Summary

Title:pflow lucid geometry--- the simulation appears offset to the emitter
Category:Crash/Critical
Status:Open
Posted By:nicoknudsen ( nico knudsen )
Date Created:18 November 2016

Problem

Description:

Hi, when i try to do a simulation with PFlow and "lucid geometry", as soon as i hit simulate the simulation happens offset to the emitter, changing to "lucid fluid" brings the simulation back to where the particles are supposed to emit. Also the geometry simulation is very slow compared to the rest of lucid. i use the latest build right now:1.3.7.9378

Also having the lucid physic button "on" doesnt seem to work at all.

link to a screenshot: https://drive.google.com/file/d/0B8eNMOre0QLSclZGdWU1bk8xakU/view?usp=sharing

Any ideas? Best /nico

Steps to Reproduce:

Thank for you the report. Any chance you can also upload the scene for testing?

Marsel Khadiyev (Software Developer, EPHERE Inc.)

Thanks for your reply, i simply did a new scene, same thing happened.

Here it is:


Attached Files:

>lucid_geometry_pflow_problem.max (933888 bytes),

Hi, any news on this? Understand you are busy, im just eager since i bougth the software for particle simulation in an upcoming project.

Best /nico

Will get back something to you on this later today. Apologies for the delay.

Marsel Khadiyev (Software Developer, EPHERE Inc.)

Fixed in next nightly (9448+)

Marsel Khadiyev (Software Developer, EPHERE Inc.)

Hi thanks for this, the particles now where generated from the right position but had other problems, the particles bounced on an "invidsible" plane higher up, multiple crashes, unpredicteble resaults and extremly slow performance compared to standard Mparticles.

Is there a stable version that works good with geometry and pflow or do you recommend the standad Mparticles?

best /nico


Attached Files:

>lucidphysic_off.max (3858432 bytes),

Thanks for the new scene, I will have a look at this bug one more time. The problem was with scene units. Since you're using inches it didn't scale properly. If you switch scene units to centimeters it should then work properly.

I am not sure why you use rigids for geometry instead of something like inflated. In fact, your particles seem small, do you need to voxelize them at all? Why not use the Lucid fluid operator instead?

If you can reproduce and send me any of the crashes it would be much appreciated.

Marsel Khadiyev (Software Developer, EPHERE Inc.)

The fix for the last scene you uploaded will be up in nightly 9453+

Marsel Khadiyev (Software Developer, EPHERE Inc.)

Thanks again.

Im doing a ridgid simulation, plastic balls spraying out from different sources and are bouncing around so i dont want fluid motion.

BTW Lucid fluid has very good performance, Lucid geometry is much slower and gives unpredictable resaults( at least for me).

Uploading another scene, this one is set to centimeters, the collision is offset to the planee position and the scene is extremly slow.

 

Do you know why? /nico


Attached Files:

>Lucid_geomerty_centimeters_off.max (1961984 bytes),

The slowness is likely due to particle generation for each sphere being created. Try switching it to inflated type for a speed up.

Marsel Khadiyev (Software Developer, EPHERE Inc.)