Squeeze flow

CFD simulation of the flow between two plates, where the top plate oscillates sinusoidally in the vertical direction (symmetry was used, video shows only right half). This was solved using STAR-CCM+ v3.02, which at the time did not have any moving mesh capabilities. In order to model the movement of the upper surface a custom Java macro was created, which exploited the software's ability to map a solution from one mesh to another.

To begin developing the model, a simple Visual Basic macro was created in Solidworks to create the geometry. This macro would take the geometry, change a dimension (based on the amplitude, frequency, and time step), save the model as a ".step" file with a unique name, and then repeat the process.

After the geometry files were created, a Java macro was created in STAR-CCM+ to mesh each geometry. The macro would open each geometry file, create the mesh, save the mesh as a ".ccm" file with a unique name, and then repeat the process.

Finally, the Java macro that modeled the "moving mesh" was created. This macro would import a mesh, map the previous results to the new mesh, change the rate of the volume source, iterate, and then repeat.

After solving the CFD model, the results were compared to an analytical solution obtained by solving the Navier-Stokes equations with an oscillating boundary. The solution obtained (as shown in the video) was valid only for Reynolds numbers much less than 1. The equation was entered into STAR-CCM+ as a field function for visual comparison. The CFD results were found to agree well with the analytical results.

For description of symbols:
http://img48.imageshack.us/img48/3599/schematicv2.png

Category:

Science & Technology

Tags:

• computation
• fluid
• dynamics
• CFD
• STAR-CCM+
• Portland
• State
• University
• PSU
• Navier
• Stokes
• squeeze
• flow
• oscillating
• parallel
• plate

[jsliu54]

Hi, could you provide the reference about the analytic solution? many thanks

[JaysonMartinez]

I solved the Navier-Stokes equation to determine the analytic solution. I used scaling to simplify the equation.

The boundary conditions were (refer to image in description): u(y=0)=0 & u(y=h(t))=0 where h(t)=H+delta*cos(omega*t)

To solve for the pressure differential (to substitute into the velocity equation) I equated an expression for the flow rate found by integrating the velocity and an equation for the flow rate found my knowing how much fluid was displaced.

[jsliu54]

Good task, could you interpret the symbols definition in the analytic solution, for example, "H, Delta".

[JaysonMartinez]

I've added a link to a schematic in the description

[minidickcn]

thanks for uploading~