Saabscene Saab Forum - Saab Technical Information Resource banner
1 - 17 of 17 Posts

·
Registered
Joined
·
1,448 Posts
Discussion Starter · #1 ·
I've had the time to do considerable research on exhaust design. A couple of books always come to mind when discussing exhaust design on a turbo car, so I will start there.

Maximum Boost, by Corky Bell, states that a 2.5" diameter exhaust pipe should be sufficient for 400 hp. That may sound contrary to what you've heard or been told. Corky does not explain much on why, except that this is because there is an exhaust gas velocity that ought not be exceeded. 3" exhaust pipes are said to be good for nearly 700 bhp.

Because he does not explain much on why, there is little else to post on his thoughts. He does not state why the exhaust gasses should not exceed a certain limit, but I'll get to that in a moment.

Turbochargers, by Hugh MacInnes. This is a book which has more explanations; I like that about it. Some explanation may be found in the scanned image of this page: http://img.photobucket.com/albums/v14/Saab...anjoHousing.jpg

Suffice to say, because of the swirling pattern of the gasses exiting the turbine, the first 18-24" of exhaust pipe does not follow the rules in the same way that the rest of the pipe does. It does still follow the rules however, just the swirling that increases the local exhaust gas velocity considerably.

Here are some of the concepts I've taken from readings I've done, and conversations with various engineers.

Boundary Layer

The thin layer of air against the walls of an exhaust pipe. This layer buffers the effect of both the temperature and speed difference between the gasses inside the pipe, and the walls of the pipe itself.

Turbulence, your number 1 flow restriction

Turbulence is any gas not moving in a uniform direction (normal to the flow field), or against the overall flow of the other gasses in the pipe. It is by far the number one restriction in most pipes. It is also more common than you probably think. Almost any bend in pipe causes significant turbulence. (The turbulence comes in the form of vorticity when the pressure and density gradients are not alligned. The greater the mis-allignment, the greater the vorticity.)

Reduced turbulence = increased flow.

The only other real restrictions are changes in velocity. If the gas must speed up quickly it will require energy to do so. That energy usually comes in the form of pressure; backpressure in the case of an exhaust system. This pales in comparison to bad turbulence, but it's still worth noting.

Believe it or not, it really isn't the size that counts.

Size is only a factor in a very complex equation when it comes to gas flow. (To know more you may wish to check out www.navier-stokes.net ) A larger pipe tends to have a more stable boundary layer because the gasses inside it are flowing more slowly. If the gasses are moving too quickly it rips off the boundary layer constantly, causing turbulence. The more stable the boundary layer, the less the turbulence, the better the flow. Corky Bell was probably referring to the boundary layer when he stated that there was a specific exhaust (and intake) gas velocity that ought not be exceeded.

Thermal coatings help stabilize the boundary layer by decreasing the difference in temperature between the boundary layer (which is about the same as the pipe) and the inner gasses (which are much hotter). Having a more even temperature reduces the effects of convection between the hot and cool gasses, which reduces turbulence, which reduces backpressure. Neato

Now to the last problem ... bends!

Whenever the gasses move through a pipe at an angle there is turbulence. This is because the diameter of the outer edge is much larger than the inner edge.

That difference in diameter means that in order to go through the bend at the same angular rate the outer gasses must move at a higher linear rate. This either causes backpressure because the outer gasses have to be accellerated, or it causes turbulence from shear forces because they don't speed up. Either way it's not ideal.

Have you ever seen those weird plastic intakes on certain naturally aspirated cars? The ones that look like flattened tubes every time they bend! That's actually intentional and has been dyno shown to improve hp! (Not to be comfused with improperly made "press-bends" which are different.)

Intakes on N/A cars are more sensitive to restriction than most exhausts, but the same thing applies. By flattening the tube the difference between the inner edge's radius, and the outer edge's radius is reduced! When done properly the total cross-sectional area of the tube remains the same. That means the gasses don't have to change velocity as much, and there is reduced turbulence. The gasses simply flatten out, which is not difficult for them to do, and doesn't cause much turbulence.

Exhausts would be very expensive to manufacture in such a shape as they are made of much more heat resistant, and expensive, materials.

Anyway, I hope this has been at least a little insightfull, if a very long read.

Dubbya~
 

·
Registered
Joined
·
4,427 Posts
Nice one Dubbya
Well explained.

Took me back to my Fluid Dynamics lectures at University (though we were looking at less interesting stuff like moving foods through pipes and heat exchangers etc
Did you know that Ketchup and Mashed Potato are thixotropic, or shear-thinning, though the viscosity of mash makes it assume plug flow with no real boundary layer?!).

I got the theory, in a qualitative sense, it's the hard sums that come with a quantitative approach I always struggled with!
 

·
Registered
Joined
·
4,215 Posts
good post and well argued

however

there is no adequate explanation on why 2.5 " ...
I see two mutually exclusive properties
one is boundary flow which likes big diameters and large radiuses but slows flow
and turbulence which is minimised by velocity which is inversly proportional to cross section
So my guess from this is that there is an optimum size
ok but...
however we are talking about pulsed flow which I imagine changes things...
we are talkng now about exhaust waves which behave differently than flow.
To maximise the pulse effect on the turbo...you need the wave arriving intact therefore high flow (thin tubes) with the wall friction minimised (ceramic coated ?)to reduce wall effect
down stream of the turbo we are looking at flow again...

So ...
Optimised system ylee #1
exhaust manifold tubular ceramic coated of x diameter all tubes of equal length and minimised radius
Flowing through a turbo where all the waves are dashed on the shore of the turbo vanes
and flow through a large diameter pipe to exit
minimising radius and maximising diameter...

or is this too simplistic...?

Also I would love to have the math to calculate x...
 

·
Registered
Joined
·
281 Posts
I say's don't overestimate the maths. Guess, stick it on a dyno and see what happens. Then change it a bit... ect..

Engineering and maths calculated that they wouldn't get water in the channel tunnel - but it flooded half way through the dig..
 

·
Registered
Joined
·
14,305 Posts
Sort of a bit with ejenner on this one... if you go back to my orginal post:

The entire exhaust system generates back pressure on the turbo. Generally speaking, any reduction in the backpressure by lowering flow resistance will give a performance benefit, so any larger section anywhere in the system will help. However, the resistance is proportional to the flow rate and nearer the turbo output the gasses are at their hottest and moving quickest, so the downpipe gives most benefit. Further along the system the gasses have cooled somewhat and are moving slower, and thus the benefit of a larger dia is less.[/b]
I think you will have to agree that it provides a reasonably acceptable model for what actually happens in practice- which is what we're all about here. ylee's right to say that we're talking pulse not pure flow here, so a lot of the conventional fluid dynamics theory is more difficult to apply.

For most folk driving a turbo car, top end power is not as noticable in perceived performance terms as quicker spool up- so I stand by my original post that lower flow resistance at any point in the system is good news
 

·
Registered
Joined
·
4,215 Posts
sorry to be a pedant here...
 I stand by my original post that lower flow resistance at any point in the system is good news  [/b]
yes and no

ai would agree that everywhere bar the port increases flow
too much porting can reduce the port resistance to such an extent you can get "tidal" flow in to the exhaust port with subsequent misfire and poor idling..
this is when you need to look at valve timing..
this is where I am at the moment methinks..
 

·
Registered
Joined
·
1,448 Posts
Discussion Starter · #8 ·
When a certain exhaust velocity is exceeded the boundary layer is continuously ripped off the walls of the exhaust piping.

This is not to say that a smaller pipe will yield more power, rather, that a larger one will not yield any significant increase in power, and will cost more, weigh more, make more noise, and take up more space.

Also keep in mind that a larger pipe only makes the gas move more slowly away from the turbine, and does not directly relate to the pressure pointed back at the turbine. If one were the follow the Bernoulli Principle (extended into compressable fluids through various Navier-Stokes equations) the faster moving fluid would have the lower internal static pressure. Static pressure in the exhaust system is the only pressure the turbine will "see" as the velocity pressure will have it's vector pointed away from the turbine discharge. As long as the exhaust is larger than the turbine discharge, there is no significant benefit in increasing it further at it's minimum cross section.

Also in regards to the exhaust pulses making their way through the exhaust system ...

The mass flux of the exhaust pulses is very little downstream of the turbine. (Rankine-Hugoniot jump conditions do not necessarily preclude stable flow.) The pulses do have some limited effect on the boundary layer, but this, along with the difference in temperature, is included in the figure of 450 ft/sec. That figure is the exhaust velocity engineers usually use as a rule not to be exceeded, and is the velocity stated by Corky Bell. Larger than necessary exhaust usually are used just to counter the negative effects of excessive tight bends, and not always effectively.

On of the main reasons people see as large a benefit from a 3" exhaust is that they go from stock to 3" all at once. There is a benefit in a 3" downpipe (remember about turbulent flow, and the spiraling exhaust gasses?), but usually the only reason any gain is seen with a muffler section is because the stock mufflers are restrictive, not because the piping is too small.

Dubbya~
 

·
Registered
Joined
·
281 Posts
So what your saying is that someone like me who has not bought my exhaust yet would be wise to invest in a 3" downpipe but to go with stock sizes for the rest of the system? I like the idea of avoiding a noisy exhaust so does the better silencing solution include 1 silencer or two and should they be stock or performance?


Emmett.
 

·
Registered
Joined
·
1,448 Posts
Discussion Starter · #10 ·
Most of the Saab mufflers were pretty restrictive. (With the exception of the Viggen, and possibly 9-5 Aero.) One could just have performance mufflers welded in place. But it's almost as easy just to replace the whole section with new pipe.

The reason I posted all this wasn't to keep people from buying 3" exhausts. Rather, I just wanted people to understand that a smaller exhaust pipe diameter does not mean that there is a bottle-neck.

On my C900 I just welded a straight pipe where the muffler used to be. Oddly, while easily as loud as the JT3" single muffler outside, inside the car, it's almost whisper quiet. In fact I can only just barely hear the exhaust inside the car, with no muffler what so ever and stock piping.


Catalyst aside, I would have little to gain in a larger exhaust until 200+ bhp. Maybe two or three horsepower. At the moment that would simply not be worth the cost. The downpipe on the otherhand (especially the battery eliminating one) is looking mighty tempting. :fawtly: Shame I have other things to do first on that car.

Dubbya~
 

·
Registered
Joined
·
16,524 Posts
It might also be interesting to see if there is any benefit in "sucking" the exhaust gases out of the end of the pipe by making sure that the end of the exhaust is in a low pressure area at the back of the car at all times.
 

·
Registered
Joined
·
281 Posts
The reason I posted all this wasn't to keep people from buying 3" exhausts. Rather, I just wanted people to understand that a smaller exhaust pipe diameter does not mean that there is a bottle-neck.[/b]
I understand that. I'm trying to find out how to get the best exhaust system. I'm definatly looking to go +250bhp so would the 3" downpipe and 2.5 center/rear section be right for that sort of power? I'm not too worried about the price as the difference isn't much and I need an exhaust because I threw the old on in the bin. Just want to buy the right one.
 

·
Registered
Joined
·
1,448 Posts
Discussion Starter · #13 ·
If you can get a 2.5" for cheaper or the same price as a 3" it should be acceptable for very high power outputs provided it uses a good performance muffler.

Whether it will be quieter depends on the muffler of the two exhaust being compared. Best way to put it is that it will be easier to make a 2.5" exhaust quiet, all other things being equal.

Look closely on pricing though. A 3" with two mufflers would flow about as well as a 2.5" with one muffler, and might very well be quieter.

Also because 3" seems to be the most popular size, it may be cheaper as well because there are longer production runs at lower cost.

Dubbya~
 

·
Registered
Joined
·
378 Posts
I knew I should have paid more attention at school
? for Adrian W ,
It might also be interesting to see if there is any benefit in "sucking" the exhaust gases out of the end of the pipe  [/b]
by sgould

Ignorance might be taking over again but, with reference to sgould's comment above,isn't this the principle of a chimney? ie; large at the begining and small at the end therefore aiding gas flow.
 

·
Registered
Joined
·
186 Posts
From memory of looking into exhaust upgrades on previous cars, as per my other post i will carry on here as this is going in deeper.
It is also true that the manufacturers of these high perform street cars spend fortunes on fine tuning the design of the engine breathing system to optimise engine power performance and fuel burn. (This is free power so they are surely going to get all they can out of it and i appreciate that they wont get every drop due to mass -production effects etc)
These fine tunings include many complex scenarios being computed and tested ( exhaust size, bends, silencers, gas velocity, pulse speed etc etc ) resulting in a high quality breathing system where the alteration of any 1 parameter will have multiple effects on the other finely tuned parameters and these may be detrimental to the engine performance unless the alteration has been given the high brow analysis that is required to get anything more beneficial than a nice sounding purr.
I think especially the gas temperature, velocity and pulse rate are crucial in determining the additional suction rate of exhaust gases from the combustion chamber to enable faster and more air intake on the next intake cycle.
This is why ive always been drawn to the hirsh upgrades in that they seem to have seriously engaged the concept of upping the VE of the engine to gain xtra bhps and this has to be the most ideal healthy performance gain avenue possible although i wish the price was lower.
I must also look deeper into what the maptun, abbott, speedpart heads are saying in regard to the inclusion of their bigger exhausts.
 

·
Registered
Joined
·
1,448 Posts
Discussion Starter · #16 ·
An excellent read on the matter (though outdated in parts if you have an older version) is "The Scientific Design of Intake and Exhaust Systems" by Philip H. Smith.

http://www.bentleypublishers.com/product.h...il=&ticket=none

Also, pulse tuning is only really prevalent in naturally aspirated engines. This is because the speed of sound (and thus the speed of the pulses) increases as pressure does. Turbo-headers would have to be more than twice as long as naturally aspirated headers to be tuned for the same RPM as they run usually about twice the pressure at the intake manifold. (At least at the first set of scavenging waves.) If you've seen many naturally aspirated headers, you'd know why that's impractical!

The large benefit from turbo-headers is that they maintain exhaust gas velocity. That keeps the velocity pressure from being converted to static pressure. As long as it's velocity pressure the velocity vector will be pointed away from the exhaust ports, but if converted to static pressure some of that pressure points back at the engine in the form of back-pressure.

Also, when made equal length, or when using a split pulse header-turbo setup, the gain isn't so much in power, but rather in spool up. Evenly spaced exhaust gasses cut down on reversion in the turbo.

(The turbine is spinning quickly, and the pressure which drives it comes in waves primarily. If the waves aren't evenly spaced there will be long periods without any pressure to drive the turbine. Because it's spinning quickly, the gasses already inside it will want to reverse direction, and the turbo will slow down, or seem laggy.)

As for putting the exhaust in a vacuum behind the car ...

I'm sure there's some small benefit to it, but keep in mind that there is also vacuum to the sides of the car, and, with a proper spoiler/diffuser setup, also vacuum underneath the car. So I think the location isn't as important.

This is turning out to be a very interesting discussion indeed!

Dubbya~
 

·
Registered
Joined
·
186 Posts
A, having browsed the chapters of that link, that looks like the most comprehensive collection of information on intake exhaust theory imaginable and i think the title should be renamed "the ontological design of exhaust and intake systems"

It would be very beneficial if all the information you and others are uncovering here could be used in the study and recording of the "so far as known" effects of altering the exhaust / intake systems on the saab turbo engine. This might lead to a more proven database of positive and negative effects of these hardware changes.
For example a table could be listed as follows

1. Engine type
2 Part changed
3 Theoretical expected change in performance
4 Subjective interpretation of change in performance
5 Actual measured change in performance
6 Any reliability change
7 Any fuel consumption change

Point 3 would contain a series of possibilitys as uncovered by research and experience.
Point 4 would be im sure contradictory
Point 5 , 6 and 7 would be the most helpful in choice of tuning products
 
1 - 17 of 17 Posts
Top