Exhaust question

[djb160]

Ako just mentioned as far as exhausts go bigger = better, I just got told by a pitstop guy that 3" will give better power in the higher range 3K+. Whereas a 2.5 will give better response up till 5K, which is cool coz according to you guys when running around 1bar the 04s cant blow past 5K anyway.

Now Im confused. Can anyone clear this up for me?

[bradc]

Just go with a 3", if you do future mods the larger exhaust will help you out more.

[djb160]

Now thats a reply I can do something with :)

[Kenneth]

Sounds like the guy at pitstop was thinking about a Naturally aspirated engine rather than a turbo engine. With turbo engines, bigger is better.

The VR-4 standard turbos spool up so quickly that its virtually impossible to loose power with a bigger exhaust

[djb160]

getting it done tommorow.

[Subaru ETA]

post some pics and how much it cost when you get it done dude

[djb160]

He said about $1000, but hoping for some change. Do I need to know anything about mufflers/silencers or just give him general instuctions about what I want eg lower more growly but not too loud at cruising revs say 2K-3K.

[d i c k i e s]

i should take some pics of mine.

[djb160]

Yes you should. Im trying to keep the VR4 tip shaped oval, just a bit bigger.

[bradc]

Go with 6" wide, 4" high :)

[djb160]

I left my digicam with him so he can take some pics

[Roadrunner]

Sounds like the guy at pitstop was thinking about a Naturally aspirated engine rather than a turbo engine. With turbo engines, bigger is better.

The VR-4 standard turbos spool up so quickly that its virtually impossible to loose power with a bigger exhaust
I disagree. If you go for a 3" exhaust (even from the cat back), you'll lose torque and the car won't be as quick. A larger-bore exhaust will definitely help at the top end but at the expense of low-end grunt. Given that most of the time I expect you'll be below 5k rpm, you'll be disappointed with a big exhaust (apart from the noise ;) ).

If you raise the boost to 1bar or more, the difference will be less noticeable, but you would still lose torque (and therefore acceleration) compared with a 2½" bore.

As an example, the works Mitsi Evo rally cars run with a 2½" bore throughout because acceleration is more important than top speed.


Go with 6" wide, 4" high :)
I assume you're talking tailpipes ... ;)

[djb160]

too late now.

[Kenneth]

I disagree. If you go for a 3" exhaust (even from the cat back), you'll lose torque and the car won't be as quick. A larger-bore exhaust will definitely help at the top end but at the expense of low-end grunt. Given that most of the time I expect you'll be below 5k rpm, you'll be disappointed with a big exhaust (apart from the noise ;) ).

Maybe true for a naturally aspirated, not for turbo engines.

The idea with a N/A exhaust is having the tubing at a length that allows the exhaust gas to flow at high speed, but with as little resistance as possible. (back pressure is bad)
Fast moving pulses of gas (as we know internal combustion engines don't provide a steady stream of exhaust gas) create a small vacuum behind them (like the one that sucks the road dirt and exhaust fallout into the back of your car) and the faster they move the bigger the vacuum effect. this way, multiple pulses joint together to create a larger vacuum and are able to scavenge the exhaust gas from the cylinder.
In reality, the vacuum is enough to leave a vacuum in the cylinder, which of course increases the suction when the intake valve opens. Result is increased HP.

If you put to big an exhaust on, then this cannot happen because the speed of the pulses will slow. As a result, getting this low in the RPM range causes back-pressure higher in the RPM range and reducing top end power. On the other hand, optimising for high RPM is bad for low end torque/power.


On a turbo engine, the exhaust pulses race up the manifold and there they meet a turbo. From here on it works much much different.

A turbo works by having a high pressure on the engine side of the turbine and low pressure on the exhaust outlet side. This pressure difference causes exhaust gas to flow across the turbine and cause it to spin.

The bigger the exhaust tube, generally the greater the pressure difference. It is this difference that causes a natural increase in boost from increasing the size of your exhaust.

[Kenneth]

here is a dyno sheet that proves this.

The blue line has trouble with the plugs at high RPM, but you can still see the torque and HP lines down low.

The red line is after having 2.5" down pipes and 3" exhaust fitted. As you can see the red line pulls away from the blue line from the start of the pull at just over 2000RPM.

The boost lines cannot be compared unfortunately as they had a leak in the connection for the second (red) pull. The torque curve shows that the boost is naturally higher though (this is all done on 100% standard boost control). While this higher boost accounts for the increased max torque, the fact that it rises much more sharply is what makes it so good.

There is nowhere on that graph that supports loss of power with a bigger exhaust on the 6a13TT engine.

[djb160]

Thanks Kenny, that makes me feel better.

[bradc]

And on my dyno when I had my exhaust upgraded, the new line was above the stock exhaust line the whole way through the rev range too

[Roadrunner]

Can't argue with you from a technical viewpoint, Kenneth. Just my experience.

When I put a 3" cat-back, 4" tailpipe on my Legacy RS-B, it destroyed the torque curve - much quicker at the top-end (over 5k rpm) but not as much fun getting there. Similar, but not so marked on my previous VR-4 with a 3" cat-back, 5" tail pipe. I've kept the standard exhaust on my current car (apart from removing the cat) and it is quick. My mate has just taken his big-bore exhaust off his Evo VII GTA and replaced it with the standard one because he couldn't keep up with me in a straight line! That's not right.

If bigger bore exhausts are better, why don't the works turbo cars (and not just Mitubishi) use larger-bore exhausts?

[djb160]

Actually having just seen the front of our downpipes, if you only get a catback system you've wasted your money, man those things are restrictive where they meet.

Can anyone tell me the difference between a mandrel bent exhaust and a non mandrel one (pics) also difference in average price performance. Asking coz Im not sure mines a mandrel bent one but need to find out asap before he tries to overcharge me.

I get it back today, theyre putting some green stuff pads on it today.

[TAR]

I think that mandrel bent means a single piece of tubing formed into the shape required rather than lots of separate bits welded together. I guess that the less joints the better the gas flow.

I might be wrong though.

[djb160]

hmmm grrrrrr in that case. Anyone confirm what TAR just said, this was my impression too?

[Kenneth]

Can't argue with you from a technical viewpoint, Kenneth. Just my experience.

When I put a 3" cat-back, 4" tailpipe on my Legacy RS-B, it destroyed the torque curve - much quicker at the top-end (over 5k rpm) but not as much fun getting there. Similar, but not so marked on my previous VR-4 with a 3" cat-back, 5" tail pipe. I've kept the standard exhaust on my current car (apart from removing the cat) and it is quick. My mate has just taken his big-bore exhaust off his Evo VII GTA and replaced it with the standard one because he couldn't keep up with me in a straight line! That's not right.

If bigger bore exhausts are better, why don't the works turbo cars (and not just Mitubishi) use larger-bore exhausts?

Not sure how the Legacy RS-B works tbh, is it sequential turbo? I am dubious as to the performance benefit of cat-back also, being that the cat is usually the most restrictive part of an exhaust.

The bigger the turbo, the longer to get boost. Whether you can loose torque in this time from a bigger exhaust, I don't know. Theoretically it shouldn't make much difference as the turbine stops scavenging and presents a big restriction in the exhaust path.
My main point (possibly not well articulated) was that in the VR-4 with standard turbos, this time is so negligible that it is not worth considering... I don't think anyone expects power below 2000 RPM.

I should point out that when I changed my exhaust, I did down pipes to muffler and it was 90% stainless (decat pipe was mild steel). This would give increased flow and better heat retention. Not so sure how just cat-back would actually perform or what, if any, benefits would be realised. I drove You See's car which I believe has exhaust done from down-pipes back, and his boost response was pretty close to standard even with all his intake mods and aftermarket ECU. Of course after 3000RPM his pulled away from mine rather convincingly.


I would be interested to know if your VR-4 has any other mods than decat, and if your mates evo VII did either. It doesn't actually take much to get the VR-4 faster than a standard Evo VI/VII. There was one at the dyno when I recorded the 183.3KW which only recorded 176KW. My only mods were intercooler and Exhaust, so was very happy.

Not familiar with works turbo cars, would be interested to find more. It depends on what your goals are, whether to compete in an event you have to stick to specific guidelines (exhaust being one of them) and stuff like that.

[Kenneth]

I think that mandrel bent means a single piece of tubing formed into the shape required rather than lots of separate bits welded together. I guess that the less joints the better the gas flow.

I might be wrong though.


Mandrel bends are the BENDS only. you can have complex mandrel bends (ie the whole exhaust could be one piece of tube with multiple mandrel bends) or you could have single mandrel bends welded into the exhaust. It doesn't make a great deal if difference other than price really. most exhaust shops when making a mandrel bent exhaust will weld multiple bends into an exhaust. This does NOT reduce the effectiveness of the exhaust unless they weld very poorly.

non-mandrel bent means they use a pipe bender which crushes the sides of the bends. it is cheap to do however and you can bend the shapes as you need them on site without having expensive equipment (ie: a CNC mandrel bending machine which would probably cost hundreds of thousands of dollars... more than most houses)

Non-mandrel bends produce restrictions in the bends, as well as causing turbulence as the gas has to make odd changes in direction over the crushed areas.

[ralliart-type-s]

The exhaust issue is a tricky one and without your own personal dyno and all combinations of exhaust available its impossible to know exactly what to fit. All engines need a degree of backpressure to run efficiently, even the full race/rally cars keep pipe sizes down as much as poss and also need a silencer in there somewhere. I reckon its down to trial and error where poss, ie whatever exhaust mods you do, make sure the connections,flanges etc are all kept the same size and in the sameplace in the system so its easy to chop and change.

[ralliart-type-s]

I reckon a good idea would be to gather ppl with say 6 different exhaust setups, get em all to a dyno 4 a weekend and try them all on one car and dyno it each time with a different combination. This would at least tell US what works on OUR setup. I'll bring the gun gum!!

[Kenneth]

All engines need a degree of backpressure to run efficiently, even the full race/rally cars keep pipe sizes down as much as poss and also need a silencer in there somewhere.

Sorry, this is INCORRECT. It is a widely held belief, but one held out of ignorance.

Internal combustion Engines (naturally aspirated specifically) are a heat pump, they work by pumping air/fuel in, converting it to hot gases and then pumping them out. We put exhaust manifolds on engines to decrease noise, not to help efficiency. Even race cars these days are required to meet noise restrictions. Except maybe hard out drag cars and F1 cars.

Anyway, back to the backpressure thing... backpressure is a byproduct of exhaust restriction. It is unwanted. What you NEED for torque is exhaust VELOCITY. The method for increasing velocity is reducing the pipe size, the byproduct being that you get back-pressure when you start trying to flow too much for the pipe.

As said in one of the previous posts, the velocity is to clear the exhaust gas and to create vacuum on the exhaust ports as the fast moving pulses of gas form a vacuum.

Increasing exhaust size means you need more exhaust gas to get this effect, so you get your torque higher in the rev range. Daily driven cars can have this too, but it would make it a real pain in the backside to drive. So we go for smaller exhausts that give us torque lower in the rpm range, but choke the engine at high RPM.

As also said, turbo engines are different... the turbine tends to stop this happening.

[djb160]

Im really glad I asked this now.

[Roadrunner]

Not sure how the Legacy RS-B works tbh, is it sequential turbo? I am dubious as to the performance benefit of cat-back also, being that the cat is usually the most restrictive part of an exhaust.
Yes, the RS-B is a sequential twin turbo, small turbo up to 4-4.5k rpm, then the larger turbo kicks in up to 7.5k. The cat in mine was removed. Sorry "cat-back" is just a term to describe the exhaust from where the cat would be to the tail pipe. The Legacy exhaust split into two pipes going to the rear silencer - one straight through the silencer and the other with a butterfly valve operated electronically by a switch from the wastegate. At low revs, the butterfly valve stayed shut to improve torque; at higher revs, it opened to provide a freer-flowing exhaust. However, as I said above, fitting a wider-bore rear exhaust killed the torque curve and made the car slower to around 5k rpm. Above 5k, with the big turbo spooled up, it was definitely quicker. It just wasn't a nice car to drive any more ... :(


I would be interested to know if your VR-4 has any other mods than decat, and if your mates evo VII did either.
Both my VR-4 and the Evo VII GTA were standard (apart from the GTA exhaust). He should have been quicker, but wasn't. Since he's refitted the standard exhaust, his car pulls much better. Maybe it needs higher boost to benefit from the larger-bore exhaust?

[Kenneth]

Yes, the RS-B is a sequential twin turbo, small turbo up to 4-4.5k rpm, then the larger turbo kicks in up to 7.5k. The cat in mine was removed. Sorry "cat-back" is just a term to describe the exhaust from where the cat would be to the tail pipe. The Legacy exhaust split into two pipes going to the rear silencer - one straight through the silencer and the other with a butterfly valve operated electronically by a switch from the wastegate. At low revs, the butterfly valve stayed shut to improve torque; at higher revs, it opened to provide a freer-flowing exhaust. However, as I said above, fitting a wider-bore rear exhaust killed the torque curve and made the car slower to around 5k rpm. Above 5k, with the big turbo spooled up, it was definitely quicker. It just wasn't a nice car to drive any more ... :(

yeah, I just got the impression that everything else was standard, so you had the cat in.
Sounds like it was slower on the first turbo. I wonder if it had some adverse effect on the operation of the first turbo and switching to the main turbo. I dont know much (nothing really) about the sequential setups on the subarus, but I have heard that they get upset fairly easily.



Both my VR-4 and the Evo VII GTA were standard (apart from the GTA exhaust). He should have been quicker, but wasn't. Since he's refitted the standard exhaust, his car pulls much better. Maybe it needs higher boost to benefit from the larger-bore exhaust?

It depends on where his performance is crap... if its performing crap on boost then I suspect something else is the issue, like some crap resonators or a muffler that really doesn't flow well.

When you say couldn't keep up with you in a straight line, were you both hard up it in each gear?

[djb160]

next question how do I make it more quiet. Bigger muffler??

[Roadrunner]

When you say couldn't keep up with you in a straight line, were you both hard up it in each gear?
Yep, pretty much from a standing start in full auto mode - I change to tip mode once it's into 2nd gear and he couldn't keep up - not much in it, but he should have been quicker than me. Since he put the standard exhaust back on, I couldn't keep with him. So I've had to increase the boost ... ;) :)

[Roadrunner]

next question how do I make it more quiet. Bigger muffler??
If it's too noisy inside the car, you'd need to fit a resonator can to get rid of the harmonics. If it's too noisy outside, who cares?!! :p :D (although fitting a bigger muffler would help).

You're lucky you're in NZ - the police in my area are now stopping and issuing tickets to anyone with a modded exhaust (that they consider too loud). £30 a time ... :(

[bradc]

not really, if the police decide a car is too loud they force you to go to an approved garage with a decibel meter and get the exhaust measured, at a cost of about $100, obviously if you fail then you have to get the exhaust modified

[djb160]

its got 2 mufflers in it plus the larger end one, I love the sound out the back but its way too loud inside for comfort. I cant hear myself talking at a normal level.

[calum]

I have never yet seen figures or RR graphs that show that a bigger exhaust makes performance worse on a turbo car.

For those that say/ believe that back pressure is a good thing, how well does your car run if you cover the exhaust outlet with your hand?

Calum

[Kenneth]

Yep, pretty much from a standing start in full auto mode - I change to tip mode once it's into 2nd gear and he couldn't keep up - not much in it, but he should have been quicker than me. Since he put the standard exhaust back on, I couldn't keep with him. So I've had to increase the boost ... ;) :)

That sounds like there was something not right and he was being choked at the top end as well. Kinda odd really, as if the exhaust flowed more freely you would think he would have AT LEAST gained top end and pulled away once boost had built.

I suppose to be accurate, I should have said a better flowing exhaust is better...

[d i c k i e s]

ill get some pics up tonight, promise! been busy.

[djb160]

Just had a we blat up to 160ish, seems slower than before but maybe that has something to do with the two guys in the car at the time. Only starts to sound AWESOME at 6500rpm, makes a very cool WRC sounding series of backfires as I slow down lots.

It'd sound way better if it was manual. Stupid autos. :(

[bradc]

I love those pops when you slow down, mine does it particularly well when the exhaust is hot, and you take it all the way up to 7000rpm in 2nd or 3rd gear, then just lift off the throttle and slow right down. It's almost orgasmic :D

[djb160]

Thats the one :evil2: it even did it at coming off the juice from 7500 in 4th

[bradc]

that would be 246kmh! unless you made a typo

[djb160]

Nah thats not right, maybe I read it wrong it wasnt over 180, must have read it wrong! Can our cars seriously do 240 in fourth!!!!. Ill even pay my membership while im paying you.

[Xeroid]

It's almost orgasmic :D

WAY TOO much information Brad, we DON'T want to know about your ....... experiences..:speechles


You definitely need a girl friend ....soon !

[djb160]

Maybe hes just easily to please. lol

[d i c k i e s]

well my pics were a lil blurred. so ill get some up tomorrow. (take some better ones)

[bradc]

djb, at 4000rpm it is 131.64kmh, 6000rpm it is 197.46, so at 7500rpm you'd be at 246 :)

As far as pleasure goes, well it's cars first :D

[djb160]

Irealised my mistake now.

[sparky_man63]

mandrle bending is a special process that stops the pipe being compressed as it is bent whereas if u look at a standard pipe bend you will notice that the pipe is narrower through the bend that at either tip

[Turbo_Steve]

When I put a 3" cat-back, 4" tailpipe on my Legacy RS-B, it destroyed the torque curve


Yah. Well known effect on the twin turbo legacies called the "Valley of Death" or VOD (do a Google :D ) . Changing the exhaust on the GTBs, B4s etc is a complete pain in the backside because of this very effect, and usually requires supporting mods to retain driveability.


Kenneth, I have to disagree about exhaust manifolds: brother in law is on a Formula one team, and was categorical that they have exhaust manifolds for exactly two reasons:
1) To prevent the car catching fire and / or spraying oil/ fuel everywhere inside if there are problems
2) Something he calls "complex scavenging" which is basically what you were describing, but using a timed pulse from one cylinder to pull gas from another.
Also I believe they ignite fuel in the manifold (massively retarded timing) which also pulls gas through the engine: perfect for a fast launch. Another benefit is that it heats the manifolds, creating a temperature gradient at the exit, which creates additional vacuum.

None of which applies to a road car, so it doesn't change the essential truth of your post: it was just the one statement about manifolds :)


Nobody has mentioned the effect of the exhaust manifold on a turbo'd car, which is also important: the headers are tuned, by length and width, to usually provide the maximum gas velocity whilst maintaining the highest flow for the application in question, with the target of getting you on boost as quickly as possible. I'm not going to pretend I know a lot about this, as the headers need to suit the turbo and the state of tune to work properly. And the maths makes my brain bendy.

So based on all this info, has anyone found a good-flowing cat / muffler combination for the 6A13TT? Assuming I wanted to build a "nearly standard" sounding exhaust that would pass a test, but flow decently. The more I search the more bamboozled I am becoming, as everyones exhaust seems to be custom made or out of production?

I hate exhaust calculations.

[Nick Mann]

If you can wait until JapFest next year, your best bet is to come there and listen to the options available.

I personally like the sound of Kierans exhaust. I think it is a Blitz system? Not too loud. Most custom made systems are LOUD!!

[Kenneth]

Yah. Well known effect on the twin turbo legacies called the "Valley of Death" or VOD (do a Google :D ) . Changing the exhaust on the GTBs, B4s etc is a complete pain in the backside because of this very effect, and usually requires supporting mods to retain driveability.


Kenneth, I have to disagree about exhaust manifolds: brother in law is on a Formula one team, and was categorical that they have exhaust manifolds for exactly two reasons:
1) To prevent the car catching fire and / or spraying oil/ fuel everywhere inside if there are problems
2) Something he calls "complex scavenging" which is basically what you were describing, but using a timed pulse from one cylinder to pull gas from another.
Also I believe they ignite fuel in the manifold (massively retarded timing) which also pulls gas through the engine: perfect for a fast launch. Another benefit is that it heats the manifolds, creating a temperature gradient at the exit, which creates additional vacuum.

None of which applies to a road car, so it doesn't change the essential truth of your post: it was just the one statement about manifolds :)


Nobody has mentioned the effect of the exhaust manifold on a turbo'd car, which is also important: the headers are tuned, by length and width, to usually provide the maximum gas velocity whilst maintaining the highest flow for the application in question, with the target of getting you on boost as quickly as possible. I'm not going to pretend I know a lot about this, as the headers need to suit the turbo and the state of tune to work properly. And the maths makes my brain bendy.

So based on all this info, has anyone found a good-flowing cat / muffler combination for the 6A13TT? Assuming I wanted to build a "nearly standard" sounding exhaust that would pass a test, but flow decently. The more I search the more bamboozled I am becoming, as everyones exhaust seems to be custom made or out of production?

I hate exhaust calculations.

I am still not sure where you a disagreeing with me. I stated that even race cars need to conform to noise standards, except perhaps F1 and Drag cars.

Exhaust scavenging is just as important on a road car as a F1 car, it is just implemented differently for a different result, a F1 car isn't built to idle at 600RPM and then have usable torque low in the RPM range.
What you are saying about scavenging is essentially what I said, in a slightly different way with specific references to F1.
I didn't think it necessary to take it to that level.

About turbo manifolds... Tuning is less important and essentially you need only get a couple of parameters correct to have a very successful manifold.

1) The manifold can flow enough exhaust gas to generate the power you are trying to achieve. (ie not creating restriction in the runners)
2) They provide the shortest path possible to the turbo.

A short path to the turbo means that the manifold will fill with exhaust gas quickly and provide an increase in pressure on the hot side of the turbine. It is this pressure that causes the turbine to spin, the higher the pressure difference the faster the turbo will start spinning. This is simple physics.

Because of this, short runners to the turbo (such as on the VR-4) are not very sensitive to runner width as the manifold builds up pressure very very quickly and 'tuning' the width and length doesn't matter past the fact that you need to have big enough runners to flow the exhaust gas required to reach your power target.

If you are unfortunate and need to have long pipework to your turbo then you do need to think about the pipe width, as increased exhaust velocity will help the turbo spool, though it will probably also choke if you try run too much power.

Turbo manifolds are easy to design compared to N/A manifolds as you are primarily concerned with creating as much pressure as quickly as possible between the cylinder head and the turbo.

The manifolds on the VR-4 are actually very good. A little porting and I expect you should be able to run quite high HP with them. Biggest restriction at this point is the feed to the turbo and then the turbo itself. (hopefully someone has a manifold they can take a picture of to illustrate)

[Turbo_Steve]

Nope, you're right, I need to learn the difference between "Except" and "Even".
Apologies, Kenny, no impugnation was intended, regardless.


In agreement with you, the manifold maths was explained to me at great length by someone calculating a set for an EJ20. As it's a flat four, all four pipes were different length, some short, some longer. His explanation was that as you can't have the cylinder pulses arriving smoothly, and it's a single scroll turbo, the best thing you can do is delay the shortest pulse to fit into the "gap" in the other three, otherwise you end up with two at once. This, supposedly is better for power output, however for getting on boost quickeryou need to have the delay earlier to try and run the pulses INTO each other, which gets you on boost faster, so you're looking for the pipe to resonate a specific frequencies around the point where the turbo starts to spool, to slow the gas down to increase the pressure.

This is supposed to be true of all manifolds on turbo'd cars, but, as you state, becomes less and less relevant the shorter the runs become.

The maths still baffles me, though.

[Fully]

Thats the one :evil2: it even did it at coming off the juice from 7500 in 4th

Were are you doing this?

[EdmundVR4]

next question how do I make it more quiet. Bigger muffler??
I had 2.5" downpipes, 3" pipe with one resonator & free flow muffler & it was noisy (booming) inside the car, especially between 2000 & 3000rpm. I fitted an Apexi ATS (Active Tail Silencer) & removed the resonator & the booming is gone !
The Apexi ATS is operated by exhaust flow pressure...at low rpms it works like a normal tip silencer, but as the exhaust flow increases the valve opens up gradually providing free flow at high rpms.
Quiet & torquey at low rpms & free flowing when winding up the engine...nice tone too !
Read more here : http://clubvr4.com/forum/showthread.php?t=13639