pyro question: sources with different emission ?

[ranxerox]

hey all, I've been delving into the pyro solver with the new masterclasses and have come up with a fundamental question. If I have 2 sources, how do I control the emission rate (of smoke, fuel or temperature) PER source ? In the pyro masterclass, it is mentioned that the parameters on the pyrosolver node under sources->emitters can be keyed, but they control the emission of all of the sources (the 'source emission mask'). How does one control the output of each source independently ?

I've been looking at this for several hours and haven't found anything. If anybody knows, feel free to chime in. If I figure something out I'll respond to my own post.

thanks

-ranxx

[edward]

I haven't tried but can't you just put down two different Pyro Solver nodes and wire one for each source before merging them together?

[Jason]

I haven't tried but can't you just put down two different Pyro Solver nodes and wire one for each source before merging them together?

I think what might be missing from the new PyroSolver is Source Divergence and Flame Divergence - two parameters from the DSD Solver that were quite nice to use. Now you must use "Gas Released" for the flame divergence, which works well - and I think it's best to model your source's vectors. This can be engineered in yourself in a couple of nodes if you know how to do it...

You can insert a lot of Fuel into the system too - and if you have radial point velocities on your source geometry and set your Source to "Point Velocity" in the Source tab, you can get some of this to happen. I'll put a little example together later on if I have a moment.

EDIT: Here you go...

shaped.mov

shaped_charge_pyro.hipnc

[ranxerox]

thank you for the hipfile, It is very interesting. I thought that you had to add 'v' to the dopimport node in order to get it to be used by the solver and I like your method of calculating the point velocities defined by the deformation.

But getting back to the original question:

I want to have 2 (or more) emitters in the same fluid box with different emission rates (keyed preferably).

this should be easy no ?

thanks

-ranxx

EDIT: I realized that maybe I wasn't clear, by emission rate I don't mean just velocity of emission, but the density of emission and whether any emission if happening from a given source at all. Hmmm, if I add a density attribute is that gonig to be used by the solver ?

I think what might be missing from the new PyroSolver is Source Divergence and Flame Divergence - two parameters from the DSD Solver that were quite nice to use. Now you must use "Gas Released" for the flame divergence, which works well - and I think it's best to model your source's vectors. This can be engineered in yourself in a couple of nodes if you know how to do it...

You can insert a lot of Fuel into the system too - and if you have radial point velocities on your source geometry and set your Source to "Point Velocity" in the Source tab, you can get some of this to happen. I'll put a little example together later on if I have a moment.

EDIT: Here you go...

Edited May 15, 2009 by ranxerox

[Jason]

Hey,

There is nothing stopping you from having two Source Relationship DOPs (with their corresponding Static Objects) in your DOPnet. You don't have to try to get a single source node setup solve for two different fuel emitters... rather leave them separate and just add nodes to the dopnet.

[ranxerox]

no I understand that, this is what I am doing now. The problem is that in order for them to emit at different times I need 2 pyrosolvers, in which case they aren't interacting in the same fluid box. Maybe I'm not understanding something then. What I want is for each of these sources to emit smoke at different times into the same fluid box (with the same properties eg. viscosity etc).

-ranxx

Hey,

There is nothing stopping you from having two Source Relationship DOPs (with their corresponding Static Objects) in your DOPnet. You don't have to try to get a single source node setup solve for two different fuel emitters... rather leave them separate and just add nodes to the dopnet.

[Jason]

no I understand that, this is what I am doing now. The problem is that in order for them to emit at different times I need 2 pyrosolvers, in which case they aren't interacting in the same fluid box. Maybe I'm not understanding something then. What I want is for each of these sources to emit smoke at different times into the same fluid box (with the same properties eg. viscosity etc).

-ranxx

Ah, yeah - I see what you mean... it's easy enough to add two sources to a single pyrosolver - (attached) - and with those it's easy to have different source velocities, but it's not easy to have the other source properties be unique, like the "Fuel Amount", and "Source Noise". I think you've found a bit of a weak spot there in the current Pyro setup - and perhaps SESI could solve this by making the fuel amount be another property in a way similar to the other Physical characteristics of a Static Object. I really think think is something that should be raised to SESI to come up with a good easy solution for.

EDIT: I've just registered a bug with SESI because you *should* be able to select your second source object and do a "Fuel From Object" and it should be added to the simulation. I did this manually.

twosources.hip

[ranxerox]

thanks for your help. I learned a lot from your setup of the shaped charge. I also have an email into support. There must be a way to do this ! I'll post any findings I come across here.

cheers

-G

Ah, yeah - I see what you mean... it's easy enough to add two sources to a single pyrosolver - (attached) - and with those it's easy to have different source velocities, but it's not easy to have the other source properties be unique, like the "Fuel Amount", and "Source Noise". I think you've found a bit of a weak spot there in the current Pyro setup - and perhaps SESI could solve this by making the fuel amount be another property in a way similar to the other Physical characteristics of a Static Object. I really think think is something that should be raised to SESI to come up with a good easy solution for.

EDIT: I've just registered a bug with SESI because you *should* be able to select your second source object and do a "Fuel From Object" and it should be added to the simulation. I did this manually.

[Jason]

There is definitely a way:) You can dive into the PyroSolver and see how they construct a source from the root DOP nodes and mimic that. But it's not always very obvious - there is use of index-fields, source-masks and all kinds of funky stuff that makes such a setup hard for the average user to understand, let alone construct for themselves. I'm pretty sure SESI just didn't consider such a setup quite yet and it'd be child's play for them to add support for it (by way of another HDA, without any C++ development).

thanks for your help. I learned a lot from your setup of the shaped charge. I also have an email into support. There must be a way to do this ! I'll post any findings I come across here.

cheers

-G

[Jason]

Apparently, the "Fuel From Object" is the wrong tool to use for this - this tool adjust the Initial fuel on the PyroObject, and that fuel isn't automatically at ignition temperature, so it doesn't burn. (my bad, but the name was a little misleading to me - perhaps it should be "Initial Fuel From Object"). What you really want is "Source From Object", which will create you an additional fuel source for your Pyro box.

[ranxerox]

right, but it still won't let you modulate the emission seperately (not in an easy way anyway). The initial response from sidefx was that this is not currently possible, but they have their developers looking at it. In the pyro masterclass Coen mentions that the parameters on the pyro solver regarding emission are just multipliers into the source mask, so maybe there's a way to modify the source mask for each object. It does seem a little daunting though. I'll take another look at it on monday

-ranxx

Apparently, the "Fuel From Object" is the wrong tool to use for this - this tool adjust the Initial fuel on the PyroObject, and that fuel isn't automatically at ignition temperature, so it doesn't burn. (my bad, but the name was a little misleading to me - perhaps it should be "Initial Fuel From Object"). What you really want is "Source From Object", which will create you an additional fuel source for your Pyro box.

[ranxerox]

OK ! So there is an easy way to do this: between the geometry you are using as a source, and the dopimport node, put down an isooffset(fog volume) followed by a volumemix (with value set to $V*scalefactor). Then on the associated static object in the autodopnetwork click on 'use forming geometry). This works well and also allows you to control the emission with a volume vop.

weee.

-ranxx

right, but it still won't let you modulate the emission seperately (not in an easy way anyway). The initial response from sidefx was that this is not currently possible, but they have their developers looking at it. In the pyro masterclass Coen mentions that the parameters on the pyro solver regarding emission are just multipliers into the source mask, so maybe there's a way to modify the source mask for each object. It does seem a little daunting though. I'll take another look at it on monday

-ranxx