Engine sound design

All that has to do with the power train, gearbox, clutch, fuels and lubricants, etc. Generally the mechanical side of Formula One.
roon
roon
449
Joined: Sat Dec 17, 2016 6:04 pm

Engine sound design

Post

In this thread I'd like to discuss:

1. Regulating cylinder count to create specific sounds.
2. Exhaust sound-affecting devices.
3. Lightweight and minimally restrictive, yet effective, mufflers or sound attenuators.

1.) As I understand it, F1 engine regulations have never been drafted with sound production as a criteria. The regs more typically specify things like swept volume, aspiration types, material use, and fuel consumption. Through such engine regulations was found the ethereal wail of a high-revving V10.

Since such a discovery was made, by happenstance, perhaps the lesson can be applied to future formulas, even ones which prioritize efficiency. For example, a change of wording in the existing regulations to specify ten cylinders instead of six, a v-angle of 72° or 144° instead of 90°, would not alter nor preclude the discovery of further efficiency gains associated with the current formula. In fact, the arbitrariness of engine layout has already been illustrated by the specification of a 90° triple-throw-crank V6, instead of a 60° or 120° V6, I-4, I-3, or an evolution of the previous V8.

Increased frictional or pumping losses from higher cylinder counts would not de-incentivize furthering efficiency in an efficiency formula. Chassis installment structural issues are not a primary concern in the era of modern composite chassis building techniques, and concerns about weight optimization should be considered alongside the fungibility of the minimum vehicle weight rule.

A V10 adaptation of the current V6 technology is one direction to go in, towards changing the color of the sound of F1's engines.

2.) Another option, perhaps more what I'd like the thread to be about, is sound affecting devices.

Are there any devices or technologies which could modulate or transform exhaust pulses? Is there any efficient means by which to convert the three exhaust pulses per crankshaft rotation, provided by the current V6s, into four synthesized pulses (or more) per crankshaft rotation? In order to simulate the sound of a higher cylinder-count engine. Ideally, without obstructing the exhaust flow excessively. I envision it as something existing downstream of the turbine exit.

Perhaps a series of reed valves and expansion chambers, or spring loaded plates which would build up and release pressure to synthesize a new exhaust pattern, or a rotor with a window or windows on it spinning at a variable rate, reflecting and permitting intermittent pulses.

3.) There are benefits to quieter engines for satisfying different venues & zoning requirements. What room for development is there for a lightweight, small, minimally restrictive muffler? Or, might other technologies be considered or created for serving the same purpose. Perhaps what solves prompt #2 could also apply here. A quiet engine can still provide an evocative sound. In fact, fans listened to a quiet, synthetic version of the old V10s every time they turned their televisions on to watch a race.

User avatar
PlatinumZealot
444
Joined: Thu Jun 12, 2008 2:45 am

Re: Engine sound design

Post

A muffler to make things even quieter? I am not so sure the fans will like that one.

The turbocharger by virtue of being a spinning wheel with "paddles" designed to remove as much energy as possible from the exhaust without restricting the engine too much, will deestroy that shrieking timbre and volumeof the exhaust sound that the fans love. Then you add to this a lower rpm limit and fewer cylinders that reduces the sound pitch.

So there three messed up things: the timbre the volume and the pitch.

roon
roon
449
Joined: Sat Dec 17, 2016 6:04 pm

Re: Engine sound design

Post

I agree with you. I mainly bring up quieting as a way for F1, or any series really, to gain access to venues in a way that seems unique to Formula E. If the engines sound good while being also being quiet, I think specific complaints about decibel levels may be limited.

Greg Locock
Greg Locock
180
Joined: Fri Jun 29, 2012 11:48 pm

Re: Engine sound design

Post

" Is there any efficient means by which to convert the three exhaust pulses per crankshaft rotation, provided by the current V6s, into four synthesized pulses (or more) per crankshaft rotation? "

Since you don't know what RPM the engine is doing you get exactly the same frequencies from a V6 at 20000 rpm as from a V8 at 15000 rpm. So it is a bit pointless. (OK I know a V8 sounds different to a V6 but you were talking about the firing order pulses)

However yes it would be possible, various forms of noise cancellation for exhaust have been tried, and then adding a loudspeaker to emit your chosen sound is obviously possible.

In an extreme case a coworker of mine took a small 4 cylinder Citreon and installed a system to make it sound like a Corvette. It was very funny watching powertrain engineers used to the LT5 dropping the clutch at 800 rpm.

roon
roon
449
Joined: Sat Dec 17, 2016 6:04 pm

Re: Engine sound design

Post

Greg Locock wrote:" Is there any efficient means by which to convert the three exhaust pulses per crankshaft rotation, provided by the current V6s, into four synthesized pulses (or more) per crankshaft rotation? "

Since you don't know what RPM the engine is doing you get exactly the same frequencies from a V6 at 20000 rpm as from a V8 at 15000 rpm. So it is a bit pointless. (OK I know a V8 sounds different to a V6 but you were talking about the firing order pulses)

However yes it would be possible, various forms of noise cancellation for exhaust have been tried, and then adding a loudspeaker to emit your chosen sound is obviously possible.

In an extreme case a coworker of mine took a small 4 cylinder Citreon and installed a system to make it sound like a Corvette. It was very funny watching powertrain engineers used to the LT5 dropping the clutch at 800 rpm.
Thanks, Greg. I'm curious though: Why do you say it is pointless? Why wouldn't we know "what RPM the engine is doing?"

A loudspeaker seems heavy, power hungry, but that's just my assumption. Is there something more along the lines of using the inherent sound energy in the exhaust to alter its own frequency? My understanding of exhaust systems is limited. It seems like they are typically reducing amplitude but not altering frequency.

Greg Locock
Greg Locock
180
Joined: Fri Jun 29, 2012 11:48 pm

Re: Engine sound design

Post

You can't 'change' the frequency of a given signal, passively (I suppose you might argue about Doppler)

You, the listener in the crowd, don't know what rpm the car is doing.

torpor
torpor
-8
Joined: Thu Jan 15, 2015 7:01 pm

Re: Engine sound design

Post

Designing an engine is all about efficiency. The actual Mercedes Formula 1 power unit is one of the most efficient engines ever built.
That means it produces the highest possible power out of least possible amount of fuel. It's a simple math: Sound is a sort of energy like heat or power. Increasing the sound reduces the power. Nobody can really want that...

User avatar
henry
268
Joined: Mon Feb 23, 2004 7:49 pm
Location: England

Re: Engine sound design

Post

I believe the quality of the sound would be improved if they reduced the number of gears.

With 8 speed boxes they are using a rev range of only 1500 rpm and changing gear every couple of seconds. If they reduced it to 5 gears we would get 2400 rpm range and changes every 3 or more seconds allowing some crescendo effect.

These engines don't need lots of gears in the way the peaky V8s an V10s did.

However there would be lots of comments along the lines of " Pinnacle of technology? My road car's got more gears than that!"
Fortune favours the prepared; she has no favourites and takes no sides.
Truth is confirmed by inspection and delay; falsehood by haste and uncertainty : Tacitus

User avatar
mclaren111
199
Joined: Sun Apr 06, 2014 9:49 am
Location: Shithole - South Africa

Re: Engine sound design

Post

henry wrote:I believe the quality of the sound would be improved if they reduced the number of gears.

With 8 speed boxes they are using a rev range of only 1500 rpm and changing gear every couple of seconds. If they reduced it to 5 gears we would get 2400 rpm range and changes every 3 or more seconds allowing some crescendo effect.

These engines don't need lots of gears in the way the peaky V8s an V10s did.

However there would be lots of comments along the lines of " Pinnacle of technology? My road car's got more gears than that!"
Totally agree !!! =D> =D>

I would suggest 6 gears

roon
roon
449
Joined: Sat Dec 17, 2016 6:04 pm

Re: Engine sound design

Post

Greg Locock wrote:You can't 'change' the frequency of a given signal, passively (I suppose you might argue about Doppler)
Semantical question: Consider an air compressor. The exhaust pulses of a pneumatic pump are stored in a chamber. Or in the context of an engine, exhaust pulses could be stored in a chamber until backpressure prohibits combustion. A valve with a resonating component evacuates the chamber producing a different sound frequency. Would this be considered passive alteration/conversion of a signal, or destruction/loss of a signal and creation of a new one? Or some other term?

bill shoe
bill shoe
252
Joined: Wed Nov 19, 2008 7:18 am
Location: Dallas, Texas, USA

Re: Engine sound design

Post

roon, the current powertrain rules were written to create (maintain) a very specific type of powertrain noise while allowing some new innovation in terms of how power was created (hybrid, MGU, MGH, and all of that). So noise design was already the main concern from day one.

The noise design was not focused on what you are (likely) thinking of, rather it was focused on maintaining the ramp-up/step-down pitch that is intrinsic to going thru multiple gears and gear-changes.

The natural form of a car with high power-to-weight is to use a large engine and a single gear. However, the current rules require 8-speed gearboxes and a fuel-curve that ramps up with rpm, these measures are to force shifting for the sake of maintaining traditional shifting noise.

To answer the question you thought you were asking-- If you wanted to further artificially alter the sound from what exists now... and you wanted it to sound more like the v-8's, then you need evenly-spaced firing order and higher-pitch.

Unevenly-spaced firing order: Current engines have uneven firing order, which is almost required by the mandated 6-cyls and 90-deg bank angle. If you required a 120 deg bank angle or maybe a 60-deg bank angle then you would naturally get even firing pulses. This is difficult to change by itself, you would either have to mandate even firing freq with current 90-deg bank, which requires offset crank journals (not easy in highly-loaded high-performance engine) or you would have to say "right, starting in year X we're switching the bank angle".

Increasing the peak pitch: This would require a separate exhaust pipe for each exhaust valve, so two from each cylinder. The first pipe would be short, the second would be long, and the different between them would be sufficient to create 12 exhaust pulses (per engine-cycle) instead of the current 6.

Last NA engines were 18000 rpm (by rule) and 8-cyl, so crank freq of 300 hz and firing freq of 1200 hz.
Current turbos are 11000 rpm (by practical experience) and 6-cyl, so crank freq of 183 hz and firing freq of 550 hz.
If you split the exhaust valves into separate pipes (6 short and 6 long) like I mentioned then you'd get 1100 hz which is almost identical to the naturally-aspirated V-8's.

Achieving the 1100 hz pitch with long and short exhaust pipes would require delaying the long-pipe pulse by ~ 0.00091 seconds. At standard speed of sound (340 m/s) for standard temp & pressure this works out to a length different of 310 mm between the short vs. long pipe. Obviously exhaust gas is not standard temp & pressure so length would vary from that, but still in the ballpark.

Put them together?: Of course, you could be clever and require an offset pipe lengths that also dealt with the uneven firing frequency without needing a new bank angle. Over relatively narrow ranges of rpm and exhaust temp/pressure, this could result in both the 1100 hz high-pitch noise and evenly-spaced pulses. You would presumably want to check everyone's exhaust temp/pressure at peak power and 1100 rpm, then set the length of the individual exhaust pipes to achieve the goals. But again, this only works for narrow range of rpm.

But remember, the ramp-up/step-down engine sound was considered in advance and is already forced in by rule. So apparently that's where everyone's heart is for sound design because that's what got done.

Tommy Cookers
Tommy Cookers
535
Joined: Fri Feb 17, 2012 3:55 pm

Re: Engine sound design

Post

12 (evenly-spaced) exhaust pulses per cycle could be seen as needing 4 turbos

isn't the existing firing pattern (at main rpm) heard as even because there's already a difference in runner lengths l to r ? (as your suggestion)
this difference incorporated to try to feed an even supply of exhaust pressure pulses to the turbo (as the single turbo Renault F1 c.1978)
though I can't remember whether the current turbo sees 6 evenly-spaced pulses per cycle or 3 evenly-spaced paired pulses per cycle

the suggestion of each cylinder having 2 exhaust lengths is ingenious but a bit daunting
we might at best have in the main rpm band 6 evenly spaced pulses ?
tuned-length exhaust system 'extractor' effects are presumably strong in these engines (as in earlier turbo F1) and too valuable to lose ?
one recent post suggested a further breakthrough in F1 exhaust systems

gruntguru
gruntguru
455
Joined: Sat Feb 21, 2009 6:43 am

Re: Engine sound design

Post

Pretty sure the current turbines have divided entry so 3 evenly spaced cylinders don't see blowdown pulses from the other three.
je suis charlie

Greg Locock
Greg Locock
180
Joined: Fri Jun 29, 2012 11:48 pm

Re: Engine sound design

Post

No, you can't just use a delay line to increase the number of pulses. If you are a mathematical type, sin (2*pi*f*t)+a.sin(2*pi*f*(t+dt)) still results in the same frequency, just a different amplitude to sin (2*pi*f*t).

toraabe
toraabe
16
Joined: Thu Oct 09, 2014 9:42 am

Re: Engine sound design

Post

henry wrote:I believe the quality of the sound would be improved if they reduced the number of gears.

With 8 speed boxes they are using a rev range of only 1500 rpm and changing gear every couple of seconds. If they reduced it to 5 gears we would get 2400 rpm range and changes every 3 or more seconds allowing some crescendo effect.

These engines don't need lots of gears in the way the peaky V8s an V10s did.

However there would be lots of comments along the lines of " Pinnacle of technology? My road car's got more gears than that!"
Williams F1 team only wanted a 5 og 6 speed gearbox due to the wide powerband of the Merc engine.