Tomfh said:
Does anyone feel the Mastan results are unreliable? E.g. that they are underestimating the true theoretical buckling loads?
1) I would have liked to have seen much better agreement between the Mastan prediction and the Nastran prediction. That concerns me. Similar buckled shapes is useful but being off by 100% is troublesome. Over the holidays, I may attempt to replicate Human909's work in a student edition of Nastran. Or I may not. Kinda depending on how onerous the install is on my virtual machine and how good the skiing is.
2) I do not presently have the wherewithal to verify Mastan's output for our test case by hand. My academic training in FEM was extensive but did not venture into the development of the kind of warping torsion endowed line elements that Mastan uses. The fellow that authored the software also authored a companion book that is freely available however. Over the holidays, I may also attempt to see if I can find out a bit more about that.
3) In the quoted statement below, I described how I actually feel that something like Mastan is a more appropriate tool than something like plate/shell element FEM for determining
theoretical buckling loads. I acknowledge, however, that more sophisticated FEM, properly used, is surely the better tool for predicting real world behavior, particularly outside of the elastic range.
4) The authors of Mastan developed it specifically for the purpose of exploring stability problems as they relate to modern code provisions and stability theory. As Agent666 mentioned on several occasions, and via some interesting graphs, Mastan has been shown to be spot on in its theoretical buckling predictions for a lot of common situations, none of which are our exact situation to my knowledge. Our situation isn't really that exotic and I've no doubt that it's been played with extensively in the past. I just don't personally know where to find any examples of it.
5) Celt83 prompted an investigation earlier where we compared Mastan results to hand calculations run using Yura's proposed Cb values that were developed for beams with the same disposition of bending moments and accounting for the beneficial effect of the L-restraints. We had to modify the loading condition to approximate a uniform load rather than a point load, however, as Yura didn't have anything for a point load. I would consider Yura's method to yield something close to a theoretically correct result for the situation that I described. The Mastan number came out at 93% of the Yura value which I consider to be very close, particularly given that the uniform load had to be approximated as 7 point loads owing to Mastan limitations.
6) A good program's only as good as the person wielding it of course. It was always my hope that Agent666 would be scrutinizing my work but, by and large, I don't believe that things panned out that way. So my work has gone largely unvetted I'm afraid. If anybody else feels like taking up the mantle with Mastan and doing some QC on my work, I would absolutely welcome that. I've posted all of my models and would be happy to post them again. That said, somebody else starting from scratch might make for a more meaningful QC check.
7) I have always been concerned about how little I/we know about the character of the Nastran modelling. In my past experiences messing with full blown FEM software (ANSYS, ABAQUS, etc) there have always been
gobs of user selected pre-processing and post-processing variable involved that drastically affected the results. And I don't know what any of those are here. Moreover, my attempts to tease that information out of Human909 have gone very badly, leading to accusations that my requests are superfluous and that I've not been carrying my share of the load when it comes to research effort. More concerning, my queries have led to Human909 withdrawing from the conversation in frustration which is not healthy for either Human909 or the discussion. So I'll not be going back to that well again.
KootK said:
In my opinion, Mastan is actually a better tool than full blown FEM for what we've been trying to do in this thread. Full blown FEM surely is a more accurate representation of reality but, I would argue, reality isn't really what were trying to parse out here. Instead, it seems to me that we're mostly trying to reconcile code provisions with the underlying LTB theory that informed them. And that underlying theory was linear elastic bifurcation buckling, just what we've been doing with Mastan but with a slightly higher degree of sophistication. In this respect, I feel that full blown FEM kind of "overshoots" things in making direct comparisons to code provisions less meaningful. That said, it is of great value here to be able to use Nastran to corroborate the modes shapes and capacities predicted by Mastan. Were Nastran mode shapes wildy different that the Mastan modes shapes then I would definitely be concerned.
![[blush]](/data/assets/smilies/blush.gif "[blush] [blush]")
![[smile]](/data/assets/smilies/smile.gif "[smile] [smile]")
Tomfh and Steveh49. Could you please stop proving me wrong and making me look bad! ![[cry]](/data/assets/smilies/cry.gif "[cry] [cry]")
(But thanks, being corrected is learning.) I got halfway through my rebuttal until I realise that my investigation disproved my own arguments. What I was going to say was:
