Is 3" large enough for my SC-61?

[Dirtboarder16]

I've heard a lot of debates about the proper exhaust diameter, without a lot of real solid conclusions. I am planning to eventually max this thing out (rated at 680hp), and I don't want to choke it on the hot side. According to Corky Bell (i think most probably know who he is), a 3" exhaust will support like 800 horsepower. 3" also works great on 400hp cars, and I've read plenty of sites that say to go as big as 5" diameter pipe (no thanks) for 600+ horsepower. How many of you guys are making big power and what size exhaust piping do you have? 3.5" is almost the same price as 3" and if it will work better, I think I'll use it. I know shep uses 3.5" but he is making around 1000hp so I'm not sure what to go with. Someone help me out.

 

[anconover]

3" is plenty. But if u wanna blow the cash on 3.5" and dont mind it hanging quite low, go for it.

 

[illeagle_talon]

Maximum boost is an awesome book and taught me a hell of a lot as well. 3" exhaust is fine for your application. I wouldn't worry about it.

 

[illeagle_talon]

Maximum boost is an awesome book and taught me a hell of a lot as well. 3" exhaust is fine for your application. I wouldn't worry about it.

 

[pboglio]

The exhaust system on a 2g AWD is 13.5 ft long 02 housing to tip. Based on widely published flow bench data, a pipe 16" long (i.e Muffler length) flows 115 cfm @ .78 psi pressure drop for every square inch of pipe flow area. So, a 3" O.D. pipe would have a cross section of ~6.5 in2. Multiply this by 115cfm and you get 748 cfm @ .78 psi drop for 16" length. The total pressure drop for the whole exhaust system is roughly (.78)(13.5 x 12")/16" = 7.9 psi of backpressure right after the turbine wheel, if your flowing 748 cfm. Thats roughly 510 h.p.

At 680 h.p. or 985 cfm your backpressure will be higher of course. To find your new pressure drop: P2 = P1(CFM2/CFM1)^2. That gives roughly 13.7 psi backpressure @ 680 H.P. on a 3" downpipe back exhaust system. Thats assuming 100% efficient straight thru mufflers and nice wide mandrel bends. That ain't good. A 3.5" pipe diameter would drop this to ~7.2 psi pressure drop @ 680 h.p. A 4" pipe would give 4.2 psi backpressure @ 680 h.p. Most guys just run 3-4" dia. and do a side dump, getting rid of 8 ft of pipe and cutting dropping exhaust backpressure by more than half.

 

[anconover]

yea, everyone runs 4" exhaust systems. Not hardly. 3" exhaust will be plenty for his setup. of coarse 3.5" will offer less back pressure, but will hang lower, and be quite a bit louder.

 

[pboglio]

Most guys who would be running 680 h.p. would have known already to switch to bigger than 3" and probably do a side dump on top of that. As in most guys running 680 h.p., not most everybody. And 3" won't be plenty for 680 h.p., not unless he likes giving up 30-40 h.p. in backpressure.

 

[thegreatms]

The exhaust system on a 2g AWD is 13.5 ft long 02 housing to tip. Based on widely published flow bench data, a pipe 16" long (i.e Muffler length) flows 115 cfm @ .78 psi pressure drop for every square inch of pipe flow area. So, a 3" O.D. pipe would have a cross section of ~6.5 in2. Multiply this by 115cfm and you get 748 cfm @ .78 psi drop for 16" length. The total pressure drop for the whole exhaust system is roughly (.78)(13.5 x 12")/16" = 7.9 psi of backpressure right after the turbine wheel, if your flowing 748 cfm. Thats roughly 510 h.p.

At 680 h.p. or 985 cfm your backpressure will be higher of course. To find your new pressure drop: P2 = P1(CFM2/CFM1)^2. That gives roughly 13.7 psi backpressure @ 680 H.P. on a 3" downpipe back exhaust system.

some think look fishy about this. the back pressure numbers seem way to high. I used a quick little program online and i got much lower pressure drops:

www.efunda.com/formulae/fluids/calc_pipe_friction.cfm

I'm going to check some of my resources when i get home to see if i can back this up...

 

[Dirtboarder16]

You guys are awesome.

Pboglio- if your numbers are right, that helps tremendously. Thanks for the in depth answer.

Thegreatms- I'll play around with that program and see what I can figure out. Thanks.

I know ideally I would just run a dump tube off the O2 housing, but this is a street-driven machine. If there was a way to run a short pipe out behind one of the front wheels that was somewhat quiet, I'd be all over it. A good muffler would be mandatory, and not many would last long being exposed to such heat. I saw the guy with the short side-exit exhaust and the cheesy aeroturbine muffler, which might do the trick, but I bet it's still loud as hell.

 

[pboglio]

some think look fishy about this. the back pressure numbers seem way to high. I used a quick little program online and i got much lower pressure drops:

www.efunda.com/formulae/fluids/calc_pipe_friction.cfm

I'm going to check some of my resources when i get home to see if i can back this up...

Yes, I know the numbers are high compared to the Moody chart and Colebrook's equation.
I posted an off the cuff calculation based on a rule of thumb formula from various muffler flow bench experiments. I trust this informal formula, even though I've never seen it posted in any textbook I've read. I've double checked it against 10 or more different muffler flow bench tests, always comes out very close to actual published flow bench tests. For instance, a 16" long 2.5" testpipe drops .78 psi @ 550 cfm, which is verified in many many experiments and used as a control for muffler pressure drop testing. How much pressure drop would there be in an exhaust system that was 10 times longer than a 16" straight pipe? About 10 times the pressure drop of the 16" long testpipe. Not to complicated, pressure drops are additive. A 3" dia pipe the length of a muffler drops .78 psi when flowing 750-850 cfm, again based on muffler flow bench tests using the straight pipe as a control. Again, a full exhaust system is roughly 10 times longer than a 16" long muffler or straight pipe, so you get 10 times the pressure drop. All based on ACTUAL flow bench testing, not theory.

Here is a linky which shows APS's flow projections for their various 3" and 3.5" exhaust systems. I say projections because the way they list their horsepower vs. pressure drop values. They don't look like flowbench results to me, everything looks like a linear extrapolation, but they are a reputable shop and its another datapoint to consider.

http://www.airpowersystems.com.au/us_spec/subaru/exhausts/apswrx-hoe-r5/apswrx-hoe-r5.htm

 

[Dirtboarder16]

Yes, I know. I have my own Pipe loss spread sheet based on the Moody chart and Colebrook's equation and simplified using incompressible flow assumptions. However, my numbers are based on countless published muffler flow bench results. Call it a rule of thumb formula based on observed flow bench results. Granted, flow benches flow air at room temperature, pipe losses would be even higher given the much higher exhaust temps in the exhaust system.

Its always the same thing, a 16" long pipe drops 20" of H20 when flowing 115 cfm per 1" square flow area at room temperature. The backpressure numbers I posted are based on that actual observed pressure drop of a straight pipe on a flow bench, not theory. The pressure drops stated are not unreasonable, "Scientific Design of Intake and Exhaust Systems" shows exhaust pipe pressure drops of about 1 psi/foot of straight pipe on a small displacement motor under high load.

This is simple stuff guys, the hard part is figuring out the exact power loss associated with a given backpressure value. Thats much tougher to figure out.

So anyway you look at it, if I run the pipe all the way to the rear of the car, it's going to have a pretty hefty back pressure? From your calculations, it seems like this could be somewhat solved by going to 4" or possibly larger diameter tubing, but that would make the car an absolute dog to drive around off-boost. The solution might just have to be a cutout.

 

[pboglio]

Yeah, a cutout looks like the best street solution. Or pull the cat or testpipe out at the track, same thing. Just take your projected horsepower divided by your current horsepower and square that term, thats how much of an increase in backpressure your going to get on the same exhaust system. So a guy running 400 h.p. on a 3" exhaust system who now wants to run 680 h.p. will increase his backpressure by 2.89 times right after the turbine wheel.

How much h.p. roughly are you putting out right now?

 

[Dirtboarder16]

Yeah, a cutout looks like the best street solution. Or pull the cat or testpipe out at the track, same thing. Just take your projected horsepower divided by your current horsepower and square that term, thats how much of an increase in backpressure your going to get on the same exhaust system. So a guy running 400 h.p. on a 3" exhaust system who now wants to run 680 h.p. will increase his backpressure by 2.89 times right after the turbine wheel.

How much h.p. roughly are you putting out right now?

I didn't want to mention it earlier, because I lurk on these boards often and I know how cruel some people can be to "outsiders", but I have a N/A 3000GT (part dsm), and am in the process of converting to single turbo. The car is currently under the knife, so I'm frantically trying to get all my questions answered so I can get it finished. Stock horsepower is rated at 222 but its closer to 200, and my goal with this setup is mid 500s at the wheels, which is near the max this turbo is capable of. It's a SC-61 (56 trim) with the T04E compressor cover and a .68 A/R T4 exhaust housing with the 76 trim wheel.

 

[pboglio]

Oh, cool setup. With 3.0 Liters of displacement, I'd say a 3.5" exhaust on your car would be like a 3.0" exhaust on a 2.0 Liter DSM, certainly not overkill for a turbo car. You've got more than enough engine displacement and exhaust flow offboost to make that a driveable setup for everyday use. I'd say 3.5" dia catback for sure, 3" dia up front if necessary where space is restricted. Go bigger later if you need to or go open dump like you said.

As you can already see, there are quite a few dickheads on this site. I'd be cautious too. Cheers.

 

[Dirtboarder16]

Oh, cool setup. With 3.0 Liters of displacement, I'd say a 3.5" exhaust on your car would be like a 3.0" exhaust on a 2.0 Liter DSM, certainly not overkill for a turbo car. You've got more than enough engine displacement and exhaust flow offboost to make that a driveable setup for everyday use. I'd say 3.5" dia catback for sure, 3" dia up front if necessary where space is restricted. Go bigger later if you need to or go open dump like you said.

As you can already see, there are quite a few dickheads on this site. I'd be cautious too. Cheers.

Thanks a lot for all your help. 3.5" it is, and probably with a cutout for the track.

 

[illeagle_talon]

Not neccesarily. I own a Stealth TT and most of the 3/S community just goes with a 3" exhaust system and they're just fine. It's not entirely about displacement, it's about how much cfm of exhaust the motor puts out. It'd kinda be overkill for your average DSM. I only know of one DSM that uses a 3.5 exhaust:.....Shep. And he's puting down a hell of a lot more power that most of us dream about.

 

[illeagle_talon]

hehe whoops...didn't notice that you were a 3/S as well. hmm....still a good 3" mandrel bent turbo back would be fine.

 

[Dirtboarder16]

hehe whoops...didn't notice that you were a 3/S as well. hmm....still a good 3" mandrel bent turbo back would be fine.

Fine isn't good enough. I made a thread on 3Si not a week ago and got some pretty good feedback. Just because a bunch of people do it, doesn't mean it's ideal. I'd say more people go with dual 3" exhaust.

http://www.3si.org/forum/showthread.php?t=249244&highlight=exhaust+diameter

 

[pboglio]

Yeah, I agree about the 3" if it were a street car. But he's talking about 680 h.p., that is a whole other set of needs and the back of the envelope numbers show this. I don't go by what some DSM god runs or what the majority does. Most people are a bunch of followers anyways.

I agree, on the street I doubt a 3" exhaust system would even come close to being maxxed, by that I mean less than 10 psi or more of backpressure. But the requirement is 680 h.p. and unless somebody can prove otherwise to me based on existing flow bench data or reasonable engeering data, it will produce unnecessary backpressure which will cost a good amount of pumping horsepower and punish the motor.

Here's a good example. I just looked at a website with a 388 w.h.p. Z06 Corvette or whatever the designation. Guy was running DUAL aftermarket 2.75" or 3" pipes. That looks awful close to a single 3.5" flowrate, maybe even more. Why somebody would be opposed to running less backpressure under absolutely critical conditions is beyond me.

 

[thegreatms]

At 680 h.p. or 985 cfm your backpressure will be higher of course. To find your new pressure drop: P2 = P1(CFM2/CFM1)^2. That gives roughly 13.7 psi backpressure @ 680 H.P. on a 3" downpipe back exhaust system. Thats assuming 100% efficient straight thru mufflers and nice wide mandrel bends. That ain't good.

Well I when home and broke out the old text books...And i still have to say that 13.7 psi of back pressure is way to high.

I calculated the pressure in a 160 inch long round pipe, 6.5 in^2 cross section, with 985 cfm of fluid flow. I used air at a temp of 300F as my working fluid, no elevation gain, no bends, adiabatic walls, smooth pipes, and external pressure of 14.7 psi.

I have results from three sources:

1. The program on efunda that I listed gives and output if .8 psi.

2. My hand calculations gives an out put of .95 psi

3. I also ran a FEA analysis using COSMOSFlowWorks (Add-in to SolidWorks 2005). Using the same input data as above it gave me a pressure change of .85 psi. (it also gave me a bunch of pretty pictures )

These seems a little low, but you must realize that this is a straight tube. A 90 degree bend can be estimated as many feet of straight pipe. If you add in the extra psi for bends (i doubt at 680hp that he will be running a cat or restrictive mufflers) I bet there will be a total back pressure of ~3 psi. with a high flow cat and mufflers maybe 5-6 psi

I think 3 psi is a more accurate number than 13.7psi

/off topic/ does anybody remember the thread that was on here a while ago where someone was talking about pressure ratio across the turbine and how the effect of lowering exhaust back pressure gets multiplied when you get back to the exhaust manifold? I can't seem to find it and I think this thread had some real world data on dsm's exhausts. //

btw I definitely think he should get a 3.5" exhaust with that big a turbo and his power goals. Even with my numbers that 3 psi could hurt you enough that you should go bigger.


Any other engineering types feel like putting in some input, real world or theory?

 

[thegreatms]

oh... i just found this site that has real exhaust backpressure measurements:

http://www.airpowersystems.com.au/us_spec/subaru/exhausts/apswrx-hoe-r5/apswrx-hoe-r5.htm

now this is on a wrx, but it is a full exhaust from "turbo to tip" should be pretty close to our situation. They list back pressure with all there models at different hp levels, with cat and with no cat.

That page shows the "APSWRX-HOE/R5" which looks like a 3.5" exhaust, at 725 hp and no cat it only has 1.0 psi back pressure.


this shows the "apswrx-hoe-r1" a 3" mandrel bend system:
http://www.airpowersystems.com.au/us_spec/subaru/exhausts/apswrx-hoe-r1/apswrx-hoe-r1.htm

at 725hp and no cat it only has 5 psi back pressure


unless he's running a cat i see no way he could get near 13.7 psi back pressure

 

[pboglio]

Do me a favor, post a source that has some actual scientific merit behind it. I will. Do a google search of: "No Loss Exhaust System(s)" by David Vizard. Now observe his rule of thumb formula based on ACTUAL flow bench results: Flow rate @ 1.5" HG = (Pipe flow area in inches)(115 cfm). A pressure drop of 1.5" HG = .735 psi. He states for example that a 2.5" diameter tube flows 564 cfm @ 1.5" HG drop, the pipe being the length of the average muffler, roughly 16-22" in length. Now head on over to Autospeed.com and do a search on muffler tests. Now use the formula I posted based on David Vizard's rule of thumb formula, now compare it to the actual flow benching of a straight 2" pipe on Autospeed.com website. Now do a google search for muffler flow bench results, now notice how they first flow a piece of 2", 2.25", 2.5", and 3" straight pipe right before the flow the various mufflers. Now use the formula I posted. Notice how the formula is within 10% of every REPUTABLE independant flow bench result that you can find on the internet.

Now put your engineering cap on and calculate the pressure drop @ 985 cfm for a 3" exhaust system 13.5 ft long or 160" in length, just straight pipe and no mufflers. Not theory, but REAL flow bench results which have been corraborated against other independant tests.

Please, post some real world data that shows test conditions, something thats been verified and correlated against other independant tests.

 

[pboglio]

Here's the link for David Vizard:

http://www.geocities.com/MotorCity/Track/6992/vizard.html

Do your own calculations based on that article. I wouldn't believe a word that came out of APS or DYNOMAX, both have test results that don't even come close to independant tests.

 

[Zex4g63]

Not all of the exhaust is going out the tailpipe if you are using an external gate. How would anybody be able to factor that in?

 

[pboglio]

/off topic/ does anybody remember the thread that was on here a while ago where someone was talking about pressure ratio across the turbine and how the effect of lowering exhaust back pressure gets multiplied when you get back to the exhaust manifold? I can't seem to find it and I think this thread had some real world data on dsm's exhausts. //

I've read that post. I also used the turbine and compressor work formulas, equated them and solved for Turbine Inlet pressure, sourced from "Internal Combustion Engine Fundamentals". Got the whole shebang on a spreadsheet. Gotta input 5-6 variables, but shows the relationship between Turbine inlet pressure and EGT, boost pressure, outside air temps, compressor & turbine efficiency, and mostly exhaust system backpressure. I've seen guys post Turbine backpressures as low as 17 psi on a .48 A/R 50 trim to somebody stating Forced Performance ran 80 psi backpressure on a 20g TD05 wheel unclipped.

By far, and the reason I place such emphasis on this guy running a 3.5" exhaust at 680 h.p., is that reducing backpressure by 7 psi would drop turbine inlet pressure by over 15 psi, depending on turbine wheel efficiency. Thats a HUGE reduction in pumping work and gives a much bigger margin against reversion and knock.