Exhaust Design Project

[Domesticated]

For the next 6 weeks I will be looking into header design from an engineering point of view. I have my stock header still and my Feel's header that's on it so I have some good comparisons I can make. I will be actually measuring temperature and pressure deltas in the manifold at different RPMS for both headers and getting a generalized head loss factor for the whole header back system. I'm also going to be using the fluids lab at the university so I can be in a bit more of a controlled environment and hopefully I'll be able to make a decent correlation between the bench results and the real results. Then I'm going to some CFD modeling and try to find correlation that way. Then finally hopefully build a math model based of the exact Navier Stokes Equation for three dimensional flow of compressible fluids using all the data to try to optimize the design for a specific setup for a certain RPM range. It's for my Advanced Fluids Analysis class. Next semester I will actually be taking an advanced CFD course and I will continue the project then in more detail.

So with that said it would be awesome if somebody had a Bisi H22 header laying around who could take some measurements for that so I could model it too and see how the math model changes. I'm just getting into now so I still have to figure out my exact approach. If anyone knows of some useful scholarly information that could help feel free to throw it in here. My research starts tonight.

[accord4me]

This is great man!

I love looking at automotive dynamics from an engineering point of view!

I wish I had a Bisi header to help you out.

I'm subscribing to this thread and can't wait to see the progress.

[sonikaccord]

You must be in mechanical engineering? I'm electrical so I have very little knowledge on gas/fluid flow. I'll have to break open my Thermodynamics book again, but I'm willing to help with the math and some calculations or just to look over work, blah blah, etc.

I'm guessing some variables would be:
Velocity
Temperature
Pressure
Pipe area
Pipe length
Time!!! (Things always get interesting when this is added)

Also, I wanted to get with you on the suspension dynamic data for the CB7 so we could design maybe a spreadsheet to allow tuners to predict their handling capabilities.

[wagon-r]

I have a Bisi F22A header, not the H22. Wish I could help. Sounds interesting.

[Domesticated]

We'll see what my instructor has to say about the project, but the analysis isn't steady state. If I consider the header wrapped I might be able to ignore a lot of the thermo stuff. I'll figure that out after Ive done some testing. Ive you're familiar with the TANS equation, this is a modified version I plan on using to control the effects between the header inlet and when averages allow d/dt to be neglected.

[sonikaccord]

I wouldn't expect it to be steady state. Are you ignoring d/dt for the entire project or just to get a rough design?

I am not familiar with the TANS equations. Something involving fluid dynamics?

[Domesticated]

Time averaged naiver stokes equation. Once the variation reduces to a negligible amount you can switch to steady state! Wiki has a pretty straightforward explanation of the TANS equation. Since we can't arithmetically solve or model the fluctuation in fluid flow or even CFD model it we found a way to to cheat and get very good results from it. http://en.wikipedia.org/wiki/Reynold...okes_equations

You won't be able to switch to steady state until the variation dies off, which I believe will be outside the header. For the next 4 weeks my instructor just wants me to analyze the "back-pressure" (bad), or better said the pressure head, vs inertia (good) and find the sweet spot for a specified flow because anything else will be to much work and I can continue my analysis next semester when I can have access to our super computer.

[sonikaccord]

Ok I see. So essentially at the exhaust valve and the exit of the header can be modeled as steady state, but as soon as the gas starts traveling is where things get a little hairy. Then you have to consider how each exhaust pulse affects the other pulses which, I'm assuming, will either add or subtract from the momentum? (not sure if that's the correct unit)

All flow fluctuations sum to zero and is a function of position and time?

[jdm92_accorn]

I wonder if the fact that some headers are stepped and some are not will have a negligible effect on thermal expansion. Also you said you are doing full exhaust analysis? Will you be using 2.25", 2.5" or 3"? Or will you do testing on each header with each exhaust piping size? I have a million more questions I'm sure will be answered once this project gets going, lol.

[Domesticated]

Ok I see. So essentially at the exhaust valve and the exit of the header can be modeled as steady state, but as soon as the gas starts traveling is where things get a little hairy. Then you have to consider how each exhaust pulse affects the other pulses which, I'm assuming, will either add or subtract from the momentum? (not sure if that's the correct unit)

All flow fluctuations sum to zero and is a function of position and time?

No, the pulses will slowly die out when the velocity is high because of viscosity and turbulence. Once the the variation gets small enough it can become steady state. Likely after the the header outlet though. I'm modeling the stock header right now and getting ready to run some cfd.

[Domesticated]

I wonder if the fact that some headers are stepped and some are not will have a negligible effect on thermal expansion. Also you said you are doing full exhaust analysis? Will you be using 2.25", 2.5" or 3"? Or will you do testing on each header with each exhaust piping size? I have a million more questions I'm sure will be answered once this project gets going, lol.

My instructor want me to keep it as simple as possible so IDK if I will be able to look into stepping this semester. I just need consideration for the cat back. For that I'm probably going to hook my exhaust up to a leaf blower and measure the pressure drop at certain flow rates. It will be a black box variable in the whole mess. My instructor really wants me to dumb it down so we'll see what happens.

[jdm92_accorn]

Damn instructors lowering the bar, lol. As far as stepped goes, pretty much every 4-2-1 header is stepped. Most 4-1's are but there are a few that aren't. I'm guessing you will end up doing ebay, feels and oem 4-2-1 and your current exhaust size for the test. Good luck.

[sonikaccord]

Your prof has a good point, I've tried to bite off more than I can chew at times and it's not pretty. You could always do the theory and number crunching on your own time after the course and such.

This is why I didn't do mechanical...I'm in for your results on the stock header

[Domesticated]

This is why I didn't do mechanical...I'm in for your results on the stock header

Haha! MEs have the most fun though! I do a lot of projects at home. I just modeled my stove top and took a bunch of measurements to correlate the model with real values lol. I also program C, VBA, and I have Arduino projects.

[sonikaccord]

I disagree! It's all about the EEs! I have a couple of Arduino ideas for the CB that I will try out later on.

What I do like about ME is that you can actually 'see' what's happening in your problems and designs. We can't really watch electrons flowing through a wire without expensive equipment.