[eno]

I am using openfoam 1.4 in cygwin to solve internal compressible turbulent flow under the motion of a piston. By modifying rhoTurbFoam I can use oscillatingFixedValue in cellMotionUX to define a sinusoidal piston motion Ux = f*(a*sin(2*pi*t)+1). The mesh deforms as expected and the simulation looks reasonable.
Now I want to drive the piston with motions (time varying, no sinusoidal displacement) measured from experiment. How can I do it?

[marico]

Hi,

if I remember rightly, this is not trivial. Maybe you can approximate the values analytically and use groovyBC. I tried just yesterday to apply the "standard piston law of motion" (way over time) and it worked perfect.

Best Regards
Marco

[eno]

Can I just modify the code to read and interpolate the external velocity at where Ux = f*(a*sin(2*pi*t)+1) is executed?

[eno]

dang... it's readily there! just modify cellMotionUx (or other initial files) like:

piston
{
// type oscillatingFixedValue;
// refValue uniform 1.;
// amplitude .2;
// frequency .5;
// value uniform 0;
type timeVaryingUniformFixedValue;
timeDataFileName "pistonVelocity.dat";
value uniform 0;
}