Volunteer for LPPX Sparkplug Simulation Team

[Rezwan]

Murali has been wanting to resolve thermal and electrical breakdown issues of LPPX’s evolving spark plug designs with finite element analysis in COMSOL or similar, but there just hasn’t been time.

Rather than continuing to put it off and ending up with a “make and break” approach, they’d like to reach out and see if they could put together a team of volunteers that could work together with Murali for spark plug simulations, similar to the team supporting plasmoid simulations.

To catch up on the spark plug issues, you may want to check this post, New Spark Plugs Pass First Test, and look at the this forum discussion.

To join the volunteer team, join the conversation on this new thread. Collaboration rocks!

[KeithPickering]

Hi, sounds interesting, and count me in. My computer is a dual-boot Linux (Ubuntu 10.4) and Windows Vista, so I can handle pretty much anything you have. I haven’t done finite element analysis, but I have done other kinds of computation-intensive work like aerodynamics.

[QuantumDot]

I don’t know quite what it is you want but i would be willing to try.

[zapkitty]

QuantumDot wrote: I don’t know quite what it is you want but i would be willing to try.

FEM skillsets can’t be distributed easily 🙂 … but perhaps the computations can?

[Brian H]

You might like to get your hands on this new linear equation algorithm presentation: http://www.physorg.com/news/2010-10-barrier-important-class-linear.html

It looks like just about any SDD-based operation is about to become exponentially faster:

The algorithm, which applies to an important class of problems known as symmetric diagonally dominant (SDD) systems, is so efficient that it may soon be possible for a desktop workstation to solve systems with a billion variables in just a few seconds.

I wouldn’t be at all surrounded if this has major implications for the Plasmoid simulation requirements, too. And once the engineering phases start in earnest, it could multiply speed of progress significantly!

[jamesr]

Brian H wrote: You might like to get your hands on this new linear equation algorithm presentation: http://www.physorg.com/news/2010-10-barrier-important-class-linear.html

It looks like just about any SDD-based operation is about to become exponentially faster:

The algorithm, which applies to an important class of problems known as symmetric diagonally dominant (SDD) systems, is so efficient that it may soon be possible for a desktop workstation to solve systems with a billion variables in just a few seconds.

I wouldn’t be at all surrounded if this has major implications for the Plasmoid simulation requirements, too. And once the engineering phases start in earnest, it could multiply speed of progress significantly!

Interesting. I’ll have to read the paper in more detail, but I suspect it has limited use for plasma simulations which are highly non-linear by nature.

On a related note Comsol are running some seminars over here (UK) I assume they’ll be doing some similar ones in the US on their new plasma module.
http://www.uk.comsol.com/events/cps/12588/. From what I understand it won’t be able to cope with the complex nature of the plasmoid formation, but should be able to model the switch spark gap breakdown fairly easily.

[DerekShannon]

For further clarification, suitable volunteers would need to already have access and experience with COMSOL or other finite analysis programs, since we won’t be able to provide software (the licenses are rather pricey!)

[KeithPickering]

There are a number of open-source FEA packages available. See: http://en.wikipedia.org/wiki/List_of_finite_element_software_packages for a list.

Does anyone have experience using any of these? (Thinking esp. of Rezwan … )

If not, I note that FRANC2D and FRANC3D are specifically designed to model crack propagation, which is (maybe?) what we’re trying to do here. Their homepage is at http://www.cfg.cornell.edu/software/franc2d_casca.htm

[emmetb]

Just installed COMSOL 3.4 is that recent enough for what you’re trying to do here? Note i have no experience with comsol (or any other FEM package for that matter) so i’m afraid i won’t be of much use myself, but i would be curious to watch over the shoulder of somebody more experienced.

[jamesr]

We have a version 4.0a licence installed on our network (with CFD, ACDC and Chemical Reaction Engineering modules). It is used a bit by someone in one of the other research groups I think. It seems to run up OK on my machine. I tried to have a go with it a while ago (just going through the basic tutorials) but never got very far, but if you send me a basic model to get me started I can have a go.

[Lerner]

Great! If we sent you some solidworks drawings and where the potential is, could you get us a plot of electric field strength?

[jamesr]

Lerner wrote: Great! If we sent you some solidworks drawings and where the potential is, could you get us a plot of electric field strength?

I’m not sure it imports solidworks files directly, but I’m sure I can convert them to whatever it needs.

But, as I said I’m very much a beginner with Comsol so don’t trust any figures I get without checking them!

[Lerner]

Hi JamesR,

Could you move this conversation to email and email me at eric@lpphysics.com?

Thanks,

Eric

[emmetb]

Would you mind CC-ing me?

[felix]

Hi everybody, I’m new on this forum.
I can colaborate with this task. I have Solidworks 10 & Comsol 4.0 and some experience with simulation of spark gaps. What’s the next step?

Felix

[Author]

Posts