V10 exhaust configuration

[JordanMugen]

I disagree with several statements here regarding length vs bends and most just sound like canned magazine or ChatGPT responses. I base mine on personal experience working with and conversations with a header fabricator who has been involved in championship winning NASCAR, World of Outlaws, NHRA, and IMSA campaigns, as well as some NASCAR contacts and knowing what their headers spec like.

First, equal lengths only matter if every cylinder has the same induction length and has the same power potential. We know this isn’t always the case due to things like manifold assymetry. The other factor is mass flow isn’t always the same… there is charge robbing / intake interference happening in the intake plenum (yes even for NA F1 cars, it’s all documented by Honda) and its effect varies depending, there is flow assymetry in the airbox from the shape, and bends will determine a primaries ultimate ability to move mass.

To quote Dr Randolph of ECR and formerly Hendricks, best power is often when blow down time losses are balanced with pumping losses, from an exhaust / exhaust timing standpoint.

On road race V8s, with peak power under 9,000rpm, it’s usually better to have 2x 4-2-1 > feeding into a single tailpipe. Usually power upstairs stays the same but mid range and low end is better. The x-pipe is partly a similiar effect but existed because NASCAR had a rule on what the pipe exit should look like and their flat “boom tubes” require it to be so.

While your long post is helpful, crossplane V8s don't even have even firing intervals on each bank. Isn't it well known to not get hung up on making a perfect 4-1 equal length collector for one bank of a crossplane V8, since it doesn't matter as the pulses aren't equally spaced anyway?

Does what you are saying really also apply equally to flatplane V8 engines typical of Formula One and non-American small block V8 engines?

For example this flatplane Mugen V8 Formula 3000/Nippon engine has (presumably) an equal length 4-1 header despite the extra bends and lengths needed (obviously it's very long due to the low rpm, IIRC 10,000rpm is the max for F3000/F-Nippon?):


Or another low-revving flatplane Mugen V8 with a 4-2-1 header which seems to be equal length:


Even with crossplane small block V8s, is equal length really not important on 180-degree crossover headers when used on small block crossplane V8 engines, like those used by Perkins Engineering in V8 Supercars between 2007 and 2012, or on the Ford GT40 in the 1960's?

Unfortunately Jack Perkins from Perkins Engineering doesn't have one of those exact era of engines in the Perkins Engineering (now a restoration company) workshop to share (they got it to fit with the 2003-on Falcon-style double wishbone front subframe on their Commodores -- parity! also for parity the Holden Motorsport/Aurora heads had equal spaced ports like the Yates D3-- by going to a front steer setup), but presumably it was something like this:


Here's a sound clip of the Perkins Engineering 180-degree header for the (HMS) Chevy small block:


(The Jack Daniels' car: it's a little hard to isolate the exhaust sound from the intake as obviously the V8 Supercars ran eight trumpets with eight ITBs and the intakes are also quite loud.)

The above header pictured above is actually a Duke Engineering header for LS-swapped Porsche 911s, in the case people don't want their LS-swap 911 to sound too American I guess:


But as you can see even with the Jack Daniels' car even in competition applications, there are some advantages despite the extra complexity. They did run the system for five years until they went to Nissan. Surely (reasonably) equal length is important on a crossover header even on crossplane USA small-block V8 engine where equal length would (normally) not be that important?

In any race car, packaging can and often does have a significant effect

Thankfully there is plenty of space in the sidepods of the Formula 3000/Nippon car to fit incredibly long and large headers (or power/torque considerations are more important than packaging, or bargeboards were rudimentary at the time and big flat sidepods to manage front wheel wake was the thinking anyway?).

as long as the cross sectional area remains the same.

Are not stepped headers now popular to obtain a variation of reverberation effects and effectively change the tuned lengths?

[JordanMugen]

F1 always had 10 into 2? what are advantages of that in power and aero aspect?

Given the powerband involved it probably wouldn't be helpful, but presumably it would make it sound more like 36,000rpm instead of 18,000rpm which would be wild!

[Or rather it would make it sound like a 18,000rpm V10 instead of like two 18,000rpm inline-fives^? Although I don't know what the acoustic effects are when you are far away and the sources are (in effect) close together for a car that's far away?]

^ On the exhaust side. If I understand correctly there is a single airbox and it's not divided for left and right banks...?

[JordanMugen]

For example; we had a header built that was the same exact length, same collector, but focus on packaging and nice big bend radii. ... Even with the same length, it made 10-12% more power everywhere and much better in the transients. This exercise was all about managing flow losses which opens up your ability to use diameters the engine actually wants and makes sense for the use.

That's a good anecdote!

It seems to match up with ~950hp F1 V10s, which presumably moved a similar flow rate of air to ~950hp NASCAR small block V8s. Big bend radii etc:


But should one not be cautious trying to compare 5.7L 1000hp small block crossplane V8s to typical 3L 500hp Formula 3000 flatplane V8s? As seen on the previous page, the Formula 3000 header systems are very long and have lots of bends, no? So the thinking there seems more like as suggested by Tommy Cookers and leblanc?

length equality of 'headers' is much more important than curvatures

Especially with naturally aspirated engines, tuned, equal-length headers have the greatest advantage.

The Formula 3000 header is very different to a Formula 1 header after all!

Even older F1 headers, when the rpm and power output were lower, are more like the Formula 3000 header. [Unless that's just outdated 1990's thinking? ]

Early 90's V10:


Late 00's V8:


Also: timing belts on a racing engine?! What possessed Renault/Mecachrome to not use gears?

[saviour stivala]

Yes. The Renault RS1 3.5L 65 DEGREE BABK ANGLE V-10 used camshafts belt drive as also did the Renault 1.5L TURBO V-6, as also did the Alfa Romeo F1 TURBO V-8.