Hi,

I need to compute max impact on glue constraints (call it max_impact), and use it to set glue constraint strengths such as: to break constraints set their f@strength<max_impact, otherwise set f@strength>max_impact. The idea here is to sim twice: first compute max_impacts on constraints, then sim a second time to glue strength either < or > than max_impacts. The problem here is each time I change glue f@strength, I get new max_impact? My initial thought is max_impact should be function of object density, speed,… and not related to glue strength?

Notes:

a-I compute max_impact by promoting impacts generating on constraints using “maximum” mode.

b-Below are 2 tests to demonstrate the “problem”:

Test 1: set Glue strength = 20e9 –> computed max impact on constraints = 2,879

Test 2: set Glue strength = 20e1 –> computed max impact on constraints = 286,801.8

As I mentioned above is max_impacts seems function of Glue strength? Attached is a simple hip file for the tests…

Thanks

Just an update: glue impacts are zero when an object, let's say a wall, is not interacting with any other objects, just holding itself under gravity –> this means that “impacts” do not measure impulse/stress generated by glue constraints (otherwise it will have some values to counter gravity pulling the wall down). Hmmm…so the whole idea of measuring impacts and using them to set constraints strength is just wrong (unless I find the right parameter to measure)! Now going back to square one: i.e. how to set constraints strength such that I guarantee some will break and some won't…

Another Update + four questions:

1) Just watched a tutorial by sidefx (https://youtu.be/lG0iZCv-sPQ?t=602) and it states if impact on constraint > glue strength then constraint will break, now according to my attached test this is not the case!? I set glue strength to 1.0 and @Frame 18, I see many impacts > 1.0 but still see their constraints are not in broken group?

2) Glue constraint number 87 seems to be always there with impact value 30510 (at frame 50), although all other constraint impacts are decaying to zero, why is that?

3) Why constraint impact = 0 when there is no collision? Should not be any internal values to keep pieces stuck together and prevent from falling under their own weight force?

4) If I increase Glue Constraint strength, I get lower impacts? Meaning, in the attached file:
If I set glue strength = 1 –> max generated impact is 30,510 (piece 87 @ frame 18)
If I set glue strength = 100 –> max generated impact is 1,946 (piece 87 @ frame 18)
My question: Why generated impact is a function of glue strength? Should not be the impact function of piece mass, velocity for example only and not function of strength defined on constraint?

Did you ever find an answer? I'm also digging into the depths of glue constraints and can't quite understand how they work… I think they only respond to “contact constraints” which are like collision etc. so they wouldn't be fighting gravity.

Has anyone figured this out yet?

It makes it very difficult to work with breaking thresholds when I don't know what the actual impact values are for my constraints. They are being scaled by the constraint strength.

If I have 5000 for my glue constraint strength, then no impact is ever higher than 5000.

I understand that the constraint is now broken since it has reached the strength of the glue and is no longer an active constraint, but it seems like the value should still show the computed impact.

OK, I found where to see it...

In the spreadsheet, in the actual rbd_object data under ConstraintGeometry the impact values are the actual computed impact.