CFD Online Discussion Forums

CFD Online Discussion Forums (
-   CFX (
-   -   4 stroke engine inlet/exhaust simulation (

flo777 March 16, 2010 00:27

4 stroke engine inlet/exhaust simulation
Is is possible to specificy the inlet boundary conditions as a sine or cos function? Would it be appropiate to model the pressure waves at the inlet and exhaust of an engine as a function of sine or cos, depending on the sound waves profile?
How would you implement such boundary conditions?
I don't know how to simulate the transient response, so can this be done in a steady state CFX simulation?

ghorrocks March 16, 2010 07:21

Yes, you can model the inlet or outlet as oscillating functions WRT time.

But keep in mind that the wave reflections at the boundaries are going to be totally wrong. If wave effects are important than this approach is not valid.

I have done lots of engine manifold simulations and I always ran the inlet to beyond the inlet trumpet and the exhaust beyond the end of the exhaust pipe to get the reflections right. Even then I could not get the reflections spot on as details in the air handling system in the lab put small wobbles in the inlet manifold wave which were diabolical to get right.

Here's a small plug - this is my PhD thesis on engine modelling from centuries ago -

flo777 March 16, 2010 10:00

Thanks for the responses
Wave effects are important, as i want to tune the engine in a specific RPM range for increase torque
This is just a 4th year project for me, i don't want to get in too many details as i am already running out of time.

My goal was to model the intake injection relative to the butterfly valve, and the exhaust gas release through a divergence nozzle. If i just run simple boundary conditions, results will not be accurate, but I assume it's ok to compare between different setups simulated at the same conditions.

ghorrocks March 16, 2010 17:44


Wave effects are important
Are you looking at applying the oscillating BC at the cylinder end or the open end? Do you have an accurate pressure versus time graph at your operating points?

As I said the boundary will give totally wrong pressure reflections. It will be a bit close if you are using an accurate pressure vs time relationship (this is what I did in my thesis 10 years ago, but my more recent unpublished work does not use this simplification).


but I assume it's ok to compare between different setups simulated at the same conditions.
I doubt this is true. I worked in engine R&D for 10 years and assumptions like that always got you into trouble.

All times are GMT -4. The time now is 16:10.