CHAPTER 6 Experimetnal study on steel-concrete-steel sandwich composite shell
6.4 Test results and discussions
6.4.2 Ultimate strength and failure mode
6.4.2.1 Ultimate strength and corresponding strains at the critical locations
The ultimate strengths of the SCS sandwich shells were recorded during the test. Both
1
Pp and Pp2were easily determined from the load-deflection curves of the structure.
Moreover, strains in three representative locations were recorded to facilitate judging the
‐ 243 - stress developments in the steel shell at the loading levels of Pp1 and Pp2 . Three representative locations were chosen that located in the bottom steel shell underneath the loading point, in the top steel shell close to the load cell both in arch centerline and width centerline directions. Moreover, a roughly estimated size of the punched cone in the bottom inner shell was assumed with a distance around 280 mm from the center point.
Strains at these two locations were also measured. Therefore, at loads of Pp1 and Pp2, strains at five locations were recorded to judge the stresses developed in the outer and inner steel shells. The ultimate strengths of the SCS sandwich composite shells as well as the strains of the five critical locations are listed in Table 6.3. The locations of the linear strain gauges (L) and three direction rosettes (R) are shown in Figs. 6.5 and 6.6. The Rx-1 means strain along the main direction along the arch or width of the shell (x is the number of the shell), and Rx-2 or 3 is the direction 45-degree off the arch or width direction.
6.4.2.2 Deformed shapes and failure modes
One major failure that occurs to most of the specimens is punching shear failure of the concrete core. The punching shear failure is a type of brittle failure that happens with a sudden drop of the load carrying capacity without large deformation of the structure (Stein et al., 2007). The shell or the plates are punched out with a conical plug of concrete.
More information on this type of failure can be referred to review by Regan and Braestrup (1985). Three means are used to recognize this type of failure modes during the test:
1) Sudden drop of the load-deflection curves, which is a type of brittle failure. This presents the sudden lost of the load carrying capacity of the structure.
2) Strains in the steel skin shell. Strains developed in the steel face shells are useful means to judge the stress state of the steel plates. However, this is a necessary condition
‐ 244 - but not a sufficient condition for the judgment. Taking the partial composite beam for example, the bottom steel plate yields due to flexural bending of the steel plate itself and the beam may also fail in brittle shear failure rather than flexural failure. In contrast, if flexural failure occurred, the flexural reinforcement of the structure must achieve yield to perform a large deformation of the structure.
3) Observations of the crack in the concrete core and steel shell.
Finally, there are three types of failure modes observed, i.e. punching shear failure of the concrete core, punching shear failure of the steel shell and global buckling of the structure.
The punching shear failure of the concrete core and steel shell took place in specimens SA2, SB1~SB7 at the first and second peak strength, respectively. At the first and second peak strength, punching shear failure and global buckling occurred to SA1.
The deformed shapes and the observed bump caused by the punched frustum of concrete of the specimens SB1~SB7 are shown in Fig. 6.11(a)~(g). Fig. 6.11(h) shows the punching shear failure of the outer top steel shell of specimens SB1~SB7. Fig. 6.11(i) shows the failure modes of the specimen SA1 that firstly fails in the punching shear failure and finally fails in the global buckling of the shell structure. Fig. 6.11(j) shows the deformed shape and punching shear failure mode of specimen SA2.
6.4.2.3 Size of the punched cone and cracks in the core material
A frustum of the core material was pushed down and formed a bump in the bottom steel shell due to the punching shear failure. Size of the top surface of this frustum is equal to the loading cell whilst the bottom size can be obtained through measuring the dimensions of the bump in the bottom steel shell. The size of the punched frustum in the core material can be used to estimate the dimension of the perimeter of the critical shear section. Based on the measured dimensions, the inclined angle of the shear failure surface can be
‐ 245 - calculated, which will be further used to analyze the punching shear resistance of the SCS sandwich composite shells.
The dimensions of the bump are denoted as La along the arch centerline direction and Lb along the width centerline direction, respectively (as shown in Table 6.4). The dimensions of the load cell are 125x125 mm2 for all specimens. Based on the measured dimensions of the La and Lb, the inclined angle of the frustum are calculated as the following;
2 2
1 / 2 2
tan / 2
s s a s s
a
a
R t R t L h t
L a
(6.1a)
1 2
tan / 2
s s
b
b
h t
L b
(6.1b)
where, a and b are the size of the load cell, herein adopt 125 mm.
All the calculated inclination angles are shown in Table 6.4. From this table, it can be observed that the inclination angle of the frustum of the punched concrete, i.e. a along the arch direction is around 40-degree excluding specimens SB6 and SB7. As the curvature increased, this inclination angle increased. Comparing the a of specimen SB6 with those of SB2 and SB7, this inclination angle increases from 31.1° to 38.5° and 45.3°
when the span-to-radius ratio increases from 0.53 to 1.41 and 2, respectively. It also can be seen that the inclination angles of the shear plane in the arch direction (i.e. a) with a range of 31.1°~45.3° are larger than the angles b in the width centerline direction with a range of 17.8°~32.5°. The mean values of a and b are 39.2 and 23.7, respectively.
These findings are consistent with the experimental observations of SCS sandwich composite shells without shear connectors tested by Shukry (1986) and reinforced or pre- stressed concrete cylinders tested by Barkel et al. (1979) and Caldwell et al. (1981).
‐ 246 - Explanation for the deeper a than b is the existence of the compression membrane forces in the SCS sandwich composite shell caused by the curvature. This explanation is further proved by the previous discussions on the change of the ain specimens SB6, SB2 and SB7 where the a increase as the curvature increases.
The cracks in the core material during the loading process were observed and marked in Fig. 6.12. The cracks and their corresponding loads are shown in Fig. 6.12 and listed in Table 6.5. From these information, it can be seen that the load of the first crack in different specimens ranges from 220 to 440 kN that are around 14~38% of the corresponding first peak resistancePp1 (See Table 6.5). This observation is identical to the findings of Shurky (1986) that the load at the first crack occurred is around 13.5~38%
of punching shear strength. It is also observed that this value is greatly influenced by the curvature of the shell. The lowest ratio of P1-to-Pp1is 14% for specimen SB6 with the high performance concrete (P1 is the load at first developed crack in the core material).