[sasanghomi]

Hi foamers ,

I have a question about using fvmotionSolver for engineMesh.
In this case we should specify movement of piston and valves by using pointMotionUz file . for example for movement of a valve we should create two columns of data ( velocity versus CAD ) in a .txt file. Actually we specify velocity of valve versus CAD But in the theory and experiment we have valvelift profile . Actually we have a specific position of valve in each time but in this case we only specify the velocity of valve But the position is a function versus velocity and timestep ( X = V* T) . So my question is that How can I specify correct timestep for guaranteeing the position of valve ? Actually I think only specifying correct velocity can't guarantee the position of valve in each time !!!
Please anybody clarify this issue for me...

thanks and best regards,
Sasan.

P.S. Also I use adjustTimeStep . So timeStep is not constant.

[mturcios777]

Hi Sasan,

I started using valve velocity profile as in previous versions you had to specify the absolute displacement rather than relative, which caused issues with our case setup. The motion solver does an interpolation so that as long as your velocity profile changes in time (CA) correctly for your valve lift profile, your valve will move the appropriate amount. I have never had issues with the valve motion using just velocity.

That said, the motion is resolved much better when your velocity profile smooth, since the interpolation is only linear. We've found that every one or two degrees is enough.

[sasanghomi]

Hi Marco,

Thank you very much for your reply. What made me confused is that the velocity range of valve is very lower than velocity range of piston . I attached two pictures of valve and piston velocity versus CAD in my case(intake stroke ). Are they Ok? Are they usual ?

Thanks and best regards,
Sasan.

P.S. I have a problem for getting convergence If you have a little time Please take a look at my question.(thank you very much)http://www.cfd-online.com/Forums/ope...nginefoam.html

[mturcios777]

The shapes look good, and the magnitudes seem in range. The valve travels a shorter distance and so the velocities should be lower. How did you obtain these profiles?

[sasanghomi]

Thank you very much . I have valvelift profile . So I have gotten the derivative of that by using MATLAB. Also I found the function of movement of piston in theory books and I plotted it .

[sasanghomi]

Hi Marco ,

Please take a look at this picture . This is CAD=36 deg . the piston moved about 12.8 (mm) But the valve doesn't move at all ( the movement of valve is very low So paraview doesn't show any movement [I checked the movements in information tab of paraview] ) . I think this picture is not a good contour of velocity at CAD=36 deg . Are you agree with me? Do you think movement of the valve is low?

Thanks and best regards,
Sasan.

[mturcios777]

You are looking at the velocity U? If you are then the valve motion won't show in there as this is the fluid velocity. If you want to know how fast the valve patch is moving, you will need to load the pointMotionU field which specifies boundary motion. When you move through the timesteps, do you see the boundary moving?

[sasanghomi]

Yes, I can see moving the piston in paraview..Do you agree with me that this contour is bad?

[mturcios777]

If you are looking at the fluid velocity it looks okay to me though a little high. When you load pointMotionU field on all the boundaries, what is the value on the valve boundary?

[sasanghomi]

I set TimeVarying..and i put there a .txt file

[mturcios777]

Sorry, I mean to say cellMotionU, since that is a volumeScalarField.