Page 4 of 4

Re: Underestimation of velocity, compressible model

Posted: Wed Jan 05, 2022 9:22 am
by Christopher
Hello,

as announced I have done a calculation run on a single operating point of our nozzle. You can see three different curves of velocity and temperature below the nozzle tip. They are the experimental data, the calculation data with the perfect gas model and the calculation data with the stiffend gas model.

As you can see, the difference between the two gas models is marginal. However, with a mesh refinement at certain points, I have achieved a much better coverage of the simulation and experimental data. So sorry, it was not an underestimation of CS, but my fault :roll:. Still learning the secrets of CFD.

So far, I don't know how to get a better fit, especially at L = 10 mm to L = 20 mm. But when I figure it out, I will let you know :D.

Yours sincerely,
Christopher

Re: Underestimation of velocity, compressible model

Posted: Mon Jan 17, 2022 10:11 am
by Christopher
Hello everyone,

finally I have done another mesh refinement in the free jet area under the nozzle. The deviation to the measured value is getting smaller and smaller. For illustration I have attached a diagram. You can see the curve from the simulation and the experiment. For better comparison, the residual (Rc) is also given there in %. As you can see, the maximum difference is about 18 %. The average difference is about 7 %.

I spent most of my time focusing on the solver pressure, physics, reference time step, and so on. However, the most challenging parameter is the mesh size. You have to keep in mind that the maximum cell size in this case is only 1/7 of a millimeter. Accordingly, the computation time was up to two entire days.

So, a simulation can be fast, robust and accurate. You can pick two of them ;).

Yours sincerely,
Christopher