Learning Forum

[Giovanni Deilura]

Hi Everyone,

I'm doing a pretty easy thermal study. I have a steady state 2D axisymmetric model where i have 2 different bodies, a first body is a rectangular region which stands for a copper coil, the second body is a plastic core in which the coil is embedded. You can imagine these two regions to be to modeled as one rectangle inside the other. Then, as the model is set up to be 2D axisymmetric, we will have a sort of ring, made by an internal copper ring which is embedded in a plastic ring. In the copper body i set up an internal heat generation according to the Joule losses I have in the coil, then i have some convection coefficient between the plastic core and the environment.

Let's say i have 10W of Joule losses. Then I have to divide these 10W by the entire volume of copper and put the number in the form of W/mm3 in the internal heat generation condition of the 2D rectangular region that models the copper ring.

What i don't understand is that in the results i plot the heat fluxes, which are in the form of W/mm2, and if I sum all the products of the heat flux times the correspondent surface i don't get the initial 10W heat source i set. Same thing happens with the reaction probes. This is strange as the study is steady state and the energy balance must be guaranteed

 

thanks to everyone open to help ????

[dlooman]

So you're doing a surface integral of the flux?  I tried this analysis on a unit cube with convection to ambient and the reaction probe correctly reported 10W.  Then I applied 10W/meter^3 to a solid cylinder with radius 14.4meters and height of 17.7meters (volume  = 10,898 meter^3).  The reaction probe reported a heat flow of 108,910 W so pretty close.  It seems a Mechanical reaction probe works correctly for an axisymmetric model.  My boundary condition is convection.  You might want to stay in MKS units to make the calculations easier.