Exhaust Tubing Diameter (Rule-of-thumb)

Joe Del · Post by **Joe Del** » Tue Mar 06, 2001 5:49 am

The last item I need for my 2270 is an exhaust system. I'm considering having a custom exhaust system made for my engine. Is this rule of thumb pretty accurate?

< 150 HP - 1 1/2" tubing
150 to 200 HP - 1 5/8" tubing
> 200HP - 1 3/4" tubing

I'm primarily looking for low end torque, not top speed.

------------------
"If it were easy, anyone could do it!"

CptanPanic2 · Post by **CptanPanic2** » Tue Mar 06, 2001 7:09 am

I have no idea, but doesn't it have to do with the size of the exaust valve? Does the carb size have anything to do with the size of the intake valve?
Brian

Joe Del · Post by **Joe Del** » Tue Mar 06, 2001 12:02 pm

I didn't think the rule of thumb would be cut and dry. I think specific info would be more accurate. This is for a 2270cc motor (78.4 x 96). I have 2.0 liter Porsche 914 heads. The oval shaped exhaust ports are about 1 1/4" at the narrowest point and 1 1/2" at the widest point.

More head details:
44mm intake
36mm exhaust
Comp. 3-angle valve job
Reshaped combustion chamber (55cc)
Full port and polish
45mm Dellorto carbs
Match ported intake manifolds
163/86B split duration cam (.500 lift, 284/300duration)

How does a 1 5/8" merged collector sound with this setup?

[This message has been edited by Joe Del (edited 03-06-2001).]

Post by **Tom Notch** » Tue Mar 06, 2001 6:51 pm

Joe, I'm going for an 1 5/8" merge for my 2.7 (104.5x78) with 914 2L heads sporting 44x38 valves. Going for a primary tube length of 37" and an flange diameter of 1 3/4" going up to a 2 1/2" pipe and muffler. I have to go full custom on mine and have most of it kinda figured out in my head and the parts laying here. The rest of the engine specs are on my web site.

I'm tuning for 1500 to 5500 rpm or so for the power band. That should complement the automatic trans,I hope. Mine so far looks like the outlet of the merge will be in the right spot under my Notch for a turbo to fit in comfortably. Someday, maybe.

------------------
Tom Notch
Tom's Old VW Home

tarnx · Post by **tarnx** » Mon Mar 19, 2001 12:29 am

I have been looking digilently for the answer to exhaust sizing for almost a year now. I haven't got to far as it seems to be an art and not a science. You need to keep up the velocity of the exhaust gases. The tubing should be uniform in diameter even thru bends. The length of the primaries determines the RPM at which maximum scavenging takes place. Once the exhaust gases cool (too long a primary) or expand (where it meets the collector) the velocity drops off and you loose the scavenging effect. There is alot of theory I have found but it would take forever to write it all down.
My best quess is size the tubing the same as the exhaust port not the valve, have it mandel bent, and use the average of the lengths of the two exhaust primaries on a stock setup. The length of a Porche exhaust would probably be close too.
I would love to get a hold of Jakes dyno for a day. The way they used to do it is take the engine up to the rpm they wanted to run and cut off the tube just after where it turned blue. The exhaust gives off its heat where its velocity drops off.
This is no way definitive just somethings I have found out.

Joe Del · Post by **Joe Del** » Mon Mar 19, 2001 5:42 am

I'll be having my exhaust system fabricated in a few weeks. I'm currently "mocking up" the long block to set the valve train geometry. Them comes the exhaust system. Since my engine is a daily driver and I'm more interested in low end torque, I decided to run a 1 5/8" merged collector with 36" to 38" equal length mandrel bent tubing. I'm installing a Suppertrapp tunable muffler after the collector so I can adjust flow rate and noise level. I have access to a Chassis dyno so I may try the "heat and cut" method that tarnx described. I'll post the dyno results after I make a few runs.

Tarnx · Post by **Tarnx** » Mon Mar 19, 2001 8:27 am

Joe I don't wish to lead you astray. I will try most anything even if it leads to a deadend. I have spent the last 4 months prototyping a set of intake manifolds to adapt 40mm kadrons to single port heads. I thought it was going to be easy. Four pair of hand cut flanges later I am finally in the ball park and I still don't know if they will work like I want.
Don't quote me or blame me , but this is how I was going to do it. JC Whitney has various size tubing with 45,90, and 180 degree bends. Weld up the stubs and run all four tubes out the back of the motor or which way it takes to clear the dyno and motor. Start with about 4 feet of tubing and crank up the motor and take a horsepower check. If the hp curve is high you need more length, if it is too low you need to cut some length. I would cut off about an inch each time with a pipe cutter trying to keep every thing the same length. Run a new hp check and see what happened. Alot of other things are also involved such as cam timing, ignition timing and especially carb jetting. As the scavanger effect increases the Carb will lean out. Once you get the rpm hp curve you want join all four together into a collector.
The diameter and length of the collector also effects scavenging. If the tubes are merged or not, how long is the tail pipe, how big around is the tail pipe, what is the back pressure on the muffler, I can't answer these questions. I would look at a set of T1 merged headers and use all the dimensions they use for the collector with my pipe diameter and length for the primaries.
My original idea behind using a supertrap was to save space as they are tunable by placing more or less disks in them making them appear to the system as a longer or shorter collector. Supertrapp might lend you assistance on where to start.
Diameter and length of the primaries, secondaries, collector, and tail pipe all affect the RPM the motor makes HP. There math formulas to determine all of this but I can't seem to find them.
Columbus sailed west not knowing where he would end up, how long it would take, or even if he could find his way back. Hopefully you won't end up floating around the ocean on a raft made of milk jugs eating sushi the rest of your life. Good luck.

kdanie · Post by **kdanie** » Mon Mar 19, 2001 2:20 pm

To muddy the waters a little more....the latest technology in circle track headers is "stepped" headers...the primary pipe increases in diameter at specific lengths to broaden the power band. Someday when I win the lottery I will buy a engine dyno and have some fun with all this type of stuff.

tarnx · Post by **tarnx** » Mon Mar 19, 2001 9:52 pm

Thanks for the info on where to find info. As I stated before I can only get you started. My personal desire is to build a tri Y configuration which was used primarily by early rodders because of its wide power band. But there is so much to consider and experiment with these days. Technology has grown and data modeling programs are probably available which take out alot of the unknown factors. I don't know of any, but you have given me leads to go on.
May the journey to the destination be as interesting as the destination.

kdanie · Post by **kdanie** » Mon Mar 19, 2001 11:00 pm

To learn more about header design check out www.burnsstainless.com and go to the tech articles. It's an excellent site with much information. Please post your results when you complete your set of headers.

Spyke · Post by **Spyke** » Mon Mar 19, 2001 11:46 pm

<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:</font><HR>Originally posted by Tarnx:
<B> <SNIP> There math formulas to determine all of this but I can't seem to find them.
Columbus sailed west not knowing where he would end up, how long it would take, or even if he could find his way back. </B><HR></BLOCKQUOTE>

But you WILL find them before you head for the dyno, right? After all, you can bet Columbus would have taken a GPS with him, had it been available!

My point being, there is a lot of information out there on exhuast system design. I think I'd forgo some of the dyno time and spend it looking for the information others have already gathered.

Just a thought, does the "Desktop Dyno" program let you model a header? Or maybe there's a more modern version of the program?

I don't think your primary tube experiement is going to pay off. As you said, once you join the tubes the gas dynamics will change. The only thing your experiment is going to tell you is what a good length for max hp at one rpm would be. There is a lot more to a useful powerband than that. For instance, most high performance four cylinder bikes use a four into two into one design header to avoid a dip in the midrange torque.

In general, peak HP numbers are not the holy grail of exhaust tuning, especially not for a street car...

[This message has been edited by Spyke (edited 03-19-2001).]

Steve Arndt · Post by **Steve Arndt** » Tue Mar 20, 2001 8:36 am

I called up headers by ed, and bought their 20 dollar info packet (catalog, lots of loose printouts, audio tape on header design). If you are serious, step up, and buy this info. Search in google for "headers by ed" and you'll find it. He has the best technical info backed by testing.

Steve

Tarnx · Post by **Tarnx** » Tue Mar 20, 2001 10:24 pm

After endless seaching and questioning today I presented my problem to my local VW guru. Robert Bently has the book I have been looking for. (www.rb.com) It is catalog #G309 The Scientific Design of Intake and Exhaust Systems. $22.95 It is in the Subject Index under Performance. They also have books on suspension, aerodymnamics, and performance design information for aircooled VWs and Porche 911s. This will be required reading as soon as my wife quites screaming about buying $100 worth of CAR books. Knowledge is power that can never be lost or taken away. Bathroom wall, Phillips 66, 1974, Lead South Dakota.