You could probably do a simple DOE on the few tests you've done so far, and greatly increase your iterative process. This is especially easy using fast CFD analysis.

 

I dont know what DOE is otherwise I would use it

 

A Design of Experiments. Its a 6-sigma tool. Basically, you have inputs (your Xs) and you have your outputs (your Ys). You are trying to optimize on certain Ys. So you do a few experiments. Each time, you input your X (say, intake inlet area, intake runner area, intake runner length, etc...everything you can put into your model)...you get an output, say fluid delta pressures, fluid velocities, etc.

So you want to optimize at a certain set of Ys (minimize pressure loss, or something like that)...you input this X & Y data into some software (Minitab, Design Expert, even Excel will do will some lengthy VB programming). Depending on how much data you do have, you can build transfer functions that simulate each X to a Y (kind of like trendlines, but can be logrithmic, etc). Now, you converge on what Y you want very fast...cause you solve for the Y you want, then validate it in the CFD. Maybe its not quite there, but you just ran another dataset that you can input into your X and Ys, and generate new transfer functions that are stronger, and keep getting more accurate. It really saves a crap load of engineering design work. It is a Design for Six Sigma tool.

I use it way to much at GE designing compressor airfoils...in that instance, we model the blade varying say, thickness and chord and several locations along the span of the blade. We have special software that runs these simulations through Ansys relatively quickly (say a few hours). We then have a bunch of varying thickness and chord sets across the span of the airfoil (these are our Xs)...Our Ys are the frequencies. We optimize the Ys to move the frequencies (or mode shapes) out of the driver boxes (natural frequencies, blade count drivers, etc).

It is really a wonderful tool. Do some digging at Boarders or Google. I'm sure there's a ton of stuff about it.

EDIT: Have a look here. It says free trial. I don't know if the software is restricted on what you can do in the trail...worth a shot.

http://www.sigmazone.com/doepro.htm

Feel free to PM me if you have any questions.

 

how do you have access to cfd stuff?

 

very nice

 

trina that is sweet, you definately got the smarts going on

 

trina your design looks like a Leehmann(sp) manifold that the turbo audi quattros used and skoda uses on their WRC cars. Dahlback racing makes one for the 1.8t.

http://www.tjmmotorsport.com/store/p...cat=263&page=1

there is some good info posted here if your interested in reading through all the crap
http://forums.vwvortex.com/zerothread?id=1731880

 

http://www.ncbi.nlm.nih.gov/entrez/q...&dopt=Abstract

http://www.synergeering.com/ebm.php

some interesting links to sintering of hard to make objects not just out of plastics.

 

I didnt even think about making the manifold from plastics. The Nylon idea is really seeming to be the winner. I need to find a friend of a friend that can make a few test manifolds from that stuff. It would be lighter, one piece, and would be insulated as opposed to the aluminum taking in the heat. It would automatically make it the best phelonic spacer around

Thanks for the find!

 

Engineers of course

 

True, but getting more flow with less boost reduces backpressure (turbo doesn't work so hard) which can increase the knock threshold.

Besides, the whole purpose of a modified manifold is to reduce the boost pressure required for a given flow. Otherwise you would just crank the boost up way high on the stocker.

You read "The Scientific Design of Exhaust and Intake Systems" by Philip H. Smith? Old and no CFD, but a lot of interesting experiments.

 

so what is going on here? will one of you brainiacks clue me in?

so you input volumes and shapes and the program outputs the flow characteristics, you improve your shapes and the program tells you what you have done?

 

I know what a DSM head looks like
The only evo head I've looked at is the BJs level 5 head sitting on my shelf, I haven't even pulled the intake manifold off the stock block yet. The floor on that head is not completely flat at the entry.

 

I see someone already tried to make a manifold that is extremely similair to this test one

I wish there were some more specs on what that guy made. From the pics it appears that his runners are WAY too long... pretty much making it a better flowing stock intake.

Anyways, I have been making lots and lots of different versions on the intake manifold and I am learning alot about it. I finally figured out also how to completely mimick the whole intake cycle for each cylinder. It takes my poor PC a VERY long time to make these computations but it looks pretty cool when it finishes.

I have also been doing alot of research on exactly how to actually produce these things. Id like to be able to make it so that I could offer them to the public for an extremely cheap price compared to most sheetmetal intakes and make them flow better. Ill keep you guys posted as I get further.

I also am going to start making exhaust manifolds that are equal length and everything else. It will take awhile to simulate the exhaust pulses accurately but I will be able to actually put a turbine wheel on the end of the manifold and simulate spool characteristics and total flow. It will take awhile but will be nice when done. Theoretically should be extremely good.

Does anyone have any requests on either the intake manifold or exhaust manifolds? Right now the intake manifold I am trying to make the best I can with the stock throttlebody up to a 75mm one and try to keep extremely close to the stock upper intercooler piping. As for the exhaust I am thinking to make it bolt up to a T3 style turbo (GT35r type) but I could make it divided (So it could work with a T4) but then let people choose the flanging on it. Anyways, just wanted to keep you guys updated. The production costs seem to be a decent amount so hopefully I will be able to make it so others can get one cheaply instead of just making a few for myself and our cars here.

 

i think for the exhaust manis that's exactly what fullrace does... i've heard mention of them working with cfd.