Hi Gunardi, I just downloaded the file, I'll take a look.

There's definitely some significant leakage!  More than I would have expected.  Did you try increasing you Eulerian domain a bit further?  You have it a few elements into the solid, but I would make it bigger.  As it stands, this model runs pretty quick.  Also you could try using the same mesh size and using brick elements.

Thanks Jandron for your interest and reply.

I had tried using a smaller eulerian element, but the leakage is still present. I will try it again.

For the cavity model (the largest part in the model), I have tried just now (after read your posting) using brick element (hex), but all the option other than bottom up are grayed out. I tried the bottom up hex already, but the mesh didn't make sense which is why I stopped there.

It happens to me frequently that sometimes there are leakage in my CEL simulation and some times in other models, no leakage. I can't explain it why.

Have you already opened the input file job-7d? The CEL simulation works out find there without any leakage.

Best regards

Gunardi

Yeah, you will need to partition the part in order to mesh it with hex elements.  Definitely worth it.  This will enable you to select structured and swept mesh choices.  I didn't look at the other model yet, but I'll try to later.

Try to make the eulerian domain larger too. You could also try to dial back the timestep using the SCALE FACTOR option of the *DYNAMIC, EXPLICIT definition (http://50.16.225.63/v2016/books/key/ch04abk44.html).

Hello Jandron,

I just tried using a bigger eulerian element and used a smaller hex element for the relevant cavity (the one that's gonna have contact to fluid). Still didn't work, still leaked through the contact of lagrange element.

I attach as well the input file here.

Thank you, Jandron, for your interests in my problem.

Best regards

Gunardi

Hello Jandron, I have tried using smaller or bigger eulerian mesh. Nothing works... I have also tried to partition the eulerian part so that one partition has the same geometry as the inner geometry of the cavity. No improvement. Still leaking.

I really wish You or Someone could help me with this.

Best regards

Gunardi

When you say "bigger and smaller eulerian mesh" - are you making the eulerian DOMAIN larger as well?  Making a smaller eulerian domain definitely won't help.

Hello Jandron,

sorry for the confussion. I meant I tried a bigger eulerian domain. I tried as well a bigger and smaller mesh global size for the eulerian domain. But nothing does help.

Thank you for your interest, Jandron

Best regards

PS. I also attach the input file here:)