Tension only analysis in ETABS 4

[stressed]

I've recently started to work with the tension/compression limit feature, and I'd like to share some of my experience so far, but I'd also like to hear from some of the more experienced users regarding their suggestions in how to best use this capability.

Ok, first of all tension only starts as a frame assignment with compression limit = 0 or some small negative number. Next, after you have your load combinations for design, go to Define>Convert combos to nonlinear cases and select the combos (the ones with lateral loads?) to convert to nonlinear. I am not sure if I should then delete my linear load combos (?)

Next, run an analysis, and after it's finished, go back again under Analyze menu and analyze "run static nonlinear analysis" which can take a while to converge. Check interactive frame member forces for axial forces to see that the braces, or whatever frames you assigned to be compression limit=0, are acting only in tension. I see that ETABS defaults to two nonlinear steps in reporting output with different results. Worse case final step seems to look reasonable.

I'm now trying to sort out if ETABS is designing properly with the nonlinear combos. I'm curious if those of you with 'Plus' version have access to this tension only nonlinear analysis or is it only available with nonlinear? I'm just getting familiar with how to use this capability, so any advice would be appreciated.

 

[WillisV]

faq802-1169

 

[masomenos]

Thanks Willis! A star for your excellent FAQ. A couple of comments though - first, tension/compression limits seem to be working in ETABS 9, so that saves you from having to manually define nonlinear hook-type links. Also, your suggestion in your FAQ to Define>static nonlinear/pushover cases seems to work well, but it appears you can also use that dialogue to define combos too by adding 2 or more load cases. In other words, the static nonlinear/pushover "case" dialogue can be used to define combos as well as individual nonlinear load cases. There is also the alternative to Define>Convert combos to nonlinear cases. This ETABS feature seems to work well now, although it's a little more complicated than doing the same thing in SAP.

 

[masomenos]

I forgot to post that I heard back from CSI technical support last week.. they confirmed that it's best to define static nonlinear/pushover cases as "combos" rather than individual cases. Otherwise, if you don't define as nonlinear combos, you risk getting in trouble with the + and - directions, and furthermore, frame members assigned as tension only (compression limit = 0) may take compression unless ALL combos are defined as nonlinear, not just lateral loads.

 

[WillisV]

Any non-linear type of analysis should always be performed on the full combination in question as superposition of loads is no longer valid.