working with remennikovs steel brace hysteresis rule in rua

huangyhg

working with remennikov's steel brace hysteresis rule in rua
i am a new user of the structural analysis program ruaumoko 2-d and will appreciate some insight into some aspects of the program.
i am at the moment exploring the efficiency of the program by trying to simulate a laboratory excitation where displacement is applied to one of the top nodes of a single story single bay steel braced frame. i have got most of the sectional and material properties of the beams (top and bottom, i section), columns (left and right, h section) and the brace (single tube brace). i represented all the members by the one component (giberson) beam element (i am not sure if this is the best thing to do, pls advise). for the beam and column, i used elasto plastic hysteresis but for the brace, i used the remennikov steel brace hysteresis rule (n=24). my main objective is to study the efficiency of this hysteresis rule so i can use it to model other braced structures. the analysis is running to completion but the result (frame's base shear-drift hysteresis) is not matching the experimental results. the compression reagion is not behaving too badly (i wld say close to what the rule is suggecting but the tension region is rather not consistent. for the displacements applied, i expect the tension curves to reach the yield force at every cycle but it looks like there is a huge pinching occuring after a few cycles and so it is reaching the applied displacements at force levels about one-fifth or even tenth of the yield force, the experimental results looks fine and consistent with no sign of strength degradation in tension until failure.
first of all, i realised that using "shape=circular hollow section" gives me a more stable results that what the section actually is (i.e. square hollow section). any thoughts or experience on the effect of the shape parameter in remennikov hysteresis rule in ruaumoko on analysis results? and if there is, how do i obtain an equivalent circular section for square hollow section to serve this purpose, taking into account their buckling behaviours.
one other thing that i was not sure of are reasonable magnitudes of the empirical effective modulus parameters e1, e2, e3, and e4 that have to be specified in this hysteresis rule. any insight into obtaining reasonable values for these for the simple frame i am dealing with?
any suggestions on the alfa and beta parameters will as well be appreciated.
also i will appreciate it very much if in anyone's experience of the use of this hysteresis rule in ruaumoko, someone could tell me the critical parameters to be careful about, i mean those parameters which influece results of the analysis.