NC,
It is clear that there are benefits to be had from both heavier and light flywheels so a compromise is usually the result however I dont want to get into a debate at this stage, rather I am interested in links to sites/info that discuss the relative merits of varying weights of flywheel, the factors a designer takes into account when specifying a flywheel etc etc. (Not talking about DMFs by the way).
Of course I am also interested in your views on the subject and would be pleased if you would articulate them.
Regards.
|
Sorry, I asked where you got it from, because it's wrong, and not helpful.
I don't know of any sites that describe flywheel design, but, at least for solid flywheels, it's not difficult.
|
|
|
|
My understanding was that heavier flywheels give a smoother, steadier idle, and even allow the idle speed to be reduced for emissions / economy purposes, but the compromise comes in terms of slower acceleration and deceleration of the crank.
Cheers
DP
|
DP - yes!, your understanding is right.
|
though is restrictive at higher rpm
That's the bit that I think was wrong. getting to higher rpm, then yes - no problem once it gets there.
|
|
|
|
My understanding was that heavier flywheels give a smoother steadier idle and even allow the idle speed to be reduced for emissions / economy purposes but the compromise comes in terms of slower acceleration and deceleration of the crank.
I also agree with DP, the point about slower deceleration perhaps being the key, i.e. if load is harnessed to a spinning flywheel then a heavier flywheel will apply more effort to the load before its kinetic energy is disipated. In the context of a vehicle - take a car running along a road at, say, 1500rpm, when the vehicle comes to an incline the kinetic energy stored by a heavier flywheel should help it to more readily climb the incline than if it has a lighter flywheel?
Here is some interesting calc:
www.physicsforums.com/showthread.php?t=180979
It seems that the benefits of the heavier flywheel are outweighed by the energy needed to accelerate it to high speed and the fact that its albeit greater kinetic at higher speeds is relatively insignificant against the total amount of energy that the engine is developing.
This discusses the benefits of a heavier flywheel:
www.torquecars.com/tuning/flywheel-lightening.php
And this a lighter one:
www.uucmotorwerks.com/flywheel/how_a_lightweight_f...m
|
Of the links you posted, only the third one is worth spending any time on. Fundamentally, uucmotorwerks get the performance aspects of lightened flywheels right (although they use perverse units, and present the results in an unhelpful way IMO).
Where I disagree with their page is the stuff about 2 schools of thought. Using a lightened flywheel doesn't change the engine's power output at all. What a lightened flywheel does do is to reduce the effective mass that the engine pushes along.
For most road cars, it's not worth bothering to lighten the flywheel, because there's only any real effect on performance in the lower gears - where you don't spend much time anyway.
The way I prefer to approach the subject is to convert everything as if it were an equivalent flywheel at the crank.
By taking the car's total mass and multiplying by the rolling radius squared, the car's mass is converted to an equivalent flywheel at the roadwheel, and so, the inertia of the road wheels can simply be added at this stage to give a total inertia that rotates at roadwheel speed.
Then, dividing the total inertia at the road wheel by the total gear ratio SQUARED, you can refer the roadwheel inertia to the flywheel, and then, you can add the flywheel inertia. This gives an equivalent total inertial load on the crank.
As you can see, dividing by the total gear ratio squared means that in first gear, the engine doesn't see a large added inertia from the car/roadwheels - the inertia of the engine/flywheelitself dominates.
In the higher gears, the inertia at the roadwheels is not reduced so much, and begins to dominate.
The vital bit is the gear ratio SQUARED - which makes lightened flywheels only important for performance in the lower gears.
|
OK NC, thanks, all makes sense, though:
>>Using a lightened flywheel doesn't change the engine's power output at all. What a lightened flywheel does do is to reduce the effective mass that the engine pushes along.>>
Makes sense if the engine power is measured at the crank, i.e. before the flywheel, though if it is measured at the tranmission or wheels then the fact that a lighter flywheel requires less energy to get up to speed would mean that more power would be available to be measured surely?
Also what about the point about kinetic energy? I thought the calcs in the first link were interesting, are there errors in there then?
Regards.
|
We're heading back towards another thread - it doesn't make much sense to measure engine power in anything other than (approaching!) steady state. From the steady state values, everything else can be found or calculated, from a transient measurement, you don't really know what you've measured. Mixing flywheel inertia into transient power estimations can only confuse other estimates like power to weight ratio.
Yes, the calcs were OK, but the question and context they were in didn't make any real sense.
So, that the mass moment of inertia of a simple disc flywheel is 1/2 * Mass * radius ^2 is OK.
That the kinetic energy stored in a rotating system is 1/2 * I * omega^2
Where I is the mass moment of inertia (kgm^2)
omega is the angular velocity in radians per second
is also OK.
|
I've got a car with a much lighter flywheel than standard. Cost me a bob or two but worth it. As said above it accelerates better in the low gears (where you spend time that matters in speed hillclimbs).
Downside. requires more care with the clutch on the road but the idle speed has not changed (already was faster than standard owing to the cam and carbs).
I like it!
|
|
|
Thanks NC, the thing is that in the after market world people can only measure the power and torque of modified engines on dynos/rolling roads hence the point about the benefits of kinetic energy at low rpm is quite relevant because a lighter flywheel can show less torque at low rpm, likewise it can show more power at high rpm because (as I said before) a lighter flywheel requires less energy to get up to speed so more power is available to be measured at the wheels.
Accordingly it seems to me that there is a point in the rev range where the benefits of the heavier flywheel at low rpm and the benefits of the lighter flywheel at high rpm cross, i.e. the lighter f/w is better above X rpm and the heavier f/w better below X rpm.
Any ideas on a calc formula for this?
Thanks!
|
It's OK to measure at the wheel, as long as it's done in steady state, against a brake rather than against an inertia.
What is shown on mickey mouse rolling roads which drive against an inertia are meaningless numbers for pub bragging only. It costs more to install a proper braked rolling road, because you need to be able to dump the energy - typically, you would see either a bank of glowing resistors, or a large water tank that heats up during the course of a day's testing.
>>there is a point in the rev range where the benefits of the heavier flywheel at low rpm and the benefits of the lighter flywheel at high rpm cross,
No. You're allowing yourself to be confused by a fundamentally duff measurement method. There is no performance advantage to a heavy flywheel at any engine speed. There are idle speed, emissions, and NVH benefits, but, not performance.
|
But NC the spinning flywheel stores kinetic energy thus (as per earlier posts) a heavier flywheel is more resistant to acceleration/deceleration (also as per the calc on the link I posted) thus if a vehicle fitted with a heavier flywheel is running at a steady speed and comes across an incline then the fact that the flywheel is more resistant to decleration the vehicle will more readily climb the incline?
Rather like a toy car that you run acroos the floor to get it up to speed then put it down and let it go, it relies on a small (though heavy relative to the size of the car) flywheel, if this flywheel were lighter then the car would take less effort to get up to speed though would not travel so far.
Any thoughts?
|
Any thoughts?
Yes; the insignificant amount of energy that can be stored in the average car's flywheel is enough to move it about two feet.
It's more of a vibration damper - and something to hang the ring-gear on - than anything else. To climb a hill would reguire a half-ton flywheel spinning at 30,000 rpm.
|
There is no performance advantage to a heavy flywheel at any engine speed.
Just to annoy you NC, the 2CV and its variants had very heavy flywheels for the power of the engine, presumably to compensate for the lightness of the short crankshaft and only two small pistons. What it meant in practice though was that changing up a gear under hard acceleration, by rushing the gearchange and letting the clutch in brutally a driver could exploit the kinetic energy stored in the flywheel to provide a surge of acceleration, often accompanied by a squeal of pain from the tortured clutch. The extra iron drum of the centrifugal clutch fitted in series on a lot of 2CVs gave even more inertia to the crankshaft assembly.
Just to annoy you you understand. Please don't tell me that the energy had to be stored in the crankshaft in the first place and that must have taken ages, that what goes up must come down, etc. I know that already from experience.
:o}
|
It's surprising when you do the calcs just how much energy is in a flywheel.
I've been a little bold and given the flywheel 20mm thickness - but, I haven't included the clutch and the pressure plate.
t=20e-3; % flywheel thickness (m) r=200e-3; % flywheel radius (m) rho=7850; % flywheel density (kg/m^3) % calculates flywheel mass (kg) M_f=pi*r*r*t*rho
M_f =
19.7292
%Calculate mass moment of inertia (kgm^2) I_f=0.5*M_f*r*r
I_f =
0.3946
% gives a range of engine speeds (rpm) n_engine=[1000 2000 3000 4000 5000 6000]'; % converts them to radians per second n_engine_rad=n_engine*2*pi/60; % calculates the KE in the flywheel (Joules) KE_flywheel=0.5*I_f.*(n_engine_rad.^2);
M_c=1000; % Mass of car (kg) % gives a ranges of speeds for the car (mph) mph_c=[20 40 60 80 100]'; % converts to metres per second v_c=mph_c*1609.3/3600; % calculates car's kinetic energy (Joules) KE_c=0.5*M_c.*(v_c.^2);
[n_engine KE_flywheel]
ans =
1.0e+004 *
0.1000 0.2164
0.2000 0.8654
0.3000 1.9472
0.4000 3.4617
0.5000 5.4089
0.6000 7.7888
>>[mph_c KE_c]
ans =
1.0e+005 *
0.0002 0.3997
0.0004 1.5987
0.0006 3.5970
0.0008 6.3947
0.0010 9.9917
So, yes, the kinetic energy of the car is about an order order of magnitude greater than that of the flywheel. So, if you approach the bottom of a hill at 6000 rpm, doing 20 mph, the flywheel might have an effect, but, for normal driving, it's not something to get excited about.
|
Taking the KE in the flywheel, and equating it to potential energy gives;
KE_flywheel(end)/(M_c*9.81)
ans =
7.9396
So, if there were no losses, the car could climb about 24 feet!
|
>>It's surprising when you do the calcs just how much energy is in a flywheel.
>>
So, yes, the kinetic energy of the car is about an order order of magnitude greater than that of the flywheel. So, if you approach the bottom of a hill at 6000 rpm, doing 20 mph, the flywheel might have an effect, but, for normal driving, it's not something to get excited about.>>
Taking the KE in the flywheel and equating it to potential energy gives; So if there were no losses the car could climb about 24 feet!
Thanks NC, intersting stuff!
It is surprising, that earlier link I posted indicated that a torque input of 203lb/ft would be required to accelerate a 20" dia 90kg flyweel at 1000 rpm per sec.
Just to ask, where you say climb about 24 feet, do you mean 24 ft along an incline of X angle? If so what is X?
Have you looked into the adoption of KERS systems in F1, braking the cars into a corner by transfering kinetic energy into small flywheels spinning at 100,000 rpm, rather than heat as per conventional brakes, and using this energy to augment the engine's power along a subsequent straight, 80 bhp has been quoted.
|
PS:
>>
the flywheel might have an effect, but, for normal driving, >>
Meant to say that the flywheel would be supplementing the engine so its kinetic energy would be spread further ... ... ...
|
I haven't looked closely at the maths, but I think its a lift of 24ft, ie up the hill so that the elevation increases by 24ft. It doesn't matter what the slope is.
|
I think its a lift of 24ft ie up the hill so that the elevation increases by 24ft. It doesn't matter what the slope is.
Thanks, not quite sure if that is what NC meant, neverthless a fair amount of energy in there then.
|
Yes, that's excatly what I meant.
I simply equated the kinetic energy of the flywheel with the potential energy of the car, m*g*h. Ignoring the losses is the vital part - in practice, you would lose lots of energy to friction in the clutch, rolling resistance, etc, etc, and you would get nowhere near a 24 foot increase in height - but, 24 feet is the best you could ever hope for starting with that amount of energy available.
|
you would lose lots of energy to friction in the clutch, rolling resistance, etc, etc, and you would get nowhere near a 24 foot increase in height
I have found it difficult to follow the text in the post
www.honestjohn.co.uk/forum/post/index.htm?t=66586&...e
due to the limitations of presenting of graphics and equations in this forum.
So taking your maths at face value, you have calculated here
www.honestjohn.co.uk/forum/post/index.htm?t=66586&...e
that the kinetic energy stored in the flywheel is sufficient to lift a mass equal to that of the car in question to a height of 7.9396 metres [ignoring all losses].
That figure just seems so counter-intuitive, that I would have suspected a decimal point error - but for the fact it has been posted by "Number_Cruncher" .
Edited by jbif on 28/08/2008 at 11:56
|
>>but for the fact it has been posted by "Number_Cruncher" .
Ohno!, quite the reverse is true! The calcs do deserve some scrutiny. I can't devote the same time and effort to checking the calcs I post on here as I do for the calcs I do in my work. The likeliehood of an error is really quite high.
Having had a brief look back at the calcs though, I don't see anything obviously wrong. Working from the sizes of a typical flywheel, and the density of steel gives about 20kg, which is at least in the right range for a car.
Phrased another way, at 7000 rpm, with a 200mm radius wheel,
r=200e-3; % flywheel radius (m)
>>% calculate angular velocity at 7000 rpm (rad/s) omega=7000*2*pi/60
>>% works out peripheral velocity (m/s)
>>v=r*omega
v =
146.6077
>>% converts to mph
v_mph=v*3600/1609.3
v_mph =
327.9610
So, the periphery of the wheel is doing over 300 mph!, and there being a lot of energy associated with that motion is perhaps not so surprising.
|
Thanks again NC, so we have established the the flywheel stores a fair amount of energy so is there anything within my previous point/question (re rolling road measurements, not crank dynos):
An engine with a lighter flywheel may show less torque at low rpm due to storing less kinetic energy, likewise it can show more power at high rpm because it requires less energy to get up to speed so more power is available to be measured at the wheels.
Hence there is a point in the rev range where the benefits of the heavier flywheel at low rpm and the benefits of the lighter flywheel at high rpm cross, i.e. the lighter f/w is better above X rpm and the heavier f/w better below X rpm."
Where I am getting to is here is where you want X to be in the rev range defines the flywheel weight.
?
?
Thanks
|
>>so is there anything within my previous point/question
Yes, the conclusion is that any dyno, whether it runs on the wheels, or on the crank is OK as long as it is a brake, and the readings are taken at steady state. Dynos which rely upon accelerating the wheels against an inertia aren't particularly helpful in serious work.
In real terms, you never want more flywheel inertia than you need, in exactly the same way that you don't see racing teams bolt extra mass into their cars unless the formula rules dictate it.
In some ways, you can view flywheel mass (hence inertia) as being doubly bad - 1) it's mass just like mass anywhere on the car that needs power to accelerate, 2) it's inertia (or rotary mass) requires engine power to accelerate in a rotary sense. There is no golden value to be aimed for.
|
In some ways you can view flywheel mass (hence inertia) as being doubly bad - 1) it's mass just like mass anywhere on the car that needs power to accelerate 2) it's inertia (or rotary mass) requires engine power to accelerate in a rotary sense.
I agree, it is clear that taking say 5kg off the flywheel is more benficial in performance terms than taking 5 kg out of the boot.
However there are perhaps some benefits to be gained from the kinetic energy within the flywheel (in addition to NVH etc), supplementing torque at low rpm when the engine is not producing much power, avoiding the need to change down a gear perhaps when cresting a slight incline etc.
I mean if the kinetic energy in the flywheel at 1500 to 2000 rpm is only a fraction of what it is at 6000 rpm then it could still be influential.
Thanks again :)
|
>>supplementing torque
>>1500 to 2000 rpm is only a fraction of what it is at 6000
Well, at 6000, you'll have 16 times more energy in the flywheel than at 1500.
Yes, the flywheel's stored energy can sometimes be helpful, just as a car's momentum can be helpful every now and again, however, for most of the time extra mass and inertia reduce performance, and the thought of adding mass or inertia as a design aim would have you kicked out of any decent engineering design office very quickly!
|
Yes the flywheel's stored energy can sometimes be helpful
>>and the thought of adding mass or inertia as a design aim would have you kicked out of any decent engineering design office very quickly!>>
Perhaps though I think that there must be an optimum weight/mass, too light causes NVH issues plus reduced flexibility (for want of a better term) at low rpm/low engine efficiency, too heavy causes other issues as discussed.
Thanks again NC !
|
>>too light causes NVH issues
Yes, there's an element of customer acceptibility here. The Vauxhall 3 cylinder engine, for a modern example, has a heavy flywheel which makes gear changes a little bit slow.
The effect of a flywheel is most keenly felt at tickover, because, say, a 50 rpm difference between the crank speed at its fastest compared its slowest is significant, and easily felt at at 800 rpm, and lost completely at 6000 rpm!
Despite us always looking at the rpm counter and accepting the number, the crank is constantly being accelerated back and forth, by compression and by power strokes - what we see on the clock is some slower moving average.
|
>>it is clear that taking say 5kg off the flywheel
Going one step further, it's much more beneficial to take the mass from the periphery of the flywheel than that near the hub.
|
>>An engine with a lighter flywheel may show less torque at low rpm due to storing less kinetic energy, likewise it can show more power at high rpm because it requires less energy to get up to speed so more power is available to be measured at the wheels.
>>at low rpm
You're mixing up power and energy here. There's no difference in how much resistance to angular acceleration a flywheel provides if the engine is turning slowly or quickly.
In direct analogy,
F=m*a
Force = mass * acceleration. The equation which expresses Newton's Second Law
Is exactly analogous to
T=I*alpha
Torque = mass moment of inertia * angular acceleration
So, in the same way that there's no optimum value of mass for a car other than the minimum possibe, the same is true for a flywheel.
In design practice, you would always make the flywheel have as low an inertia as possible while still retaining good NVH, idle quality, idle speed, and emissions**. If you have an engine with many cylinders, you don't need as large a flywheel, because a) there's more rotating mass on the crankshaft already, and b) the engine runs more smoothly anyway, and needs less smoothing.
** for racing, rallying, hill climbing, etc, these considerations don't apply, and making further reductions from an as manufactured flywheel inertia is a common modification.
|
>>Just to annoy you NC
Always a pleasure Lud, never a chore!
I think you've answered cheddars question though - although a heavy flywheel spinning quickly might give you a temporary jolt forward, you've already paid for it in poor performance elsewhere.
Now, if you had a device whereby, you could keep the engine at full power and full power rated speed all the time, and divert energy into either moving the car, or charging a flywheel for later deployment, this could be useful in motorsport - aren't F1 teams playing with toys like this now?
But, to re-iterate the point, in road cars, there's no performance gain from a heavy flywheel, and no serious gain to be had from a light one.
|
From time to time some perpetual motion geek tries to promote energy storage by flywheel, forgetting among other things the tiresome gyroscopic effect... or is there a gimbal fantasy?
Of course under the right circumstances a flywheel can be used like that. They have a concrete one at the Culham fusion reactor, a massive turntable that goes quite slowly, under 100 rpm I believe, but weighs hundreds of tons and can be turned into a generator for the fusion tests which demand more juice than the national grid can manage all at once.
I had the privilege, as a newspaper hack, of seeing the JET device while it was still being built. Like something in a Dan Dare strip, really.
|
>>or is there a gimbal fantasy?
You would use 2 contra rotating wheels. The gyroscopic torque from one would cancel out that from t'other.
|
|
Both wheels would resist any angular deflection of their common axis. The direction of their rotation would make no difference.
|
>>The direction of their rotation would make no difference.
Yes..., and no!
While some degree of gyroscopic stiffness would remain, the cross coupling that, for example on rotating a front wheel during a steer angle change results in a camber torque being applied would cancel.
So, you could locate the two contra rotating wheels with their c of g coincident with the car's and with a vertical axis. Then, you would get a car stiffer in roll and pitch (without using stiff springs, or beastly anti-roll bars), without any cross coupling between them.
|
So you could locate the two contra rotating wheels with their c of g coincident with the car's and with a vertical axis.
So that would be under the front seats.....? Scary.
|
>>Scary.
Knowing of the fretting fatigue problems on the taper ends of the cranks of original Minis has always had me worried while driving them - the flywheel spins directly in line with the driver's nether regions!
|
|
The short, highly-stressed transmission of an ERA passes within centimetres of the driver's much-prized parts, and the gearlever too is right where you don't want it to be in the event of a very sudden deceleration... some cars with rear transaxle transmissions and long whippy propshafts turning at engine speed (I am thinking of certain fairly recent rwd Alfas) were a bit worrying in principle too...
|
>>some cars with rear transaxle transmissions and long whippypropshafts turning at engine speed (I am thinking of certain fairly recent rwd Alfas) were a bit worrying in principle too...
No need to worry, it's a standard prop size & has rubber donuts front, middle & rear with loops in case of a donut breaking. I've never heard of even the most abused one coming adrift.
|
Once had a propshaft make a bid for freedom while driving a Triumph Spitfire. When I subsequently saw what it had managed to do to the underside of the car I was very glad I had managed to stop very quickly.
There is not a lot of anything between your fundament and the end of a maverick propshaft in a Spitfire.
|
|
This is all beginning to make my head spin, never mind my car's flywheel!
Edited by L'escargot on 28/08/2008 at 08:56
|
|
|
|
|
|
