This section presents the analysis results and performance assessment of case study buildings for each performance objective.
Occupancy Comfort and Operational Performance Levels
Time histories of acceleration at the center of the roof of case study buildings for one of the time histories are shown in Fig.
9 for along-wind and across-wind directions, respectively.
The ensemble average and coefficient of variation (C.O.V.) of maximum acceleration at the roof level of case study buildings for 1-year MRI wind speed are reported in Table
8. Acceleration in torsion (equivalent translational acceleration) is defined as
, where
is the distance from the center of torsion to the objective point (corner point in this study) and
is the angular acceleration of the torsional vibration [ISO 10137 (
ISO 2007)]. The results are presented independently of the value of
, because the initial stiffness and consequently demand for both cases are the same.
Based on the results, aerodynamic modification is very effective in reducing acceleration. Accelerations in along-wind, across-wind, and torsional-wind of Case 2 are about 72%, 91%, and 73% of Case 1, respectively.
Accelerations in along-wind, across-wind, and torsional-wind directions of Case 3 are about 75%, 74%, and 84% of Case 1, respectively. Since the maximum acceleration is in the across-wind direction, it can be said that the efficiencies of both Case 2 and Case 3 are comparable, and aerodynamic modification can reduce acceleration by about 25%. The ensemble average of peak accelerations is compared with the frequency-dependent acceleration limits of the prestandard for 1-year MRI and office occupancy, as shown in Fig.
10. According to the results, the value of demand acceleration is smaller than the limits, and acceptance criteria for occupant comfort performance level are met for the buildings.
The ensemble average and coefficient of variation (C.O.V.) of maximum displacement at the roof level of case study buildings for 50-year MRI wind speed are reported in Table
9. Note that results are presented independently of the value of
, because the initial stiffness and consequently demand for both cases are the same.
Based on the results, aerodynamic modification is very effective in reducing displacement, as well. Maximum displacement in the along-wind and across-wind directions of Case 2 is about 81% and 72% of Case 1. Maximum displacements in the along-wind and across-wind directions of Case 3 are about 81% and 67% of Case 1, respectively. In comparison with Case 1, the maximum displacement under along-wind load for both Case 2 and Case 3 is about 19% smaller. Under across-wind load, the maximum displacements of Case 2 and Case 3 are about 28% and 33% smaller than those of Case 1, respectively. Thus, the efficiencies of both Case 2 and Case 3 are comparable as the difference is not significant (about 5%). Since the maximum displacement is in the across-wind direction, it can be said that aerodynamic modification can reduce maximum displacement by around 30%.
In accordance with the prestandard, the peak displacement should be less than 320 to 400 mm (
to
). So, the maximum displacements of buildings (Table
9) are less than the limits. Also, strength limits (
) were satisfied during the design procedure, since service wind loads were included in the load combination. Therefore, all acceptance criteria for operational performance levels for buildings are satisfied.
Continuous Occupancy Performance Level
The story force and shear for Case 1 based on nonlinear time-history analysis (NTHA) at the moment of the minimum and maximum overturning moment in along-wind and across-wind directions are compared with design values in Fig.
11. Since values at the minimum overturning moment in the along-wind direction were very small, they are not depicted in the figure. It can be seen that the results by NTHA are very close to those of design values based on ESWL. The differences between design along-wind forces and results by NTHA are due to the assumed distribution for the resonant component of ESWL (similar to the mean component). The ensemble average of story force and shear for case study buildings are compared in Fig.
12(a). It can be seen that the results for Case 1 with
and
are almost the same, and the difference is less than 1%. The reason lies in the fact that there is a difference between some influential factors in the initial design procedure and performance assessment. In the design procedure, specified material properties and strength reduction factors of less than 1 were used. Meanwhile for time-history analysis, expected strength and strength reduction factor equal to 1 were used. The difference results in a larger initial yield strength of the structures than expected values from the initial design. Thus, the structures remain almost elastic, which implies that dynamic properties will not significantly change, and induced dynamic loads remain the same in both structures. In comparison of the aerodynamically modified model (Case A, average of Case 2 and Case 3) with Case 1, forces in the across-wind direction of Case A are a little smaller, and forces in the along-wind direction are comparable.
The maximum overturning moment and base shear of case study buildings in along-wind and across-wind directions are reported in Table
10. In comparison with Case 1 with
, responses of Case 1 with
are a little larger in the along-wind direction and smaller in the across-wind direction.
Since some elements in Case 1 with yielded, the natural frequency decreases a bit and demand increases. Because the structure experiences full cyclic deformation due to a small mean value in the across-wind direction, total damping increases by hysteretic damping. Demand is a consequence of both effects, and for the case study building the reduction effect by hysteretic damping is more dominant, resulting in the smaller demand.
In the along-wind direction, deformations are not in a full cyclic form due to the large mean value, and hysteretic damping does not contribute to the total damping of the system. However, damping increases through unidirectional plastic deformation. For this building, the amplifying effect is more dominant, which results in the larger demand. However, the differences between the two structures in both along-wind and across-wind directions are negligible.
The difference between Case 1 and Case A in the along-wind direction is not significant (about 2%); however, Case A has about 13% smaller responses in the across-wind direction.
The C.O.V. of responses for Case 1 with and is about 8% to 13%, though the values for Case 1 with in the across-wind direction are about 1% larger due to inelastic behavior. The C.O.V. of responses for Case A is about 14% to 40% larger than Case 1. Notwithstanding is that the values of C.O.V. of all cases are relatively small, which confirms that the uncertainty in the ensemble average of peak values is quite reliable.
Ensemble averages of maximum story displacement and drift are compared in Fig.
12(b). Since minimum values for along-wind direction are small, they are not shown in the figure.
It can be seen that the maximum drifts or displacements of Case 1 with and are almost the same in both the along-wind and across-wind directions. The ensemble average of maximum story displacement (and story drift) in the along-wind and across-wind directions are 210 mm (0.172%) and 234 mm (0.190%), respectively. For aerodynamically modified models, the ensemble average of maximum story displacements or story drifts in both the along-wind and across-wind directions are almost the same as 203 mm (0.167%). Compared to Case 1, the ensemble average of maximum story displacement, as well as story drift, of Case A in the along-wind and across-wind directions is around 3% and 13% smaller, respectively.
The C.O.V. of ensemble average of maximum story displacement or drift is quite small, confirming that the uncertainty in the ensemble average of peak values is relatively low; i.e., the results are reliable (9% to 11% for Case 1 with
; 9% to 12% for Case 1 with
; 13 to 19% for Case A with
). In addition to the peak drift value, residual drift was measured for Case 1 with
. Since the defined ramp-down part at the end of time history is quite long (about 100 s), residual drift was defined as the value at the last increment of each time-history analysis, and the absolute values are shown in Fig.
13.
Since the C.O.V. of the ensemble average of residual drift is very large in this case, the maximum values in the along-wind and across-wind directions (14 and 60 mm, respectively) are considered for conservatism. Based on the suggested allowable value of drift by the prestandard, the peak displacement should be less than 533 to 800 mm ( to ). In addition, residual displacement at the roof level and each story should not be more than 160 mm (). Therefore, acceptance criteria for both transient and residual drifts are satisfied in all cases.
No plastic deformation was observed in any building members designed with . Since all members of Case 1 with and Case A with remain elastic, the demand capacity ratio (DCR) for all members is less than 1, and strength limits are satisfied.
Several coupling beams of Case 1 with
experienced small inelastic deformations, as shown in Fig.
14(a), and all other members remain elastic. It can be seen that maximum plastic rotations in the across-wind direction are relatively larger than those in the along-wind direction. Only a few elements yield in the along-wind direction, while many elements yield in the across-wind direction. The force-displacement curve of the most critical element is shown in Fig.
14(b). As shown in the figure, elements in the along-wind direction almost remain elastic, and maximum inelastic deformation is very small.
Based on the results, all coupling beams are in immediate occupancy level. The plastic deformation in both directions is very small and comparable with the reported values by Aswegan et al. (
2017) and Jeong et al. (
2021).
According to the ASCE prestandard, to limit the risk of low cycle fatigue failure, the number of cycles with inelastic strain at 1.5 times section yield should be limited to about 10 cycles. The section yield of coupling beams varies between about 0.4% and 0.7%, and the corresponding inelastic rotation at 1.5 times section yield is in the range of 0.2% to 0.45%. These values are close to the acceptance criteria of coupling beams in ASCE 41-17 for immediate occupancy performance objective (0.5% to 0.6%). According to the results in Fig.
14, plastic rotations are much smaller than the limit value of 0.2% to 0.45%. It can be concluded that failure due to low cycle fatigue is not expected. Based on the above evaluations, all acceptance criteria for continuous occupancy performance objectives are met for all three case study buildings.
Safety Evaluation
The intention for the proposed enhanced continuous occupancy objectives is to ensure those acceptance criteria for continuous occupancy objectives can be satisfied under wind load with relatively larger MRI. For this reason, first, the story force and shear, as well as maximum (transient and residual) drift and displacement, of Case 1 with for different MRIs are compared.
The overturning moments in the along-wind and across-wind directions for MRI of 900 years are about 14% and 11%, respectively, larger than those of MRI of 500 years. Overturning moments in the along-wind and across-wind directions for MRI of 3,000 years are about 35% and 50%, respectively, larger than those of MRI of 500 years. Therefore, if the system satisfies the criteria under these significantly larger wind loads, it implies that there is a large margin of safety under design wind load with MRI of 500 years.
Based on the results, both maximum and residual story displacement and drift increase with the MRI significantly, especially in the across-wind direction. However, the maximum values are still considerably smaller than the limit of 533 to 800 mm ( to ) for maximum transient drift and 160 mm () for residual drift. This evaluation implies that there is a large margin against the acceptance criteria for transient and residual drifts.
The force-displacement curve of some critical coupling beams for different MRIs is shown in Fig.
15. It can be seen that more elements yield by increasing wind load and maximum plastic rotation is significantly increased, especially in across-wind directions. The amount of increasing the plastic rotation in the along-wind direction is considerably smaller than the across-wind direction.
Maximum plastic rotations of coupling beams for MRI of 500, 900, and 3,000 years are shown in Fig.
16. Because of the large C.O.V. of maximum plastic rotation, the maximum case among the 10 time series for each MRI is shown in the figure and was considered for comparison. Based on the results, both the number of yield elements and the maximum plastic rotation increase considerably by MRI. Furthermore, the difference in the across-wind direction is more significant. According to results, plastic rotations even for MRI of 3,000 years are much smaller than the limit value of 0.2% to 0.45%, and it can be concluded that failure due to low cycle fatigue is not much expected.
Since all acceptance criteria for continuous occupancy objectives are satisfied for MRIs of 900 and 3,000 years, which have significantly smaller annual probability, it can be concluded that there is a large margin of safety against failure mechanism.