CHAPTER 6 Experimetnal study on steel-concrete-steel sandwich composite shell
6.4 Test results and discussions
6.4.6 Effect of variables on the punching shear strength of the shell
The test results, including first and second peak resistances, their corresponding deflections and strains, are listed in Table 6.6. The investigated variables influencing the strength of the SCS sandwich shells are:
1) Bond improvement by introducing mechanical shear connectors to the SCS sandwich composite shell structure;
2) The thickness of steel shell or ratio of the steel shell in the section of the structure;
3) The strength of the filled core material;
4) Spacing of the connector;
5) The curvature of the SCS sandwich shell i.e. ratio of /R hs.
Influences of these parameters are shown in Fig. 6.17. Introducing mechanical shear connectors increases the resistance of the SCS sandwich composite shell (shown in Fig.
Fig. 6.17(a). Increasing the thickness of the steel shell leads to thicker effective depth of the SCS sandwich composite shell. Therefore, the punching shear resistance of the sandwich shell is increased (shown in Figs. 6.17(b) and (c)). Moreover, the punching shear strength of the steel shell itself is also increased when thicker steel shell is used (shown in Figs. 6.17 (b) and (c)). Varying the spacing of the shear connectors changes the quantity of the shear connectors linking the shear crack in the core material that finally changes the punching shear strength of the structure (shown in Fig. 6.17(d)). Using higher strength core material definitely leads to higher punching shear strength of the
‐ 252 - structure due to higher shear stress of the concrete (shown in Fig. 6.17(e)). The curvature of the shell also influences the punching shear strength of the structure. Changing curvature influences the stress state of the inside core material and thus change the fracture surface, which finally influences the punching shear strength of the structure (Shown in Figs. 6.17 (f) ~(h)).
6.4.6.1 Effect of using mechanical connectors in SCS sandwich composite shell
From Fig. 6.17(a), it can be seen that using mechanical connectors greatly improves the structural resistances. SA2 exhibits the first peak value of 1440 kN that is 1.53 times of 944 kN for SA1. The second peak resistance is increased by 36% from 1218 kN to 1657 kN when the connectors are used. The elastic stiffness of the structure was also increased from 265 to 358 kN/mm. The reason for these improvements both on strength and stiffness of the sandwich shell is that the introduced J-hook connectors provide a higher composite action between the core material and steel face plate. This higher composite action offers higher stiffness to the section of the structure. Moreover, the introduced shear connectors increase the shear resistance of the cross section that permits the specimen SA2 to take larger punching shear load.
The introduced J-hook shear connector in SA2 also prevents local buckling of the steel shell in the compression zone that occurs to SA1. This is due to that the J-hook shear connector in SA2 reduced the length of compression zone. Another problems exposed in the test is that the selected J-hook shear connectors made assembling problems. For a pair of interlocked J-hook connectors, if they were not welded in the exact normal direction to the shells, it will cause the assembling difficulties for the shell skeleton. This will greatly reduce the construction efficiency and make the fabrication costing.
Therefore based on the lesson learned in the first stage, in the second stage of
‐ 253 - experimental study, the headed shear connectors are used for all the SCS sandwich shell specimens. The headed shear studs exhibit advantages of easy installation and equivalent strength to J-hook connectors, and it will also reduce the critical requirement on the welding of the J-hook connectors on the curved shell surface and thus increase the construction efficiency.
6.4.6.2 Effect of thickness of steel surface shell
Fig. 6.17(b) shows the load-deflection curves of specimens designed with the same dimensions and materials but different thickness of surface skins. 5, 8 and 12 mm thick steel face skins are used for SB1, SB2 and SB3, respectively. From Fig. 6.17 (b), it can be observed the first peak resistance of the SCS sandwich shell increases from 1083 to 1363 and 1737 kN when the thickness increases from 5 to 8 and 12 mm, respectively. The first peak resistance of structure is increased by 26% and 60% for specimens with 8 and 12 mm thick steel face skins compared with the one with 5 mm thick surface skins, respectively.
These increments of the first and second peak resistances are shown in Fig. 6.17(c).
Another interesting observation is that specimen SB1 with 5 mm thick steel surface shells exhibits lower second peak resistance than the first peak resistance. However, both sandwich shells with 8 and 12 mm thick steel plates exhibit contrary phenomenon. As the thickness of the surface skin steel plates increases, the Pp1 increases almost linearly whilst the second peak resistance increased faster than the first peak value. Explanations for this effect on increasing of structural resistance are: 1) the thickness of the top steel shell contributes to the punching shear resistance of core material as well as the steel surface shell; 2) however, the increment of the thickness of the steel shell on Pp1 is less
‐ 254 - significant than on Pp2; 3) the first peak resistance of the shell is mainly contributed by the core material whilst the second peak resistance is determined by the punching shear resistance of the top steel shell. Generally, using larger thickness steel shell both increase the first and second peak resistances. In this case, SCS sandwich shell with 125 mm thick core, the thickness of the steel shell is recommended larger than 8 mm to guarantee a enhanced second peak resistance.
6.4.6.3 Effect of spacing of connectors
The effect of spacing of the connectors is shown in Fig. 6.17(d). The connectors provide composite action, link the shear crack in the core material and thus provide shear resistance, and minimize the length of the compression zone to prevent local buckling.
Shear connectors in the SCS sandwich structure act as the shear reinforcement in reinforced concrete structure. Increasing the spacing of shear connectors reduces the quantity of the connectors linking the shear failure surface of the structure and therefore reduces the shear resistance of the structure. From Fig. 6.17(d), it can be observed that increasing the spacing of the headed stud connector from approximately 120 to 220 mm decreases the first and second peak resistance by 18% and 28 %, respectively.
6.4.6.4 Effect of strength of in-filled core material
Fig. 6.17(e) shows the effect of the core material strength on the ultimate strength of the SCS sandwich composite shell. This influence is studied through tests on specimen SB2 and SB5. These specimens were designed with the same dimensions except the core material. ULCC is used in SB2 whilst HPC is used in SB5. The properties of these two types of materials can be found in Table 6.2(a). From Fig. 6.17(e), it can be observed that using HPC greatly increases the punching shear strength especially the first peak
‐ 255 - resistance of the SCS sandwich shell. The punching shear strength of SB5 using HPC is 2.25 times of SB2 with ULCC. The stiffness of the SCS sandwich shell is also greatly increased by using higher strength concrete. After the core material is punched, the second peak resistance of the two specimens provided by the steel surface shell and connectors are very close. Because, the second peak resistance is determined by the punching shear resistance of the outer steel shell.
These phenomena can be explained by that higher strength of core material provides higher punching shear resistance as well as elastic Young’s modulus. Therefore, both the punching shear resistance and the stiffness of the structure are greatly increased.
6.4.6.5 Effect of curvature of the shell
Fig. 6.17(f) shows the load-deflection curves of the SCS sandwich composite shells with different curvatures. The curves of the first and second peak resistances versus different curvature (ratio ofh Rs / ) are plotted in Fig. 6.17(g). From these two figures, it can be observed that as the curvature of the SCS sandwich composite shell increases, the punching shear resistance of the structure increases firstly from 1166 to 1363 kN when the /Rhs ratio increases from 0.06 to 0.16. However, this strength decreases from 1363 to 1210 kN as the /Rhs ratio increases from 0.16 to 0.22. This implies that the increment of the curvature on the strength of the sandwich shells has both positive and negative influences on the punching shear resistance, which is greatly depended on the curvature.
If the curvature is smaller than 0.16, this influence is positive. Once the curvature exceeds the limitation, this influence will be negative. The generalized stress is shown in Fig.
6.17(h). From this figure, it can be seen that the generalized shear stress acted on the failure surface increases as the /Rhs increases. However, from Table 6.4, it is known that
‐ 256 - the critical perimeter of the punched cone greatly decreases in SB7 that was designed with the highest /Rhs ratio of 0.22. Therefore, the following conclusions can be drawn from the observations in Table 6.5 and Figs. 6.17 (f)~(h):
1) Increasing the curvature of the shell increases the inclination angle of the shear crack in both arch and width direction. This increased inclination angles of the shear crack minimize the controlled perimeter of the critical section.
2) Increasing the curvature of the shell increases the shear stress that acts on the shear failure surface (shown in Fig. 6.17 (h)). This is caused by the compression membrane forces acted on the shear failure surface in the core material. Similar observations were found in the RC reinforced concrete shells (McLean et al., 1990).
3) Though the stress acted on the shear failure surface increases, the ultimate strength decreases due to reduced controlled perimeter of the punched cone that is caused by the curvature of the shell (shown in Table 6.5). Compared with the influence of the curvature on the control perimeter, the influences of the curvature on stress is much smaller.