Ansys Discovery

Starting from the workbench under Component Systems, we’ll call on the Discovery application. Launch that, and there are a couple ways we can import. The simplest I find is to drag and drop, so we’ll just pull the hot msh file into drag to open the msh file. Loaded on the bottom right under Advanced Selections, we’ll see that our name selections have come in. So the Hotol interface, Hot Default, Hot Outlet at the bottom, and Hot Inlet at the top. So go ahead and start to assign our boundary conditions. We’ll do flow inlet. We’ll say mass flow rate 0.25 kg/s. Do the same thing for the outlet.

It’s just a few button clicks. We have our model set up. We could come in and change the material if we want to. We can add temperature to the inlets, and in a couple releases to come, hopefully, we can do conjugate heat transfer analysis with faceted bodies. We’ll extract the monitors that we want to look at and click play.

So pretty quickly, we get our results. We can see our Max Velocity, Pressure Drop. We can change between different Contours, so we can look at the Iso Surface. We can look at the Contours, turn off the streamlines, add particles, and so on and so forth. Last thing I want to call out is the Fidelity Slider at the bottom right. Now I just said it’s its default. I’m going to put it closer to its highest Fidelity or most accurate. And if you’re unaware, the slider will always be available. What changes, however, is the voxal size that is available, and that voxal size is determined by the power of your graphics card. So if you run an instance of Discovery on a laptop with Graphics Card A and run an instance on a desktop that has a superior graphics card, you’ll get a more accurate result with the better graphics card. That being said, the trend that you see regardless of your graphics card is going to be accurate, so you can still use it for rapid design iterations. The more high-powered your graphics card, the closer it’s going to match Fluent and real-world testing as well. In some of the cases that we’ve done, it’s been within about 5% of results from Fluent, so it took quite a bit longer to solve. But we can see that the Max Velocity and Pressure Differential did change.

Looking at the flow distribution, we can see that we could improve the plenum and flow distribution into the core quite a bit more. We have a bit of recirculation in the top chamber, and in the top quarter or so of the TPMS structure, it looks like we have a little bit of a dead zone. So there are some strategies that we can use, and maybe I’ll create another course on how to improve the performance by changing these types of characteristics. But for now, we’ll continue on to Fluent.