Light flywheel: Worth it?

Light flywheel: Worth it?

Author
Discussion

Evoluzione

Original Poster:

10,345 posts

249 months

Tuesday 20th November 2012
quotequote all
I'm coming to the conclusion that light flywheels are just a waste of money unless they are attached to an all out competition or track car. Here's my thoughts:

Does the engine rev any quicker? Does it make any more power? Is the car any quicker?

Ok, so you say it must because it has less work to do, but hang on, the engine is trying to pull (lets say on average) a ton of metal down the road, so is a (say) 5kg saving going to do much?

The lack of engine inertia cause three problems:

Difficult to set off from standstill, especially with a stronger clutch fitted

Idles rougher

The revs drop quicker in between each gearchange, does this last point alone work against anything gained?

My main question is this: Has anyone ever measured the effects?

stevieturbo

17,489 posts

253 months

Tuesday 20th November 2012
quotequote all
Any benefits will also depend on gearing. The higher the gearing change...(ie 1st gear will see the biggest difference, top gear the least ) the more it can be felt.

But IMO on an engine that makes decent power and in a heavyish vehicle, the gains would almost be impossible to measure.
I certainly wouldn't go out of my way to source a light flywheel. If the clutch setup came that way, then so be it.

PeterBurgess

775 posts

152 months

Wednesday 21st November 2012
quotequote all
You can measure the difference even in fourth gear on an inertia dyno. I attach a link to Steve's dyno page, it makes interesting reading and will upset a few folk who argue whose dyno is accurate etc etc, as I have always said, to me repeatable is perfection and if you improve it shows more and if you 'dun it wrong' it shows less, whether flow bench or dyno! We have found the inertia dyno very repeatable and, unlike Steve's recommendation od do it all on the same day (or conditions), the Dynocom logs temp, pressure and humidity and then converts to what ever correction type we want, we use SAE J1349 which is usually closest to no correction (what it is on the day). Our Dynocom also has a pau in the form of an eddy brake should we need to hold steady state conditions or run against a load as in turbos that need a little load to spool up, the best of both worlds?

http://wotid.com/dyno/content/view/19/38/

Peter

Edited by PeterBurgess on Wednesday 21st November 06:48

350Matt

3,761 posts

285 months

Thursday 22nd November 2012
quotequote all
the problem here is the power you measure on an engine or chassis dyno is usually holding the engine at a steady speed so you won't see any benefit to a lighter flywheel

however if you coudl measure 'transient ' power ie the power the engien makes during acceleration then you would see the difference

as while a 5Kg difference in flywheel weight is very little on a car that weighs 1200kg that 1200kg isn't trying to spin up at 5000rpm/second so lightweight flywheels are worth doing and usually help with gearshift speed

however yes idle quality will be worse

stevesingo

4,864 posts

228 months

Thursday 22nd November 2012
quotequote all
If you did a before and after dyno test at fixed intervals (steady state) and plotted the curves, there would be no difference.

If you did a transient dyno test at 200rpm.min acceleration, the lighter flywheel would show a gain. Repeat at 500rpm/sec the gain would be larger.

If you tried a transient test on a rolling road/hub dyno, referencing wheel speed and did a back to back at say 10mph/sec in 4th gear. The light fly wheel will show a gain. Repeat in 2nd gear, the gain will be larger.

Energy saved by not accelerating flywheel mass is passed on to the wheels/hubs. The faster you accelerate the engine the larger the saving from the reduced inertia of the flywheel, therefore the greater the wheel/hub output.

Hope that makes sense.

Steve

PeterBurgess

775 posts

152 months

Thursday 22nd November 2012
quotequote all
I thought most up-to-date rolling roads were Inertia (and maybe Eddy brake pau as an addition)? Therefore not a problem to measure transient power. I do have a feeling that using a pau does not reflect actual power delivery as the pau has to hold the load, let go, hold the load let go etc etc. We can notice very quickly the pau does not allow repeatable testing as the unit gets hot and this alters the calibration. We are advised, if we use the pau, let it cool as much as we use it, ie use it one minute then cool it one minute. Mostly we just use inertia testing which is gentle and very rapid not to mention spookily repeatable!

Peter

Pumaracing

2,089 posts

213 months

Thursday 22nd November 2012
quotequote all
I can calculate the exact effect of altering flywheel weight on vehicle acceleration because I programmed moment of inertia calculations into my vehicle performance simulator program. The effect depends massively on gearing, how high the engine revs, whether engine power exceeds tyre grip anyway in the low gears and other factors but using my friend's Citreon AX rally car revving to 8500 as an example gives the following results.

1 kg removed from the rim of the flywheel at a radius of 5 inches drops 0-60 mph time by 0.17 seconds and 0-100 mph by 0.25 seconds. To get the same reduction in 0-100 mph time by removing weight from the chassis takes 15 kg.

Weight removed from closer to the centre of the flywheel obviously has a much lesser effect.

Going from a stock iron flywheel weighing maybe 8 or 10 kg to a billet steel or aluminium one weighing half that can therefore be a pretty significant change.

Running the same calculations for 1 kg at 5" radius on my 2.0 Ford Focus with about the same 140 bhp but revving much lower and in a heavier chassis reduces 0-60 mph time by 0.13s and 0-100 mph by 0.18s. Equivalent chassis mass is about 10 kg.

Trying to equate flywheel inertia reduction to some sort of bhp figure using an inertia dyno is a fool's errand for reasons too lengthy to go into.

Arkle

41 posts

211 months

Thursday 22nd November 2012
quotequote all
Theory is great, but when I changed the dual mass flywheel on my 964 C2 for the RS version the difference was phenomenal. So much more fun to drive,especially the lower gears, the heel and toe response much better which made trackdays much more satisfying, no change in the tickover, slightly tricky in traffic but nothing you don't get used to. Lazy driving with the much quicker engine response is easy, any decent driver will adapt just like you do when you jump from one car to another.
I only changed because the dual mass had failed, but I'm so glad I did.
I'm not sure I'd source an unknown part just for the fun on another model, after all Porsche have had a bit of experiance with drivers cars.

PeterBurgess

775 posts

152 months

Thursday 22nd November 2012
quotequote all
Dave

As you say you are simulating not measuring.

Why fools errand, if you owned and operated a dyno day in day out and had run water brake, inertia and Eddy brake as I have done everyday for 25 years you would maybe write a little differently.

Are you saying all the stuff on the link I posted above is incorrect?

We have already seen the difference in acceleration with lighter flywheels on an otherwise same car. It is interesting to note that Silkolene came to the same conclusions as we did, but you will probably say they are on a fools errand as well.

Maybe if you get to buy and run an inertia dyno you might share with the forum as I am happy to do? I am amazed how much I have learned running the inertia/eddy dyno for 2 and a 1/2 years, just goes to show when we think we know it all we actually know very little smile

Peter


Pumaracing

2,089 posts

213 months

Thursday 22nd November 2012
quotequote all
PeterBurgess said:
Dave

As you say you are simulating not measuring.
Exactly. I'm calculating with mathematical precision something that CAN be calculated with precision because it relies solely on basic Newtonian mechanics and there are no unknowns. F=ma and inertia can be equated directly with mass after allowing for gearing. I'm not estimating, guesstimating or otherwise fudging a change in one thing by trying to present it as a change in something else.

What, on the other hand, you think you are "measuring" (as if that were somehow better than calculating something properly) when you change the mass of something but the machine gives you a different bhp number is impossible to say. It certainly makes no scientific sense. You are in effect taking the equation F=ma, changing the value of m but trying to claim that what actually changed was the F. Meaningless.

A steady state dyno at least makes scientific sense if it's properly calibrated. An inertia one uses parameters that can be changed willy nilly and fudge factors to try and compensate for those. Whether it's repeatable is irrelevant. Whether the vehicle actually accelerates faster with less mass or inertia (obviously it does) is also irrelevant. The bhp number guesstimate is simply not any sort of scientific number that properly relates to the actual power of the engine if that number can change when you alter something completely unrelated to the power production part of the engine.

That's as far as I'm going to let myself get sucked into this.


PeterBurgess

775 posts

152 months

Thursday 22nd November 2012
quotequote all
I said the inertia dyno will show a difference when a lighter flywheel is fitted. It is as simple as that. if you owned a rolling road inertia dyno you would discover the same.

You may say all you like about accurate measurments or whatever, why not take up the argument with all the dyno manufactureres, do you have the arrogance to say they are all incorrect?

I asked you if all the info on the link I posted above was incorrect, do you think that is the case?

I use my dyno to win championships on real cars not simulated on a pc. The results speak for themselves.

I am not trying to suck you in to an argument but you make sweeeping aggressive
statements and then try and say a simulated pc is better than real world measurement and tuning. A simulator is ok for armchair tuning in my humble opinion but not for real world tuning. If you owned and opereated a dyno and then told me how much more real the simulator is I would have respect for what you say but to be honest it comes across as sour grapes cos you havent got a dyno.

Peter



Pumaracing

2,089 posts

213 months

Thursday 22nd November 2012
quotequote all
Just for people's edification I'm going to put some numbers on how big an effect engine component inertia is and therefore how much it affects vehicle performance. To go back a bit, what my vehicle simulation program does is take the full steady state power curve of the engine, transmission losses, the vehicle weight, gear ratios, tyre size, drag coefficient, frontal area, tyre grip coefficient, gear change time and a load of other things and then it calculates every propulsive factor, every retarding one and hence the acceleration. Obviously some factors can be known with precision like the weight and gearing and others have to be estimated to within reasonable limits so clearly no one is going to claim that you can simulate reality with total precision when estimates have to be made.

However from any given baseline position you can then change any factor that IS known with precision and see how the vehicle is affected.

To the basic chassis mass two things must be added.

1) A factor to allow for the fact that wheels and tyres absorb rotational as well as linear kinetic energy.

2) A factor to allow for the inertia of engine components that rotate or reciprocate. This element is affected by gearing so it produces a different "chassis mass equivalent" of extra weight in every gear.

Taking my Ford Focus as an example it weighs about 2900 lbs with just the driver in it. If the engine components were all made of unobtanium and weighed zero then the whole of the engine's power would be available to accelerate the car. Sadly the engine components are quite heavy and require a lot of accelerating. Using the car's actual gearing and tyre size and engine inertia factors obtained from actual weights, diameters, calculation and some sensible estimates the additional amount of mass that must be added to the chassis mass in each gear calculate as follows.

1st - 570 lbs
2nd - 193 lbs
3rd - 89 lbs
4th - 45 lbs
5th - 25 lbs

So in 1st gear the engine rotational inertia mass equivalent is nearly 20% of the car's own mass. Obviously it's massively higher than the actual weight of the engine components. That's a huge factor to not allow for in a simulation. You could equally say that of the engine's approx 140 bhp, one sixth of that or 23 bhp is being used up in accelerating the engine itself and only the remaining 117 bhp is available to accelerate the car.

In the higher gears the effect diminishes rapidly and by 5th it's negligible. That's also of course why trying to "equate" a flywheel inertia reduction with some single number of extra engine bhp is a fool's errand because the effect is different in each gear and as close to zero as makes no odds in 5th.

Now of course my Focus is quite heavy and you could put the same engine in a much lighter chassis. The inertia factor stays the same but if the whole vehicle only weighed 1500 lbs the inertia factor in 1st would be more than a third of the car weight. Now you're starting to see just how much lightweight engine components, mainly the flywheel, can help the vehicle performance. In very high revving engines like bike ones or in F1 the inertia factor will be far higher still and every gram saved becomes crucial.

Edit:

Just one last addition to reinforce the points above. Let's say I replaced the 140 bhp 2 litre engine in my Focus that revs to 6000 rpm with a 140 bhp 1 litre bike engine revving to 12000 rpm. Same power, so theoretically at first glance the same performance, but half the torque at twice the revs. To get the same speeds in each gear but at 12k instead of 6k rpm would require the final drive ratio to double numerically, say from 4.0 to 8.0.

Because the inertia effect is proportional to the square of gearing, if this new engine had a flywheel and other parts with the same inertia as the Ford engine the vehicle mass factor in each gear would quadruple. In first it would be equivalent to 2280 lbs or almost as much as the car. The acceleration would be crippled. That's why bike engines, or ideally any other high revving engines, have multiplate clutches with as small a mass and diameter as possible. That's also why big capacity low revving V8s are not very sensitive to flywheel inertia at all.

Edited by Pumaracing on Friday 23 November 02:01

PeterBurgess

775 posts

152 months

Friday 23rd November 2012
quotequote all
Lots of good maths there Dave, you always were most excellent at maths.

You do not, however, address any of the 'real world' points I have raised in my posts. That is why I was mentioning 'armchair tuning' and 'real world tuning'. I hasten to add I am not knocking armchair tuning as it is vital to development but it is not and cannot be the end game. Thesis (armchair tuning), antithesis(doesnt work as you expect, real world tuning), synthesis (lets move forwards using all the tools we have), that is my take on Philosophy with regard to the logic of scientific method.

Peter

andygtt

8,345 posts

270 months

Friday 23rd November 2012
quotequote all
I lightened the flywheel/clutch in my noble massively... Basically I changed to a triple plate 7.25' clutch and halved the flywheel so almost all the reduction in weight was on the outer side of flywheel for maximum result. Lost nearly 10kgs!

The change is enormous, the box is noisy due to lash because of the now uneven idle, the revs drop like a stone on gear change and literally to idle if your slow changing on the road.

However the way the engine builds revs is also transformed I love it... It's also really easy to balance the car on the throttle on track which is a real bonus in a mid engine car.


Pumaracing

2,089 posts

213 months

Saturday 24th November 2012
quotequote all
andygtt said:
I lightened the flywheel/clutch in my noble massively... Basically I changed to a triple plate 7.25' clutch and halved the flywheel so almost all the reduction in weight was on the outer side of flywheel for maximum result. Lost nearly 10kgs!
I'd be interested to model that in my simulation program as an extreme example of flywheel lightening if you have sufficient details. Is there somewhere on the web with detailed car specs - gearing, tyre size, weight, drag etc?

tr7v8

7,287 posts

234 months

Saturday 24th November 2012
quotequote all
I lightened the flywheel on my TR7V8 & lost around 3kg (the SD1 flywheel is massive!) It vastly improved the pick up & MAY have improved the acceleration. I say may as there were many other changes at the same time.
But a bit of fag packet maths, show that the 3kg reflected through the gearing meant it was the equivalent of losing around 30kG off the cars weight or around 2.5% of the kerbweight.

stevieturbo

17,489 posts

253 months

Saturday 24th November 2012
quotequote all
tr7v8 said:
I lightened the flywheel on my TR7V8 & lost around 3kg (the SD1 flywheel is massive!) It vastly improved the pick up & MAY have improved the acceleration. I say may as there were many other changes at the same time.
But a bit of fag packet maths, show that the 3kg reflected through the gearing meant it was the equivalent of losing around 30kG off the cars weight or around 2.5% of the kerbweight.
Perhaps in 1st gear, not in top gear.

tr7v8

7,287 posts

234 months

Saturday 24th November 2012
quotequote all
stevieturbo said:
tr7v8 said:
I lightened the flywheel on my TR7V8 & lost around 3kg (the SD1 flywheel is massive!) It vastly improved the pick up & MAY have improved the acceleration. I say may as there were many other changes at the same time.
But a bit of fag packet maths, show that the 3kg reflected through the gearing meant it was the equivalent of losing around 30kG off the cars weight or around 2.5% of the kerbweight.
Perhaps in 1st gear, not in top gear.
Correct, only 9kg in 5th. But acceleration normal infers lower gears.

stevieturbo

17,489 posts

253 months

Saturday 24th November 2012
quotequote all
tr7v8 said:
Correct, only 9kg in 5th. But acceleration normal infers lower gears.
Not really. In most motorsport, how often are you ever in 1st gear ? Not that often really. And most cars with any decent level of performance will be traction limited in 1st gear anyway ( unless 4wd )

So it's 2nd gear upwards that matters, where the possible benefits of a light flywheel are much diminished, almost negligable.

Every little bit can help, but is it worth it ? is a different matter.

For a light low powered vehicle then yes. For something heavier or making a respectable amount of power, any differences will be small.

I know I wouldnt go out of my way to source a light flywheel, but if a clutch kit happened to come that way, so be it.

PeterBurgess

775 posts

152 months

Saturday 24th November 2012
quotequote all

'Every little bit can help, but is it worth it ? is a different matter.'

Spot on, for sprints and hiilclimbs, if allowed, it would be daft not to do it, same as for circuit racing.

A different reason we came across for negligible flywheel was landrover trialling. The owner had the smallest/lightest flywheel he could fit. Why we said, isnt steady idle better? No says he, when I back off the throttle I want it to stop dead so I am in control. This guy brings his Landy down on the back of a lorry some 5 feet off the ground, very short ramps and just long enough to take the Landrover. This guy is amazing, we all come out to watch as do our neighbouring unit occupants. The owner lines up at the bottom of his ramps, floors the throttle then almost instantly backs off...car is parked on the back of the lorry and absolutely effortless, claps from all of us smile

Peter