What causing this simulation so long to start?

James

https://prnt.sc/W0kLzE2IpksH

I don't know what this is, but it keeps doing it  for so long, and not starting the simulation.
All I do is as usual, import obj and add dynamic dforce.
The operaton can't be stop either. So annoying.

felis

That is springs, i.e. edges, violating collision distance.

Meaning the mesh is too dense, or the collision distance should be set shorter.

With many/much it is a higher risk of explossions.

Else you could set Self Collision to Off.

LZY

In the Surface Pane under simulation set the collision offset value to a (much) smaller number.   the default is 0.2 you want to go much lower than that on a dense mesh.

crosswind

You may see those data of vertices rest length < defined collision offset (0.0xx < 0.20) ? the offset distance is much less than default 0.2. As others said, you may set a much smaller value (maybe 0.01 or 0.005 in your case...) or even just 0 in Collision Offset to avoid long time spring check.

However, setting the value too small may bring other issues... (too close to the figure or other colliders, explosion....)  So, still, as each base vertex will participate in this check, do not make your garment with too dense mesh. From your screenshot, I'm afraid there're appr. 20K base vertices on the garment... If it's a simple garment, this value is pretty high... Base mesh, no more no less, with SubD, you'll get good enough results.

James

It can't be the collsion offset is less than 0.20.
Because I just use default settings, I haven't change anything. So the value should be 0.20
The garment is only a head scarf, covering head neck and a bit of shoulder, where in MD the particle distance is only 15mm.
It can't be 20K

https://prnt.sc/qo52cvoORm11

Hemm... is it because the MD stiches?

Richard Haseltine

0.20 is the default, which is too large for your model. If the edges, which dForce uses as springs, are shorter than the minimum offset then dForce will stretch them to respect the minimum ofset, adding energy to them, which dForce will then try to diffuse away, making the springs too short again so they need to be stretched, adding more energy to the sim, which is likely to lead to a run-away process that will break the simulation.

Yes, modelled stitcting would be an issue.

crosswind

James said:

It can't be the collsion offset is less than 0.20.
Because I just use default settings, I haven't change anything. So the value should be 0.20
The garment is only a head scarf, covering head neck and a bit of shoulder, where in MD the particle distance is only 15mm.
It can't be 20K

https://prnt.sc/qo52cvoORm11

Hemm... is it because the MD stiches?

Okay ~~ you didn't mention that you also exported topstitch as OBJ. While using MD stuff in DS, especially with dForce, you should always export topstitch as Texture (set export Type in Frabric settings...)... the difference will be huge.

You better check the Node Info or Scene Info with the scarf selected in DS... see how many vertices there are... (see my SHIRT example in screenshot 1 ~ 2. SS1 is stitches exported as Texture, SS2 is stitches exported as OBJ) .

There're two stages in spring check and each dynamic vertex participates in each stage of checking (screenshot 3 ~ 4).  The mesh density of "physical Topstitches" is usually pretty small, so if the Collision Offset is kept as default 0.20, the check will be very time-consuming ~~