2D simulations of hydrofoil profiles

Hi,

Fine Marine is definitely the software of choice here! :-)

Best regards,

Benoit

Thank you,

there is some documentation about 2d simulation in fine marine? or maybe tutorials? all I find is for 3D simulations...

Best regards,

R

Yes, absolutely: Tutorial 3 - Advanced - is a 2D case of a falling prism and Democase 2 is also 2D for a rolling decay test.

They should be a good starting point to know how to manipulate 2D projects.

Last point: if you go directly to the HEXPRESS documentation, you can find a documentation page: HEXPRESS > User Guide > Mesh Wizard > 2D/3D Mesh Generation. It gives several details about 2D case preparation.

Best regards,

Benoit

Hello,
thanks for the answer, I was able to get very good results by simulating NACA profiles with a round leading edge.
I am trying to simulate profiles with a sharp leading edge for high Reynolds numbers (10^6 - 10^7), unfortunately, however, the residuals are always very high and only through HEXPRESS I cannot thicken the mesh adequately. Is there anything I can do?
thank you

R

Hello,

great to hear you managed to get good results with a round leading edge!

For a sharp one, I see there are two items in your message: 1) solver divergence and 2) viscous layers insertion if I understood correctly.

For 1), the best is to save the solution before it diverges and check where the pressure/velocity spots are coming from. This advice comes from the FAQ here: "FINE/Marine > FAQ > My calculation stopped abnormally. How can I manage?" Maybe the check list could also help. If you are still in 2D, you should check the mesh quality and also the tessellation itself.

For 2), what is your problem precisely? are you inserting viscous layers with inflation method and you observed a relatively low height compared to expectation?

Best regards,

Benoit

Hello,

sorry for my late reply, I ran some test simulations, saving the results more frequently, in this way, as you advised me, I evaluated the points where the pressure and speed (so also the turbulent viscosity) varied the most and I improved the mesh at those points. I read the reference you sent me and it helped me.

I probably did something wrong also in the simulation setup, I changed the boundary conditions like this:

- x max, y max, y min -> far field, with Vx equal to the speed of the fluid in the case of study.

-x min-> prescribed pressure.

And imposing an initial solution with Vx as initial velocity equal to the speed of the fuid in the case of study.

It seems to work residuals are drastically reduced.

best regards,

R

Hello,

sorry for my late reply, I ran some test simulations, saving the results more frequently, in this way, as you advised me, I evaluated the points where the pressure and speed (so also the turbulent viscosity) varied the most and I improved the mesh at those points. I read the reference you sent me and it helped me.

I probably did something wrong also in the simulation setup, I changed the boundary conditions like this:

- x max, y max, y min -> far field, with Vx equal to the speed of the fluid in the case of study.

-x min-> prescribed pressure.

And imposing an initial solution with Vx as initial velocity equal to the speed of the fuid in the case of study.

It seems to work residuals are drastically reduced.

best regards,

R

ok great! Thanks for the information. The setup seems correct.

Best regards,

Benoit