CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   CFX (http://www.cfd-online.com/Forums/cfx/)
-   -   transient, discrete , repeating boundary condtion (http://www.cfd-online.com/Forums/cfx/105825-transient-discrete-repeating-boundary-condtion.html)

Far August 10, 2012 01:51

transient, discrete , repeating boundary condtion
 
Hi all,

I am modelling intake for formula racing, where I have one intake and four outlets (ports). These ports are open for 90 deg each, in other words at one time only one port is open. The pressure at these outlets is also varying with time(i.e. with each degree of crankshaft motion).

I am modelling this using transient profile and hence creating the new variable. I have couple of questions

1. It this the good strategy or I should use the CLL or even CEL to get this condition work?

2. In all examples I can see that at different spatial locations (on same boundary) values are changing, which is not true in my case. What should I do? Should I create the discrete points and apply same bc or or only take one point and repeat it for different time? In simple words, how can I apply uniform boundary condition at each outlet with the time and boundary (port 1 0-90 deg, port 2 90-180 deg, port 3 180-270 deg and port 4 270-360deg as variable

3. Do you have any other idea that how I should implement this boundary condition, that varies in time and also to different boundaries.

Thanks in advance.

ghorrocks August 10, 2012 06:37

First of all you better work out what you actual know about the outlet boundary condition. Do you know the pressure or mass flow rate? If you have a meassured pressure trace then you can use that. If you do not know anything then you better use a mass flow rate, with the mass flow rate being determined by the rate the piston sweeping the cylinder, and adding factors for valve losses and other effects if you are brave. Note that this second effect will not take into account dynamic effects such as inlet manifold supercharge.

It is easy to implement this. Can you either define 4 curves versus time, or one curve versus time and call it with the valve timing different, ie variable(t), variable(t+cyl1), variable(t+cyl2) etc where cyl1 and cyl2 are the timing offsets of each cylinder relative tothe first.

And why is your valve timing open for 90 degrees? No engine on the planet uses that timing. Optimum valve timing is usually far from the geometric ideal due to dynamic issues.


All times are GMT -4. The time now is 06:47.