HELP! RSA Analysis Errors.

[breyrw]

I was trying to perform a pushover analysis on a structure in RSA 2010. The pushover analysis never seemed to work and the total lack of help documentation on the matter made things even more difficult. Now no type of anlysis will run. It gives error number 100, incorrect data. Does anyone have an idea of what this means and how to fix it? The other issue now is that all of my static load cases automatically default to a non-linear analysis type, even without specifying non-linear. Anyone have some suggestions?

Thanks

[stephen.lewis]

Re: Defaulting to Non-Linear; if you have "non-linear behaviour" type elements defined in you rmodel then any analysis you subseqeuntly run will defualt to non-linear. This will definitely apply to cable elements. I think it may also apply to; tension only, compression only elements. Switch the elements back to normal bar type behaviour (i.e. elemnts can take both compression/tension/flexure) and the analysis should go back to linear.

[breyrw]

So I bet non-linear hinges fit into that category as well...that could be why it is not going back to linear. Thanks.

[breyrw]

You wouldn't happen to know how to get plots from a non-linear analysis? I go to Results-Advanced-Diagrams and pick what I want to plot and choose increment for the horizontal axis, but the apply button never becomes available. It remains grayed out. I don't think the non-linear analysis is working properly eventhough it says that results are available.

[stephen.lewis]

sorry, i havent had that problem before. Are you sure your model converged properly? If the model is ill-conditioned you may have excessive and onerous deflections which will give problems when plotting deflections etc. have you run a lienar anlaysis first to make sure you have a stable model before diving in to a nonlienar analysis?

[breyrw]

The model ran fine as a linear analysis. Sometime after the NL analysis, the deformed structure is crazy, but other times it looks just fine, but still no results. I have even tried the non-linear analysis on a simple frame and even just a simply-supported beam and still no luck. There has to be something that I am overlooking...

[stephen.lewis]

The model ran fine as a linear analysis. Sometime after the NL analysis, the deformed structure is crazy, but other times it looks just fine, but still no results. I have even tried the non-linear analysis on a simple frame and even just a simply-supported beam and still no luck. There has to be something that I am overlooking...

Im not a push over analysis expert but I think you are basically increasing the applied load until collapse so may be try and factor up the applied load load in your static linear case to this value and see if you get ill-conditioning. Also run a modal analysis with no applied loading and check the animation/deflected shape for first few mode shapes to see if any memebrs are not connected properly which may explain ill-conditioning

[breyrw]

Ok, I finally got it to obtain reasonable results for a non-linear analysis on a consistent basis. A problem still remains when trying to create a diagram of results for the non-linear analysis. After going Results menu>Advanced>diagrams, the apply button never becomes available. Is a trick to getting this button to become available? I noticed that in the help for this topic, the picture they show of the dialogue also has the apply button grayed out.

[waldemar.okapa]

The diagrams, and the Apply button, are available for multiply component cases, such as time-history, push-over and nonlinear analysis.
However in case of nonlinear analysis only the last component (load incement) results are saved by default. In order to get the diagrams check 'Save results after each increment' in the nonlinear case parameters (Analysis > Analysis Types > dbl click on a case in the list)