Huge Difference in ETABS vs Tekal Structural Designer 3

[Al Jazari]

I have been using ETABS for ages, and the results are always books perfect. I was introduced to Tekla Structural Designer, i like the interface and many aspects about it, but one thing is amazing me, The huge difference in deflections between ETABS and TSD, am i doing something wrong ? or is TSD getting it completely wrong ?

both the models are here. i have verified forces again and again but the difference is huge in both software.
tekla : 1.8in
ETABS : .14in

Link

ETABS

 

[Trenno]

Having not downloaded the models, what is the actual discrepancy here? Column loads? Beam moments? Slab deflections?

 

[Al Jazari]

Its written on diagrams,

floor/story displacements

 

[Agent666]

Is it just this model or tested over simple test cases and seeing the same inconsistencies?

For example can you show the same differences with a simple vertical cantilever where its easier to check things are 100% the same?

Even in your current model, replacing all the members with new ones of another size with similar stiffness reductions, material properties, etc, could highlight some error in the modelling that is causing this.

Being out by a factor of 10 could be an extra zero creeping into your young modulus value. I've seen people be out by a factor of 1000 before and never realised.

 

[Al Jazari]

yes i tried few buildings, the inconsistency is same

 

[Agent666]

Even simple contrived test cases?

CSI used to publish a whole series of test cases in their documentation perhaps find an appropriate case to check and then contact your software vendor if you're seeing the same significant differences.

Also try checking the fundamental period by hand using the Rayleigh method. This will highlight which program (or model) is apparently in error.

One other thing that can affect things, is how modelling of the floors using shell elements can arbitrarily stiffen the beams. In etabs you can get around this using membrane elements for the floors. One test to see if you are getting this occuring is to take a simple 2D frame only model and compare. I'm not really familiar with tekla, so not sure how you model floors and whether it is susceptible to this effect.

 

[jhardy1]

I haven't checked either of your models, but I would suggest that a seismic drift displacement of just 0.14 inches (3.6 mm) sounds unrealistically small for any sort of "normal" 4-storey building; 1.8 inches (46 mm) sounds far more plausible as a design displacement.

http://julianh72.blogspot.com

 

[Al Jazari]

its a theoretical model not a practical building with oversized members,

 

[jhardy1]

It's probably easier to debug your models and applications if you use "realistic" loads and member sizes, because at least then you should have some sense of which answer is closer to the "right" answer.

Are you using equivalent beam member formulations in both programs? (e.g. is one program accounting for shear displacements, while the other is not?)

http://julianh72.blogspot.com

 

[Al Jazari]

This thread gets me soul searching advisory answers,

i am 17 years in this business, i know what to do , the problem right now is mismatch in two softwares, doesnt matter practical models or not ! it is not a problem of typology, i can model most nonsense structures, but the RESULTS SHOULD MATCH !!!

practical help would be open the models, review them and get to the bottom of it !

 

[Agent666]

WELL, not all of us, have access to both software packages to compare. You are likely the only one who does who is participating, and the hunch is that the tekla model is the one with the issue (which I've never used).

I've had a few suggestions to 'hand' check some simple scenarios but you have not responded to the points I made.

Everyone makes mistakes (yes including you), my suggestions are trying to determine if it's something obvious and/or silly you may have done, or if it truly is a software issue. Both are possibilities.

 

[klaus.]

I do not know both of these programs
But in that case I would compare first with a very simple model where the solution is known by hand calculation

 

[Trenno]

Don't get defensive when you provide next to no information in your question, a few screenshots of deflected shapes wouldn't go astray either. Ask stupid questions, get stupid answers...

Absolutely agree with Agent666, when working on a complex analysis model, I will often create a separate one to double check the software is doing what I want it to in the complex model. I've not used Tekla before, but there is a reason why people don't use it for tall buildings - from what I understand it's not a true/pure FEA analysis package.

 

[Al Jazari]

@Klaus : Good suggestion, did it and ETABS is correct.

@Trenno
It is fairly straight forward question,
Complete information is there, even the models are there , as for deflected shape you ask for ETABS does not make a deflected shape diagram on model with VALUES, the above plot i posted is the deflected shape with values. It is not a stupid question is fairly simple, software A and Software B results don't match. WHY ? they don't match is the question. IF you don't have the software then i understand you cannot answer the question. As for Complex, this is not a complex model i have evaluated values and checked manually ! and ETABS values are correct.
Your understanding is wrong! and so is your assumption, people are testing it here in Dubai for tall buildings, TEKLA is not stupid to launch such a platform when they know their competitors.
FYI it has both FEA and Linear DUN11 analysis capabilities i would suggest you download a trial version and try it !

 

[Trenno]

This is fairly pointless argument. There's a reason why ETABS is used for complex/tall buildings and Tekla is not. Horses for courses. Trimble know their analysis engine is nowhere near as powerful as CSI's. In our office, anytime you need to depart from the standard low-rise rectangular box building, Tekla is useless as you need to manipulate the analysis model to get it do what you want and/or expect.

As a side note: ETABS being more of a true/pure FEA program, displacements are reported at nodes. Add additional nodes along your frame members to get additional displacement values.

At the end of the day, you need to use engineering judgement to satisfy yourself the structure is sound.

 

[Agent666]

Can you post the results of your simple test?

Have you contacted tekla and asked them to investigate?

 

[Al Jazari]

Yes i did and i got the answer.

 

[IRstuff]

And what did they say?

I read from their website that TSD creates an analytical model; which raises a bunch of questions, since not only do the material properties have to be correct, the analytical model needs to accurate reflect the true mechanical and structural reality of the design. It apparently can be good enough to design big buildings, but that would seem to be highly dependent on much the user understands the exact modeling being performed.

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529 Entire Forum list

http://www.eng-tips.com/forumlist.cfm

 

[Al Jazari]

They explained it that it is because of modification factors in setting, I am trying to work my way for a solution , i ll post it for every one, once solved.

 

[Al Jazari]

So here is the solution to this problem for anyone else starting off with Tekla Structural Designer,

First of all, TSD is a not new in the game of structural analysis softwares, it is evolved from a very respectable old software Orion which was being maintained CSC, after buyout by Tekla, it has been upgraded to a higher level, and personally i find it at par with ETABS.
So through Tekla Help,( which was very responsive and helpful in this matter ). Problem was with Modification factors that are set by default

When i set all the factors to 1 the floor displacement matches, i did a simple structure as Agent666 suggested. A four story single bay structure with nodal load of 10 Kips. This time i modeled in STAADpro. and here are the results

and the results in TSD :

One more thing : if you observe the first picture i posted the Supports were Pinned, this time they were Fixed

THANKYOU all in participating.