Operating system: Linux-5.4.0-7642-generic-x86_64-with-debian-bullseye-sid 64 Bits
Graphics card: GeForce RTX 2080 Ti/PCIe/SSE2 NVIDIA Corporation 4.5.0 NVIDIA 440.100
Broken: version: 2.90.0 Beta, branch: master, commit date: 2020-08-21 15:35, hash: `rBebf10b72b05f`
**Short description of error**
Fluid simulation at small scale seems to react very violently to minimum movements of colliders.
**Exact steps for others to reproduce the error**
First of all, I'm not very expert in mantaflow so I'm sorry if this is just a waste of your time.
As a test for an animation we're producing I'm trying to simulate a glass at real world scale (20cm domain).
I did the tests in 2.83.5 LTS to familiarize me with mantaflow, and yesterday began trying the same in 2.90 beta.
The behavior of the fluid is so different that I think I may have found a bug.
Please see the attached files.
I did the same basic setup in both 283 and 290 and rendered the results.
I also tryed very high timesteps, but even if it was better the fluid was still jumping around crazily.
At last I did a slowmotion test (timescale 0.1) and the fluid started gaining energy so much that it started spawning particles from the domain border and transformed into a red tornado.
Also, another strange thing I noticed in the 290 test is that at frame 2 the particles overshoot the glass going outside the bottom of the domain, and at frame 3 they went back inside ready to fall in the glass.