Ansys Fluent: False Gear Pump With Dynamic Mesh: Setup

Simulation of a 2D gear pump using the dynamic mesh method. In the video you can see that the gears do not touch each other, obviously this is not physically possible, hence what i mean by "false".
The contents of the UDFs are:
UDF1:
#include "udf.h"
DEFINE_CG_MOTION(gear_1_new,dt,vel,omega,time,dtime)
{
real t = CURRENT_TIME;
NV_S (vel, = , 0.0);
NV_S (omega, = , 0.0);
omega[2] = 60.0*(1-0*exp(-500000.0*t));
}
UDF2:
#include "udf.h"
DEFINE_CG_MOTION(gear_2_new,dt,vel,omega,time,dtime)
{
real t = CURRENT_TIME;
NV_S (vel, = , 0.0);
NV_S (omega, = , 0.0);
omega[2] = -60.0*(1-0*exp(-5000000.0*t));
}
These are to set the motion of both gears.
#Ansys #AnsysFluent #CFD #AutoDeskInventor

[@vivekstreme1994]

This works on real gear pump too but make sure you have a really good system as the number of cells become very high and computation time increases a lot. This is mainly because a real gear pump has very fine clearance and your domain needs to remesh regularly or else you get negative cell volume error. if you are planning to do research on gear pump, please be aware of this issue.

[@usamatahir1353]

good work man

[@HappyJourney3003]

I get No valid body found in geometry error while starting the meshing window. Is there any solution to this? I've followed all the steps as shown

[@desikanv8894]

its looks nice is it possible to make rotating of two concentric tube to get the output of the neccessary velocity to rotate the inner cylinder

[@03_alfihidayathutasuhut87]

can you give the udf files sir?

[@DineshBabuCFD]

Nice work.. Congrats.. Can we apply same process for 3D GeRotor analysis?

[@akndenizman8391]

Hey, did you find any info about 3d gerotor pump analysis?

[@deniengineering]

The UDF file works in center geometry of motion wall in axis Z. Anybody know how to solve that?

[@vivejk4443]

i didnt understand what you mean, what do you want to solve? Basically, you are making the gear_1 body rotate about its centre of gravity and defining an axis around each body for rotation and likewise for gear_2. There is a tutorial by ANSYS regarding UDF of two rotating gear. please refer it. the youtuber has imported the same udf

[@deniengineering]

@@vivejk4443 now, i know how to define it. Its just my fault and my ignorance

[@vivekjkumar5797]

can this be used to simulate a real gear pump? Can i still follow these steps to simulate a real gear pump?

[@fluentsetup1232]

Yeah it could, I'm not sure if the dynamic mesh would be able to handle it since i haven´t tried it myself.

[@vivekjkumar5797]

@@fluentsetup1232 thanks

[@vivejk4443]

@@fluentsetup1232 i tried doing for real gear pump(with involute spur gear) but i get negative volume cell error after some time steps. i will try reducing element size further, maybe that will help

[@futuretech42]

Nice work

[@vivejk4443]

what do i do if i get dynamic mesh :negative volume error?

[@fluentsetup1232]

Without looking at it I'd say try reducing the time step or increasing the mesh elements size. If that doesn't work it's probably an issue with the dynamic mesh tool.

[@deniengineering]

​@@fluentsetup1232can you make video to solve thats?

[@vivejk4443]

@gineering i dont know if im late but i got to know that my issue was that the clearance between both addendum and dedendum of the gear was very less, so i would get frequent errors. Another cause of this issue could be time step and mesh element size not being able to capture the motion between the two gear. Make sure you use a powerful computer or workstation to do this simulation if you have very low clearance between the two gears

[@boblin9263]

I want to ask is it possible to rotate one gear and simulate the response of fluid (Navier-Stokes eq.) and the other gear (rotates accordingly due to the fluid pressure or gear contact ).

[@fluentsetup1232]

Yes it is possible, the only problem i see is that you need to program a udf to tell the other gear what to do in case the two gears touch each other, which is a little more complicated. I might do it for a future video.

[@boblin9263]

@@fluentsetup1232 Is it also possible to rotate the second gear based on the net torque coming from the fluid pressure or stress distribution? This is something like "fluid-structure interaction" problem.

[@fluentsetup1232]

@@boblin9263 I thought that was what you meant with your first coments lol. But FSI is not necessary, you can just use dynamic mesh with an UDF to set the motion for the first Gear and another dynamic mesh(not necessarily with UDF) to make the second gear respond to torque transmited by the fluid.

[@maxpower4798]

could you show the udf files you used?

[@fluentsetup1232]

I just added them to the description. I Thought they had been there this whole time lol.

[@usamatahir1353]

can you explain this rotation ? omega[2] = 60.0*(1-0*exp(-500000.0*t)) for angular velcloity and please tell me units of rotation as well

[@vivekjkumar5797]

yes, i need to know too

[@fluentsetup1232]

@@vivekjkumar5797 I think the default units are [rad/s], although i could be wrong on that. On the expression for the angular velocity, ignore the exponential, it was a mistake that i forgot to erase before uploading the udf

[@vivekjkumar5797]

@@fluentsetup1232 yes i was surprised that the exp was multiplied with 0. anyway, thanks for the clarification. also, can i do the same thing with sliding mesh?

[@yeico01]

@@fluentsetup1232 I dont understand the equation. Could you eplain, please ? thanks

[@DineshBabuCFD]

@@yeico01 Its supposed to be
omega[2] = 60.0