PARAMETRIC MODELING OF TIMBER LIGHT-FRAME SHEAR

WALLS USING OPENSEES: PRELIMINARY RESULTS

Giorgia Di GANGI
Research Assistant
Department of Structural and Geotechnical Engineering, Sapienza University of Rome, Rome, Italy
giorgia.digangi@uniroma1.it*

Cristoforo DEMARTINO
Assistant Professor
Zhejiang University / University of Illinois at Urbana Champaign Institute (ZJUI), Haining, China
cristoforo.demartino@me.com

Giuseppe QUARANTA
Assistant Professor
Department of Structural and Geotechnical Engineering, Sapienza University of Rome, Rome, Italy
giuseppe.quaranta@uniroma1.it

Giorgio MONTI
Full Professor
Department of Structural and Geotechnical Engineering, Sapienza University of Rome, Rome, Italy
giorgio.monti@uniroma1.it

Abstract
This contribution illustrates the main features of an original parametric Finite Element (FE)
model developed using OpenSEES in order to study the response of timber light-frame shear
walls under earthquake. Within the proposed FE model, the framing system and the sheathing
panel are modeled as elastic elements while the connections are simulated through zero-length
non-linear elements. In particular, it is assumed that the overall nonlinear seismic response of
the wall rests on the cyclic behavior of the connections. The presence of openings has been
taken into account, as well as the use of more than one panel to brace the wood frame in case
of walls with low aspect ratios. An exemplification numerical analysis is proposed at the end.
Keywords: Energy dissipation, Finite element model, Openings, Sheathing-to-framing
connections, Timber light-frame shear wall.

1. Introduction
Timber light-frame shear walls are structural members typically employed within medium- or
high-rise platform framing buildings in order to withstand in-plane lateral actions, such as wind
or seismic forces. In particular, they are widespread in Northern Europe, North America and New
Zealand in 90% of residential buildings with one or two storeys. The typical layout of these walls
(considering the structural part only) consists of an assembly of vertical studs and horizontal
joists, which are connected at their ends with internal constraints. The latter, in turn, are typically
assumed as hinges (Figure 1). Sheathing panels are connected by means of metal fasteners (e.g.,
nails, screws or staples) in order to sheathe one or both sides of the main frame. The dimensions
depend on the size of the sheet, which can be realized using different materials, such as Oriented
Strand Board (OSB), plywood, gypsum, fibreboard. Commonly, as pointed out in Ref. [1], the
size of a shear wall is 1.22 m × 2.44 m or 2.44 m × 2.44 m, whereas the framing elements (studs
and joists) cross-sections are about 38 mm × 89 mm and 38 mm × 140 mm for internal and
external wall studs, respectively.

208
 Figure 1. Typical structural configuration of a fully anchored
timber light-frame shear wall sheathed on both sides.
The cross-section size of external framing elements is often chosen to accommodate minimum
building requirements for thermal insulation (see Figure 1).
Generally, common fasteners (6D, 8D and 10D) with thick shank that gives greater strength,
are employed for framing connections. The fasteners are placed both on the perimeter studs
(typically with spacing equal to 50, 75, 100 mm) and on the intermediate studs. In the latter
case, they are two or three times spaced with respect to the fasteners placed on the perimeter
studs, only to prevent buckling of the sheathing panel [2]. Experimental tests have demonstrated
that timber has, in general, a poor dissipative capacity whereas the steel connections – such as
sheathing-to-framing joints [3][4][5], base [6] and stud-joist joints [4] – ensure a good amount
of energy dissipation and cyclic ductility notwithstanding their significant pinching, strength
degradation and softening. These pieces of evidence are well reflected in the numerical models
available in the literature, where the non-linear wall response is basically related to the load-
deformation relationships of the connections [7][8]. In general, elastic beams are implemented
in order to represent the behavior of the framing elements whereas sheathing panels are
modeled with plane-stress elements, assuming an elastic behavior in compression and an
elastic-brittle behavior in tension [9]. Sheathing-to-framing connections and base connections
are usually modeled with non-linear springs [10]. Because of the lack of extensive parametric
analyses of timber light-frame shear walls [11][12], an original parametric FE model has been
developed within OpenSEES [13] in order to assess the influence that geometric variables have
on the global seismic performance of the wall. The main features of this parametric FE model
are illustrated in the following.

2. Model implementation within OpenSEES

The original FE model developed using the open-source software OpenSEES [14] herein
presented is, to the best authors’ knowledge, the first attempt of modeling a timber light-frame
shear wall within the TCL environment. The software OpenSEES was chosen for its capability
to obtain in a simple environment a parametric model of a timber light-frame shear wall and
because it’s opensource software freely available to the community.

209
 Figure 2. Overall process to build the parametric FE model of the wall without openings.

2.1 FE model without openings

A common way to simulate efficiently the dissipative behavior of timber light-frame shear walls
is to model framing system and sheathing panel as elastic elements while representing the
connections as zero-length non-linear elements. This approach has been adopted previously in
different studies [7][8] and is implemented herein. The developed parametric model requires
the following geometric parameters (see Figure 2): panel size (panel height H and panel width
L), horizontal and vertical nails spacing (sph and spv), number of vertical studs (nstuds) and
framing elements cross-section size for internal studs (bis and his) and boundary elements (i.e.,
bes and hes for external studs, bp and hp for horizontal joists).
The nodes are created on a regular grid according to the nails spacing sph and spv. They are
labeled as follows: the first digit indicates the layer they belong to, whereas the two next couples
of digits indicates the ith (in the x direction) and jth (in the z direction) position of the node on
the grid, respectively ( Figure 3). Two layers are employed for the nodes. Layer 2 includes the
fictitious nodes used to insert the internal releases between the end of vertical studs and the
horizontal joists. Layer 4 includes the perimeter nodes belonging to the frame. An example of
the FE model generated for 5 nails along the horizontal direction and 3 nails along the vertical
direction is illustrated in Figure 3, where base and height are aligned with the x axis and the z
axis, respectively. Once the nodes are placed, the elements are generated. The framing elements
have been modeled using Elastic Beam Column Elements whereas the sheathing panel is
modeled by means of ShellMITC4 Elements whose mesh size is adjusted automatically based
on the nails spacing. In order to consider the framing joints acting as hinges [10], two
ZeroLength Elements are inserted at both ends of the vertical studs, with zero stiffness for the
rotation along the y direction and infinite stiffness for the remaining degrees-of-freedom.
Finally, CoupledZeroLength Elements have been employed in order to represent the non-linear
behavior of the sheathing-to framing connections. In this way, the overestimation of stiffness
and force of the nail is avoided because, under non-linear loading, the yielding of the element
occurs on a circular surface [15]. All these elements are labeled by considering the layer they
belong to (first digit), together with first and second nodes of the element itself along x and z
directions (subsequent couples of digits) as illustrated in Figure 3.

210
 Figure 3. Layout of the shear wall FE model implemented in OpenSEES
(in this example, it includes 5 and 3 nails along the horizontal and vertical direction, respectively).
Layers 2, 3, 5, 6 and 7 refer to vertical studs, horizontal joists, hinges within the frame, panel
(together with its nodes) and sheathing-to-framing connections, respectively. Finally, since the
frame is double sheathed (Figure 1), the symmetric shell and zero-length elements are included
in Layer 8 and Layer 9, respectively.
The SAWS mechanical model – originally proposed in Ref. [16], then developed in Ref. [17]
and modified in Ref. [18] – is employed to represent the mechanical behavior of a single nail
using the CoupledZeroLength Element. The 10 parameters that define the SAWS model have
been identified considering the experimental data of a single ring nail Φ 2.8 provided in Ref.
[9] by means of the non-classical identification techniques presented in Ref. [19]. For more
details about the FE model and its validation, the interested reader can refer to Ref. [20].

2.2 FE model with openings

A parametric FE model of the shear wall able to take into account the presence of openings has
been also implemented (see Figure 4). By setting the dimensions of the openings (bdim and
hdim) and their distance from one edge (dist), the position of the vertical studs around them is
automatically updated. Nodes and shell elements at the opening are then deleted by means of
the command remove. Moreover, the header of the opening is inserted with internal releases for
the connection with the intermediate studs, in order to represent the real behavior of the joints
connecting framing members (that behave like perfect hinges). To validate the accuracy of the
FE model for a wall with openings, experimental data provided in Ref. [21] for specimen 2
have been considered.

211
 Figure 4. Overall process to build the parametric FE model of the wall with openings.

Figure 5. Comparison between experimental and predicted force-displacement curves for the reference
wall configuration (specimen 2 in Ref. [21]).
As it is shown in Figure 5, the predicted force-displacement curve (black line) is in good
agreement with the experimental data (red line) and the numerical results (blue line) provided
in Ref. [21]. In the adopted model, the shell elements have size equal to 50 mm × 50 mm
whereas nails spacing – according to Ref. [21] – is 100 mm and 300 mm on the external edges
and on the horizontal elements of the central section, respectively. It is remarked that different
(finer) mesh sizes have been also used, and no significant variations have been observed.
Actually, nails Ф 3.1 /80 (mm) have been used to connect the OSB plates and timber framing
elements. Without an experimental load-displacement curve for this nail size, the performances

212
related to a single fastener have been obtained by multiplying times 1.08 the ordinates of the Ф
2.8 ring nail force vs. displacement curve, following the procedure in Ref. [9]. Specifically, the
amplification factor has been estimated as mean value between the stiffness ratio and the
resistance ratio of the two nail types, by exploiting the simplified analytical relationship
proposed by EuroCode 5 [22] to predict the stiffness for timber-to-timber connections,
considering nails without pre-drilling:
1.5m  d 0.8 K ser , 3.1   3.1 0.8
K ser =  =  (1)
30 K ser , 2.8   2.8 
where  is the nail diameter. The difference between the experimental and the numerical curve
can be explained because of the amplification factor (i.e., Eq. 1) is not fully capable to catch
the mechanical response of the nail.

3. Analysis of walls with low aspect ratio

As it was pointed out in Ref. [11], the response of non- or partially-anchored timber shear walls
strongly depends on the aspect ratio (i.e., height-to-width ratio). The relative rigid rotation of
the sheathing panel with respect to the frame mostly stresses the nails near the corners (Figure
6). As it can be inferred from the experimental tests presented in Ref. [23], a flexural behavior
of the timber shear wall can be observed if the width is significantly less than the height (aspect
ratio equal to or larger than 3). Otherwise, the contribution of shear deformation to storey
displacements increases, with a growth of stiffness and racking capacity. This is due to two
factors, namely: i) a higher number of vertical studs is required (and thus the whole system is
stiffer); ii) the wider the base of the wall, the higher the number of horizontal nails. Due to the
available size of sheathing panels, a timber light-frame shear wall with low aspect ratio could
be comprised of more than one panel to brace each side of the timber frame. In order to quantify
the difference in terms of wall response, the parametric FE model has been developed to take
into account this condition (Figure 6). The results show that the racking capacity for a single-
and multi-panel wall is fairly constant whereas the decrement of the panels inertia for in-plane
actions led to a reduction of the global secant stiffness.

Figure 6. Global force-displacement curve of a wall with aspect ratio equal to 0.7 (right) considering two
adjacent panels for both sides of the wall.
The displacement field of the nails is magnified by a scale factor equal to 100 (left); black arrow: fastener

213
 displacement smaller than the one corresponding to peak force; blue arrow: fastener displacement
exceeding the peak force; red arrow: fastener displacement at failure.

4. Conclusions
An original parametric FE model for timber light-frame shear walls implemented within
OpenSEES has been described in this work. First, the numerical model of double-sheathed
timber light-frame shear wall without openings has been presented. Then, a parametric FE
model taking into account the presence of openings has been illustrated and, finally, further
considerations about walls with low aspect ratio have been provided. Preliminary results and
validations have demonstrated the reliability of the model, which can be employed to perform
extensive sensitivity analyses and to support the search for optimal configurations.

5. References
[1] WANG, J., DEMARTINO, C., XIAO, Y., LI, Y., “Thermal insulation performance of
bamboo- and wood-based shear walls in light-frame buildings”, Energy and Buildings,
Vol. 168, June 2018, pp. 167-179.
[2] KÄLLSNER, B. AND GIRHAMMAR, U.A., “Analysis of fully anchored light-frame
timber shear walls-Elastic model”, Materials and Structures, Vol. 42, No. 3, April 2009,
pp. 301-320.
[3] SARTORI, T. AND TOMASI, R., “Experimental investigation on sheathing-to-framing
connections in wood shear walls”, Engineering Structures, Vol. 56, November 2013, pp.
2197-2205.
[4] HUMBERT, J., BOUDAUD, C., BAROTH, J., HAMEURY S., DAUDEVILLE, L.,
“Joints and wood shear walls modelling I: constitutive law, experimental tests and FE
model under quasi-static loading”, Engineering Structures, Vol. 65, April 2014, pp. 52-
61.
[5] GERMANO, F., METELLI, G., GIURIANI, E., “Experimental results on the role of
sheathing-to-frame and base connections of a European timber framed shear wall”,
Construction and Building Materials, Vol. 80, April 2015, pp. 315-328.
[6] GAVRIC I, FRAGIACOMO M, CECCOTTI A., “Cyclic behaviour of typical metal
connectors for cross-laminated (CLT) structures”, Materials and structures, Vol. 48, No.
6, March 2014, pp.1841-1857.
[7] TUOMI, R.L. AND MCCUTCHEON, W.J., “Racking strength of light-frame nailed
walls”, Journal of the Structural Division, Vol. 104, No. 7, 1978, pp. 1131-1140.
[8] GUPTA, A.K. AND KUO, G.P., “Behavior of wood-framed shear walls”, Journal of
Structural Engineering, Vol. 111, No. 8, August 1985, pp. 1722-1733.
[9] GATTESCO, N. AND BOEM, I., “Stress distribution among sheathing-to-frame nails of
timber shear walls related to different base connections: experimental tests and numerical
modelling”, Construction and Building Materials, Vol. 122, September 2016, pp. 149-
162.
[10] YASUMURA, M., “Racking resistance of panel-sheathed shear walls with openings”,
World Conference on Timber Engineering, WCTE, Riva del Garda, Trentino, Italy, 2010.
[11] SALENIKOVICH, A., DOLAN, J. D., “The racking performance of light-frame shear
walls with various tie-down restraints”, 6th World Conference on Timber Engineering,
July 2000.
[12] DHONJU, R., D’AMICO, B., KERMANI, A., PORTEOUS, J., ZHANG, B., “Parametric
evaluation of racking performance of platform timber framed walls”, Structures, Vol. 12,
November 2017, pp. 75-87.
[13] DI GANGI, G., “Structural Analysis and Design of Timber light-frame shear walls”, PhD
thesis, Sapienza University of Rome, February 2019.

214
[14] MCKENNA, F. AND FENVES, G., “The OpenSEES Command Language Manual:
version 1.2”, Pacific Earthquake Engineering Center, Univ. of Calif., Berkeley, USA,
2001.
[15] CASAGRANDE, D. AND ROSSI, S. AND TOMASI, R. AND MISCHI, G., “A
predictive analytical model for the elasto-plastic behaviour of a light timber-frame shear-
wall”, Construction and Building Materials, Vol. 102, January 2016, pp. 1113-1126.
[16] FOSCHI, R.O., “Load-slip characteristics of nails”, Wood Sci, 1974, pp. 69-76.
[17] FOLZ, B. AND FILIATRAULT, A., “Cyclic analysis of wood shear walls”, Journal of
Structural Engineering, Vol. 127, No. 4, April 2001, pp. 433-441.
[18] DOLAN, J.D., “The dynamic response of timber shear walls”, PhD thesis, University of
British Columbia, October 1989.
[19] QUARANTA, G., MONTI, G., MARANO, G. C., “Parameters identification of Van der
Pol-Duffing oscillators via particle swarm optimization and differential evolution”,
Mechanical Systems and Signal Processing, Vol. 24, No. 7, October 2010, pp. 2076-
2095.
[20] DI GANGI, G., DEMARTINO, C., QUARANTA, G., VAILATI, M., MONTI, G.,
“Timber shear walls: numerical assessment of the equivalent viscous damping”, 9th
International Conference on Computational Methods, ICCM2018, August 2018.
[21] TOGAY, A. AND ANIL, Ö. AND KARAGÖZ IŞLEYEN, Ü. AND EDIZ, İ. AND
DURUCAN, C., “Finite-element analyses of light timber-framed walls with and without
openings”, Institution of Civil Engineers-Structures and Buildings, Vol. 170, No. 8,
August 2017, pp. 555-569.
[22] EN, “1-1: 2004 Eurocode 5: Design of timber structures – General – Common rules and
rules for buildings”, 1995.
[23] ANIL, Ö., TOGAY, A., IŞLEYEN, Ü. K., DÖNGEL, N., SÖĞÜTLÜ, C., “Effect of
timber type and nail spacing on the hysteretic behavior of timber-framed shear walls with
openings”, International Journal of Civil Engineering, Vol. 16, No. 6, February 2017,
pp. 629-646.

215