How accurate is Staad finite element analysis.In case of water
tank resting on soil, can we give springs only KFY value, not
releasing MX, MZ.e.g.
supports
1 to 200 KFY 24000
We get very less moments in bottom slab.
Please respond.
Everybody...do you notice that on Eng-Tips we get a lot of STAAD questions? But very few of other software vendors...is this because there are many more STAAD users or is it the program?
JAE...I think it is both. STAAD has been popular despite its numerous shortcomings. I think a lot of people have made the investment and don't want to throw that away, so they stick with a program that they at least feel comfortable that they know many of its problems. Kind of "the devil you know" syndrome.
Having said that, we can see from the numerous posts that many are just finding out how error prone this software can be. You would think that Research Engineers would take to heart some of the feedback they have gotten over the years. In my experience, they view the problems to be mostly with the user, not the software.
It has to be
FIXED BUT MX MY KFZ 12000.
because if you do not allow rotation about X and Y axis then
you should have all nodes prevented from rotaions and you may have
only one level of displaced nodes.
I know this does not answer your question but i see from the other responses that they have had similar experiences with Staad that i have had. They blame all error or inconsistencies on the user. This is the case even after i have shown them programming bugs that arise.
I've invested the money and don't care to spend more for another engineering design system.
Dear 1161,
Staad will not prevent you from defining springs as
FIXED BUT KFY 12000
..BUT, it is not realistic and does not represent reality.
in my view, this is wrong. it has to be
FIXED BUT MX MY KFZ 12000
in order to represent plate supported on springs.
Dear nades, I think what 1161 says about the soil spring to be KFY is correct, as staad considers y direction as the normal gravity direction,the direction in which all soils support the structures resting on them.
STAAD is the most popular structural analysis program because it is much cheaper than the more comprehensive (and more accurate) programs such as STRUDL or ANSYS.
If you are performing an analysis which requires the accuracy of finite elements, it would be best to use a package better than STAAD. STAAD has a very limited library of elements and provides very poor description of its available elements and their strength and limitations. I have known several engineers who have used STAAD for finite element analysis and had disasterous results. Also, the user support services for STAAD is very poor compared to other products if you run into problems.
For basic structural analysis STAAD is probably all you need. It is easy to use and relatively accurate. However, you would be fooling yourself to think it does advanced analysis properly. It will give you an answer but the answer may be wrong.
I am a registered user of STAAD. It does have several shortcomings,
however, if you stick to "standard" problems with finite elements, you
can get good results. I have, for instance checked the plates against
an analytical solution for vertical walls with varous loadings and found
STAAD to be accurate.
As for MATS or TANK BOTTOMS on the ground you need to use:
(Joint No) FIXED BUT FX FZ MX MY MZ KFY XX.XXX.
You must release all the other DOF's or you'll induce bogus membrane
stresses, and it will affect your moment results(this is what you're after,
I presume). Under no circumstances do you want to fix moments.
Since you have now left all the other DOF's free in all the
joints, you will induce instabilities in the solution unless you FIX
FX and FZ at one of the joints in your mat and FX OR FZ(not both)
in some other joint in your mat, away from the first. This will give
you accurate moments(not very sensitive to subgrade modulus, by the way)
and a stable solution. If you are interested in lateral soil resistance, then
use lateral SPRINGS, do not FIX.
wmccain, how do you evaluate your results with STAAD? What is the baseline that enables you to say you've got good results?
In my experience finite elements (true finite elements)must be evaluated for several factors to encompass the behavior or response to the change in those variables. Not the least of which is the element type itself. So, in my interpretation, you cannot just run one file and consider the matter complete. And since the heart of the finite element method is a numerical approximation there is an inherent amount error in the analysis - how does STAAD report this error or even better how does STAAD minimize this error?
wmmcain, I would also like to know how many engineers out there ever typically deal with "standard" designs. It seems that with so many owners altering or adding onto their existing buildings, there is not a lot of standard problems out there.
Thanks, born2build. I've never seen so much whingeing
over one word. Maybe I should have said "flat plate
related problems." I do not used STAAD for tubulars,
tubular junctions, or multiple curvature shells. I do
think, however, and I think the STAAD verification problems
bear me out, that mats on Winkler foundations, rectangular
tank walls, and other "flat" problems with rectangular
elements can be solved accurately assuming the user's
boundary conditions are correct. On this last point see
nades, above and my response. I am not defending STAAD.
If anyone has any further problems about the use of the word "standard", then let them scream and yell to their
hearts' content.
Gee, if I had a verification problem and I posted/published the results I guess I would make sure the problem was such that any comparison looked favorably on me too.
However, Gentlemen, we are asking you to step out into the unknown, to really apply the product to engineering problems and not textbook examples. In this area, I ask you, are you completely satisfied with the STAAD product?
I have used STAAD for a number of years and have always felt that the problems I analyzed were appropriate for the software. In that respect, I got exactly what I needed from STAAD. But I will not perform complex dynamic analysis or FEA with the program. Which that program is marketed to do but doesn't live up to those expectations.