I'm attempting to create a cannon fire and currently using a sphere as the pyro emitter. I've had to scale the sphere down to a fairly small size and this seems to create stupidly long simulations even though scaling it up makes it sim faster?
What would the best method be to create a cannon fire? i'm assuming using a sphere scaling it down and shoving it down the barrel isnt the most effective way to do so. Also by scaling it down the explosion seems to be fairly pathetic so how can i scale up the fuel whilst keeping the size of the sphere? Im attemting to scale the explosion small enough so it doesnt touch the sides of the cannon due to the cannon having no “thickness” and hollow so the rigid body doesnt seem to work.
Advice setting up a pyro sim
6039 5 2-
- wjm
- Member
- 39 posts
- Joined: Nov. 2014
- Offline
-
- PradeepBarua
- Member
- 443 posts
- Joined: Sept. 2012
- Offline
Did you check this:
https://www.sidefx.com/index.php?option=com_forum&Itemid=172&page=viewtopic&t=34827&highlight= [sidefx.com]
There is a file by Jeff.
https://www.sidefx.com/index.php?option=com_forum&Itemid=172&page=viewtopic&t=34827&highlight= [sidefx.com]
There is a file by Jeff.
-
- wjm
- Member
- 39 posts
- Joined: Nov. 2014
- Offline
-
- PradeepBarua
- Member
- 443 posts
- Joined: Sept. 2012
- Offline
-
- old_school
- Staff
- 2540 posts
- Joined: July 2005
- Offline
Yes it's being used as a pump to shove velocity in to the sim. It mixes in velocity in to the simulation's velocity.
With fast moving colliders, you either substep the sim or manufacture velocity to approximate the fast moving collider.
Many ways to pump in velocity:
- use a sculpted piece of geometry with velocities applied as in the first example
- pre-simulate interesting velocity volume sequences and save to disk then in future simulations, use these as pump fields.
- pre-simulate simulations such as RBD's, particles and cache to disk. Then use these geometry objects with their velocities to pump in to future simulations.
I attached an example file of the pre-simulation approach. The idea is to simulate your fast moving collider but slow down time a lot. Then save the velocity volumes to disk for later use. Change the seeds and generate several versions of the same explosion for variation for several cannons.
I have cached many volume sequences to disk over the years for later re-use. Just like saving your favourite texture maps and texture sequences for later re-use.
All of these methods allow you to highly direct your simulation for predictable results as you increase the resolution of your simulation which is critical for production.
With fast moving colliders, you either substep the sim or manufacture velocity to approximate the fast moving collider.
Many ways to pump in velocity:
- use a sculpted piece of geometry with velocities applied as in the first example
- pre-simulate interesting velocity volume sequences and save to disk then in future simulations, use these as pump fields.
- pre-simulate simulations such as RBD's, particles and cache to disk. Then use these geometry objects with their velocities to pump in to future simulations.
I attached an example file of the pre-simulation approach. The idea is to simulate your fast moving collider but slow down time a lot. Then save the velocity volumes to disk for later use. Change the seeds and generate several versions of the same explosion for variation for several cannons.
I have cached many volume sequences to disk over the years for later re-use. Just like saving your favourite texture maps and texture sequences for later re-use.
All of these methods allow you to highly direct your simulation for predictable results as you increase the resolution of your simulation which is critical for production.
There's at least one school like the old school!
-
- Jebbel
- Member
- 1 posts
- Joined: Jan. 2015
- Offline
-
- Quick Links