Include Page | ||||
---|---|---|---|---|
|
Include Page | ||||
---|---|---|---|---|
|
...
Author: Benjamin Mullen, Cornell University
...
Verification & Validation
Verification
Adapt the Mesh
In order to test our simulation for convergence, we will refine the mesh. Refining the mesh will allow use to make sure that the results we are calculating are independent of the mesh. However, instead of refining the mesh everywhere (which would be wasteful, as most of the area of the domain far away from the shock has constant values), we will use our results to refine our mesh. Specifically, we are going to use the gradient of the pressure to determine where to refine the mesh. First, let's take a look at our mesh. In the Outline window, select Graphics and Animations, under _Graphics, select Mesh, then press Setup. Select all of the surfaces (except y=0.35) and press Display. This will display the current mesh.
...
The most striking difference between the two results is the thickness of the shock. Notice that for the refined mesh, the shock is less thick that for the original mesh. This shows that the refined mesh is converging towards the real case.
Comparison to Analytical Solution
In order to verify our simulation, we need to compare our results to either an analytical solution or an experiment. Below is a table comparing the values from the simulation with the calculations from the pre-analysis.
...
As we can see from the table, we are getting fairly good matching between the computation and analytical approaches. From this we can build our trust in our simulation.