Problems with Static Linear Analysis when applying thermal load

[Srki1982]

Hi everyone,

I'm trying to understand where the problem is in my model because of the result I receive.
1. First I apply thermal load 10 to 100 Celsius on my model in order to see the thermal distrubution with steady-state heat transfer analysis.
2. Second analysis is linear static analysis with all my static loads. The result is good. I receive approx. 100 MPa as a maximum value.
3. Third analysis is when I additionally apply thermal load on all nodes by command: Model -> Load -> From Output. When the analysis is finnished I receive almost 500 MPa as a maximum value.

I think I have correct material data input in femap (Stainless steel EN 1.4404, according to EN 10088-2) (I work with mm):

Young's Modulus: 200000
Poisson's Ratio: 0.3
Thermal Conductivity: 15
Thermal Expansion: 16 x 10-6
Specific Heat: 500000000
Mass Density: 8 x 10-9
Translational Accel./ Gravity: 9810
Default temperature: 20

Is there any other place I need to check in order to obtain correct result? Is there any additional check in analysis setup window or somewhere else?

Thank you in advance

 

[BlasMolero]

Hello!,
• Thermal Conductivity of steel is 15 W/mºC = 0.015 W/mmºC, then your values are wrong, not consistent.
• Check if "over-constraining" exist, it you do not allow thermal expansion in your structure, then thermal stress would be quite higher.

Best regards,
Blas.

~~~~~~~~~~~~~~~~~~~~~~
Blas Molero Hidalgo
Ingeniero Industrial
Director

IBERISA
48011 BILBAO (SPAIN)
WEB:

http://www.iberisa.com

Blog de FEMAP & NX Nastran:

http://iberisa.wordpress.com/

 

[Srki1982]

Thanx BlasMolero,

I have tried that value but it doesn't change anything. The values are the same. The model is not overconstraint, it is only constraint on two smal plate parts which symbolize a connection plates for swivel castors (Frame construction with four castors).

 

[yellowflight]

Hi,

The thermal-conductivity is not an issue in your calculation as there is no thermal flow by convection for example.
So correct that this does not not change your data.
Maybe you can make a rough check by removing the heat-data and replace this by an average thermal bodyload.
The whole constuction is set on that temperature then.

Paul
Yellowflight.com