Dear Seismosoft,
I built a model representing nearly the half of an industrial precast frame with a planar (in xz plane) frame. I have two restraints which are assigned to the base of the column and to the mid-span of the beam. The first restraint disables the freedom of all translations and rotations. The second one restrains the translation in vertical (z) and out-of-plane (y) directions and out-of-plane translations about x and z axes. The mass is assigned to the elements with their own weights. The additional mass is assigned to the top of the beam. The neoprene pad is represented with linear link elements which are located between the beam and the column elements. The dynamic time-history load is assigned to the base of the column as a function of acceleration.
I’m running a dynamic time-history analysis and the base of the column moves while the beam remains fixed. Have you got any suggestions about correcting the general behavior of this model?
Best regards.
Unexpected behaviour of a model in DTHA
Re: Unexpected behaviour of a model in DTHA
Soydan, your second restraint, according to your comment, is fixed for x and z translation. Did you mean x and z rotation? If the beam restraint include x and z translation then that would explain why it doesn't move relative to the column base.
If you did, in fact, mean translation for the second restraint, then you might need to investigate the appropriate scale factor on the aplied base acceleration. If your input accelerogram is in units of g's, then you need to specify the appropriate factor on the applied loading (for example, the factor would be 981 if your distance units are set to cm). An incorrect scale factor on an accelerogram can also explain why there is very little movement relative to the base.
Another issue which can arise involves the use of eccelerograms which need to be baseline corrected. If double integration of the acceleration time history produces an artificial drift in the ground displacement, then the accelerogram needs to be baseline adjusted using SeismoSpect or SeismoSignal to obtain zero ground displacement at the end of the time history.
If you did, in fact, mean translation for the second restraint, then you might need to investigate the appropriate scale factor on the aplied base acceleration. If your input accelerogram is in units of g's, then you need to specify the appropriate factor on the applied loading (for example, the factor would be 981 if your distance units are set to cm). An incorrect scale factor on an accelerogram can also explain why there is very little movement relative to the base.
Another issue which can arise involves the use of eccelerograms which need to be baseline corrected. If double integration of the acceleration time history produces an artificial drift in the ground displacement, then the accelerogram needs to be baseline adjusted using SeismoSpect or SeismoSignal to obtain zero ground displacement at the end of the time history.
Tim Huff
-
seismosoft
- Posts: 1276
- Joined: 06 Jul 2007, 04:55
Re: Unexpected behaviour of a model in DTHA
Soydan,
One other issue that you did not mention is if you are talking about real-time visualisation of the deformed shape (in which case you probably need to "fix" the appropriate node in the Deformed Shape Settings) or about the actual analysis output, in which case issues such as those raised by huffte may apply (though there could be many others).
Seismosoft Support
One other issue that you did not mention is if you are talking about real-time visualisation of the deformed shape (in which case you probably need to "fix" the appropriate node in the Deformed Shape Settings) or about the actual analysis output, in which case issues such as those raised by huffte may apply (though there could be many others).
Seismosoft Support