Pressure unit in 2D moving box

mzbr1g12

Hi,

I tried case with moving box to create waves and I obtain pressure exerted on the box form PartMoving.csv. I tried to get forces out from the pressure by calculating (pressure x initial particle spacing x normal direction) which I believe I get N/m.

However, as I am doing it in 2D, I am wondering how SPH define the length,m if the unit is N/m. Is the pressure unit is N/m2 or just N/m. If pressure is in N/m2, then I get the force in N/m. But if the pressure is in N/m, I will get only N for the force.

Can someone help me.

Thanks.

mzbr1g12

Just to add.

Usually for volume method like star ccm. although we model it in 2D, the model still have at least 0.2 thickness for area to be able to calculate force from there. In case for SPH, we only have 1 line of particle without thickness which means its not possible to get area. By multiplying N/m2 to initial particle spacing, I still got N/m instead of N.

Alex

If you need to compute forces, you must skip the step of computing pressures and go ahead to compute forces:

Use PartVTK to choose the particles of the moving box (-onlytype:-all,+moving) compute aceleration (-vars:+ace,+press) and you can create different ASCII files (-saveascii). Then you only have to , for each output file, do the summation of all ACE values (note that these ones are distributed in AcexPos, AceyPos, AcezPos, AcexNeg, AceyNeg, AcezNeg, so you can compute the magnitude) and multiply the summation by MASS and you will obtain the resulting force on the box for that instant (the intant of the ouput file Part_0004.asc).

Regards

Alex

mzbr1g12

Hi,

Thank you for your reply. I understand that there are 2 ways to compute the force in a forced oscillation motion of a box. Either using pressure integration over the body surface or force required to sustain the motion (by ACC times MASS).

So, is it possible to use pressure integration to obtain the force in 2D oscillating box?

kindly regards,
Zahir Ramli

mzbr1g12

sorry again..for my second doubt, in the case of forced oscillation by a moving box where we already define the velocity/motion, what is the acceleration actually define/represent ? Is it acceleration of the moving box or acceleration of moving fluid particle around the box.

Alex

"Use PartVTK to choose the particles of the moving box (-onlytype:-all,+moving) to compute aceleration (-vars:+ace,+press) and you can create different ASCII files (-saveascii)"

During the SPH simulation, only aceleration values of fluid particles are computed and used to calculate velocities and positions of the fluid particles in time. There is no need to compute the ace of the boundaries since we will move them as we want. Using PartVTk for your output files, you will compute those ace values of boundaries. Those values represent the force exerted by the boundaries which in fact is the force exerted by the surronding fluid particles.

Regards

mzbr1g12

Hi,

Thank you again for your explanation.

Just to clarify. I have this experimental result where I got the horizontal force and vertical force. So, I am planning to compare the forces(x and z) from SPH with this experimental result.

From what I understand from your previous post; In order to get the forces(horizontal n vertical);

Fx = mass x Xtotalace
Fz = mass x Ztotalace

where,

Xtotalace = sqrt(totalaceposx^2+totalacenegx^2) ----> magnitude
Ztotalace = sqrt(totalaceposz^2+totalacenegz^2)

I am somehow got confused with negative acceleration and what its actually represent. Hope you can help me out.

Kindly regards

Alex

only the sum of pos and neg.
forces are split in different directions since it is interested for some applications
but you can skip this and use:

Xtotalace = totalaceposx+totalacenegx
Ztotalace = totalaceposz+totalacenegz

mzbr1g12

Thank you for your reply. Really appreciate it.

mzbr1g12

I am sorry. Just need to confirm one more thing. If I calculate the total force by using ace, I will get the unit in N. However, if I pressure integration for the pressure I will get force in N/m. I tried to calculate from both perspective and it seems that it cant be compared. Something is missing when I tried to calculate force using pressure integration.

Do you have any idea on this particular problem?

Alex

Force can not be computed starting from pressure values, not correct. Values of pressure of boundary particles are not physically correct and pressure of fluids close to the boundary are slightly corrupted by previous fact. The best way to compute forces is using the ACE values of boundary particles which are the particle forces exerted by the fluid on those boundaries.

lssndr

Hi,

I have a case of rectangular box and I need to compute the pressure on the left boundary of the box.
I created the ascii file with this instruction:

%partvtk% -dirin %dirout% -savevtk %dirout%/PartMoving -onlytype:-all,+moving -vars:+press -saveascii %dirout%/PartMoving

I read, in PartMoving_ijkl.asc, for each particles the values of PressxPos, PressyPos, PresszPos, PressxNeg, PressyNeg, PresszNeg.

How can I plot the pressure values along the height of the wall?

Regards

Alex

You should create CSV files, no ASCII files.
There you can plot the values against height.

However you should use MeasureTool to compute pressure values at a given point using the information of the fluid particles.

Regards

lssndr

Thanks

I have another problem:

I should compute the force along a moving piston, so I computed "ace" choosing the particles of the moving box (-onlytype:-all,+moving) and creating different ASCII files (-saveascii).

I have to multiply the summation (totalaceposx+totalacenegx) by MASS of the boundary moving particles. how can I find the values of the mass?

Regards

Alex

how can I find the values of the mass????????????????????????????????????????????????????????????????????


1) mass=volume*density... so mass=dp*dp*dp*rhop (=dp*dp*rhop in 2D)
2) look in the XML file where this value appear in the CONSTANTS section

lssndr

Thanks,

I had a problem with the units when I imported the files into Excel

solved!