Discussion Groups

Simulation Mechanical and Multiphysics

New Member
Posts: 2
Registered: ‎03-21-2013

Stress analysis of steel elements.

321 Views, 1 Replies
07-23-2013 01:45 PM

Hello mates,


I'm new in Autodesk Simulation and I want to make a quick consult if it is possible, because this are my first attempt.

In the moment I work on my master's graduate project - Analysis of Stadium Roof Structure.

So I have two very critical steel element, that i need to analysis - stress and crack/fracture. I make the elements in AutoCAD -> Fusion -> Multiphysics 2013.


Some information for the roof. It almost the same as the roof on Green Point Stadium (at Cape Town, South Africa).




From outside is beam compression ring from big box sections 1,2m x 2,2m and length from 10 to 11,50m.

The inter tension ring is from 8 locked-coil strands, 98mm in diameter. The radial cables are the same.


So.. for element One: parts of two box sections with structural bearing and "ears" for the cable and the steel farm.


Original section :




Made by me.



I define the the light green and the darker purple parts as Plates, all other as Brick.

Default contact type: Welded. Contact between the two plates is Surface. Material: AISI type H13 Hot Work tool Steel. (No European format's or I just can't find them)

I have not define the holes for the bolts in this model. Is this wrong or not?


I define fixed supports/constrains to this faces :



The results are :




I think that I have some issue because is not normal to NOT have any forces in the right box section.

This is for the element One.


Some comments and recommendations?







Element two is more tricky, because is only supported from the cables.


Original element :



My element :




Here the contact is Bonded, same steel AISI type H13.

The Forces will be applied to the local axes on showed in my model whit vector/custom direction.


1st. Can the element be solved without supports/constrains ?

2nd How to define the forces from the cables on the element ?

3th Is it right to define Gravity load to this element ?

4th the cables are been additional hold and from mini steel plates hold with bols, as you can see on the original photos. Do I need to model them too?


Just use it.
Please use plain text.
*Expert Elite*
Posts: 561
Registered: ‎08-30-2012

Re: Stress analysis of steel elements.

07-29-2013 06:57 AM in reply to: wrbst

Hi wrbst,


For element one, are you sure that Plate elements are the proper choice for the light green and purple parts? Considering that everything else is brick elements, my guess is that the CAD model was created with all parts being solid. See the documentation page "Model Mesh Settings" to learn if plate elements is the proper choice.


Also, the documentation says "Automatic contact between plate parts is not supported" in the section about Surface contact. So your model does not have contact between the two plate parts, and therefore the two plate parts are not connected together. So the load cannot be transferred from the one side with stress to the other side without stress. See the page "Types of Contact" if you want to get into the details.


For element two,


1st, the model cannot be solved without constraints if the analysis type is linear static stress. Even if the analysis type is MES (which can solve models that are statically unstable), it is unlikely that the sum of the forces equals 0 out to the precision of 15 or 16 decimal places. So without constraints, the model tries to accelerate. So please add some constraints to the model. For example, one set of cables is "constraining" the motion caused by all of the other cables.


2nd,  if you know what the forces are in the cable, you can apply those forces to the surface (much like you have shown in your figure with the arrows). If the forces are unknown, then you could use MES and model the cables.


3rd, this is just a guess, but the weight of the model is probably small compared to the forces in the cable. So gravity is probably negligible.


4th, you will have to decide whether those components are necessary or critical to the analysis or not.


John Holtz, PE
Mechanical Engineer
Pittsburgh, PA

Simulation Mechanical user since Dec 1997
Please use plain text.