A question about the relationship between rigidbodies and model scale

[HentaiTree]

Recently I used fluid simulation on blender, which required me to enlarge the model to get a more accurate simulation. This model is imported as 0.08 scale. After importing, click build and rigidbodies runs correctly. but when I enlarged it to x12.5(over x3), rigidbodies messed up. so I tried to import as 1.00 scale, and click "build", it works well

then I enlarged it to x10 and shrinked it to x0.08, both operation didn't harm to rigidbodies' movement. Can I conclude that only import model at the scale of 1.00, that I can adjust its scale in free and keep the rigidbodies run correctly? Or are there any ways that allow those models which imported in 1.00 work well after enlarging? Sigh, I've imported many models by the scale of 0.08, I have no idea how to solve it. T^```T

[nathanvasil]

Blender's physics don't tend to work well with unapplied scale. Whenever you're running any physics simulations, I'd recommend applying scale (from the ctrl-a menu for me) on anything you've scaled.

Although there's the potential for a lot of tricky corner cases, it looks like it works well if you simply select the MMD model's parent empty, scale it, apply scale, and then build physics. But like I said, there are a lot of tricky things to think about in that regards, and I've only done the barest of testing on it.

Note that if you do this, you will probably want to change your gravity settings as well.

[HentaiTree]

sigh.. seems it doesn't work well after applying scale :(

[nathanvasil]

Oh, I see, applying the scale just turns it into unapplied scale on the next child.

I guess you'd need to scale (and apply scale) on the individual rigid bodies. Possibly, on the armature and mesh as well, separately.

[HentaiTree]

thx :)I will try it