How to Select Header Primary Tube Diameter – Summit Racing Quick Flicks

How to Select Header Primary Tube Diameter – Summit Racing Quick Flicks

Hi my name is Norm welcome to another Summit
Racing Equipment Quick Flicks. Today we are going to talk about primary tubes
okay, not so much collectors today but we are going to talk about primary tube diameter
od and the primary tube length. Before we go into that though I would just
like you to think for a moment about the combustion process that four stroke event which most
modern internal combustion engines share okay. I am particularly focusing on the exhaust
side of things today if we get into more theory down the road we will come back and think
perhaps about the intake side and the actual combustion side of things. But I want you to think about the fact that
combustion has just occurred okay so we have had the big bang, the controlled explosion
in that chamber. That explosion, mechanically forces the piston
to move towards bottom dead center at or near bottom dead center, your exhaust valve is
going to open as the exhaust valve opens your gases are going to be evacuated the reason
for that and the reason for more efficient evacuation is to make more power. The efficiency behind the evacuation of that
cylinder allows more or hopefully all of the spent gases all the byproducts form that previous
combustion event to be evacuated from that cylinder. This allows the intake portion of combustion
to fill a fresh voided cylinder so the more I can pack more power molecules for lack of
a better word into that cylinder, more carbon hydrocarbon and more air, more air fuel mixture
into that cylinder. Okay so that is what my primaries hopefully
do, what my header system will hopefully do. Now I want to base that information around
my build okay my cubic inches and we are going to see here that when we look at primaries
in theory were probably going to look at cubic inches per cylinder. Also around cubic inches and RPM so I want
to know where my peak RPM for my build is going to be. Very important to have this kind of information
in your mind or at least as a plan as we come into this theory and we look at primary tubes. Okay so todayís formula I want to solve for
a primary pipe area, in doing so I will know what type of primary tube diameter I am looking
for. Once I discover and find out my primary tube
diameter I can then look at primary tube length as an enhancement to that information. First of all, our initial equation tells me
that peak torque RPM equals the primary pipe area times a constant of 88,200 divided by
the displacement of one single cylinder now to simplify things, let me say this, I am
going to work with a three fifty cubic inch motor I have a 5000 RPM in mind for my peak
torque or peak torque enhancement so I am actually going to rearrange my equation to
solve for the unknown for primary pipe area or PPA. So PPA would equal the peak torque RPM divided
by 88,200 my constant again, then times my single cylinder displacement. Now the primary pipe area is going to be the
area of the ID so it would be area equals pie R squared and in this case I am measuring
the OD and I am taking away the wall thickness twice to come up with my primary pipe area
for my 1 and 5/8ths, one and three quarter and one and 7/8ths primary tubes. I wrote that information down already to simplify
our being together today just to make things easy. A 1 and 5/8th is 2.07 inches squared, a 1
æth is 2.19 inches squared, a 1 7/8th is 2.53 inches squared. My single cylinder displacement which I also
need to know based upon my 350 cubic inches divided by 8 cylinders per those 350 works
out to like 43.75 cubic inches per cylinder. Okay so when I start doing the math and plug
everything in, now remember I am looking for the fact that I have a primary tube diameter
I am solving for my RPM because I want to get into this 5K range for my racecar, I want
to peak torque around 5 grand. I will know in my mind that horsepower should
be beyond that. So I am looking for 5K. When I do the math, I find that a one and
5/8th primary tube peaks around 4175, a one and æ peaks around 4415, still not quite
to my 5000 RPM, when I look at my 1 and 7/8 primary tube diameter, I notice that theoretically
I am at 5100 so really very close to what I am after. Now bear in mind that for your application
you are only going to have so many possibilities. It might be ideal that a one and 13/16th would
put me right at 5000 to the best of my knowledge; nobody makes a one and 13/16th primary tube
diameter. It will be a compromise when you start looking
at what till fit your chassis and serve your purpose, unless you have really deep pockets. After solving the equation, we can see that
my 1 and 7/8th primary tube diameter is going to give me the results that I am after. Okay so I can tailor what is going to happen
with my 1 7/8th primary tube by thinking about length at this point so if I extend and make
my primary tube longer I am going to add more emphasis below the 5000 or 5100 RPM and add
a flatter torque curve on that side. If I shorten it and extend, pardon me donít
extend the primary tube length I am going to emphasize a flatter curve and more torque
available on the higher RPM side of 500 or 5100. So I hope you have picked up on this, next
time we will look at collectors and maybe some exhaust wave theory like that, hopefully
explain some timing events possibly as it involves your exhaust system. If you like this quick flick, please post
below and thank you for watching .

About the Author: Michael Flood


  1. Is this how one can tune for peak performance at desired rpms? Or is this how one would enhance the constant performance curve of an engine? I just got a small block 350 engine, so this really helps since I am new to this and eager to learn. Thanks

  2. I converted the single cylinder displacement of 364.25 cubic centimeters over to the common units of the formula – inches. The formula works with displacement in cubic inches all the way down to square inches for surface area and even inches for primary pipe diameter.
    Plugging you in, I have single cylinder displacement at 22.2 cubic inches and I used your supplied 3000 RPM to finish the equation and solve for the primary pipe area and then subsequently the primary pipe diameter.

  3. Primary pipe area = (3000 x 22.2) / 88200
    Primary pipe area = 0.7555 inches square
    Doing some more math I used 0.040 for the wall thickness for my mild steel circular tubing. The formula for the area of a circle = pi x radius x radius; and I pulled the wall dimension out.

  4. For example 1.25 inch OD pipe (a circle) minus the wall thickness twice (2 x 0.040 = 0.080)…so, 1.25 inches minus 0.080 = 1.17 my corrected inside diameter… so, 1.17 inches / 2 = 0.585 inches for my radius. Now 0.585 x 0.585 x 3.14 = 1.07458 inches square.

  5. I tried 1.25 inch OD and got 1.075 inches square for the pipe area (above), remember we are trying to solve for primary pipe area. Then 1.125 inch OD which yielded 0.86 inches square; followed by 1.0625 inch OD which got me very close with 0.76 inches square. So it looks like a 1 -1/16 inch OD tube with a wall of 0.040 inches is your theoretical primary pipe diameter for your particular single cylinder displacement and peak torque RPM.

    I hope this helps; let me know what you think.

  6. got it i missed a lot of steps there but now i got the formula and realize that i need to convert to cubic inch for it to work (duh). thanks for taking time to explain it i think i needed to see it with my application plugged in to understand it. thank summit!!

  7. What would be the best primary tube OD for a 2012 SS L99 Camaro? Was looking at a 2-1/2 in with a collector of 3 in. With the exhaust of 3 in all the way out. Want the best hp output. Please Help.

  8. At what RPM does your peak torque occur? And will this change in the future? Let me know which one you want us to plan for and we will get you a tube size 🙂

  9. Dear Roman,

    Sorry for the delay. Primary tube diameter is used to enhance the amount of torque developed at the peak torque RPM. That’s why knowing the actual or theoretical RPM where the peak in torque occurs is built in to the equation.
    Do you know the actual RPM or the theoretical RPM where your 350’s peak torque occurs? Let me know.

  10. I plugged the 4900 RPM that you supplied into the formula and come up with a theoretical PPA (primary pipe area) of 2.61 inches square. Doing the math for the surface area of a circle and using 0.040 inch for the wall thickness I have a 1 -7/8 inch primary tube outside diameter coming in at 2.53 inches square and a 2.0 inch version coming in with 2.94 inches square. I would use the 1 -7/8 inch set unless you plan on a dramatic change in displacement and/or peak torque. Hope this helps good luck!

  11. Absolutely most informative video I've ever seen about tube diameter. How would you figure the math on a stepped header application. I run tri-y stepped headers? 

  12. Hi, this is really one of the best summit quick flix I've seen. Norm is absolutely the right guy. I've seen a couple of other vids from Norm and there just simply awesome. Congrats Norm!!

  13. Could you elaborate on the last part? When you stated that depending on length torque distribution will vary, what are the reasons for this?

  14. Hi Norm, I was just trying to do the numbers for my 327 engine. Is it possible that there might by a mistake on your whiteboard? 2,07 in2 for the 1 5/8" tube seems to be considering the OD instead of the ID. Shouldn't the PPA consider ID instead of OD? 
    Please correct me if I'm wrong.
    Take care, 

  15. Hi guys.

    This link explains in more detail, and these are the formulas. For street motors.

    Peak rpm (torque)= ppa*88200/individual cylinder displacement

    Ppa=peak rpm torque*indiv cylinder displacement/88200

  16. Also, this guy and this video is one of the best.  Great for car geeks/engine building geeks, and people who don't have a more popular engine but still want the most out of their car

  17. hi there how do you computer for the  1 cylinder displacement? supposed i have a 4 cyl 1600Cc engine? do i divide 1600 with 4 ?

  18. i have oversized my  piston, done port and polish  should i change my headers? a 1.5 liter engine turned out to be 1.6 liter output right now, is there a need to increase the pipe diameter of my headers>?

  19. Has this engine been dynoed and found that peak torque is at 5000 rpm?  Otherwise this video makes it sound like primary tube diameter solely dictates peak torque rpm, which is not true…
    If I'm understanding this equation correctly, the calculated rpm is something like the "characteristic rpm" of the primary tube? Say for the given displacement and given primary tube diameter it is able to flow (without significantly increasing restriction) up to the calculated rpm?

  20. How do we calculate muffler size. I am making a gokart and so i need the muffler size. Since i am participating in an event i need to show the calculations also. Plz help

  21. You didn't do the second equation correctly when solving for PPA. If you want to solve for PPA based on the first equation you would multiply PtRPM X 1cl displacement and divide by 88,200.

  22. Good video, but it seems like the guy didn't really review the material beforehand. Uh, Uhmm, lack of better word.. Seriously man just think of what you need to say before the short video.

  23. First, 1-5/8" is 2.07 but the other two are wrong; 1-3/4" = 2.40 & 1-7/8" = 2.76

    Next, the second formula is incorrect; PtRPM x 1 cyl. Displacement / 88,200 = PPA

    Finally, an alternative formula to pi r squared is Diameter x Diameter x 0.7854
    (Example: 4.00 x 4.00 x 0.7854 x 3.480 x 8 = 349.85 (350))

  24. Didn't bother to read all the comments but the second formula is not correct..

    It should be:
    PTRpm x 1CylDisp
    ——————————– = PPA

  25. what would be the best size for honda accord 2.0 lt vtec head and the ruuners how long i want to use it for drag. i would be so glad sir if you will reply to this comment thank you sir

  26. The math is wrong for tube area;
    1 5/8" he didn't subtract any wall thickness.
    1 3/4" & 1 7/8" he only subtracted the wall thickness once (.080") if that is the thickness of the tubes..he should have subtracted .160".

    1 5/8" OD pipe with .080" wall thickness has an area = 1.686" (( 1.625 – (.080*2))= 1.465ID /2 = .7325 for the radius, squared = .53655625 * PI (3.14159) = an area of 1.6856 square inches.

    1 3/4" OD tubing (.080 wall) = 1.986 square inches
    1 7/8" OD tubing (0.80 wall) = 2.31 square inches

  27. The math for find Primary tube area PPA is also incorrect:

    It should say, (TQ peak rpm * 1 cyl. displacement) / 88,200

  28. I have a 1.4 liter turbo charged 4 cylinder car that has a complete down pipe through exhaust and muffler of 2.5 inches, mandrel bent. I would like to know if 2.5 inches is enough for my 205 WHP and 257 FT/LB of TQ to the wheels. It is a single pipe. My redline is 7k. 83 cubic inches. my guess it that peak torque is around 2k rpm.. This car is modified. It was suposedly at peak torque from the factory at 1850 Rpm's. When it was stock it had 140 bhp and 150 ft/lbs of base torque. That is 120 WHP..

  29. Pretty simplistic, but informative to get you in the ballpark… But your peak torque RPM would be afffected more by the cam profile and cylinder head flow/intake manifold type… It's the whole combination that should work together.

  30. You choose the wrong guy to play with smart guy !!!!! You'll ave fun with me !!!!!!! Is that anought exclamations ? Ill make sur everybody knows how smart and mature you are ,try to laugh of the polite peoples and KIDS hwo do some good comments ! Stabing a kid's mom's not to smartAndres .C GO Show your peoples are smart you are with !!! kids !!! INmature guy

  31. Norm, your is jacked. If your PTRPM equation is correct, manipulating for PPA = (PPRPM x displacement)/88,200

  32. how did you calculated the length of primary pipe? also please help me with my engine. I have EX500 engine of Kawasaki ninja and I'm using it for my formula racing car . will you please give me the calculations for primary and collector length ?

  33. Now if you had those tuned length headers put on, ie: each size, i bet my life the difference in power would be so little that in the car you really would not notice it,especially after a tune for each one, so the most important fact is this, if you had a 350 chev engine which can rev out to over 8,000 RPM then what you need to go over 1/ 7/8 " DIA headers i dont think so, because its all about when your peak torque comes in and ends nto how far your engine will rev out too ?
    Also if you have a street driven car like say the new Holden Chev SS which has a 6.2 lt engine and also considering everything under the bonnet which can get very hot and also gets in the way, then you can only do so much as far as the design of those tuned length headers, if you have not restrictions under the bonnet then no worries at all, then you would have the collector of each header pipe start out in the same direction as each exhaust exit angle and then mandrel bend the pipe so to fit and collect at a point to fit under your engine bay and not touch anything and more importantly easy access to spark plugs, other wise like most companies out there they simply copy each others designs and thats what is on the market right now, in australia and overseas as well ?


    if im doing exhaust piping fully 3" and in the middle reduce to 2.5" and then increased back to 3'" again towards the back end (tips), does this mean the whole flow is changed to 2.5"effect or 3" .

    or for example i want full 3"piping but im missing a 3" flange so i reduce one small part to 2.5"just for the sake of using 2.5"flange. but the rest is back to 3".. NOTE that one flanged part is reduced to 2.5" .

    can you please explain the effects. ty

  35. Labels! Labels labels, always label your units! B- and go back and redo your homework. The formula is wrong too.
    3500 RPM / (88200 * 44 ci for a Harley) for example gives us 9.0^-4 or 0.009. Really? What? LOL

  36. sir can i get the reference of the formula that you have used for calculating header pipe diameter. if possible please share reference with me

  37. First, I want to start out by saying that I love Summit for my 1977 Corvette! But some constructive criticism is that you keep distracting overly loud intros out of the video. Additionally, distracting words like "uh, and like" reduce the learning when you are teaching Math. Lastly, the only diameter I found on summit for my 77 350sb was 1 5/8". Thanks for the videos though!!!!

  38. great job Norm. I'm going to do a feature on Holleys for the Australian market on YouTube any advice would be appreciated
    Regards Smithy

  39. Great video ! To your calculations, are you suggesting a short header will have more torque it it leads to smaller collector per say 1 7/8 to a 2" than a long tube header 1 7/8 to a 3" collector

  40. I love this series…great platform. Please recommend and link additional reading on the subject. You’re making it easier to buy from Summit.

  41. Some right info, some wrong info…. First (top) formula is correct. Second (bottom) formula is mathematically incorrect. Lastly, the calcs for the PPA are wrong (did not take into account wall thickness).

  42. Wouldn’t you rather have peak torque low in the RPM range and use different gear ratios to move the torque curve?

  43. What can I do to help my lower end torque ? I have 2in fenderwell headers with 3.5 outlet. Can I build a collector to help keep the speeds up. Also the primary tubes are 42 inch long & I'm running a stroked big block mopar and looking at near 600 hp.

Leave a Reply

Your email address will not be published. Required fields are marked *