Seismic ratcheting of steel low-damage buildings
Thesis DisciplineCivil Engineering
Degree GrantorUniversity of Canterbury
Degree NameDoctor of Philosophy
During earthquake shaking some structures tend to deform and yield more in one direction than in the other. This phenomenon is sometimes termed “ratcheting” and the displacement demands may become significantly larger than for structures without a ratcheting tendency. This thesis explores the numerical and experimental studies on performance of steel low damage buildings with ratcheting tendency under seismic demands.
Numerical studies are used to develop simple methods to estimate the displacement demands of such structures with different periods (T) and force design reduction factors (R). Shake table studies of two storeys half-scale steel moment frame with asymmetric friction connections (AFCs) at the column bases and at the beam ends were carried out. A tested structure with residual displacements due to earthquakes was strengthened/stiffened using several methods to minimize the possibility of increase in peak/residual displacements in the residual displacement direction due to aftershocks. Experimental tests were conducted with i) no brace, ii) a buckling brace with slackness, (iii) a ratcheting brace, and (iv) gapping braces, which had a ratcheting brace in conjunction with a buckling brace with a displacement gap.
Time history analysis of steel structures with initial out-of-plumb showed that buildings with greater initial out-of-plumb and force design reduction factor tended to have larger residual and peak inter-story drifts. It also showed that for high out-of-plumb, the ratio of residual-to-maximum possible peak drift tends to unity, indicating that the buildings are yielding predominantly in one direction.
Nonlinear time history analysis of steel structures also showed that a sequence of realistic shakes from actual earthquake recording tended to increase the median peak and residual drift response of the structures. The structures tended to experience ratcheting in predominantly one direction. The tendency for ratcheting increased with increasing force reduction factor (R) and design drift.
Time history analysis of the single degree of freedom elastic structures with different stiffness in each direction showed that greater displacements generally occurred in the direction of lower stiffness for elastic structures. Peak drift in the stiffer direction was able to be predicted by the spectral displacement associated with the period in that direction and the peak drift in the flexible direction estimated from the displacement in the opposite direction using energy considerations.
For yielding structures with different stiffness/strengths in opposite horizontal directions, when strength proportional to stiffness, the inelastic displacement in each direction could be estimated from the elastic response in that direction using standard modifications for inelasticity. For long period structures peak inelastic displacements were similar to the peak elastic displacements considering the stiffness difference.
Shaking table testing of a half-scale two-story steel moment frame with asymmetric friction connections (AFCs) at the column bases and at the beam ends showed that, when beam ends and the base-column joints were modelled by appropriate trilinear and bilinear hysteresis loops respectively, the response with time matched the numerical simulations well. Residual drifts were less than 0.2% for peak inter-storey drifts up to 3%, and less than 0.7% for peak inter-storey drifts of 6.0% indicating desirable seismic performance. It was also found that it was possible to obtain repeatable peak and residual displacements with a variation of less than 2.0% for straight structures subject to same record. Since there was no significant member damage, these friction structures may be considered to be low-damage.
Shake table study of strengthening/stiffening of a two storey low damage half scale steel structure using tension braces showed that adding a buckling brace (BB) reduced the residual drift by 75% and did not push the frame in the opposite direction. The ratcheting brace (RB) was very effective of straightening the structure with a residual displacement change of -260% implies that it caused a 60% greater residual displacement in the opposite direction. The gapping brace (GB) also changed the residual displacement by -150%. Both the buckling brace (BB) and gapping/buckling brace (GB/BB) combination had the desirable characteristic of limiting further drift in the residual displacement direction without pushing the frame in the opposite direction.