12th Canadian Masonry Symposium

Vancouver, British Columbia, June 2-5, 2013

ENGINEERING MODELS FOR SHEAR WALLS MADE OF MASONRY

S. Ortlepp1 and W. Jäger2
1
Research Assistant, Chair of Structural Design, Faculty of Architecture, Technische Universität Dresden, 01062,
Dresden, Germany, [email protected]
2
Professor, Chair of Structural Design, Faculty of Architecture, Technische Universität Dresden, 01062, Dresden,
Germany, [email protected]

ABSTRACT
Walls of buildings do not only bear vertical forces but also act as stiffening members such as
shear walls. Applied horizontal forces result primarily from wind, as well as imperfections and in
earthquake-prone areas from earthquake actions.
In the case of capacity verification of masonry shear walls, where the bearing horizontal force
depends mainly on the size and location of the normal force, the partial safety concept leads to
unfavourable results and smaller load bearing capacities than the approach of the global safety
concept using service loads.
For the common model to calculate internal forces of the shear wall, a simple cantilever beam
with vertical forces from slabs and horizontal forces from wind is used. Due to this, the
maximum bending moment is situated at the bottom of the cantilever. But the resulting cutting
forces of a shear wall can be decreased by means of improved engineering models like frames
and interactions between wall and slab. Consequently, a possibly former unstable shear wall can
now be verified by using new models.

KEYWORDS: engineering modelling, shear wall, model comparison

INTRODUCTION
This report shows several options for the mobilization of structure-relieving recourses primarily
considering the assessment of the state of strain on shear walls. This is based on a truss model
[1], which is modified by different approaches. Consequently a shear wall, which gets too high
bending moments from the cutting force determination on the classical cantilever beam with
horizontal wind loads, can be brought closer to the real state.
The simplest determination for the capacity of a shear wall is to calculate the load eccentricity e
(see Fig. 1 and Eq. 1) and to verify that it remains within the cross section of the wall. For the
Ultimate Limit State the value of e should be smaller than half of the wall length t, otherwise the
wall overturns, acc. EN 1996-1-1 [2].

M
e ; e  t / 2 (1)
N
a)
FSlab
½FWind ½FWind

slab

flange wall stiffening wall

FSlab
FWind FWind

slab
stiffening wall
flange wall

FSlab
FWind
slab

flange wall
stiffening wall

b) N M e

Figure 1: a) multi storey shear wall, b) simple model and cutting forces

For the following approach, the strength of the wall is not of interest. This should be a general
proposal for the determination of internal forces, close to the reality.

CANTILEVER WITH RESTORING FORCES

The conventional cantilever model remains unchanged. In addition, restoring forces of
components (e.g. present flange walls) are considered. These forces create an opposite bending
moment to the moment resulting from the slab supports and wind, see Figure 2.
 FSlab

Fflange wall

Fflange wall

Fflange wall

a) b) c)
Figure 2: a) Model, b) Structural system with restoring forces (red), c) Deformation

Steps
• determination of vertical and horizontal loads including possible restoring forces,
• load eccentricity resulting from the normal force and moment summation downwards,
• verification of the shear wall and
• verification of restoring forces in the reinforced concrete elements

CANTILEVER WITH SPRINGS

For the determination of the restoring forces, the attachment of the slab is considered. Additional
springs (torsion and / or slide springs) take into account the load transfer from the slab. Since the
model of the cantilever is applied again, an easy programmability is given. Nonlinear calculation
algorithms (gaping of the joint between the slab and the wall) can be taken into account [4]. To
replace the slab stiffness by an acting spring is still problematic, because the degree of fixation
between wall and slab is unknown. The slab might deform more than the wall. Thus, the slab
might act more in a loading than a supporting manor onto the wall (Figure 3). Consequently, this
kind of reserve is only to be considered with reasonable certainty if slab and wall rotate
differently and at minimum one full storey above is present (Figure 4 and Figure 5). This ensures
that the slab is clamped sufficiently in the stiffening wall and can be used for the reduction of the
bending moment.
Figure 3: Wind from the left – model and structural system. There are no restoring springs
applicable

or
Figure 4: Wind from the right – model and structural system with springs and respectively
with connected floor strips
Figure 5: Hinged support of the top floor slab and restrained floor slab in all lower storeys
(sketch without scale)

Due to its static indeterminacy, the model is unsuitable for manual calculation, however this
system can be solved by using simple 2D truss programs that are available as freeware. For the
common cantilever beam, software based on Excel is published in [3].

Steps
• determination of vertical and horizontal loads including possible restoring forces,
• determination of the spring stiffness of the slab,
• perform calculations
• verification of the shear wall and
• verification of the restraining bending moment of the slab

FRAME WITH BRUSHES FOR CONTACT CONNECTIONS

In addition to the consideration of the framework effect, the separation of the connecting wall-
slab is considered separately within this model. Currently this model can only be used in a truss
model, where single bars fail due to tension.. Many truss models can already consider this type
of bar failure by default (Figure 6). The optional support effect of a flange wall is achieved by a
further strut on the outside of the "brush". Dead loads from this flange wall are only active as a
backward rotating moment, if the cross section lifts up, whereupon the outer tie rod fails.
Otherwise, the dead load will be transferred by this rod directly down to the foundation.
 FSlab

Fflange wall

Fflange wall

Fflange wall

a) b)
Figure 6: a) Model, b) Structural system with brushes for contact connections to the
stiffening wall

Due to the high degree of non-linearity of this model it is not suitable for manual calculation. It
requires software for modelling the strut-and-tie structure which supports the failure of tension
rods. For the input data, further details of boundary conditions (stiffness of the brush, spacing,
length, etc.) are required, if the brushes are not generated automatically. This model includes
structural nonlinearities of wall-slab-node, whereby the restraining of the slab can be accurately
mapped.

Steps
• determination of vertical and horizontal loads including possible restoring forces,
• entering data (including brush) and perform calculations
• designing the slab for shear forces

COMPARISON OF THE MODELS ON AN INNER SHEAR WALL

The comparison of the different models will be shown on the example of a single-family house
(see [3]). In each case, the internal force values from the various models are compared. The
calculation of the cantilever requires the supporting forces of the slab.
The load determination of the frame is described in detail in [3]. Figure 6 shows the final loads in
vertical and horizontal direction.
 10.5 13.3
5.60 5,50
6 ,37 5,32

10.5 13.3
5.60
2,75
5 ,59 2,57

0,0

30 1,25 3,15 30
5,00

Figure 7: Loading scheme of the whole frame (Load unit [kN] respectively [kN/m]) [3]

USING THE COMMON CANTILEVER BEAM

The cantilever is assumed as fully fixed into the wall, because above the upper floor slab exists a
gable and the lower floor slab is loaded by the wall above. The bending moment of the slab,
which is restrained at the left could be determined with a value of 18 kNm, see Figure 7a. This
bending moment acts on the cantilever in addition to the support load.
2 6.6

30.0
6 .37 5 .6 0 6. 37 31.39
18.0

Forces in
2 6.6

[kN], [kNm] and [kN/m]

30.0

5 .59 5 .6 0 5 .59 31.39

=>
18.0

10.5 13.3

18.0
26.6

16.5

a) b) c)
Figure 8: a) Supporting forces from the floor slabs, b) Load on the cantilever beam, c)
Transformation into node loads
The restoring forces out of the gable are assumed with a maximum of 23 kN at the upper floor,
and 20 kN at the lower floor (see [2]).

23.0
24.6
31.39
6.37

24.6
20.0
5.59 31.39

Figure 9: Cantilever beam with restoring force

APPLICATION OF THE FRAMEWORK MODEL

By considering the entire floor frame the deformation of the slab has an influence. The amount
of the restraining for the slab is now determined by the node stiffness, which leads to a lower
bending moment. Here, the actual fixation of the slab into the wall is considered (but without
taking into account the nonlinear behaviour). Thus, this model is more realistic than the simple
cantilever beam, where a 100% rigid restraining of the slab was assumed.
 10.5 13.3
6 .37 5. 60

13.3
10.5
5. 60
5 .59

Figure 10: Frame model including shear walls and slab strips
With this model the exact determination of the actual position of the load introduction into the
wall is not possible. The node of the model can transfer tensile stresses, which is not the case in a
real structure. The cracking of the slab-wall connection is not considered.

USING THE BRUSH MODEL

The brush model can consider both, the restoring forces from the gable and the loading from the
slab. Simultaneously bars, which fail due to tensile stresses, can take the exact state of the
support area of the slab to the wall (gapping of the "joint" is possible) into account.
23 .0

10.5 13.3

6.37 5.60
20. 0

10.5 13.3

5.59 5.60

a) b)
Figure 11: Cantilever beam with brushes a) Model, b) Deformation and failing tension bars
MODEL COMPARISION
Depending on the used model, different values with an enormous variation are achieved.
Figure 12 shows the comparison of the values of several cutting forces for all three different
models.

Normal force Moment load path

6 6 6

[m] 5 5

4 4 4

Height [m]
Height [m]

Height [m]
3 3 3

2 2 2

1 1 1

0 0 0
-150 [kN] -100 -50 0 [kNm] -100 -50 0 50 -1,0 -0,5 0,0 0,5 1,0 1,5 [m]

Cantiliever Cantiliever + restoring Force Frame Frame+restoring Force Brush l. border r. border

Figure 12: Internal forces due to wind from left in comparison of different models

It is to be noted that in some cases the load eccentricities are close to the cross section border or
even outside of it, hence in this case no verification of the shear wall is possible. Strictly, only
the models ‘frame with restoring forces’ and ‘brush model’ can satisfy, that the load path
remains inside the wall cross section.
By the stepwise activation of the system reserves, the design-dependent load eccentricity is
successively reduced in the shear wall. As a prediction for the model closest to the reality, the
brush model (according to [3]) is suggested, which can take into consideration the load
distribution and nonlinear properties of the entire system.

CONCLUSION
By considering the wind load from the left one achieves results, which are acceptable and useful.
Using the common cantilever beam model, the used example cannot be checked in case of wind
load because the resultant force is situated outside of the wall. With the enhanced models ‘frame
and restoring forces’ and ‘brush system’; the verification condition improves satisfyingly, as the
force resultant of the wall is shifted into the cross section.
The enhanced models can achieve better results, which are nevertheless on the safe side.
However, both wind directions must consider finding the worst cutting forces for the
verification.
Strictly, the cantilever beam model is far away from reality. The enhanced models approach
better the reality, but still do not reach the real state of the shear wall. This shows the modelling
of the entire building [3] with finite elements with nonlinear material and contact properties.
REFERENCES

1. Beck, H., Schäfer, H. (1969) „Die Berechnung von Hochhäusern durch

Zusammenfassung aller aussteifenden Bauteile zu einem Balken“ Der
Bauingenieur 44, pp. 80–87.
2. EN 1996-1-1 Eurocode 6 - Design of masonry structures - Part 1-1:
General rules for reinforced and unreinforced masonry structures.
3. Jäger, W., Ortlepp, S., Bakeer, T., Vassilev, T., Motazerolghaem,
M., Richter, C., Bergander, H. (2010) „Schnittkraftermittlung für
aussteifende Mauerwerkswände“ research report, Technische
Universität Dresden, Faculty of Architecture, Chair of Structural
Design
4. Vassilev, T., Jäger, W. (2010) „Nichtlineare Berechnung von
Aussteifungsscheiben mit dem Übertragungsmatrizenverfahren“ EXCEL-
program X-WAND. Technische Universität Dresden, Faculty of
Architecture, Chair of Structural Design