Jump to content
HybridZ
Daeron

Exhaust Tube Sizing: I did your Arithmetic for you!!

Recommended Posts

I was reading a thread about a Twice Pipes setup, and wanted to post a link to something I had done in the past. I couldn't find it, so I thought I would post this up here and nominate it for s Sticky for reference purposes.

 

We all like to HEAR our engines. The lure of big, loud pipes calls gearheads of all ages, no matter what kind of machine their motor is attached to. Dirtbikers, riverboat racers, weekend lawyer Harley riders, Pontoon Boat captains, helicopter pilots, racecar drivers.... we ALL love our pipes.

 

Unfortunately, though, due to a tricky arithemtic process called squaring, whereby the radius of a pipe is multiplied by itself (and pi) to obtain the cross-sectional area (and thus, the flow capacity) of a circular pipe... Many of us find it very difficult to easily visualize and compare the differences between a 2" and a 3", and dual 2" vs single 4", and dual 1.75" versus single 3" (or was it single 2.5"?) pipe setup.

 

Rephrase: Comparing potential exhaust/intake tubing sizes is tricky because you have to square the difference in size. Our brains are used to comparing the difference in size in a linear manner. Thus, two falsehoods easily take root in EVERYONE'S mind (until education prevails!!):

 

A: The difference from 2"-3", and the difference from 3"-4", would seem to be the same;

AND

B: The capacity of a single 3" pipe should be roughly twice that of a pipe roughly half the size, 1.5".

 

Unfortunately, both of those statements are dead wrong.

 

Let me go do the math for you!!

:wc::wc::windows::wc::hs::windows::mparty::cool:

 

So!!!!!!!!!

 

Here you go.

tubingsizeversusarea.jpg

 

The numbers in the two columns on the left are pipe diameter. This is how pipes are sold usually. The numbers in the columns on the right are cross-sectional area. THIS is the number that you compare apples-to-apples with between different size pipes when determining flow. Double the cross sectional area, you double the flow capacity. However, lets look at the cross sectional area of a pipe that is 1.75". At 2.405 square inches, we want to know what to compare our nice, new dual system to, so we double that number to get about 4.8 inches. Scroll down on the chart a few lines and there is one at 4.909 square inches. Scan back over to the left and we discover that the twin 1.75" pipes have just a hair less flow potential than the single 2.5" pipe. Further inspection reveals that it takes twin 3" pipes to beat a single 4" pipe!

 

 

These things are not as intuitive as they might seem, and the further away from Zero you get, the more drastic the changes make themselves felt. I hope this helps as much as I imagine it could.

Share this post


Link to post
Share on other sites

I was reading up on exhaust stuff, got bored, but felt like doing something semi-productive. This graph shows the relationship between pipe diameters and cross-sectional area, and the differences between a single and dual arrangement in terms of area.

 

Choose a pipe diameter, find it on the graph, and follow the horizontal lines to find an equivalent diameter. The abscissa represents single pipe diameter (even for twice pipes) and the ordinate is total pipe area.

 

For example, it can be seen that choosing 1.75" twice pipes (MSA) is equivalent in area to a roughly 2.5" single pipe.

 

Enjoy...

 

image002.gif

Share this post


Link to post
Share on other sites

Nice chart! This is why we were supposed to learn math in school ;)

 

A couple observations;

Obviously you need big(er) exhasut to make big power but even if you are not nearing the theoretical choke size of the tube an oversized exhasut can still help a turbocharged setup. The turbo spools almost entirely because of a pressure differential between the inlet of the turbine housing and the outlet. A nearly perfect turbo exhaust would be a cone coming right from the turbine outlet (exhaust manifold pressure ---> atmosphere). That's usually a pretty bad idea on a car though...so, a big huge exhaust is a good solution to make the biggest pressure differential across the turbine and get the fastest spool up and boost response while not burning your car to the ground. So the guys not making huge power on small engines are still benefiting from a 3" exhasut. Sorry if that was rambley, but i get excited about turbo cars and making exhausts for them.

 

Another thing to remember is that the exhaust flow "sticks" to the walls of the tube so even though two tubes may have the same crosssectional area, the internal surface the exhasut is "dragging" on is quite different. The single 4" has WAY less internal area than two 3". For this reason a round tube flow much better per square cm than square one.

 

Just some ideas to think about. I build exhaust and other plumbing for high performance and custom cars all the time, so this is what I like to think about while I sit shirtless on my couch...after work...sigh

Edited by Snailed

Share this post


Link to post
Share on other sites

Nice chart! This is why we were supposed to learn math in school ;)

 

A couple observations;

Obviously you need big(er) exhasut to make big power but even if you are not nearing the theoretical choke size of the tube an oversized exhasut can still help a turbocharged setup. The turbo spools almost entirely because of a pressure differential between the inlet of the turbine housing and the outlet. A nearly perfect turbo exhaust would be a cone coming right from the turbine outlet (exhaust manifold pressure ---> atmosphere). That's usually a pretty bad idea on a car though...so, a big huge exhaust is a good solution to make the biggest pressure differential across the turbine and get the fastest spool up and boost response while not burning your car to the ground. So the guys not making huge power on small engines are still benefiting from a 3" exhasut. Sorry if that was rambley, but i get excited about turbo cars and making exhausts for them.

 

Another thing to remember is that the exhaust flow "sticks" to the walls of the tube so even though two tubes may have the same crosssectional area, the internal surface the exhasut is "dragging" on is quite different. The single 4" has WAY less internal area than two 3". For this reason a round tube flow much better per square cm than square one.

 

Just some ideas to think about. I build exhaust and other plumbing for high performance and custom cars all the time, so this is what I like to think about while I sit shirtless on my couch...after work...sigh

 

Those are good points, and are topics that people usually don't consider when talking about exhaust design. I agree that large exhausts definitely help a turbo spool faster, but I also think that they help naturally aspirated engines.

 

There is a lot of evidence that a properly tuned engine makes more power with open headers vs. having a pipe connected to the exhaust (contrary to internet myths). I believe the reason for those results is that a lot of the cylinder scavenging occurs in the headers (given that they are adequately long and merge into one pipe), and the exhaust pipe is mostly just there to route the gasses to the back of the car. There may be some resonance effects dependent on pipe length, but as I've implied, given the header is adequately long (length is dependent on where you want your powerband), any effects generated by the pipe are likely negated by frictional losses.

 

Thus, this long skinny pipe increases frictional losses in the exhaust flow, increasing pumping losses and decreasing volumetric efficiency while not really helping anything. So, you want as little friction loss as possible, while maintaining ground clearance. Essentially, you want the pipe to act as the atmosphere.

 

 

Another point that was made is that friction in the pipe(s) is dependent on surface area. The flow sticking to the pipe is a good way to visualize it. This is the concern I have in using MSA twice pipes, as that is what I plan on getting when my Z is built. The cross sectional area is roughly equivalent to a single 2.5" but the twin 1.75" pipes have more combined inner surface area, and thus there will be more losses. Whether performance is affected much, I don't know. When I'm ready, I do want to do some exhaust dyno testing to show myself which setup works best.

 

Keep in mind, my post is based on a combination of theoretical knowledge and experimental data. If someone has done some relevant dyno testing that would be great! If I missed anything, feel free to contribute.

  • Like 1

Share this post


Link to post
Share on other sites

There is a lot of evidence that a properly tuned engine makes more power with open headers vs. having a pipe connected to the exhaust (contrary to internet myths). I believe the reason for those results is that a lot of the cylinder scavenging occurs in the headers (given that they are adequately long and merge into one pipe), and the exhaust pipe is mostly just there to route the gasses to the back of the car. There may be some resonance effects dependent on pipe length, but as I've implied, given the header is adequately long (length is dependent on where you want your powerband), any effects generated by the pipe are likely negated by frictional losses.

 

 

I disagree. You can gain additional scavenging benefits and power band tuning with piping after the header collector. Its very common to gain horsepower and broaden the torque band by adding tuned megaphones and secondary merge collectors.

 

559631212_6XTd4-M.jpg

 

even F1 cars with their 18,000 rpm 2.4L engine run very specifically tuned pipe lengths.

 

379194530_f267f2e545.jpg

Share this post


Link to post
Share on other sites

John, think you missed leons point since you are really just elaborating on what he said. I would hardly call a 13" megaphone a full exhaust, but yes they do help scavenging on NA cars.

 

An interesting thing on those F1 headers and a good example of things we can learn from $$racing$$ and apply to our street cars... As counter intutive as it is, a steeper (15* or more) merge angle on the collecter is actually MORE efficent than a shallow one at high rpms. I beleive this was figured out when F1 was dominated by turbo hondas in the 80s. That and prioritized wastegates are two great F1 concepts that I apply to my turbo manifold builds 25 years later haha

Share this post


Link to post
Share on other sites

John, think you missed leons point since you are really just elaborating on what he said. I would hardly call a 13" megaphone a full exhaust, but yes they do help scavenging on NA cars.

 

An interesting thing on those F1 headers and a good example of things we can learn from $$racing$$ and apply to our street cars... As counter intutive as it is, a steeper (15* or more) merge angle on the collecter is actually MORE efficent than a shallow one at high rpms. I beleive this was figured out when F1 was dominated by turbo hondas in the 80s. That and prioritized wastegates are two great F1 concepts that I apply to my turbo manifold builds 25 years later haha

 

I dunno, he said "open headers."

 

Merge collector angle is very dependent on the cam. On the 310 duration cam in my old race car I ran a 26 degree merge collector. On a customer's ITS prepared 2.4L (stock cam) I built a 14 degree merge collector. Both designs provided the best overall horsepower for their particular use based on engine dyno tests.

Share this post


Link to post
Share on other sites

I dunno, he said "open headers."

 

Merge collector angle is very dependent on the cam. On the 310 duration cam in my old race car I ran a 26 degree merge collector. On a customer's ITS prepared 2.4L (stock cam) I built a 14 degree merge collector. Both designs provided the best overall horsepower for their particular use based on engine dyno tests.

 

Yeah, I see your point. Either way, an 8 foot exhasut pipe isn't helping anyone make big power.

 

I'm sure cams change things quite a bit when you are focused on scavanging. I'm more used to turbo stuff where stock cam cars just get more boost to make the same power ;)

Share this post


Link to post
Share on other sites

I disagree. You can gain additional scavenging benefits and power band tuning with piping after the header collector. Its very common to gain horsepower and broaden the torque band by adding tuned megaphones and secondary merge collectors.

 

even F1 cars with their 18,000 rpm 2.4L engine run very specifically tuned pipe lengths.

 

 

I agree with that John, and I wasn't trying to imply that open headers is the "best" exhaust design. I think I should have clarified my statement a bit better. A stock car (i.e. a Z) will make more power with a 6-into-1 header without a "typical" pipe (say 2.5" with the tip at the rear bumper) than with the pipe. That's what I meant by my vague description of a "pipe connected to the exhaust."

 

Now when you get into exhaust tuning, of course you tune for certain lengths which give you resonance effects at certain engine speeds. The megaphone can be thought of as a bell-mouth for the exhaust. Megaphones help broaden the bandwidth, if you will, of the pressure wave reflection (resonance effect) as the taper of the megaphone pipe can be thought of as infinitesimal "steps" with a pressure wave reflection at each step. It's the same deal with intake bell mouths.

 

Megaphones are surely beneficial, but sadly you don't find megaphones attached post collector on most factory cars, and they are even excluded from many custom built street exhausts! Hence, taking off the long exhaust pipe, and leaving an open header increases torque.

 

Now, this leaves the option of building a new exhaust that can make performance even better, but the original point I wanted to make was that a long "typical" (no megaphones, no specifically tuned lengths, etc.) exhaust pipe is detrimental to performance. You gain torque by removing the pipe and running open headers, and then even more torque can be found when you add things like megaphones and tuned lengths post collector, as John mentioned. Thus as already mentioned above, when people are having 2.25" or 2.5" "typical" NA exhausts built, they are still making a performance sacrifice (and no, you don't lose low end torque :lol: ).

Share this post


Link to post
Share on other sites

Exhaust systems are a necessary evil for all cars. The exhaust gasses and noise need to be directed away from the driver if the driver is to perform at an optimum level for any length of time. In a properly designed racing exhaust system the most evil part of their installation is the additional weight.

Share this post


Link to post
Share on other sites

Exhaust systems are a necessary evil for all cars. The exhaust gasses and noise need to be directed away from the driver if the driver is to perform at an optimum level for any length of time. In a properly designed racing exhaust system the most evil part of their installation is the additional weight.

 

Right. As I've said, we need the pipe because its purpose is to route the gasses to the back of the car. We need exhaust pipes, and you are right in that a "properly" designed exhaust system won't have a performance hit, besides the weight. I've been jumping around it, but what I'm trying to do here is construct an explanation dispelling the myth that a big exhaust pipe will hurt low end torque, or that you need a stock style exhaust or else you lose power. It will do nothing but help torque over the entire rpm range when compared to a smaller pipe.

Share this post


Link to post
Share on other sites

Its obvious an exhaust system can choke an engine if it's too small, but is there a too big?

I've always taken exhaust systems the same way as header primary diameters, smaller to keep up velocities for low end torque and bigger for the higher RPM horsepower.

Share this post


Link to post
Share on other sites

Its obvious an exhaust system can choke an engine if it's too small, but is there a too big?

I've always taken exhaust systems the same way as header primary diameters, smaller to keep up velocities for low end torque and bigger for the higher RPM horsepower.

 

As discussed in the thread, there is a significant distinction between the job of the header and collector and the long exhaust pipe.

 

If you size your exhaust pipe to "keep velocities up" you are robbing yourself of torque, since frictional losses are increased by both decreasing pipe diameter and increasing velocity. Frictional losses go up with velocity squared. Add in the long length of the pipe and some bends, and even more torque is lost. Thus, with a properly designed exhaust, there is no "too big" when it comes to performance. Otherwise, there is a "too big" and it's when you can't fit it under the car anymore! Of course, there comes a point of diminishing returns (upping diameter has negligible gain) and that's when you've sized the pipe properly.

 

Scavenging is the purpose of the header and collector, and possibly a tuned length of exhaust, not your long pipe that goes out back!

Share this post


Link to post
Share on other sites

If you size your exhaust pipe to "keep velocities up" you are robbing yourself of torque, since frictional losses are increased by both decreasing pipe diameter and increasing velocity. Frictional losses go up with velocity squared. Add in the long length of the pipe and some bends, and even more torque is lost. Thus, with a properly designed exhaust, there is no "too big" when it comes to performance.

Hopefully what you are leading up to is mentioning tuned parameters/dimensions for a certain RPM you want to work in. If I say I want peak power at 5000 RPM I won't be running 1 3/4" primaries. Velocities would be too low to take advantage of. Same reasoning for carbs. I'm not going to run a 50mm with a 47mm choke if I'm not going past 6500 RPM. Velocities and volume of the intake charge with large ports and runners wouldn't work in my favor.

 

Other ways to keep velocity up is to keep the heat in. Ceramic coat, wrapped headers, whatever. Some people ceramic coat the entire length of the exhaust so heat isn't lost. Also look into some thread I made a while ago showing some header porn from the motherland. One thing I noticed right off the bat is that a lot of them had their primaries going straight out and rounding over rather than straight down as soon as the exhaust gas exits the port. Unfortunately I don't have any pictures because my computer is dead. :[ But just imagine the gasses entering the header and hitting a wall as soon as it enters, because the primaries are directed downward. The expensive headers I saw had the gasses continue a couple inches outward and a nice big 90º bend was made. Of course, right hand drive you have more room but there is still plenty of space with LHD. Just for fun I'm going to create my own now that I'm doing some decent welds.

 

Compare MSA to others to see what I mean.

MSA where the connection between the primary and the flange, the primary basically goes straight down:

large156012a.jpg

 

The others, note its never a sharp tight radius angle, always large. From the translations they all mentioned tidbits of velocity and as you can see one mention of a cyclone effect:

vo24g8.jpg

34hfaq8.jpg

 

Helps explain the functionality of a megaphone (pulse jet motor):

1hqrnd.jpg

 

So I think this kind of works along with what you're saying, a little. Small enough to keep velocity up higher than needed for your performance needs, is obviously restricting. However, there is a too big, once again lowering the velocity to where it may not be good for your application. For instance on a race prepped Spitfire, the tuned length and size of a proper pipe, when standing behind it from 5 feet away I can feel it on my legs. For my Z with a 3" pipe at idle, I can't feel it more than 2 feet away. The only use of my 3" pipe would be if I started winding it up past 6000 RPM, which I don't so shame on me.

 

My point is, and we all already know this is if you take it to the dyno with a set length, cut off 6" at a time and watch what happens to the power. When the power drops after so many 6" pieces have been cut off, return to the length 1 cut before the drop, then insert a sliding piece of pipe so you can slide it in and out in to find where within that 6" length the power is at its peak to finely tune. Once you find your peak, that is your optimum length for that setup (pipe size, primary size, header bends, cam, displacement, RPM) Change any single thing, and you will need to go back and do it all over again, if you wanted to be picky. A guy running in vintage with a 2.4L motor has his pipe coming out just behind the seat. His butt dyno claims it to be better than all the way to the back.

 

IMG_6936.jpg

 

I need to shut up and sleep. Tomorrow I will read over this and be like.... why did I say that.

Share this post


Link to post
Share on other sites

Anecdotal Dyno Remberances:

 

L28 on Bonny car lost 20 HP when we bolted on the full exhaust needed to run the full belly pan. TWIN 2.5" exhausts. 8500 shiftpoints.

 

on the same dyno...

 

L20A did a baseline tune without the exhaust same as we did for the L28, then hung the exhaust in anticipation of having to reconfigure the map slightly to optimize for the restriction of the exhaust. We lost no horsepower, and ran the test several times because this really surprised us. As a result when we run the 2 Liter engine we ALWAYS run the full exhaust for the reasons John C mentioned, the car is WAY quieter with the rear exit exhaust.

 

But with the L28, 20HP is 20HP and if we don't need to run it, we won't!

 

Conclusions:

 

For a 2L engine pumping to a shiftpoint of 9500 rpms, headers and a twin 2.5" exhaust did not change our horsepower at all.

For our 2.8 L engine running to only 8500, that same 2.5" twicepipe system costs us 20HP (it was more before the fuel and timing maps were tailored to the different exhaust characteristics.

 

So the 'nothing is 'too big' argument' I would tend to agree with---but the size of tubing required for an exhaust to NOT drop ultimate horsepower is far larger than most people would think!

 

There are other considerations, but I was personally shocked to see our drop running twin 2.5's. I figured that would be PLENTY of flow. It is, in a sense....but only for a 2-liter! :blink:

Share this post


Link to post
Share on other sites

Josh, you may want to re-read the thread, as I think you missed the point of the post you quoted.

 

Here are the parts of post #13 that you didn't quote:

 

As discussed in the thread, there is a significant distinction between the job of the header and collector and the long exhaust pipe

 

(...)

 

Scavenging is the purpose of the header and collector, and possibly a tuned length of exhaust, not your long pipe that goes out back!

 

Therefore, you want to correctly size the header and collector in order to tune the exhaust system. Again, you're very focused on velocity but that's not the only, nor the biggest thing that drives exhaust (or intake) tuning. System resonance is what really drives this phenomenon. I think I already mentioned it earlier but I'll reiterate.

 

In the exhaust system, as the valve opens a compression wave is sent downstream to the header and collector. At a discontinuity, such as the collector, the compression wave reflects back upstream as an expansion wave. If this expansion wave reaches the valve right before it closes (after the peak in the valve's mass flow) it creates a secondary peak of mass flow-rate through the valve by essentially sucking out (technical term ;) ) more gasses. Therefore, you size the length of the header and collector based on parameters such as valve duration, pipe diameter, exhaust temp, etc.

 

This stuff is more complex than just "keeping the velocity up" as many think. Tony's car is a prime example of this, where the header/collector is tuned and varying the long pipe out back at best doesn't affect anything (but noise), and at worst starts robbing more and more power.

 

I highly recommend getting this book and reading and understanding as much as possible. If you don't "get" something, then post! :)

Share this post


Link to post
Share on other sites

I would refer you to Kas Kastner then. The information he writes about on Triumphs specifically notes that you want your exhaust pipe to be exactly this long with this setup. Granted my translations of the Japanese pages were... barely helpful, the subject of keeping velocity up was found multiple times. Doesn't necesarily mean keeping it up by restriction. It was just an observation, it may mean nothing.

 

I agree with you that the header is the major part of the system, however I honestly believe the pipe matters too. Hundreds of Triumphs are out there to prove it, especially my fathers. The method is also used for gocarts. One thing I will admit is perhaps the tuned length is more-so for smaller displacements considering a Spitfire is 1.3L and a gocart motor is 100-200cc. I wish Don Potter were still around so we could give him a call and see whats up. The vintage car I posted, has Potter exhaust stuff on it.

 

Now, Potter was not his engine builder, however he uses a guy who had close contact with Potter. Tony, did you ever try varying lengths or did the tests end in the conclusion that full length X diameter had a loss of 20hp and no pipe showed the 20hp improvement? Was there any middle ground where there was more power to be had?

 

Once again we hit a point on Hybridz where books and calculations say one thing but tests show another. Once again I can only relate to what I've seen with my dad and those are Triumphs, perhaps Z's are completely different.

 

I wouldn't take calculations as gospel though. The rules and patterns of the periodic table are there but there are many exceptions. The Lewis Dot Structure explains the majority of atomic/molecular structures but there are some instances where it doesn't explain.

 

Regardless of where this discussion goes if I were serious about making all the power I could get, then I would go dyno test varying lengths even if its said it doesn't do anything. Tony, for the 2.8L with the pipe, where was peak power compared to without a pipe?

Share this post


Link to post
Share on other sites

Whether it's a Z or a Triumph or a bike engine, the same principles apply. As this thread discusses, the header and the collector pipe are the tunable aspects of the exhaust system. You're talking about tuning the secondary/collector length. This is what you would do with any engine's exhaust. Trust me, buy the Heywood book and read it. You will have many more revelations reading that book than browsing the internet or consulting engine builders...

Share this post


Link to post
Share on other sites

Our initial testing was done on open headers and the header was optimised (including the header collector) to get best HP under the curve. After that, we grabbed the largest pipe we could fit under the car and still have ground clearance. We wanted as little flow restriction as possible after our tuned header collector combination. We didn't run the 2.8 with the belly pan that often, only two classes, and with the hassle of installing / removing it we concentrated on open header performance.

 

The exhausts collectors were 2.5" diameter, and we couldn't make megaphone 2.5 to 3" transition pieces fit due to the angles involved, so we concentrated on mandrel bent 2.5" tubing with as little bending as possible to get the turbes to exit the back of the car. I think our backpressure under full load at full rpms was less than 4" W.C as I recall...but this was 10 years ago. It was near nothing. We could not explain the 20HP drop with such a low restriction other than there had to be some sort of anti-reversionary effect of dumping the exhaust pulses out of the collector which scavenged better than dumping it into a tube. We theorized for a while about putting 4" sections on a stub of 2.5" piping then merging in the horizontal plane by the tranny to the 2.5 out to the back...but that was not to be. We made our record and stuck with what worked.

 

Running the 2 liter on headers with primaries that were (I forget now exactly) 2 or 3" shorter than the L28 we did not experience the loss in the same piping. Hookup was identical almost all the shortening of the pipes was done in vertical header primaries, so the exhaust just 'tucked up higher' than previously giving us slightly better ground clearance at the transmission bellhousing area. That it worked on our existing piping and we didn't have to make a completely new set to run out back was an unexpected plus that went along with not losing any HP!

Share this post


Link to post
Share on other sites

I'll just say I'm wrong and continue on. Lol

 

This was a header system I was wanting to base mine off of. The bends don't look as complicated as others. It would be a fun project. I could see potential difficulty in the merge engle that you guys discussed.

24ce152.jpg

Edited by josh817

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now


×