# Does way of applying load step effect FEA results

• 153 Views
• Last Post 08 September 2018
• Topic Is Solved
Rashi posted this 29 August 2018

Hi again,

I'm trying to simulate effect of inertia and thermal expansion of an impeller which is connected to a shaft by a bolt under pre-tension.

I have used 1/6th of the model to reduce the simulation time.

For same exact settings(I have copied the simulation with method 1 and only changed the load step for method 2) I have done two simulations with two different load step methods as mentioned below,

Method 1,

step 1- Room temperature, no rotation, pre tension load applied.

step 2- T degree C, X krpm, pre tension load locked.

Method 2,

step 1- Room temperature, no rotation, pre tension load applied.

step 2- Room temperature, X krpm, pre tension load locked.

step 3- T degree C, X krpm, pre tension load locked.

The issue is two methods give two different radial displacement/maximum stress for the impeller at similar load step for the two different methods.(Method 1, step 2 and Method 2, step 3)

So my question is does the way of applying the load step affect the stress results?

jjdoyle posted this 29 August 2018

Does your model have path dependent nonlinearities in it?

i..e Large deformation, path dependent plasticity or nonlinear path dependent frictional contact .

Try plotting the force vs deflection curve of each method to see how they differ.

Rashi posted this 30 August 2018

Dear Sir,

The component has around 0.15 mm deformation. I have turned on large deformations also. Also it contains frictional contact. I do not know whats nonlinear path dependent frictional contact means.

I will check the force vs deflection curve and share there results.

Thank you.

peteroznewman posted this 30 August 2018

Dear Rashi,

Frictional contact is an example of a path dependent nonlinearity.

For example, if a thermal expansion was followed by a bolt pretension to engage the frictional contact surface, that creates a different result than a bolt pretension followed by a thermal expansion.

When a model has only linear effects in the model, the order of application of loads doesn't matter and the result is the same. The order of application of loads over time is considered the "path" to the final load condition.

Regards,

Peter

• Liked by
Rashi posted this 08 September 2018

Dear @Peter Sir, Thank you for the explanation.