Rigidbody Mass

Unity Engine
1

So, as of this moment, I’m kind of agitated…

I always operated under the belief that Mass was in KGs. Maybe it is, however someone on reddit noted the value should be between 0.1 and 10.

At first I thought bs… but…

Is this a new thing? I admit I havnt looked at Unity Docs around rb mass in years, but I dont recall ever reading anything like that previously.

Im looking for comment from someone who works at Unity ideally and has access to the source… Wondering if I need to go back and revise EVERY rb setting Ive used in my games. The values work, but if they are making the engine work harder for nothing they probably need sorting.

2

I think it’s just poorly worded. 0.1-10 is probably the ratio that should be maintained between two or more colliding/connected rigidbodies. Mass ratios greater than this will result in things flying into space.

3

Yes, I do recall reading something along those lines back a ways that mass shouldnt vary by large numbers, that I already knew, and was fine with, because I based everything on real world’ish numbers.

ie.
Jeep:1000 (assumed kgs)
Artiliary: 2500

4

Yup, I’m sure it was worded better in earlier docs.

5

The docs have always said exactly that. As they say, large values can lead to instability (due to floating point limitations) so ideally mass should be somewhere around 1.0.

–Eric

6

Strange, Anthony Yakovlev said in the WheelCollider thread that the default WheelCollider settings work best with a mass of 1500kg.

7

I use real masses in my stuff and it’s fine: http://www.speedboatsim.com

As someone else said, it’s more about ratios of interconnected or colliding bodies. You can’t expect to set a Mach truck on top of a ping-pong ball and have it work nicely.

8

I too use real masses, and its fine. Its not what the docs suggest you should do though. Could just be a case of the docs being up to the typical docs standard (lacking).

However, my main concern is working the system harder than it needs to. Be nice to have official comment from someone who works in the physics engine, on whether this is a docs are wrong (poorly worded) issue or not.

1 Like