In one sim I am using glue to cluster shattered pieces with a cluster to cluster glue strength of 0 and an intraCluster strength of -1; in the other I am simming the clusters themselves as individual objects. I get vastly different results

The RBD state node, physical tab, Computer Center of mass pop up help mentions glued sub objects.

I want the motion of the clustered ‘objects’, but I also want pieces to be able to break off with glue. Due to pipeline constraints i cannot dynamically shatter.

I'm wondering if someone would have a look at this file and explain the wildly different results when effectively the sims are identical.

Thank you for your support,

</Ross d.Y.>