cyclic loading modelling

[mortymory]

I have experimental results for Cyclic loading–unloading tensile tests(stress-strain diagram) on austenite superelastic wires and bars. the test was carry out at 0.01 HZ frequency. i want to make a similar model with seismostruct and compare the result. actuly i dont know how can i create cyclic loading with this computer prpgram in accordance with stress-strain diagram which i have. i cant create a static time-history curve for this. please help me to do that.

[seismosoft]

Have you not tried to transform the strain time-history into a displacement time-history (in your case you need simply to multiply the strain by the test specimen's length) and then use the 'Static Time-History' analysis type of SeismoStruct with the derived displacement time-history as the applied load?

Seismosoft Support

[mortymory]

thank you. you're right. i have transformed the strain time-history into a displacement time-history and I've done a cyclic analysis. now i have another problem i want residual strain at the end of action but when i run a displacement time history for an inelastic frame element in tension, the stress at the unloading time(when the displacement is zero) does'nt disapear.i have tried truss elemet for this cyclic tension test too and in this case the residual displacement is zero which is not true. can you guide me once more please?

[seismosoft]

If you are imposing displacements (hence strains) on your structural element, and if your displacement time-history ends with a value of zero, then it is impossible for you to have residual displacements/strains, since you are forcing these to be zero.

And if, when subjected to the displacement/strain history, your structural element was pushed into its inelastic range, then you are indeed likely to have residual stresses at the end of your imposed-nil-deformation analysis.

Seismosoft Support