Designing an Intake Manifold: Books & Software

[morry]

Hi guys,

Im thinking about designing an intake manifold for my car for fun. I don't necessarily want to make it or anything but I think it would be a good project to do.

Can you guys recommend me some books or software that would help me out?

I don't have much of an idea with CAD, but I am starting to learn.

 

[Cliff_J]

Well that could really be an interesting project. Is this for an inline engine configuration or a V engine? The inline should be far easier since they don't normally support coolant passages and other extra passages.

If you don't actually plan on making it you could simply draw the design as a solid and then 'wall' it using the software if it has that option. Or make a copy of it a percent smaller and then a boolean subtraction to remove the material from the center of the solid.

 

[morry]

Well, I suppose it would be great if I could actually design something that could be manufactured, but I doubt this is possible at the moment.

So Ill settle for something that looks the goods at the moment. It will be for a V engine.

I have access to solidworks, will this program be able to do the job?

Is there a "bible" of manifold design in the industry?

Thanks for the help. :D

 

[Averagesupernova]

You could study manifold design for engines a very long time and only scratch the surface. What are the goals? Where do you want the power in the RPM range? How is your camshaft ground? All of these things tie together.

 

[morry]

Ive had a look at a couple of books at uni and I've seen its pretty complicated. Theres massive books dedicated to it.
Id be more interested to know about the effects of things such as plenum volume, runner length, throttle size etc. Without going through the derivations and all that background stuff.

 

[morry]

Any tips for using Solidworks or other forms of CAD? I am just starting out, I have no experience whatsoever. Any good sites to visit etc?
Cheers guys.

 

[FredGarvin]

To do a complicated part like a manifold, you are going to have to get good with surfacing. You're not really going to be able to do a part like that with just solids. You have tackled a tough project for a first timer.

 

[morry]

Hmm, this is sounding more and more like a challenge with every post.

Im hoping to work on the FSAE team when I am in 4th year, preferable in the engine department, so designing manifolds etc will be part of what Ill have to do.

Thanks for the input guys.

 

[faust9]

The reference guide: http://files.solidworks.com/supportfiles/Reference_Guide/2004/ReferenceGuide_ENG.pdf

A lot of manifold design in CAD boils down to using pipes. You decide on a path and tell your CAD software to route a pipe of x diameter and y thickness along the path. SolidWorks can do this.

Now as mentioned above, manifold design is voodoo science. You can come up with the most amazing design on paper. Your CFD software might tell you your design is top notch. Then you put it on a car---it doesn't work nearly as well as you hoped it would. What's worse, the piece of junk manifold already on your car ends up outperforming your latest and greatest design. I've been there, done that.

When you have a design the only true test of is performance is on an actual engine. The dyno(s) will tell you if your design is good or not. The FSAE drivers will tell you if they like the intake or not.

You can start to look into Helmholtz frequencies or resonance. Talk to the FSAE team members. They will probably tell you what works and what doesn't. They probably have a good feel for optimal runner length and diameter and airbox volume already.

Well good luck.

 

[Cliff_J]

Hmm, I was picturing something simple like the 'log' style intake on an older BMW inline 6 that wouldn't be too hard to model with one horizontal pipe, 6 curved pipes to the head, and a mount for the throttle butterfly. But you want something far more complicated!

It seems the fastest way to prototype and test different intake combinations for the desired powerband is something along the lines of the tuned velocity stack with a pipe for each cylinder that ends with a bell shaped mouth into a common oversized plenum. Its even used in production with the Mustang Cobra R and the M5. (last pic on the links)
http://www.soligen.com/parts/cobrar-man.shtml
http://www.schatzandkrum.com/gallery/gallery.htm

I also attached a couple pics from Holden's experiments in next door NZ.

 

[faust9]

Hmm, I was picturing something simple like the 'log' style intake on an older BMW inline 6 that wouldn't be too hard to model with one horizontal pipe, 6 curved pipes to the head, and a mount for the throttle butterfly. But you want something far more complicated!

It seems the fastest way to prototype and test different intake combinations for the desired powerband is something along the lines of the tuned velocity stack with a pipe for each cylinder that ends with a bell shaped mouth into a common oversized plenum. Its even used in production with the Mustang Cobra R and the M5. (last pic on the links)
http://www.soligen.com/parts/cobrar-man.shtml
http://www.schatzandkrum.com/gallery/gallery.htm

I also attached a couple pics from Holden's experiments in next door NZ.

The problem with large plenum designs is throttle response gets slower. This is especially noticible on FSAE cars because the engines used there are around 600cc spin at about 10,000 to 15,000 rpm and have to breath through a 20mm restrictor. With large plenum intakes you get a very pronounced response lag.

The way to determine optimum length and diameter it to make a plenum with changable velocity stacks inside. You swap in velocity stacks of various sizes and try to get as much power out as possible. The same holds true for diameter. I'm sure the FSAE team at morry's school have done this at least once, so they have a good feel for what should and should not work.

One of the reasons you don't see large plenum/short runner designs more often in street apps is because these designs push the power/torque bands up. It becomes more difficult to drive a car with the band pushed into the 6k RPM range than a car with the power band in the 3k5 range. Grandma doesn't want to have to dump a clutch at 4k to pull away from a stop light; however, Joe midlife crisis does.

 

[Cliff_J]

Good point, but in large plenum designs in the M5 or Holden apps the butterflys are at the end of the runner so throttle response is not affected, but yes with a butterfly at the front it would be lag. I'd just like to know what the effect the close butterflys have on the pumping losses.

While on this topic, I always assumed the reason the 60s era tunnel rams, known for their high end powerband, had such long runners was to compensate for their large cross sectional area in addition to cooler charges by sheer distance from the head. Yes/No? Kinda odd to to think would could be done today with composities and their extremely low heat transfer coefficients, on the GM LS_ varieties you can touch the top of the intake without being burned.

3k5 - read a lot of schematics lately? How about some of these cars today referred to on the internet as 'dyno-queens' where they can push absurd numbers of RWHP but its such a narrow untractable powerband they can barely eek out 1/4 runs with cars of half the peak power. I'm quite impressed with the Jegs motor challenge where its average power and torque over a fairly broad RPM range, so much more practical than some small peak at the top with sacrifices everywhere else.

 

[morry]

Couple of good looking manifolds there cliff. That M5 one looks crazy. I've seen a lot of those trumpet style manifolds on the V8s before. I like the look of that mustang manifold. Something like that doesn't look too hard to design.

Ill hopefully be having a chat to the FSAE guys this week so hopefully they know a bit about how to go about manifolds too.