Heatflux goes beyond the axis of an axisymmetric thermal FE model

[lookforward]

Hello, Colleagues.

I was building a transient thermal model. It was a 2-D axissymetric model. I defined the initial temperature and frictional heat on an inner surface as the heating source. The results looked good except for the heat fluxes of several nodes along the inner frictional surface has magnitude along the directions beyond the center axis of the model.

I heard in axisymmetric model, the center axis is considered as insulated by default. No heat fluxes should go to the direction beyond the center axis, which means all the heat fluxes around the center axis should go toward insider the geometry rather than outside of the geometry. If this is correct, my results must have errors. Why? Is it to do with boundary condition or thermal load application?

 

[corus]

I'm not sure how you can apply a frictional heat source on an inner surface unless the inner surface isn't at R=0. If it is hollow then are you sre it's not just that the length of the arrows cross the line R=0. Shorten them in the view options if that's the case.

Tara

http://tinyurl.com/4ydjg7m

 

[lookforward]

Thanks Corus.

I apply frictional heat source in the format of a mathematical function throughout the radius at the inner surface including R=0. The function gives zero heat when R=0. This model is not hollow. I modeled a pin and plate friction test rig. I built all parts in one model and differentiated the different components by assignning different material parameters to them, which including polymer pin, metal plate, fluid lubricant, steel bath containner etc. The inner surface for frictional heat is pin and plate surface.

The boundary conditions include:
1. a pre-definied filed to define the initial temperature of the model under room temperature
2. a thermal load applied by "concentrated heat" option. This thermal load is equal to a dynamic/transient/time-dependent frictional heat and is proportinal to the pin radius
3. three outter edges but the axissymetric axis was constrained at room temperature since these edges represent that the outsurface of the test rig is around by air. (the model is a four-side square although different inside componenets.)

You are making a good point about the arrow lengths. I will check the arrow length. But I am sure the heat flux at R=0 and another node near R=0 are big and the vectors (both length and arrow) at these two nodes goes beyond the axissymetric axis.

What I can do?

 

[corus]

If you've assigned the heat flux to be zero at R=0 then it will be zero, unless you've made an error. Check the values using Tools/Query/Probe values

Tara

http://tinyurl.com/4ydjg7m

 

[lookforward]

I checked the heat flux at R=0 in my model, it was not equal to zero although I have assigned heat flux to zero in a function defined in "Fields"-"Analytical fields". I don't know where might go wrong to cause the inconsistent input/output.

Shall I try to add additional constrains, say tie the node at the thermal load surface with a R=0 with the other nodes with the same R=0 but far off the thermal load surface?