Thermal Analysis of High Power RF Coaxial Filter Using ANSYS AEDT (Two Way Coupling)

Hello, this is Ibrahim Nassar with Ozen Engineering. In this demo, we will be performing electrothermal analysis of a high-power low-pass coaxial filter.

The EM simulation will be performed using HFSS and then the thermal simulation will be performed using the Ansys optiSLang tool is not used here, instead the Ansys IcePak tool, which is a CFD solver, or all integrated inside the Ansys Electronics Desktop.

In this example, we will be setting up a two-way coupling between HFSS and IcePak to pass EM loss from HFSS to IcePak and pass back to the temperature distribution calculated by IcePak to modify the material definitions in HFSS and then recalculate the EM loss.

This approach should improve the accuracy of the electro-thermal simulation. Here is the HFSS design of a coaxial filter. This coaxial filter was generated using the SynMatrix tool.

We will be using in this demo 2025 R1 and here is the response of the filter S11 and S21 which works as you see up to 5 GHz. Okay, let's expand this with the 3D Modeler window. Okay, in order to enable the feedback from the IcePak solver, we need to do basically two things.

The first thing is that we need to do is to modify the material properties to have a thermal modifier. To do that, we select the aluminum material which is assigned to the coaxial inner part.

So by right-clicking on Aluminum and select Properties, then View Edit, we need to check this box for the thermal modifier. And here you select what properties need to change with this thermal modifier and we can click here and select Edit.

So we can use this expression and you can modify it if you want. You can use a quadratic form where you define the basic coefficients or the advanced one where you define the other properties here.

You can also use a temperature-dependent dataset where you import it and assign it to this material if you have it from measurements or you got it from a manufacturer. Okay, close this window. So we need to have a thermal modifier to adjust the bulk conductivity of aluminum. And we hit OK.

The second thing that we need to change is the Teflon which is the inner dielectric of this coax. So if I right click on Teflon properties and similarly view edit material we make sure that the thermal modifier is checked.

And here we can have the thermal modifier to modify the relative permittivity of the Teflon, and we can also set it up to change the dielectric. Let's expand this window, the dielectric loss tangent of the Teflon. And let's use the default expression here.

Okay, the second thing that we need to do is to set the object, the temperature, the object temperatures.

By right-clicking on the design and selecting object temperature, we need to check this box to include the temperature dependence and we also need to check the box to enable the feedback from IcePak. Here you can change the temperature if you need or we can just leave it at 22 degrees Celsius.

And we hit OK. Then let's solve the HFSS design again since we modified the material so the solutions got invalidated. Okay, so it's doing the frequency sweep now, it's almost done. And now we have the results.

So if we go to HFSS Fields Calculator, we can calculate the volume loss density and the surface loss density. and those are the losses that will be passed into the IcePak solver.

So we need to do the volume loss density over the Teflon material, which is the volume of outer dye, 2K, and we integrate an eval, and this is basically the volume loss density.

If we clear that, now we select the surface loss density, select the geometry, the surfaces of the cylinder 8 which is the aluminum, hit OK, integrate, eval, and this is the surface loss density calculated by HFSS. So we will see that those numbers will be passed into IcePak.

You can also adjust the input power to this model so here it is just assigned to port 1 to be 1W but you can change that to an adjusted calculation. And this is just a post step. Now we need to create the IcePak design and simulate it.

To do that we will use the automated workflow and the latest release of Ansys. By right-clicking on the design and selecting Create Target Design, the target design will be IcePak. We want to pass this last adaptive solution from HFSS.

If we go to the IcePak tab, we will select force convection and let's have a flow speed of 1 along the positive Z axis. And we hit OK. So now the IcePak design got created and we can see that the EM losses got mapped properly. And we illustrated how to map this in another demo.

So here we will just go over the design settings here. So opening one of the other thermal boundaries that we have here is an opening one from the top and opening two from the bottom.

So if we open the one in the bottom, so it's an inlet, so there's a velocity of one as we selected when we created the IcePak target design. and the one in the top is a pressure with ambient pressure and we hit OK.

We can now generate the mesh by right click on mesh and select generate mesh and now we can view the mesh by over the cut plane here or we can select the geometry and select the object and we see the mesh. Okay, now let's double check the setup that got created.

So it's temperature, flow, and let's do the flow regime to be turbulent. And let's do the radiation model to be discrete, ordinate, and we want to include the gravity.

Convergence, let's keep the defaults and here yes, number of iterations can be adjusted, so let's leave it to the default of 100. Here's the solver settings, so the Z velocity is 1. Radiation, let's keep the defaults and hit OK. And now, this is set for one-way simulation.

So now to make a two-way coupling here, Here you can adjust the number of iterations. The number of iterations here means that the EM loss will be calculated by HFSS passed to IcePak. IcePak will be used to simulate these losses and calculate the temperature distribution. That's an iteration.

Second iteration will be that the temperature will be passed back into HFSS and HFSS will redo the calculation of the EM losses with the adjusted temperature and then pass back again to IcePak the EM losses. So let's just to have it quickly done here.

So let's just keep the number of coupling iterations to be 2 and the maximum IcePak iterations per coupling. This will also reduce it and keep it to 20 and we hit OK. So now we have everything done so we can just right click on analysis and say analyze all.

Okay, so now the simulation is done, so we can right click on Setup 1 and look at the residual. And so this is the residual over the number of iterations that were done. And if we go to Profile, we can see that the EM loss that is passed is very similar to what we calculated in HFSS.

And you can see that it has been calculated multiple times because we have requested two iterations. So this was the first one time and this is coupling iteration number two. So we see that the trials have been calculated another time.

Okay, it's closed now we can plot the temperature by selecting the geometries and right click and select plot fields temperature and let's plot it on the surfaces only and this is the temperature distribution. We can also plot the velocity and let's plot it along one of the planes.

So let's select the XC plane and right click on the 3D Modeler, select Plot Fields, Velocity, and let's plot the velocity vectors. And let's keep the defaults. We can also go to HFSS and look at the profile and verify that there was a temperature feedback.

So if we open the profile, we see here that the adaptive meshing has been repeated with a feedback temperature from IcePak. So that's all for this demo and thank you for watching.