to look at the flow lines, which depict the wind turbulence that is generated by the
building. Although this adds an extra step to the process, it is beneficial to ensure that there is no undesired turbulence created
by the building form. As CFD programs are improved in the future, the accuracy of this process will increase.
During these iterations, the CFD software is also used to evaluate the aerodynamics of the building form by depicting how
much wind pressure is exerted over the model's surface. The amount of wind pressure that acts across the surface is represented
by colour gradients in Flow Design. After each iteration is tested within the simulated wind conditions, the designer makes
observations and adjusts the form for the next iteration to improve the form’s aerodynamics. These adjustments to the form may
Figure 9. Annual wind rose for site in Regina. be designed by referring to the wind effects library that was mentioned in section 4.2 and included in the first author’s Master of
Architecture thesis. These observations and adjustments are done concurrently with the adjustments to improve the surrounding
wind conditions for the exterior programs.
4 CFD PROCESS To test this method, a building form was developed through nine iterations, improving the surrounding wind conditions or
4.1 Process Overview aerodynamics of the form with each iteration (Figure 11). The first few iterations were used to increase the width and length of
the channel of increased wind speed between two buildings, and a later iteration refined the building geometry so that the
For each building design iteration, CFD software is first used to simulate the speeds and patterns of wind flow around the required wind conditions would be created when the wind blew from either of the two predominant wind directions. The ninth
building form. This tests the appropriateness of the surrounding wind conditions that have been created by the building form for iteration accommodates all of the exterior programs in a wide variety of wind conditions (Figures 12, 13, 14). After this
the exterior programs that must be accommodated. The speed with which these results are provided allows the architect to refine iteration, the CFD process was stopped and the resulting building form was then used to develop the FEA process.
and re-test many iterations of the design until the building form creates the desired wind conditions. The CFD software also
evaluates the aerodynamics of the building form, as the software provides the architect with information about the wind pressure
that is exerted on each building face.

14th International Conference on Wind Engineering – Porto Alegre, Brazil – June 21-26, 2015 14th International Conference on Wind Engineering – Porto Alegre, Brazil – June 21-26, 2015

201
APPENDIX A | PAPER FOR ICWE14

7 8

Figure 13. Wind speed data slice of ninth iteration at 2m Figure 14. Wind flow lines around ninth iteration at 2m
above the ground. above the ground.

5 FEA PROCESS
5.1 Process Overview
There are many structural systems that can be used to resist wind loads [15]. However, for the purposes of this paper, a steel-
frame system has been chosen and the FEA process has been developed specifically for the planning of this type of structural
system. The simplicity of the system provides an appropriate base with which to develop a methodology that could potentially
be used for other structural systems. The modular nature of the system also allows the architect to design small units of structure
that are repeated throughout the building. This creation of a single unit may be applied to buildings of various sizes, as the units
may be repeated throughout the building as many times as necessary.
The FEA process is used to develop a digital model of a single structural bay. The wind pressure that is provided by the CFD
Figure 11. Iterations to improve wind conditions and aerodynamics of form. software, combined with gravity loading, is input into the FEA software to predict how the bay will respond to this combined
loading. Based on the deflection animation and the colour gradient that represents the displacement of each point of the model,
the architect can see where and how much the bay will deflect in the wind and under gravity. The architect can then stiffen the
bay against this deflection, and run the finite element analysis on multiple iterations of bays, with the goal of increasing the
stiffness of the assembly each time.
5.2 Pressure Zones
The first step in the FEA process is to divide the building mass that has been created in the CFD process into different pressure
zones, based on the colour gradients representing wind pressure on the model’s surface that are obtained from the CFD software
(Figures 15, 16).

Figure 15. Colour gradient representing wind pressure Figure 16. Colour gradient representing wind pressure on
on the model’s windward surface. model’s leeward surface.

The building is divided into six pressure zones (Figure 17) based on the areas that are subject to the same amount of wind
pressure, as depicted by the colour gradients from the CFD software (Figures 15, 16). Then, the average positive and negative
wind pressure values for each zone are obtained from the CFD software for each of the site's two predominant wind directions.
For the purposes of this method, each zone is assumed to be subjected to the highest combined wind pressure and suction, out of
the values obtained from both wind directions. These values are shown in the pressure zone matrix (Figure 18), although their
accuracy depends on the level of accuracy that may currently be obtained from the CFD software.
Figure 12. Ninth iteration with surrounding exterior programs.

14th International Conference on Wind Engineering – Porto Alegre, Brazil – June 21-26, 2015 14th International Conference on Wind Engineering – Porto Alegre, Brazil – June 21-26, 2015

202
 APPENDIX A | PAPER FOR ICWE14

9 10

deflection under wind loading. This may be accomplished by reducing column spacing, increasing member sizes, or adding
bracing. This process is repeated until a bay is developed that has appropriate column spacing, member sizes, and bracing to
resist large movements in the wind. Four iterations are shown to develop a structural bay for zone 1A (Figure 22), and the
process is repeated for zone 1B, as well as zones 2 through 6 to develop a structural bay for each pressure zone within the zone's
specified wind pressure conditions. Reducing the scope of the structural design to a single structural bay allows the architect to
avoid the time-consuming task of modeling and testing the entire building's structure, while still being able to understand the
column spacing, member sizes, and bracing that will be required throughout the building.

Figure 22. Iterations to develop a structural bay for zone 1A, showing amount of displacement of each point under load.

This method does not intend for the architect to replace the structural engineer. In a later design phase, the structural engineer
would perform a more thorough structural design and analysis and would likely make adjustments to the structural bay that is
developed with this method. The importance of this method is that it provides the architect with a sense of the approximate
Figure 17. Building divided into six pressure zones. structural spacing, as well as an understanding that some form of bracing will be required. The architect can account for this
while designing the building interior, so that no design decisions compromise the approximate structure that must be
accommodated. This method provides a way of integrating structural considerations into the initial building design, which
improves collaboration between the architect and the structural engineer throughout many design phases.

5.5 Structural System

After a structural bay has been developed for each zone and sub-zone, the bays are inserted within the massing to create the
building's structural system. This schematic structural drawing (Figure 23) provides the architect with a comprehensive visual of
the required structural density throughout the building. The architect can use this visual to design the building’s interior and the
exterior facade. If larger member spacing would be beneficial in certain areas to create larger open spaces, the architect can test
new bays with the FEA software, using the same iterative process to increase the column spacing, while also increasing member
Figure 18. Pressure zone matrix with highest pressure and suction for each zone. sizing or adding bracing to compensate for the longer spans. Once such a bay has been developed so that its deflection in the
wind is acceptably low, these larger bays can be inserted where they are needed within the building. The architect can repeat this
process as necessary until the structural and architectural designs work together, and neither compromises the other.
5.3 FEA of Building Massing
The wind pressures and gravity loading are then input into the FEA software and applied to the massing model of each
pressure zone, as shown for zone 1 (Figure 19). The colour gradient that represents the displacement of each point of the model,
as well as the deflection animation (Figure 20), reveal areas within the zone that may deflect more while under wind load than
other areas within the zone. In the case of zone 1, the form deflects the most at the right side, where the massing is the thinnest
(Figure 20). As such, zone 1 has been divided into two sub-zones: 1A and 1B (Figure 21). These sub-zones are subjected to the
same amount of wind pressure, but require different structural configurations due to the difference in geometry thickness.

Figure 19. Wind pressures and Figure 20. Deflection animation of Figure 21. Zone 1 sub-zones.
gravity loading applied to massing pressure zone 1.
of pressure zone 1.

5.4 Structural Bays

A structural bay is then modeled for zone 1A, and tested with the FEA software by inputting gravity loading and the wind
pressure for zone 1. The deflection of the bay in the wind is observed with the displacement colour gradient and deflection
animation, and if there is too much deflection, the architect adjusts the model of the bay to increase its stiffness and reduce its Figure 23. Drawing of the building's structural system.

14th International Conference on Wind Engineering – Porto Alegre, Brazil – June 21-26, 2015 14th International Conference on Wind Engineering – Porto Alegre, Brazil – June 21-26, 2015

203
APPENDIX A | PAPER FOR ICWE14

11

5.6 CFD Analysis of Modified Form

The form of the building can then be adjusted to reflect the inclusion of the steel structure (Figure 24). For example, curved
sections could be faceted, with each side length equal to the length of the column spacing. This modified form is then re-tested
with the CFD software to ensure that the change to the form did not undesirably alter the wind conditions around the building
(Figure 25). If the wind conditions are found to have changed in certain areas, this would affect the wind pressure that is applied
to that area of the building. A new pressure zone would be created and a new structural bay for that zone would be developed
with the same iterative FEA process. If the wind conditions are altered so much as to become undesirable, the form itself would
need to be adjusted so that it fosters the intended wind conditions, yet is also an appropriate shape to be able to accommodate a
steel-frame structure. This step completes the feedback loop between the CFD and FEA programs, and at this point, the design
method may be repeated as many times as necessary to develop a building that both shapes and resists the wind.

Figure 24. Adjusted building form to reflect the inclusion of a Figure 25. CFD software test of adjusted form.
steel structure.

6 CONCLUSION
A design method in which architects use wind effects and loading as a design generator may be carried out with a pairing of
computational fluid dynamics software and finite element analysis software. Using CFD and FEA software in a way that is
appropriate for architectural applications allows architects to integrate wind and structural engineering considerations into the
early design stages of their current working practices. As CFD and FEA programs continue to be improved, the accuracy and
usefulness of this method will increase. It is the hope of the authors that software programmers may draw from this method to
create CFD and FEA programs that are conducive to architectural applications, so that in the near future, architects may use this
design methodology to integrate wind, structural and architectural design processes.

REFERENCES
[1] D. Gans, ed., Bridging the Gap: Rethinking the Relationship of Architect and Engineer, Van Nostrand Reinhold, New York, USA, first edition, 1991.
[2] R. G. Black and S. Duff, “A Model for Teaching Structures: Finite Element Analysis in Architectural Education,” Journal of Architectural Education,
48(1), 38-55, 1994.
[3] P. M. Kurowski, Finite Element Analysis Design for Engineers, SAE International, Warrendale, USA, first edition, 2004.
[4] “Autodesk Vasari,” Autodesk, Accessed January 28, 2015, http://autodeskvasari.com.
[5] “Flow Design Features: Flow Simulation Tools,” Autodesk, Accessed January 28, 2015,
http://www.autodesk.com/products/flow-design/features/all/gallery-view.
[6] M. Freytag, “Scan-and-Solve for Rhino,” Intact Solutions, Accessed October 2, 2014, http://www.scan-and-solve.com.
[7] “Effective Uses of Scan&Solve: Resolution and Convergence,” Intact Solutions, Accessed January 28, 2015,
http://www.intact-solutions.com/sns_documents/sns_2014/resolution.htm.
[8] “Wind & Weather Statistics Regina Airport,” Windfinder, Accessed January 28, 2015, http://www.windfinder.com/windstatistics/regina_airport.
[9] E. Hussein, “Dean's Message,” University of Regina, Accessed January 13, 2015, http://www.uregina.ca/engineering/about-us/deans-message.html.
[10] J. Bennet, Wind Design Guide, Architectural aerodynamics course reading, University of Wellington, 2007,
http://www.victoria.ac.nz/architecture/centres/cbpr/publications/architectural-aerodynamics/pdfs/BBSC_433_Jessica-Bennett_Wind-Design-Guide.pdf.
[11] L. Cochran, ed., Wind Issues in the Design of Buildings, American Society of Civil Engineers, Reston, USA, first edition, 2012.
[12] M. Krautheim, R. Pasel, S. Pfeiffer and J. Schultz-Granberg, City and Wind: Climate as an Architectural Instrument, DOM Publishers, Berlin, Germany,
first edition, 2014.
[13] T. Stathopoulos, “Introduction to Environmental Aerodynamics,” in Environmental Wind Engineering and Design of Wind Energy Structures, ed. C.C.
Baniotopoulos, C. Borri and T. Stathopoulos, 3-30, SpringerWienNewYork, New York, USA, first edition, 2011.
[14] E.C. English and F.R. Fricke, “The interference index and its prediction using a neural network analysis of wind-tunnel data,” Journal of Wind
Engineering and Industrial Aerodynamics, 83(1999), 567-575, 1999.
[15] M. Salvadori and R. Heller, Structure in Architecture: The Building of Buildings, Prentice-Hall, Upper Saddle River, USA, third edition, 1986.

14th International Conference on Wind Engineering – Porto Alegre, Brazil – June 21-26, 2015

204