Behavior of Shear Connected Cavity Walls
K. Papanikolas, M. Hatzinikolas, J. Warwaruk
September 1990, Department of Civil Engineering, University of Alberta
DISCUSSION AND ANAYSIS OF TEST RESULTS
This Chapter consists of two parts. The first part examines and discusses the test results and especially the effects of the parameters under investigation (Sections 18.104.22.168 and 22.214.171.124) on the performance of a cavity wall.
In the second part I comparison between the finite element analysis (described in Section 3.2) and the test results (Chapter 5) is presented for the case of masonry cavity walls. The objective was to confirm the validity of the finite element model (F.E.M.) in computing the internal forces transferred by the shear connectors. Since an elastic finite element analysis was conducted the comparison will be made only in the elastic range.
6.2.1 Shear Connector Plate
From the average values obtained by the wall segment tests (Section 5.2) it can be concluded that the capacity of the shear connector depends upon the type of load. The obtained average values were:
- Axial Compression: 5.8 kN
- Upward Vertical Load: 4.39 kN
- Downward Vertical Load 3.26 kN.
From the small values of the standard deviations (Section 5.2) it can be seen that the hole at which the tie is placed does not significantly affect the capacity of a connector under a given load. The mode of failure, which was identified as yielding of the metal plate around the hole at which the connector is placed, confirmed that the connector is fixed within the concrete block.
From the pull-out tests conducted on the simplified shear connector (see Sections 4.3.4 and 5.5) it was found that this connector is well confined within the mortar joints of the concrete block wall. The steel plate around the hole used by the connector, (or all connector types, will yield before the connection of the connector with the concrete block fails. The simplified shear connector type D, which consists of corrugation and holes, had the best behavior and was able to resist an average of 7.12 kN before the connection failed (note that the steel plate will yield at an average tensile force of 3.92 kN).
6.2.2 Effect of Concrete Block Width
By comparing specimen S1W2 with S1W3 (Figures C-3 and C-5) it can be concluded that for the case of hollow concrete block backup wythe the width of the concrete block does not significantly affect the overall behavior or the cavity wall, since the moment of inertia of the backup wythe does not increase much when the width of the block wall is increased. The poor performance observed for both specimens is attributed to the low tensile strength of ungrouted and unreinforced concrete block walls.
That is not the case for the specimens with reinforced concrete block wythe. Comparing S1W2 with S1W4 (Figure 6.1) and S2W3 with S2W4 (Figures C-13 and C-15), it can be seen that increasing the width of the concrete block results in an increase of the load carrying capacity of the wall system in the order of 50% to 90%. In addition, the maximum deflection at failure decreased by an average of two times due to changing from a 150 to 200 mm nominal block size.
For the case of 150 mm standard concrete block wythe (as presented in Sections 5.3.5 and 5.3.9) the grouting was poor, resulting in large deflections and premature failure. It is therefore important for the case of reinforced concrete block walls to ensure proper grouting.
6.2.3 Use of Vertical Reinforcement
A comparison of S1W1 and S1W3 with all the other specimens demonstrates that vertical reinforcement in the block backup wall results in a composite wall system with a load carrying capacity at least twice that or a hollow block backup wythe. That can be attributed to the following reason. For the type of load under consideration, the concrete block wythe is subjected to tensile stresses and therefore reinforcement is essential in carrying these tensile forces. The effect of grouting and reinforcement is shown in Figure 6.2 which compares specimen S1W1 with S1W2.
6.2.4 Effect or Shear Connector Arrangement
The variation of the connector pattern was investigated throughout this study and it was found that the spacing of the connectors and their location within the assembly must be considered as a relevant factor in performance of a shear connected cavity wall.
High axial and shear forces are attracted by the connectors near the supports. That can be confirmed by comparing specimen S1W2 with S2W1. The load carrying capacity of S2W1 was improved by 30% just by adding one shear connector at the top middle course of the failed mortar joint of wall S1W2.
As expected, by decreasing the spacing between the connectors the performance of the wall assembly is improved. A comparison of S1W2 and S1W4 with S2W3 and S2W4 showed that by reducing the distance between connectors in half the ultimate load capacity is doubled. The lateral deflections at comparable pressures were also decreased dramatically as a result of reducing the distance between connectors.
For the case of cavity walls with metal stud backup system it was found that the distance between the connectors affects significantly the composite performance of the wall assembly. For the case of connectors placed every 800 mm in both directions no composite action between the two wythes was observed. For a given pressure the deflections of the brick veneer were at least twice those of the backup system. By placing the connectors according to configuration C (see Figure 4.6 (b)) composite action between the two wythes is achieved. In addition to that the ultimate capacity increases significantly for the second case. Figure 6.3 shows the effect or the shear connector arrangements by comparing specimens S3W1 with S3W2.
6.2.5 Deflections of Masonry Cavity Walls
The allowable deflection for a masonry cavity wall is related to the crack width. In order to limit the maximum crack width to 0.5 mm it was found (Ref. 8) that the allowable deflection should be of the order of L/720. For the tested specimens this corresponds to an allowable deflection of 4.2 mm.
The deflections of the reinforced full scale specimens, corresponding to a pressure of 0.8 kPa (conservative value for wind load) were compared with the allowable deflection (4.2 mm). An average safety factor of 3 was found. That reduction in deflection will minimize crack width and moisture penetration, aspects that are very important for the design of masonry structures.
6.2.6 Effect of Backup System
Two different types of backup systems were used throughout this study: concrete block wall and metal studs-gypsum board assembly.
Figures 6.4 and 6.5 show the effect of the backup system for different shear connector patterns. As it can be seen from Figure 6.4, which compares specimen S1W2 with S3Wl, for connector arrangement type A better performance is obtained when concrete blocks are used for the backup system although the ultimate capacity of the metal stud cavity wall was higher. As was stated in Section 6.2.5, for this type of connector pattern composite action is not present when metal studs are used for the backup system.
When connector arrangement type C is used very good performance for both concrete blocks and metal studs backup systems were observed. Comparison of S2W3 and S3W2 showed that both specimens acted compositely up to failure. At the serviceability limit state better performance was observed for the masonry cavity wall (smaller deflections). From the point of view of ultimate capacity the metal stud cavity wall resisted higher pressure than its masonry counterpart.
6.2.7 Effect of Temperature Difference Between Two Wythes
From the cavity wall that was exposed to the climatic conditions it was found that internal stresses are generated due to material properties and environmental factors. More specifically, Figure 5.8 showed that there is a trend between the temperature difference and the shear forces at the connectors. Increasing the temperature difference increases the shear force.
Two different approaches exist in comparing the test results with the analysis. The one is by comparing the deflections for a given pressure and the other by comparing the internal forces in certain elements for a given pressure.
It is believed that for this study the most appropriate way of comparing the results was the second approach for the following reasons. The deflections in the elastic range (for both wythes) that were obtained by the analytical and experimental studies were very small (less than a 1.0 mm) and therefore cannot be used for a comparison (since a difference of deflections on the order of 0.5 mm will yield a 50% error). Secondly, since a cavity wall system is very stiff as can be concluded from the small deflections, the internal forces must be large and therefore that is the parameter that first concerns an engineer.
Since this research is focusing on the study of the load transferred by the connectors to the backup system, the comparison was carried out for the axial force at the bottom shear connector which is the critical one. The comparison was made only for the specimens that had their shear connectors instrumented with strain gauges. These are the specimens with 100 mm cavity (see Table 4.2).
The strain gauges readings were reduced by assuming a linear strain distribution. Then using a linear constitutive law with E = 200 GPa a linear stress distribution was found. Finally, using equilibrium equations the corresponding internal forces were calculated.
The comparison is presented in Table 6.1. The maximum axial force transferred by the connectors is reasonably well predicted by the analytical model. The average test-to-predicted ratio was 1.069 with a standard deviation of 18%.
For the completeness of the study Figure 6.6 shows a comparison of deflected shapes at different pressures for specimen S1W2. The deflections obtained by the F.E.M. were found to be half the corresponding experimental values.