CFD Online Discussion Forums

CFD Online Discussion Forums (
-   Main CFD Forum (
-   -   intake manifold simulation (

uday November 11, 2009 01:24

intake manifold simulation

I am trying EGR mixing in the intake manifold to compare the cylinder to cylinder distribution in fluent. The manifold have two inlets (in this one is for fresh air & another is for EGR), two outlets for two cylinders.

Iam using pressure inlet & pressure outlet boundary conditions for manifold inlet & outlet.I have pressure values for inlet and outlets. but these are varing with crank angle. Like one out let open means another one is closing means acting like a wall. For this iam trying to write UDF.

Is this my approch is correct for mu prolem to find out EGR mass fraction in each cylinder.Other wise there are any methods to do this.

please give me suggestions on this matter.

Thanks in advance.

ganeshykr December 7, 2009 11:15

boundary conditions for EGR problem

Instead of writing UDF for applying time varying boundary conditions, you should use "tabular transient profile" method available in Fluent. In this you'll have to specify all the variable values varying with crank angle in a text file (journal file) and open that after opening case file. Crank angle can be converted into time. Search in fluent users guide, you'll get all the details. Apply mass flow boundary conditions at the inlets and outlets. Let me know if i am not clear.

uday December 12, 2009 03:52

Thanks for u r replay.

I have one question when we are giving boundary conditions as a tabulated data what is the time span (means time period or iterrations) we have to give for different values:

sec pressure

1 0.2bar
2 0.3bar
3 0.5bar

here i take time difference means increse 1 sec so on which parametere it will depend.(Is it on timeperiod or iterration)

please give me suggestion on this matter.

ganeshykr December 12, 2009 04:11

EGR simulations
if i understood u clearly, you are asking about the time difference between two values. it depends on how accurate you want the calculations to be.
For eg. you may feed the data at every degree of crank rotation from 0 to 720 degrees (you will have 721 values) which will take long time to finish else you may feed data at every 10 degrees (you'll have 73 values). So, you can fix the time step size also accordingly i.e. dt = 1 or 10 degrees

you can convert degrees into seconds as

1 degree = 60 / (N x 360) seconds

ganeshykr December 12, 2009 04:14

While creating the journal file for tabular transient profile specification, you need follow a format:

ProfileName NumberOfColumns NumberOfRows WhetherPeriodic(1)OrNot(0)
time profile1 profile2 profile3 ..................................

uday December 13, 2009 00:46

Thanku for u r reply.

uday December 21, 2009 02:39

morning friends,

iam doing simulation of twin cylinder intke manifold i got the converged solution.but the problem is it not showing any teparature distribution.when iam cliking mouse(left) in the working space it showing the distribution.
there is no this type of problem with pressure.
So please give me suggestions in this matter.
Thanks in advance.

All times are GMT -4. The time now is 22:32.