Validation for wave propagation

ShadiShadi
in DualSPHysics v5.0

Dear Dualsphysics user,

I am using Dualsphysics to simulate a subaerial landslide wave test. The water height a a specific point (wave Probe 1)has been compared with the experimental test. Initially, I observed a satisfactory agreement between the simulation and the experiment for the first wave. However, I could not track the subsequent waves!

The comparison graph and physical model have been attached.


comparison.png

I would appreciate any guidance or suggestions you can offer to address this discrepancies.

Best,

Shadi

Comments

  • AlexAlex

    Are you completely sure that your numerical setup is the same as in the experiments?

    Can you share with us your XML file?

  • ShadiShadi

    Hello Alex ,

    Sure, here is the numerical set up for the simulation.

    Best regards,

  • AlexAlex

    Please send the xml with properties of "acrylic" as well

  • ShadiShadi

    Thanks Alex,

    Here is the file including acrylic properties:

    Best,

  • AlexAlex

    Dear Shadi


    Here my suggestions:


    1) do not use <masspart value="distributed_bodymass" />

    no need, the option by default of same mass of floating particles tha fluid particle should work properly


    2) you should impose the mass of the object, instead of relativeweight

    that will be more accurate

    do you know the 3-D actual mass of the object?


    3) the width of the tank so that distance to lateral walls should be very important here

    have you double checked if the experimental and numerical width are the same


    4) can you share the reference paper of the experiemnt?


    Alex

  • ShadiShadi
    https://forums.dual.sphysics.org/discussion/comment/5850#Comment_5850

    Hello Alex,

    Thank you so much for your suggestions.

    I removed the " <masspart value="distributed_bodymass" />" and calculated the mass of object using its volume and density.

    The width of object is 0.3 m, so the volume would be 0.02106 m^3 . therefore, the body mass = 1200*0.02106=25.27 kg.

    However, despite these changes ,the results did not improve.

    image.png

    Here is the XML file and the reference that I used for validation.

    Any suggestions would be appreciated.

    Best regards,

  • ShadiShadi

    Hello @Alex

    I wanted to discuss my finings regarding the usage of both "massbody" and "relativeweigh" in my simulation. I have encountered two peculiar results .

    When utilizing " relativeweight" , I found a better agreement with the experimental data(as shown in previous post). However, upon closer examination of the rigid body motion, I noticed that the box stopped moving and did not reach to the end of slide. I conducted additional tests with different geometries, and the box seemed to be suspended, unable to reach to the bottom of the tank.

    image.png

    similarly, when using "massbody", I encountered the same issue where the box does not reach the end of the slide. Additionally, the fluid exhibited a rather turbulent behavior.

    image.png


    I would greatly appreciate any suggestions or insights you may have regarding this matter.

    Best regards,

    Shadi

  • AlexAlex

    The only difference when using "relativeweight" (and "rhopbody") compared to "massbody" is that with the first ones the GenCase code will compute internally the value of massbody and this will depend on the resolution. So that, higher resolution means that massbody will converge to the actual value. However if you are not using a super high resolution and you already now the actual mass of the body then you should impose the value using "massbody".


    This means that massbody is different in your two situation, you can check the value in the output XML file generated by GenCase.


    Regards

Sign In or Register to comment.

© Vanilla Keystone Theme 2024