0
$\begingroup$

I was trying to simulate an melting ice cube by letting a fluid object (cube) slowly spread in its container. Doing this i ran into some weird glitches.

The glitch that bothers me the most is that the fluid stretches along the x axis until it hits the domains edge. It even goes through obstacles. enter image description here In this case the fluid domain is close to the size of the pot my ice cube is melting in. If i stretch the domain any further the liquid also extends (only on the x axis tho) until it hits the border of the domain.

This is what i tried so far to solve the problem (nothing really helped):

  • Change the fluids volume initialization to shell: This fixed the deformation and stretching of the fluid but made the animation look very unrealistic. Looked more like the ice cube is collapsing instead of melting
  • Change the size and shape of the fluid object: This did not show any effect
  • Increase the fluids resolution: This did not show any effect

Here is my fluid object enter image description here

Here is my .blend file

$\endgroup$
2
$\begingroup$

It's the normals. You need to select your ice cube, go into edit mode, select all vertices, and choose Mesh-->Normals-->Recalculate Outside. The sim then sims as it should.

$\endgroup$
  • $\begingroup$ Big thanks, you just spared me a lot of grey hair :) $\endgroup$ – Basti Feb 12 '18 at 15:36
2
$\begingroup$

The container looks like it's doubled-walled, and is probably too thin for the fluid collision detection to work with. I would suggest making a thicker pot just for the simulation and swap in the current model for rendering.

You can make the pot thicker by adding a displacement modifier on top of the fluid sim modifier, which will push out the geometry along the normals.

enter image description here

$\endgroup$
  • $\begingroup$ Do you also have an explanation for the stretching part? I already thought of the walls being to thin, but that wouldn't explain why the actual fluid is later on way bigger than the initial fluid object. As i mentioned the fluid will stretch into infinity if i expand my fluid domain. This shouldn't depend on the obstacle or can it? $\endgroup$ – Basti Feb 11 '18 at 18:20

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.