Why does the speed and direction of the model fluctuate?

Hello, everyone:
I used WorldBuilder to design a model of the mid-ocean ridge, with the lithosphere temperature following the half-space cooling model. However, the velocity arrows in my result were jittering severely, and the magnitude and direction of the velocity kept changing over time, as shown in the first animation. My expectation was for a relatively stable state. I don’t understand why this is the case. I would be very grateful if you could help meThe size and direction of the speed arrow fluctuate!

@sxj It’s quite difficult to say what exactly is happening because the visualization program chooses the scale used to show the errors separately in each time step. It is quite possible that in one time step, the max velocity is 0.99 and so the program shows velocities on a scale from 0 to 1, where 0.99 is a long vector. In the next time step, the max velocity may be 1.01 and the program shows velocities on a scale from 0 to, say, 1.5. A velocity vector for a speed of 1.01 is then substantially smaller than it was before, even though the velocity has not really changed very much.

A much better way to understand output is if you choose a fixed min and max for the display across all time steps. Can you try that?

Best
W.

Thank you for your suggestion. I fixed the maximum and minimum values according to your suggestion, but the size and direction of the speed arrow are still changing.Fixed maximum and minimum speeds

Hello, I also made a comparison. I compared the models previously written using the prm file, and the speed was relatively stable. Later, after designing my model using WorldBuilder, there was a fluctuation in speed. Therefore, I extracted the speeds at the same points of these two models and drew a comparison chart as follows. I don’t understand why the speed changed so much after using the wb file (the red line represents using the wb file, and the blue line represents using the prm file. The reason for switching to the wb file for design was that my subsequent model design was quite complex and I needed to use wb for design). Thank you very much!

@sxj Thanks for checking my suggestion. I think I’ll have to let others help with this question.
Best
W.

@sxj - The following would be helpful to diagnose the issue:

  1. The PRM file for both models and the WB file for the second model case
  2. Images showing the difference in initial conditions between the two models.
  3. Analysis of any other properties that are varying through time. For example, when switching over to using the WB is it possible that one of the boundary conditions changes by accident?

Thanks,
John

Thank you every much!

Thank you for your suggestions. I will follow your advice to conduct the check.