Discussion about the Pressure Probe Location and Pressure Oscillation - mDBC & DualSPHysics v5.0

edited August 2020 in DualSPHysics v5.0

Hi Everyone,

In DualSPHysicsv5.0\examples\mdbc\04Dambreak, the model is based on [Kleefsman, K.M.T.; Fekken, G.; Veldman, A.E.P.; Iwanowski, B.; Buchner, B. A volume-of-fluid based simulation method for wave impact problems. J. Comput. Phys. 2005206, 363–393.] There are 4 pressure sensors on the obstacle side towards the reservoir, locations as shown below P1, P2, P3, P4:

The exact locations of the P1, P2, P3 and P4 sensors are: (0.8245, 0.471, 0.021), (0.8245, 0.471, 0.061),(0.8245, 0.471, 0.101), (0.8245, 0.471, 0.141). In the xml file setting, the location of the obstacle, boundary particles and the pressure probes are shown as below: (please ignore the scale of DP/2 and Dp shown in the figure)

You will see in the figure, the pressure probes are located inside of the boundary surface, and the tank bottom level starts from Dp/2=0.005m actually, then the height of the 4 pressure sensors should minus this 0.005m. So the questions will be:

  1. In the simulation, why the pressure sensors are located inside the boundary?
  2. In the simulations of applying mDBC, where should the pressure probes be located, if the pressure sensors are just on the boundary surface?
  3. The height of the pressure probes are 0.005m lower when compored with the paper, so if I want to keep consistent with the paper, the location of the pressure probe should be increased by 0.005m? Which means the height should be 0.026m, 0.066m, 0.106m and 0.146m?

After running the example case, the pressure filed data in the first 2.0 second for the P1-P4 is shown as below, (did not show the peak value above 16kPa, these could be removed from the results, as too high)

Compare with the data in the paper, P1 and P3, even I did not plot them together, it can be seen that the value generally coincided well, but there are still some tiny oscillations in the curve, so the question will be:


  1. how to eliminate the pressure oscillation and the high peak value, by increasing the number of particles, or change the location of the pressure probes?

I have not tried adjusting the parameters, which will take long time on my laptop. Post the questions here, hope to get some advice first. Thank you all!

Comments

  • Pressure sensors should be always at the boundary limit, which is at dp/2 of the outer layer of boundary particles.

    In this case, we try to follow the experimental position of the pressure sensors, but we also considering our reference system....

    For example upper layer of boundary particles may be created at z=0 m therefore our bottom boundary limit is at z=0+dp/2 instead of z=0... that will explain the difference in z positions.... similar to X...

    So what I mean is that you have to consider the experimental locations but also the numerical domain we are creating here.... In any case working with dp=0.01m so that dp/2=0.005m, the differences are very small.


    Regards


  • Hi Alex, thank you very much for your clear explanation! Then I see, in my simulation, I will try to locate the pressure probe on the boundary limit. Yes, you are right, when change the location of the pressure probe by 0.005m, the results remained almost the same.

    By the way, do you have any suggestion on eliminating the tiny oscillation on the pressure curve? Thank you again!

Sign In or Register to comment.