Why round exhaust pipe?

[Bob Holmes]

Its interesting Bill. I've been reading Gordon Blair's book on 4-stroke engines. Apparently going a bit too small on the exhaust is not nearly as punitive to horsepower as going a bit too big.

The trick is knowing which one you've got. LOL.

[Beagle]

playing with primary diameters on a simulation software does interesting things Below the torque peak. Above torque peak, too small always seems to lose bad. If you're only worried about torque peak and above, bigger to the point of too big doisn't seem to be a problem since nobody offers 2.5" primary headers for a 350 that I know of, center cylinders be damned.

Since I care a lot about things happening before peak torque, I got amused looking for "just right". Too small can make huge differences 1000 rpm below torque peak and moves it down several hundred RPM. Chasing averages like Bob has been doing could really burn up some dyno hours with just the primaries, then you toss in step headers and it makes my head spin. I'd be most concerned about things happening 0-48" from the exhaust ports. Has anybody done a D shape to match some of the exhaust ports, and then transitioned to a circle?

[MR P-BODY]

playing with primary diameters on a simulation software does interesting things Below the torque peak. Above torque peak, too small always seems to lose bad. If you're only worried about torque peak and above, bigger to the point of too big doisn't seem to be a problem since nobody offers 2.5" primary headers for a 350 that I know of, center cylinders be damned.

Since I care a lot about things happening before peak torque, I got amused looking for "just right". Too small can make huge differences 1000 rpm below torque peak and moves it down several hundred RPM. Chasing averages like Bob has been doing could really burn up some dyno hours with just the primaries, then you toss in step headers and it makes my head spin. I'd be most concerned about things happening 0-48" from the exhaust ports. Has anybody done a D shape to match some of the exhaust ports, and then transitioned to a circle?

I make my headers D shaped to match the port then within about 2" it transitions back
to round... when ever I can I try to have 1 cylinder volume in the tube before it reaches
the first tangent point but in alot of cases thats not possible

[CDMBill]

My exhaust ports have a slight d-shape on the short side radius and I went with round flanges as: A. that was what was available at a reasonable price and, B. the reversion dam theory of breaking up the flected wave alightly before it hits the back of the valve. And no I have not tested a set of otherwise identical headers with D shaped flanges. It took me forever to build the first set as it was.

The Trick Flow A460 v3 heads have a BBC exhuast flange pattern and I used .375 thick laser cut flanges that were 2.25" in the raw form and used 2.125" tubes for the first step. These were tig welded in side and out after I tacked them up and built the rest of the tubes down to the collector flanges with the Mig.

[skullbucket]

You never see a river with a square bend in it.

[MR P-BODY]

You never see a river with a square bend in it.

Nature usually has the best designs... tried and trued

[jcharliem]

playing with primary diameters on a simulation software does interesting things Below the torque peak. Above torque peak, too small always seems to lose bad. If you're only worried about torque peak and above, bigger to the point of too big doisn't seem to be a problem since nobody offers 2.5" primary headers for a 350 that I know of, center cylinders be damned.

Since I care a lot about things happening before peak torque, I got amused looking for "just right". Too small can make huge differences 1000 rpm below torque peak and moves it down several hundred RPM. Chasing averages like Bob has been doing could really burn up some dyno hours with just the primaries, then you toss in step headers and it makes my head spin. I'd be most concerned about things happening 0-48" from the exhaust ports. Has anybody done a D shape to match some of the exhaust ports, and then transitioned to a circle?

Interesting post. Many years ago, I owned two Boss 302 Mustangs (the real deal - not today's version). One of 'em was pretty juked up. I played around with different header pipes. 1 5/8" pipes seemed to perform better than 1 3/4". And that was buzzing the thing deep into the +7k rpms. Of course, by today's standards, a 302ci mill is relatively small. Yet, when I built the 350ci mill in my Chevelle, I opted for 1 5/8" pipes feeding into a 3" exhaust. Never bothered to look back. With some things - not all - I think smaller may be better.

[MR P-BODY]

Interesting post. Many years ago, I owned two Boss 302 Mustangs (the real deal - not today's version). One of 'em was pretty juked up. I played around with different header pipes. 1 5/8" pipes seemed to perform better than 1 3/4". And that was buzzing the thing deep into the +7k rpms. Of course, by today's standards, a 302ci mill is relatively small. Yet, when I built the 350ci mill in my Chevelle, I opted for 1 5/8" pipes feeding into a 3" exhaust. Never bothered to look back. With some things - not all - I think smaller may be better.

You may have seen the added torque in the low end and got the car moving but if
you spend all the time in the upper RPMs they would hurt performance... thats the
whole thing... where do you spend most of the time.... headers are really designed
to work at a controlled RPM to have the best performance at that point.... we would
love to have headers that could move the primaries and diameter as the RPMs changed
along with collectors..... but thats not happening without MAJOR bucks and a lot of weight

[yellomalibu]

You may have seen the added torque in the low end and got the car moving but if
you spend all the time in the upper RPMs they would hurt performance... thats the
whole thing... where do you spend most of the time.... headers are really designed
to work at a controlled RPM to have the best performance at that point.... we would
love to have headers that could move the primaries and diameter as the RPMs changed
along with collectors..... but thats not happening without MAJOR bucks and a lot of weight

That's one of the things that John Lingenfelter stressed in his book "on modifying small block chevy engines". It's excellent advice for people with street cars - don't build for peak hp because you spend very little time at that rpm in a street car.
Some of the info in his book is dated, but the principals remain the same.

[Bob Holmes]

Man isn't that the truth. Folks keep building engines to make "hero" horsepower numbers, for cars that will never be driven in the power range of the engine. Then they complain about the doggy performance and blame the engine builder or some part for the problem.

It constantly happens with turbo newbies. Get me the big turbo, it'll make the car better...LOL.

[SuperBuickGuy]

Man isn't that the truth. Folks keep building engines to make "hero" horsepower numbers, for cars that will never be driven in the power range of the engine. Then they complain about the doggy performance and blame the engine builder or some part for the problem.

It constantly happens with turbo newbies. Get me the big turbo, it'll make the car better...LOL.

that's why this discussion interests me so much - no one asks why. The engine is a system, and you can buy all sorts of hero parts; but if you don't design the system from air intake to tailpipe, you'll never get what you expect. I still think, and evidence supports, that square pipe could work quite well for exhaust pipe. That the reason it's round is convenience.

Consider - how many actually believe the "vortex" air intake swirl device (aka money-from-fool-remover-device) works? yet, that same argument is used here - that a vortex creates better flow in the exhaust. Square is devilishly hard to make go in the same directions as round, but for the same available space (again with the system approach), my hypothesis is (now) that square is better from all points after the turbo flange.

While the original idea - square on the vette - isn't going to happen (I have side pipes now), the Spider is begging for square because I still need to fit 10# of stuff in that 5# italian bag.

[JeffClark]

Whenever I read a topic of exhaust on a Corvette I get light headed and run for the fridge looking for a beer to calm my nerves... I had a similar problem and thought of the side pipe thing but just couldnt do it, this was my Bracket car I built in the 90's now a Dana 60 and not the IRS I had a cool grand back then in mandrel bent 3 inch tube and a week in the air fitting note the scrapes in the driveway behind car.... On your car with the wheel combo and how it sits I think the Side pipes work, nice looking car.

[TC]

Pretty sweet car there Jeff!!!!!

[SuperBuickGuy]

Whenever I read a topic of exhaust on a Corvette I get light headed and run for the fridge looking for a beer to calm my nerves... I had a similar problem and thought of the side pipe thing but just couldnt do it, this was my Bracket car I built in the 90's now a Dana 60 and not the IRS I had a cool grand back then in mandrel bent 3 inch tube and a week in the air fitting note the scrapes in the driveway behind car.... On your car with the wheel combo and how it sits I think the Side pipes work, nice looking car.

Nice car, I should have the side pipes on it tonight (presuming the honey-do list isn't too long).

[JeffClark]

Thanks that was the last car I raced I sold it in 99 as a roller to a guy in Tucson, it was fast for the era no so much by todays cars, I have quite a few engines and parts left and been getting the urge to go racing again so a month ago I purchase a well used 72 SS Chevelle old bracket car ironically it is a silver car originally so I will be going through it this winter and back to racing again next year.