Two step thermal analysis

[Anonymous]

Hi!

Im trying to do a two step thermal stress analysis,  im using the analysis type 'non-linear static' and im applying a 'body temperature' of 370K in 'subcase 1'. Assigning expansion parameters (on material card) to only one section part of the bolt (and also only in one direction) should cause thermal expansion and thus thermal stress. In 'subcase 2' im applying the same 'body temperature' as in 'subcase 1' and in addition a pressure load.

The 'body temperature' should work as a preload, however the registered forces from the preload does not match the expected preload using F/A=alpha*deltaT*E.

Ideas? (The reference temperature is set to 293,15K on material card)

[John_Holtz · Accepted solution]

Hi @Anonymous

First, I think you are using Nastran In-CAD (or the stand alone Nastran solver with some other pre-processor). You do not appear to be using Simulation Mechanical. Is that true?

Your formula for the load versus contraction is correct, but it assumes that the bodies that you are bolting are infinitely rigid. If the bodies are elastic and can compress, then the contraction will create less stress than calculated by your formula. (Another way to think of it is this. If your material has 0 strength, the bolt would be free to contract and would create a stress of 0.)

The required contraction (temperature change) needs to be larger than you calculated. It changes with different arrangements, so you need to adjust the temperature change appropriately to get the desired preload for each model.

______________________________________________________________

If my post answers your question, please click the "Accept as Solution" button. This helps everyone find answers more quickly!

[Anonymous]

Thanks for your reply John!

The relaxation explaines the low contact force, I want to know the total contact force on the surface of the head of the bolt. Ive managed to group the nodes included, however im unsure of which result data to extract, the application is sealing the contact surfaces shut. Is the "SSHL CONTACT NORMAL FORCE" the best option? The "SSHL CONTACT PRESSURE" is also calculated using node data, however how can a pressure be computed using node data?

/Filip Andersson

[John_Holtz]

Hi,

You can use which ever contact result is available. (Sometimes I see something like "Contact > Force", and sometimes I see "SSHL Contact > Force". I do not know why one results appears at times but not the other result.)

A pressure can be calculated from the force: P = F/A where A is the contributing area from the surrounding elements shared by the node. It is the reverse process of when a pressure is applied to the model, the solver calculates F=P*A and applies a force to the nodes.

[Anonymous]

Hi John!

Im having a split solid with bonded contact in compression, however when im picking the nodes at any of the contact surfaces the contact normal force is always 0. The stressfield seems to be OK but the contact forces seems to be lost or ignored?

kind regards

Filip