CFD For Engines 5

[Panther140]

Hello all.

I have been on a quest to put more of my ideas into reality, and do it efficiently.

Right now, I need advice on choosing a CFD program to learn with. My main intents and purposes for this are to analyze engine intake systems on an engine, prove/illustrate concepts, then proceed with a more concise prototyping process. I don't need to analyze with precision. I simply want to observe an approximation of the flow so that I can possibly size the system close to the engine's demands.

Are there any high level programs that would allow me to make simple models? I have the system modeled in Autodesk inventor 2014, but I have not yet tried any CFD modeling. Let me know of any programs that stand out as the best CFD program for somebody who learns alone.

Thank you.

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

 

[Tmoose]

You may qualify for a free single cylinder version of Lotus engine simulation, which claims to model the gas dynamics in the engine manifolds and enables the complex operating modes used in modern engines to be simulated.

http://www.lotuscars.com/engineering/engineering-software

 

[Panther140]

Thanks for the link. That program has the option to change the cam profile, which is an integral part of what I'm doing. I'm going to look into it further

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

 

[RodRico]

If you happen to be an active duty or veteran of the US armed forces, you can get the full Solidworks suite including simulation for $20. It's an amazing bargain for which I am very grateful !

https://store.solidworks.com/veteran/default.php

 

[MikeHalloran]

Wow, RodRico, that is an amazing bargain. Almost makes me want to sign up.


My cheapo former employer wouldn't spring for the full simulation package, even for just one seat.
I was able to do some useful comparative simulation using the demo packages that come with a plan vanilla seat, but the FEA demo is less limited than the CFD demo.

Panther, SolidWorks is not super easy to learn, especially alone.
... partly because of its Gallic logic.
... and partly, as of SW2010, because of phase errors between the included training materials and the actual program. I hope that's been corrected.
... and partly because my then employer wouldn't allow viewing YouTube over their network.

Lots of SW tutorials are on YT.
Most are at least as good as the official SW stuff; many are better.
You probably need one computer just for SW and nothing else (and it must have a supported video card and lots of memory),
and another computer just for viewing YT.

It would be good to get friendly with your local Value Added Reseller (dealer), and attend whatever free workshops they offer.



Mike Halloran
Pembroke Pines, FL, USA

 

[RodRico]

Mike, I imagine you might know a veteran who would be willing to lend you use of "their" copy

 

[Tmoose]

Or a student. Especially if you are paying for their tuition etc. I think that version expires in a year or something

 

[Panther140]

Id like to dig into solidworks. I've only used autodesk programs so far. I believe they have a CFD program. Has anybody tried it? I'm hoping that one would be easier to transition into for a user that is familiar with Autodesk. I've been using Autodesk since I was a young boy and my dad was in college (making stick figure houses at age 6!) That one might work well in my situation, but that is just background information. I'm not fixated on Autodesk, its just always been what I have on my computers. "the best CAD program is the one you already have downloaded"
.

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

 

[MikeHalloran]

I started using AutoCAD around version 9.

It doesn't help at all with Solidworks; they are very different tools.

Mike Halloran
Pembroke Pines, FL, USA

 

[Panther140]

Oh I see. Does anybody have experience with Autodesk's CFD program?

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

 

[dgallup]

I've had a little exposure to Solidworks CFD package. Not as bad as I expected. Most analysis packages built in to CAD systems have very limited capabilities and simplified user interfaces. While the Solidworks program is easy to use, there seems to be a pretty good bit of capability under the surface.

However, I really have to ask what you expect to get out of CFD. If you just want to look at say an inlet port under steady state flow then a dumbed down CAD integrated CFD program should be fine. If you want to investigate effects of compression ratios, spark timing, etc., then a 1D simulation program like Tmoose linked above is probably the better tool. If you really want to model a running engine, it's going to be very difficult as you not only need lots of moving parts but also heat release, etc. For that you are going to need an advanced CFD tool like FIRE and a CPU farm to run it on.

----------------------------------------

The Help for this program was created in Windows Help format, which depends on a feature that isn't included in this version of Windows.

 

[Panther140]

I don't need to simulate a running engine right now. I would like to be able to demonstrate the concept of this system at a very high level. This will be a tool used to communicate the idea to my team/boss. Any 4 cycle engine would be sufficient to demonstrate this concept.

What I need to simulate:
- Piston going up and down in cylinder in a typical fashion found in engines.
- Intake valve opening/closing to my desire. 1 single non-variable cam profile/timing
- Exhaust valve that opens and sufficiently evacuates exhaust from cylinder.
- Intake plenum/runner/unique-stuff that I have modeled in Inventor.




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

 

[RodRico]

Take a look at OpenFOAM at

http://www.openfoam.com

. It's open-source, free, and often used for internal combustion engine modeling. If you Google "OpenFOAM Internal Combustion Engine" you'll see a number of papers on the topic as well as a tutorial on the OpenFOAM Wiki. There's also a video demonstrating Inventor with OpenFOAM at

http://www.you2repeat.com/watch/?v=sTAKstAJ4eM.

I haven't used OpenFOAM, so I can't comment on it, but it seems pretty popular. I had planned to use it when I started my project, but then found out I could get SolidWorks for next to nothing. OpenFOAM remains an option if I run into problems.

 

[JayMaechtlen]

Looks like veterans are eligible for the student package, which runs for one year.
A great deal if someone wishes to learn the software, of knows a bit and wants to do some hobby work.
You'd not want to be halfway into a project, even if a hobby project, and hit the end of the license.


Jay Maechtlen

http://www.laserpubs.com/techcomm

 

[RodRico]

A student stops being a student after a while, but a vet remains a vet for life. Hopefully they'll only charge me $20 again for another year.

 

[braddles90]

Hey Panther,

For those operating simulations you've noted above I think you need a mix of both simple CFD and the 1D simulation software.

Trying to run CFD on a full cylinder cycle with moving intake/exhaust valves is the stuff of PHD theses/heavy heavy calculations. YOu need things like those fancy 'stretching' meshes to take into account the changing flow domain size, or to remesh the domain at certain time-steps, etc - that kind of thing is incredibly computationally expensive.

You could try to use a mix of steady state, fixed domain simulations in 3D CFD (Solidworks) to get an estimate of valve flow coefficients at different amounts of lift, then use that data in the 1D simulation to see how that affects the whole engine cycle. I get the feeling this approach is the exact reason 1D software like GTpower exist - it's easier to take this approach than to try and model a full working cycle in 3D CFD like Solidworks and ANSYS. Probably better to try to start with the simplistic models and software as well, since the major CFD packages to do this kind of work are super expensive.

I'm certainly no expert in this (I haven't used the 1D softwares before but read about them where I can), but when I was doing CFD stuff at university some years ago, all of the trainings we got from ANSYS were showing just how 'cool' their moving/stretching mesh stuff was; but to get to the point where those models work and are accurate enough would be an incredible amount of

Good luck though!

 

[mohawk2]

I'd like to recommend the Yoshimura Dual Stack (

http://shop.yoshimura-jp.com/files/img1/765-T2--0600.jpg),

as an intake upgrade. I made my own, but that shows the concept.

From my own research & development using the idea, it is possible to have the shortest possible intake runner length, i.e. the metal head intake tract with just a nice rounded/ellipse feed lip. Then a dual or even triple stack beyond this. When the inner end of one is in the correct position relative to the outer of the one below, the gas pulse bridges the gap & thus it acts as a longer runner at low rpm & a shorter one at high rpm.

Adding these to my project bike, added 6hp at peak power rpm, with no other mods, due to the shortening of the intake tract, but with the dual stack, there were NO losses at lower RPM. The other advantage was the the torque curve that flattened & dropped after 8500rpm, stayed flat to 10,500rpm before falling more gradually. Thus there was a larger area under the torque curve !

 

[EHudson]

Another SW option is to join the Experimental Aircraft Association ($40/yr) and get SW as a bonus:

http://www.eaa.org/en/eaa/eaa-membership/eaa-member-benefits/solidworks-resource-center

 

[Tmoose]

It appears the EEA version may be the student design kit, which includes SW Premium.
There look to be some basic CFD and FEA capabilities.

http://www.solidworks.com/sw/docs/3DS_2017_SWK_Premium_Datasheet.pdf

 

[Panther140]

Thank you guys. I didn't know that about EAA memberships. I went to the EAA airventure airshow in Oshkosh, Wi last year. It was fascinating. I'm going to join that organization.


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