CFD Online Discussion Forums

CFD Online Discussion Forums (
-   FLUENT (
-   -   Wankel 2D simulation (

amcfd July 16, 2007 06:59

Wankel 2D simulation
Hi, I'm having troubles dealing with the dynamic mesh for a modelisation of a 2D Wankel engine combustion chamber. Did someone already simulate the flow (unsteady with rotor movement) in a rotary engine ?

Joe July 16, 2007 08:21

Re: Wankel 2D simulation
What sort of problems are you having? Be specific ...

amcfd July 16, 2007 08:53

Re: Wankel 2D simulation
I can actually deal with the motion of the rotor in the engine and with the dynamic mesh associated to it without any problem as long as I don't take the joints of the rotor into account. Let me explain it better : on the rotor of the engine, they are joints on every apex of the rotor (triangle)to seal each one of the three chambers of the engine. The problem with those apexes is that they are suppose to follow the inside of the stator of the engine and at the same time rotate with the rotor. In the real engine, the joints are maintained in contact with the surface of the stator with a spring, so they can move "up and down" to follow the stator properly. I've tried many possibilities to deal with these apexes in Fluent but none of them works, so I wanted to know if anybody already tried to simulate a rotary engine to know the way he (she) dealed with this problem...

Joe July 16, 2007 09:59

Re: Wankel 2D simulation
I am actually setting up a similar class of Fluent problem at the moment. A few things to take into account:

The grid topology must remain constant i.e. your apex seals cannot cut the fluid domain off totally.

Your apex seals can maintain the grid topology if they come close to, but dont touch the stator wall. If I recall correctly this will require the apex seals to reciprocate in the radial direction. You will obviously have the specify this reciprocation a priori as part of the mesh movement boundary conditions.

If this bring-the-apex-seal-as-close-as-possible-to-the-stator-without-actually-touching-it approach still results in too much leakage between the chambers you may need to use localised dynamic momentum sources to retard/stop the leakage from chamber to chamber. This will require quite advanced mesh marking and will probably induce numerical instability that will have to be handled carefully.

If you like, post a zip file of your current case and I will have a look at it and comment ...

amcfd July 17, 2007 03:26

Re: Wankel 2D simulation
Thanks, works fine...

cz5224043 March 7, 2012 09:11


Originally Posted by amcfd
Thanks, works fine...

Hi, currently I am also working on a 2D wankel engine modelling project and having the same problem as you had before. It seems you sorted it and completed the simulation. Could you please help me out a little bit?

I am stuck in finding a way to seal the gap between adjacent chambers for months and it seems like importing the momentum source would be the best way to deal with it. Since you nailed this problem, can I have look at your DEFINE SOUECE UDF file and how did you mark the cells ?

Many thanks in advance.

All times are GMT -4. The time now is 10:49.