CFD simulation for aux fuel kit

[OGR34]

What I have is an auxiliary fuel kit that is to be installed between the throttle body and the intake manifold. What I want to do is simulate where the atomized fuel would end up in the system at any given air flow rate. Would anyone be able to recommend me a good CFD program to accomplish this? And is this possible just using SolidWorks? Any help would be great!

 

[PJGD]

As a first approach, I would probably try OpenFoam:

http://www.openfoam.com

I have no idea if this will provide the answers you are seeking or even if what you are looking to do is even sensible, but it may be worth a try.

PJGD

 

[BrianPetersen]

Unsteady turbulent two-phase flow with evaporation will be a challenge!

Look at nitrous plates that are available at hot-rod shops for the traditional carbureted intake manifolds. Usually they have a tube extending across the width of the opening with holes drilled in it to hopefully distribute the flow more-or-less evenly across the width. I doubt if anyone did any CFD on these ... just old-fashioned ruff calc followed by cut and try.

 

[OGR34]

Thanks for the input guys. I'm really actually only interested in the results at WOT now that I think about it. Would it be less of a challenge if I have a set air flow rate at WOT? The reason why I'm trying to do this is because I want to have a visual comparison on how fuel will be distributed depending on how I point the injectors (angled, straight, upstream, downstream etc.). I believe if I could simplify the scenario enough, I would be able to come up with some viable results that will apply to the real world without unnecessarily over-complicating the simulation model. I'm not very familiar with CFD so I've been trying to find some tutorials online, but unfortunately all I've found are oversimplified 2D scenarios. What do you guys think?

 

[braddles90]

CFD generally has two paths - prohibitively expensive for software that is mostly set up for you (CFX, FLUENT), or free, but you'll spend weeks learning how to code to set it up to run properly before even starting to model your problem.

If you're looking at a simplified method, check the software has the ability for Lagrangian particle tracking. As mentioned, this is a pretty complex case (disperse 2-phase flow, gas/liquid interactions, probably evaporation, etc) - but if you just want to get a feel for fuel distribution, you can use particle tracking as a 'poor man's 2-phase flow'. Instead of solving two lots of the N-S equations for each fluid phase, particle tracking basically just injects particles at a point, and tracks how they flow through the single phase fluid. This can give a reasonable idea on the spread of fuel, but you'll have to check whether you can set 'injection patterns' to mimic injector spray shape, etc.

As I mentioned though, it all depends on how much cash you have - from memory Fluent was a reasonably-priced system, but probably still pretty pricey. Some distributors provide leasing costs though, so you can just hire the licence for as long as you need it. Others offer 'cloud based' versions, so you only use the modeller licence and then pay for server time when you want to run your model. There might be something available for Solidworks (not sure sorry).

 

[gruntguru]

"As mentioned, this is a pretty complex case (disperse 2-phase flow, gas/liquid interactions, probably evaporation, etc)"

Dont forget intermittent flow from the plenum into 6 individual runners - in firing order. The flow/time profile for each runner will be complex and rpm dependant although it could probably be adequately approximated after researching the literature.

je suis charlie

 

[Panther140]

The idea of calculating all of these details sounds nice, but I think testing could be your most effective tool. You might be able to calculate things like directional change, then consider mass/momentum of the fuel within the air, temp, and low pressure zones. Then form a hypothesis based on that, and test it on a wet flowbench. You could make observations based on that test, find correlations in your calcs, and continue learning. You could also have a camera lense (like a bore scope) within the wall to actually watch the fuel mixing with the air.

Another thing that I think would be interesting is using some sort of dye on the inside of the ports to give you an idea where the points with most flow friction are. The fuel would “wash” the dye more where those points happen. Of course, I made that idea up as I went along. I have no idea how/if that would work. But it could be possible to allow the fuel to leave evidence wherever it is most concentrated


"Formal education is a weapon, whose effect depends on who holds it in his hands and at whom it is aimed." ~ Joseph Stalin