Downloaded from ascelibrary.org by University of Waterloo on 11/03/15. Copyright ASCE. For personal use only; all rights reserved.

In-Situ Evaluation of Structures Using Load Testing

Nestore Galati, PhD1 and Tarek Alkhrdaji, PhD, PE2

ABSTRACT
In-situ load testing is commonly used to assess the safety or the serviceability of
an existing or new structural system or part of its elements for a particular external
load condition. The most frequently used load testing procedure includes the in-situ
load test method described in Chapter 20 of the ACI 318 - Building Code
Requirements for Structural Concrete and the Cyclic Load Test method in ACI 437
and ACI 437.1R. This paper describes the load testing programs used for two field
projects to evaluate the load capacity of existing and upgraded structural components.
The main objective of this paper is to describe the process for determining the test
load level, loading procedure, instrumentation, evaluation criteria and load test
outcomes. The first case study is for the evaluation of the sidewalks for a historical
steel bridge in which the supporting beams lost up to 70% of the original cross-
section due to extensive corrosion. The load test was performed in order to assess
whether the sidewalks were able to safely support the pedestrian load. The second
case study is for the evaluation of a typical bay in an old reinforced concrete (RC)
pan-joist floor system. The floor was being considered to support new
telecommunication equipment that required an increase in the capacity of the floor
system. Load tests were performed for typical existing members and members
upgraded using externally bonded FRP reinforcement as well as bonded RC overlay.

1
Design Engineer, Structural Group, Inc., Strengthening Division, 7455 New Ridge Rd., Suite T, Hanover, MD,
21113, ngalati@structural.net

2
Engineering Manager, Structural Group, Inc., Strengthening Division, 7455 New Ridge Rd., Suite T, Hanover,
MD, 21113, talkhrdaji@structural.net

657

Copyright ASCE 2009 Forensic Engineering Congress 2009

Forensic Engineering 2009
 658 FORENSIC ENGINEERING 2009

INTRODUCTION
Load testing of concrete structures in the United States is a century old tradition
Downloaded from ascelibrary.org by University of Waterloo on 11/03/15. Copyright ASCE. For personal use only; all rights reserved.

with some applications carried out as early as in the 1890s [Birkmire, 1894]. In the
early days, in-situ load testing was the most direct proof of performance of, at the
time, proprietary and novel construction materials and methods. The American
Concrete Institute (ACI) began formalizing load test procedures for concrete
structures in 1920 [ACI, 1920]. Subsequent codes [ACI, 1936] defined the deflection
evaluation criterion as a function of the span and the total depth of the member. This
formed the deflection criterion that is still in effect to date [ACI, 2005]. Notable
investigations into load testing of concrete structures documenting the practice of the
last few decades can be found in a variety of literature [FitzSimons and Longinow,
1975; RILEM, 1984; and Bungey, 1989].
Today, in-situ load testing is commonly used to assess the load carrying capacity
of structural members as affected by design or construction errors, verify new
methods of repairs or strengthening, verify if the existing members can carry loads
different that those used for the original design, and demonstrate the safety of a
structure that has experienced damage due to corrosion and degradation. There are
currently two methods used for in-situ load testing of concrete structures: The 24-
hour load test procedure, as prescribed in Chapter 20 of the ACI 318 Building Code
[ACI 318, 2005], and the Cyclic Load Test (CLT) method, as prescribed in ACI 437
Strength Evaluation of Existing Buildings and ACI 437.1R Load Tests of Concrete
Structures (ACI 437R, 2003; ACI 437.1R, 2007). Each of these documents is
published by the American Concrete Institute (ACI).
The ACI 318 load test method and its evaluation criteria are widely referred to as
the 24-hour load test because the test load is held in place on the structure for a period
of 24 hours. Because residual displacements must be measured 24 hours after the test
load has been removed, the actual test duration (not including setup) takes at least 48-
hours to complete.
The CLT method includes testing the structure with repeated loading/unloading
cycles with a load magnitude that increases as the test progresses, rather than a
constant load applied for a predetermined length of time. This provides improved
information regarding the behavior of the structural system.
Both the 24-hour and the CLT methods allow the use of dead weight (water
bladders, sand bags, steel plates, etc.) or hydraulic jacks to apply load to the structure.
However, the use of dead weights may not be safe since the load cannot be removed
quickly, and the use of hydraulic jacks is typically considered to be safer since the
applied load can be quickly removed at the first signs of overstress or failure.
Analytical models are typically used to design the load test and to evaluate the
effects of structural characteristics such as load sharing with adjoining or adjacent
members, effects of the magnitude and location concentrated loads, effect of using
concentrated loads versus original distributed loads, and the effects of support fixity
on load distribution. In some instances, the analytical models are refined after the test
is completed by adjusting the boundary conditions and stiffness of the adjacent

Copyright ASCE 2009 Forensic Engineering Congress 2009

Forensic Engineering 2009
 FORENSIC ENGINEERING 2009 659

elements to match the actual measured deformations of the structure. Once the model
has been calibrated, the analytical model may also be used to accurately determine
Downloaded from ascelibrary.org by University of Waterloo on 11/03/15. Copyright ASCE. For personal use only; all rights reserved.

the internal forces caused by external loads.

Determination of the test load level and loading procedure are not entirely
straightforward and are discussed next using two field case studies. A general
description of the loading levels, cycles, and the number of steps are given in each
case study.

CASE STUDY 1: LOAD TESTING OF HISTORICAL STEEL BRIDGE

The bridge under investigation was constructed at the beginning of the century
and has experienced extensive corrosion of the steel members, as showed in Figure 1.
Due to the age of the bridge, no engineering drawings and construction records were
available. The original sidewalks on both sides of the bridge were 10 ft wide.
Concrete-filled steel buckle plates, which were riveted to supporting steel stringers
and floor beams, supported the sidewalk decks and the full area of the roadway decks
(see Figure 2).
The buckle plates are supported by three longitudinal lines of stringers under the
sidewalks consisting of riveted, built-up, steel members. The stringer lines 1 to 3 are
referred as “fascia stringers,” “sidewalk struts,”and “curb stringers,” respectively.
The transverse floor beams under the sidewalks (brackets) cantilevers
approximately 4 ft (1219 mm) out from each vertical post, which are spaced every 7
ft - 3 in (2210 mm) on center along the length of each span.
In the 1960s, the eastbound roadway was widened by decreasing the width of the
upstream sidewalk. This was accomplished by reducing the length of the anchor
spans of the sidewalk brackets by 4 ft (1219 mm) and moving the curb stringers
closer to the sidewalk struts. The roadway buckle plates adjacent to both sidewalks
were removed and the roadway decks were reconstructed as reinforced concrete
slabs. The buckle plates, concrete sidewalk slabs, and all the other original structural
members outboard of the sidewalk struts were retained at both sidewalks.
Due to the extensive corrosion and complexity of the structure, the development
of numerical models would not be reliable unless calibrated with field measurements.
For this reason, the consulting engineer on this project requested a cyclic load test to
be performed on the structure in order to asses the existing capacity and to collect
experimental data that can be used to calibrate a 3-D FEM model of the structure.
A total of six (6) tests were performed on the two sidewalks, three tests on each
side of the bridge. The first test area was located at the downstream side of the bridge
(downstream load tests). This area included two sidewalk bays with a total area of
about 14 ft - 6 in (4420 mm) by 9 ft (2743 mm). The second test area was located at
the upstream side of the bridge (upstream load tests). This area also included two
sidewalk bays with a total area of about 14 ft - 6 in (4420 mm) by 4 ft (1219 mm).
The evaluation comprised of three load tests for each test area.

Copyright ASCE 2009 Forensic Engineering Congress 2009

Forensic Engineering 2009
 660 FORENSIC ENGINEERING 2009

The first load test involved testing the sidewalk slabs for negative bending
moment by simultaneously loading two adjacent spans. The second test involved
Downloaded from ascelibrary.org by University of Waterloo on 11/03/15. Copyright ASCE. For personal use only; all rights reserved.

testing the sidewalk slabs for positive bending moment by loading only one span.
The third test involved testing of the steel brackets by applying the load to the
sidewalk directly over the bracket.
The test loads, which were to be applied as concentrated forces, needed to create
internal forces in the structure that should be at least equal to those produced by the
applicable load combination per ACI 318-05. The design live load is 100 psf (4.8
kPa) uniform live, which is uniformly distributed over the entire sidewalk area, as
prescribed by the State Building Code where the bridge is located. The load test
magnitude for the bracket at the upstream side was performed with a load intensity of
1.3D + 1.6L, corresponding to the full design-load prescribed by the state where the
bridge is located.
Equivalent Point Loads
The test loads were determined by considering the sidewalk as a continuous beam
spanning above the steel brackets for the upstream load-test area, and between
brackets and their anchor-span extension for the downstream load-test area. After
determining the flexural and shear demands, the analysis was repeated with the
introduction of concentrated loads that produced similar flexural and shear demands.
The analysis considered loading configurations with pairs of point loads, spaced at 3
ft longitudinally. The load configurations for the downstream load tests are
illustrated in Figure 3.
Load Test Setup
The loads were generated by 30-ton-capacity hydraulic jacks, connected in
parallel, which reacted against a system of steel counterweights placed at the ground
level. The load was transferred through ½” (12mm) post-tensioning cable, passing
through 1 in. (25mm) diameter holes drilled in the slab, and locked against the
counterweight systems and to the hydraulic jacks (See Figure 4)
During the load tests, loads, displacements and strains were continuously
monitored in real-time by load cells, direct-current-variable-transformer (DCVT)
transducers, and strain gauges (SG) connected to a data-acquisition system. The data-
acquisition system recorded data at a rate of 5 Hz from all devices, and displayed in
real time the load vs. deflection curves of critical locations on a computer screen. The
DCVT transducers measured vertical displacements at 14 locations. Strain gauges
measured strains at 6 critical locations on the steel buckle plates and steel bracket.
The use of hydraulic jacks, and continuous monitoring of the structure during the load
tests, ensured that the loads could be rapidly removed in case signs of distress
developed, preventing permanent damage to the structure.
Load Test Procedure and Results
The structure was tested following the cyclic load testing protocol of ACI 437
(ACI 437.1R-07). For each load test, repeatability, permanency and deviation from
linearity were evaluated, even though they were not developed for steel structures.
The most important criterion for a steel structure would be the deviation from

Copyright ASCE 2009 Forensic Engineering Congress 2009

Forensic Engineering 2009
 FORENSIC ENGINEERING 2009 661

linearity because failing of this parameter would be a clear indication that steel
yielding has occurred.
Downloaded from ascelibrary.org by University of Waterloo on 11/03/15. Copyright ASCE. For personal use only; all rights reserved.

No evident signs of distress were detected in the structure during the load tests.
The load-deflection envelopes indicated that the test sidewalks and brackets had a
linear behavior, passing all the ACI-437 acceptance criteria.
Discussion of Results
The load tests indicated that there was a substantial longitudinal stiffness in the
sidewalks, which was attributed to the interaction among adjacent deck slabs (buckle
plates and concrete fill), brackets, and railing. Load-sharing mechanisms were
evident based the following observations:
1. At the downstream side, the recorded deflections in the center bracket are
similar to the two adjacent brackets for both the Negative-Moment and
Positive-Moment tests, as shown by the deflected shapes of the sidewalks
along Load Lines 1 (See Figure 5).
2. The mid-span deflections generated by the Negative-Moment tests are larger
than the deflections caused by the Positive-Moment tests. This is an indication
of load sharing between adjacent spans.
3. The recorded deflections at the free-end of the brackets are only slightly
greater than the deflections measured at the mid-region of the bracket (see
figure 6).
4. A more pronounced slope between the end and middle points, and consequent
larger deflection, would be expected if the bracket end would be entirely
disengaged from the sidewalk deck.
The load sharing mechanism previously described allowed the structure to sustain
the entire design load and to meet the acceptance criteria of ACI 437 even though the
structure was substantially corroded.

CASE STUDY 2: LOAD TEST EVALUATION OF UPGRADE FLOOR

SYSTEM
The following case study discuses the strengthening and cyclic load test that
was performed on the second-level structural floor of a building located in downtown
Cleveland, Ohio. The nine-story building was constructed in 1917. Originally, the
second floor of the building was occupied by a department store. With the high
percentage of internet traffic between New York and Chicago running through fiber
optic lines, downtown Cleveland is becoming a prime location for the telecom
industry. To meet the demand of this industry, the owner of the building decided to
upgrade the floor on the second level to house telecommunications equipment.
Telecom hotels are essentially warehouse spaces for internet related
telecommunications and other high-tech equipment that usually requires structural
floor capacity in the range of 125 psf (6.0 kPa) to 175 psf (8.4 kPa).

Copyright ASCE 2009 Forensic Engineering Congress 2009

Forensic Engineering 2009
 662 FORENSIC ENGINEERING 2009

Building Description
The building was constructed with masonry over a concrete incased steel
Downloaded from ascelibrary.org by University of Waterloo on 11/03/15. Copyright ASCE. For personal use only; all rights reserved.

frame and a reinforced concrete floor system. The ceiling height is approximately 14
ft (4270 mm), and the column spacing varies from 19 to 23 ft (5800 to 7000 mm).
Due to the age of the building, very limited construction and maintenance records
were available. The existing engineering drawings provided only floor plans and
geometry of the members. No information regarding material properties and
reinforcement details were available. All dimensions and section geometries were
field-verified.
The floor system consists of reinforced concrete joists supporting a concrete slab
monolithically cast with the joists. The concrete slab is 3 ½ in. (89 mm) thick and
reinforced with #3 (#10) steel bars spaced at 18 in. (457 mm) on centers. The joists
are 27.6 ft (8410 mm) long, 6 in. (153 mm) wide, and 15.5 in. (394 mm) deep. Field
investigation of the joists revealed that they are typically reinforced with two 1-in (25
mm) square bars at mid-span, one of which is bent up at approximately 5.5 ft (1676
mm) from each end and extend above the support. There was no information on
reinforcement details at the supports. The field investigation also revealed that an
additional straight bar is provided over the support at the location of the bent up bars.
No information of material characteristics was available for the joists. Due to time
constrains, a nominal concrete strength of 3,500 psi (24 MPa) and steel yield strength
of 33,000 psi (227 MPa) were assumed for the preliminary analysis of the joists.
These values were based on the typical material strength recoded for structures with
similar age. The assumed values were later verified through testing.
Localized destructive investigation of the joists revealed that the joists do not
have any steel shear reinforcement. Preliminary analysis of the existing floor system
indicated that the existing beams and slab are capable of carrying the proposed loads.
The joists, on the other hand, were deficient and their capacity was limited to
approximately 96 psf (4.6 kPa), governed by the shear strength of the joists. To house
telecommunication equipment, the floor needed to be upgraded to carry its own
weight, a super imposed dead load of approximately 25 psf (1.2 kPa), and a service
live load of 150 psf (7.2 kPa) acting over the entire surface of the floor. The super-
imposed dead load is a result of the concrete overlay required to level the existing
concrete slab.
Strengthening and Load Testing
The cyclic load tests were designed to locally verify the performance of some
typical joists that appeared to be the “weakest link.” To this end, the joists were
loaded near their ultimate strength and their response was measured in terms of
deformation and crack width. The cyclic load testing procedure involved applying
concentrated loads to the test joists at predetermined locations to simulate the effect
of maximum flexural forces at mid span and maximum flexural and shear forces at
the supports of the test joists.
Analytical modeling of the joists indicated that the maximum moments and shear
forces could be produced using two-point loads spaced 6 ft (1830 mm) apart. The
load was applied using hydraulic jacks that pulled against a reinforced concrete

Copyright ASCE 2009 Forensic Engineering Congress 2009

Forensic Engineering 2009
 FORENSIC ENGINEERING 2009 663

micro-pile that was installed on the ground floor below. A high strength steel bar was
used to connect the jacks to transfer the load to the micro-pile. Linear variable
Downloaded from ascelibrary.org by University of Waterloo on 11/03/15. Copyright ASCE. For personal use only; all rights reserved.

differential transformers (LVDTs) were used to measure joists deflections at five

locations along their lengths (see Figure 7). A load cell was used to measure the
applied load. Measurements were collected using a data acquisition system that
allowed for real-time monitoring of the applied load and the behavior of the test
joists. During the test, deflections and crack width were monitored for stability.
Two load tests were performed on the joists. Test 1 was performed on two joists
that were isolated by saw-cutting the concrete slab along a line between the joists.
Prior to testing, the joists were strengthened for shear using carbon FRP
strengthening systems applied in the form of U-strips 12 in. (305 mm) wide and
spaced at 18 in. (457 mm) on center. The load test was terminated when the mid-
span deflection became unstable and inelastic behavior was observed. Large residual
deflections were measured when the load was removed.
Results of the first test indicated that failure of the joist was governed by yielding
of reinforcement at the support. No failure signs were observed at joists mid-span.
Based on the test results, the joist was rated for a super-imposed dead load of 25 psf
(1.2 kPa) plus a live load of 135 psf (6.5 kPa). The shear performance was adequate
with no shear cracks or failure signs observed.
Bonded reinforced concrete overlay was used to address the observed joist
deficiency. The overlay consisted of 3 in. (76 mm) thick concrete reinforced with
steel wire mesh. To ensure adequate bond between existing and new concrete, the
slab surface was prepared by aggressive abrasion blasting to remove all weak
concrete and provide an open-pore structure. Pull-off tests were performed on the
overlay to verify that failure will not occur at the interface.
After the overlay was applied, Test 2 was performed on the same two joists to
verify the overall behavior of the strengthened system. The test was performed
following the same procedure outlined earlier. However, the joists were only loaded
to 85% of the ultimate as recommended by ACI 318 to prevent excessive damage to
the upgraded joists. During testing, a number of flexural cracks developed on top of
the slab (through the overlay) at both ends of the joists (at the negative moment
region) (see Figure 8). The number and distribution of the cracks indicated that
sufficient bond existed between the existing and new concrete. This behavior
confirmed that a composite behavior was achieved. The load test results
demonstrated the improvement stiffness and deflection behavior of the joists after
strengthening (see Figure 9). Based on the test results, the strengthened joists were
rated for the self-weight plus 36 psf (1.7 kPa) super-imposed dead load (RC overlay)
and 150 psf (7.2 kPa) live load.

CONCLUSIONS
This paper presented the process of load testing using the cyclic load testing
(CLT) method that were used to perform in-situ evaluation for two projects: a
historical bridge sidewalk and a reinforced concrete pan-joist floor system in a

Copyright ASCE 2009 Forensic Engineering Congress 2009

Forensic Engineering 2009
 664 FORENSIC ENGINEERING 2009

building. For both structures, the load test was conducted without disrupting the
normal operation of the structures.
Downloaded from ascelibrary.org by University of Waterloo on 11/03/15. Copyright ASCE. For personal use only; all rights reserved.

Load testing is a method of strength evaluation that is more representative of the

performance of the strengthened structural member than analytical approaches and
can provide valuable information regarding the health and performance of a
strengthened structural element.

REFERENCES
American Concrete Institute (1920), "Standard Specification No. 23 – Standard
Building Regulations for the Use of Reinforced Concrete".
American Concrete Institute (1936), "Building Code Regulations for Reinforced
Concrete," ACI 501-36T.
ACI Committee 437 (2003), “Strength Evaluation of Existing Concrete Buildings
(ACI 437R-03)”, American Concrete Institute, Farmington Hills, Michigan,
28 pp.
ACI Committee 437 (2007), “Test Load Magnitude, Protocol and Acceptance
Criteria”, (ACI 437.1R-07), American Concrete Institute, Farmington Hills,
MI, 2006, 83 pp.
American Concrete Institute (2005), “Building Code Requirements for Structural
Concrete”, ACI 318-05, American Concrete Institute, Farmington Hills,
Michigan, pp. 443.
Birkmire, W. (1894) “Skeleton Construction in Buildings,” John Wiley & Sons, New
York, 80pp.
Bungey, J., (1989), The Testing of Concrete in Structures, Second Edition, Chapman
and Hall, New York, New York, 228 pp.
FitzSimons, N., and Longinow, A., (1975), “Guidance for Load Tests of Buildings,”
Journal of the Structural Division, American Society of Civil Engineers; pp.
1367-1380.
RILEM Technical Committee 20-TBS, (1984), “General Recommendation for
Statical Loading Test of Load-Bearing Concrete Structures In Situ (TBS2)”,
RILEM Technical Recommendations for the Testing and Use of Construction
Materials, E & FN Spon, London, England, pp. 379-385.

Copyright ASCE 2009 Forensic Engineering Congress 2009

Forensic Engineering 2009
 FORENSIC ENGINEERING 2009 665
Downloaded from ascelibrary.org by University of Waterloo on 11/03/15. Copyright ASCE. For personal use only; all rights reserved.

Cantilever
Sidewalk Bracket

Stringer Line 1
Back-span
Bracket Lattice Member
at Stringer Line 2

Figure 1 – Extensive corrosion at the downstream side of the bridge

Figure 2 - Bridge cross-section

Copyright ASCE 2009 Forensic Engineering Congress 2009

Forensic Engineering 2009
 666 FORENSIC ENGINEERING 2009

CL CL

CL CL

5’-6” (1675 mm)+/-

Downloaded from ascelibrary.org by University of Waterloo on 11/03/15. Copyright ASCE. For personal use only; all rights reserved.

5’-6” (1675 mm)+/-

3’-0”
(915 mm) CL
CL Load Line 2
3 kips 3 kips 3 kips 3 kips
(13 kN) (13 kN) (13 kN) (13 kN)

(1219 mm) +/-

(1219 mm) +/-

4’-0” +/-
4’-0” +/-

CL 5.5 kip (25 kN)

CL Load Line 1
2.25 kips 2.25 kips 2.25 kips 2.25 kips
(10 kN) (10 kN) (10 kN) (10 kN)

7’-3” (2210 mm) +/- - Point load application 7’-3” (2210 mm) +/-
- Point load application
(Typ.) (Typ.)

Figure 3 - Load Configuration for Downstream Load Tests

PT Chuck Plus Wedges

PL ½”x6”x6” Load Cell
Hydraulic Jack
4”x4” Cribbing 4 x 6 x ¼ Steel Tube

Bridge Sidewalk
DCVTs
Strain Gages

½” PT Cable

Weights

4”x4” Cribbing PL ½” x 28” x 48”

(4 Sides)
PL ½”x10”x10” PT Chuck Plus Wedges

Figure 4 - Schematic of the Load Test

EAST SPAN WEST SPAN EAST SPAN WEST SPAN

0 0 0 0
Cycles 1-2
Cycles 3-4
0.02 Cycles 7-8 0.6 0.02 0.6
Deflection (in)

Deflection (mm)
Deflection (in)

Deflection (mm)

0.04 1.2 0.04 1.2

0.06 1.8 0.06 1.8

Cycles 1-2
0.08 Cycles 3-4 2.4
0.08 2.4
Cycles 7-8

0.1 3
0.1 3

a) Negative Moment b) Positive Moment

Figure 5 - Downstream Test: Deflection along Line 1

Copyright ASCE 2009 Forensic Engineering Congress 2009

Forensic Engineering 2009
 FORENSIC ENGINEERING 2009 667

(2) (2)

SOUTH SOUTH
Downloaded from ascelibrary.org by University of Waterloo on 11/03/15. Copyright ASCE. For personal use only; all rights reserved.

0 0 0 0

0.02 0.6 0.02 0.6

Deflection (in)

Deflection (mm)

Deflection (in)

Deflection (mm)
0.04 1.2 0.04 1.2

0.06 1.8
0.06 1.8

0.08 Cycles 1-2 2.4 Cycles 1-2

Cycles 3-4
0.08 Cycles 3-4 2.4
Cycles 5-6 Cycles 5-6

0.1 3
0.1 3

a) Negative Moment b) Positive Moment

Figure 6 - Downstream Test: Deflection along Line A

LVDT LVDT
LVDT Measuring LVDT LVDT
Deflection
Jack

Reaction Beam

Figure 7 - Diagnostic Load Test Setup

0 0

-0.1
-5
-0.2
Deflections [mm]]

-0.3
Deflections [in]

-10

-0.4

-15
-0.5

-0.6
-20
-0.7 Before Strengthening
After Strengthening
-0.8 -25
0 25 50 75 100
Fraction of Span Length [%]

Note: P = 19.6 kip (87 kN) before strengthening, P = 14.7 kip (65 kN) after strengthening

Figure 8 - Cracks in the negative Figure 9 - Effect of strengthening on

moment region joist profile (Test 1 & 2)

Copyright ASCE 2009 Forensic Engineering Congress 2009

Forensic Engineering 2009