[zbuffer]

Hello!

I am a researcher in underwater robotics. I have to identify hydrodynamics of an underwater vehicle and I wanted to get some idea about what values I should expect to see (a kind of a ground truth).

I know CFD can give you all kinds of wild results when you don't know what you're doing so I am reading some book now to learn, but I have already made a simple test, based on some article. to check if I will get the added mass value of something as simple as a sphere moving through the fluid.

The added mass effect appears when a body moving in fluid is accelerating. So I conducted two simulations. One simulation with a constant flow velocity equal to 1m/s and one simulation with a constant acceleration of the flow at 1m/s2. Then I have subtracted the force calculated in the first case from the force in the second case after 1s.
According to common sense and the article I have read this should give me the added mass from F=ma.

The problem is I get a result which is not correct according to the theoretical solution of the problem based on potential flow. I know the results will never be exactly the same because CFD will account for more than potential flow but it should be similar and I get around 3 times the value. I started thinking that I might be missing PI somewhere but it doesn't look like....

The theory says: 2094.4 N (fluid density 1000, sphere radius 1m).
The CFD says: 6384.3 N

The theoretical result can be found here:
http://web.mit.edu/2.016/www/handout...ion_050916.pdf

I have used SimScale platform: https://www.simscale.com/workbench?p...d-ac9c06a09a16

I have created a box with a sphere inside. The simulation is using the SIMPLE solver with a Laminar flow. Two sides of the box are velocity inlet and pressure outlet, others are symmetry. The ball is a no-slip wall.

Do you have any idea what I am missing?
Maybe there is some general error in my reasoning....

Best,
Patryk

[vatavuk]

Hi Patryk,

You said that you used the SIMPLE solver. If you mean simpleFoam, this may be a problem, because this solver is for steady simulations. For transient flows you should use pisoFoam or pimpleFoam. For laminar cases you could also use icoFoam.

Best Regards,
Paulo

[zbuffer]

I was not specific enough.
I used SIMPLE for the constant velocity and PIMPLE for the constant acceleration.
Then I also used PIMPLE for the first case which gives the same results as SIMPLE...

[kncujo]

Originally Posted by zbuffer

I was not specific enough.
I used SIMPLE for the constant velocity and PIMPLE for the constant acceleration.
Then I also used PIMPLE for the first case which gives the same results as SIMPLE...

Hi zbuffer,
I know it's been a while since the post was published.
I'm dealing with a similar issue to yours. I'm calculating the added mass of a body immersed in a fluid. The technique I'm using is different from the one you were using, but in my case as well, the results from the CFD significantly overestimate the value of the added mass.
Did you ever find an explanation for your error?