Direct calculate the water pressure inside DualSphysicsV4 for each time step
Hi,
I want to confirm some question about calculation of the water pressure. As I need to analysis the water impact pressure with high frequency, it may be impossible to output the DualSphysicsV4 calculation results for each time step to calculate the impact pressure by MeasureTool4_win64. So I tried to modify the DualSphysicsV4 code to directly calculate the pressure (based on the equation in manual “HOW TO NUMERICALLY COMPUTE, p14”) during the running of DualSphysicsV4 and output the pressure for each time step.
The calculation results for pressure seem reasonable. By comparing the pressure results based on direct calculation and postprocess MeasureTool4_win64, almost same results are confirmed. However, for some cases, there is a little discrepancy between the direct calculation and MeasureTool4_win64. As there are some special coefficients used in MeasureTool4_win64 (-kcusedummy -kclimit), I am not sure if these coefficients can provide some influence on the pressure calculation. Could someone please explain the physical meaning of -kcusedummy –kclimit. Can I regard the directly calculation as a better solution than MeasureTool4_win64.
Thank you.
I want to confirm some question about calculation of the water pressure. As I need to analysis the water impact pressure with high frequency, it may be impossible to output the DualSphysicsV4 calculation results for each time step to calculate the impact pressure by MeasureTool4_win64. So I tried to modify the DualSphysicsV4 code to directly calculate the pressure (based on the equation in manual “HOW TO NUMERICALLY COMPUTE, p14”) during the running of DualSphysicsV4 and output the pressure for each time step.
The calculation results for pressure seem reasonable. By comparing the pressure results based on direct calculation and postprocess MeasureTool4_win64, almost same results are confirmed. However, for some cases, there is a little discrepancy between the direct calculation and MeasureTool4_win64. As there are some special coefficients used in MeasureTool4_win64 (-kcusedummy -kclimit), I am not sure if these coefficients can provide some influence on the pressure calculation. Could someone please explain the physical meaning of -kcusedummy –kclimit. Can I regard the directly calculation as a better solution than MeasureTool4_win64.
Thank you.
Comments
SO THAT, HELPS TO NORMALISE THE KERNEL VALUE WHEN THE NEIGHBOURHOOD IS NOT COMPLETE; AT FREE-SURFACE, CLOSE TO BOUNDARIES... THIS AVOIDS OVERESTIMATION OF THE VALUES
-kcdummy:
Defines the dummy value for the interpolated quantity if Kernel Correction is not applied (default value =0)
WHEN THE KERNEL IS NOT ENOUGH COMPLETE, KERNEL CORRECTION IS NOT APPLIED AND VALUE OF 0(DUMMY) IS SHOWN. VALUE OF 0 OR -999 CAN BE USED DEPENDING ON THE MAGNITUDE YOU ARE COMPUTING
Yes. I found that the pressure based on direct calculation is almost identical with the MeasureTool results when the measured location is far from free surface or boundary. However, for the case closed to free surface or boundary, the direct calculation shows much larger results than the MeasureTool.
As I want to calculate the impact pressure on the water tank when sloshing occurs and the measured location is above the still water surface in each time step, it seems that I have to include this kind of kernel correction in my direct calculation subroutine. Could you make it clear that how I can include the kernel correction as similar as –kclimit and –kcdumm. Or could you please recommend some papers about this kind of kernel correction.
At present, I added the pressure calculation subroutine at the end of the main loop (after “RunCellDivide(true)”). By calling KerGetInteractionCells, I can get the interaction particles list and then, calculate the corresponding Kernel value and the pressure in each interaction particle. Finally, the pressure on the measured point is solved as: P=sum(Pb*Wab)/sum(Wab).
Best Regards