Anyone had experience with this ?

[97 MXV]

I wouldn't want to be the one to find out

aka "Don't try this at home".

Basically, flywheels can let go, and when they do it's usually pretty dramatic (at 7000rpm there's a lot of centrifugal force on them).
Personally, I wouldn't use that flywheel for anything more than decoration, The bellhousing won't stop it, and in a lot of cases the trans tunnel does stuff all too. (at least in an MX5 the clutch is in front of your feet)

aka "Be Mindful of the Consequences".

I think the flywheel would warp to a point that would render it almost undriveable before it failed.

Without a reasonable analysis of stress due to centrifugal force, versus strength of material at the temperature it was at when the stress was applied, this could be a dangerous assumption.

aka "Don't try this at home".


No worries guys Thankfully the poor wretch of a flywheel will live on with no mean nasty Consequence from any potential hot cracking.
Sailaholic will see to that and put it where heat can't get to it......happy ever after behind quad throttles and in front of a 4.78 diff.

Thanks Keith.
I have the facilities through my contacts to mag particle / dye penetration test it to look for cracks.

Now that flywheel is sorted, lets pose the question of how to look at this Risk...by Probability or by Consequence ? How much weight do you give dire Consequence ?
Is the Consequence real or imagined ? If the Consequence is imagined, then saying "Who Cares" or "To Hell with the Consequence" or even "IDGAF" is perfectly reasonable.

Imagine if the earth was flat, the Consequence of sailing over the edge would be dire. But if the world was round, who cares ? Consequence would not matter.
So maybe it is worth wondering do we judge this poor wretched flywheel with flat earth thinking around Consequence or with round world thinking around Probability ?

Imagine the old days when all that could be envisaged were the Consequences.....Before Japanese engineers applied modern 4 cylinder learnings to the aging V8.....the Toyota UZ.
New World 4 valves, quad cams and a forged steel bottom end, deep skirt with 6 bolt mains, not Old World 2 valves, single cam, pushrods and cast iron bottom end, no skirt with 2 bolt mains.
Imagine a flywheel attached to these:


____________________Jap V8_____________________________________Yank V8_________________________

Now look at the imbalanced cast iron flywheel the Yanks attach to their flimsily held unbalanced cast iron crank.

_________Deep undercut Step __________________________Big Imbalance Weight Step

Then you find Yankee tuners make an art form out of front pulleys to try and sort their crank balancing/resonant crank mess out.
No other country in the world builds such fancy pulleys.....because other countries without the Consequences of old world V8s don't need a fancy pulley industry.



So given this and that big heavy flywheels store launch energy, drag racing and catastrophic flywheel ruptures tend to be synonymous.....ie their earth is flat....our earth is round.
Thank you Japan. Thank you for transforming cheap flat earth technology that was thought to be "the status quo" to affordable round world technology we all now can envisage.

BTW it is hard not to notice how these flat earth tuners are now even making their V8 pulleys an expensive 4 Cylinder Solution in search of a Problem elsewhere.
This is done by ignoring Probability and applying flat earth Consequence to the rest of the world which happens to be round. ("Round means stress range is kept under the endurance limit at the limits)

V8 flywheels with their cast iron construction, stepped cross sections and imbalanced construction, are a balancing nightmare for those running NASCAR, Drags or Moonshine.



However in a modern inline four world that is already balanced and round, the evidence is not there to give weight to incipient catastrophic failure mode of flat earth cast iron.
Take a look at the crank comparison on a 351 Ford verses a Mazda BP (5750 cc Vs 1840 cc). Easy to imagine the V8 crank and its flywheel having a few more torsional issues.


_______________________Ford 351 V8 (5750 cc) SG Cast Iron ________________________________________________Mazda BP (1840cc) Forged Steel

Just from a size/weight perspective the poor old V8 crank carries 300% more torque and suffers 60% more torque impulse in a cast iron crankshaft.
For serious stuff it is obviously a throw away crank at just half the endurance limit of the forged steel BP crankshaft despite being of malleable quality, and having 25% more unit mass.

Furthermore the unbalanced V8 flywheel is held on by 6 x 7/16 (11.1mm) bolts on a 75mm PCD compared with 6 x M12 bolts on a 65mm PCD holding on the BP flywheel.
Doing the sums means the BP flywheel is 1% more strongly fastened than the 5.7 litre V8 flywheel.
Despite everything going for OEM bolts on a BP crank, it seems ARP bolts are still necessary to qualify for NASCAR status (on the earth's flat side).

The above argument and evidence makes the case that Consequence is real for Ford USA fans but imaginery for Mazda Japan fans because one thinks flat and the other thinks round.


Getting back to our tragic and sad CrMo flywheel, the Probability of incipient catastrophic failure mode in low power road or circuit use provided there is no cracks present is practically zero.
This is because of a belief in an ultimate ductile failure mode, distorting not disintegrating itself and its surrounding protective structures. It is so clean and simple you can imagine the FEA.
The material toughness and the smooth 2D transitional cross sections gives stress nowhere to concentrate itself to give rise to heat cycle induced fatigue.
Stress wants to just flow nicely in this flywheel, because it has no stiff constraints attracting stress or holes/transitions concentrating stress, except of course around the bolt holds in the centre.
The worst failure I could imagine is unacceptable TIR (Total Indicator Runout) from heat induced warping found when it is bolted up to a crank.

The question boils down to this....Ultra lightweight flywheels are clearly unsuited for certain applications involving high power clutch slip, whether previously hot spotted or not.
It is obvious from the appearance of this one, it was too light for the turbo motor and its modulation through the clutch.
The worst hot spotting consequence I could find in Google search was warping and clutch chatter. A heavier duty one is better for turbo use.

This all begs the question, what about aluminium ? Given all the crank balance/vibration issues in the USA, maybe alloy flywheels fall into the category of "Don't try this at home".
Since aluminium has an endurance limit near zero, it becomes an eventuality that unless the fatigue load spectrum is properly managed someone must face the Consequences.
Aluminium is soft and score marks in areas of high stress initiate cracks. The question is when does it get inspected, tested and scrapped ? Is it serial numbered and pre-failure maintained ?
Funnily enough the manufacturer's website promoted Replacement Steel parts but not Replacement Aluminium parts.
The manufacturer is Mindful of the Consequences though, and so too will the customer be if they care to "Read the Instructions" !

Extract from Manufacturer's Instructions.
FOR ANY PERFORMANCE/RACE USE VEHICLE: Inspect flywheel whenever possible for fatigue, cracks,
damage or adverse wear. Some of the most critical areas to inspect are: (1) The crankshaft register (2)
Flywheel to crank mounting holes (3) Ring Gear. Extreme heat can adversely affect the dowels and ring gear.
Extreme heat can, as with any flywheel, affect the ring gear causing it to grow and not return to its static
diameter, possibly causing eventual ring gear failure. Precautions must be taken in performance use
vehicles to avoid this dangerous situation. The use of a scatter shield or safety blanket for the clutch and
flywheel area is a MUST in all performance/race use vehicles.

[StillIC]

Nice work Keith. I think you probably correct in your latest post, however, based on seeing the flywheel you had, I still have a concern......

Stress wants to just flow nicely in this flywheel, because it has no stiff constraints attracting stress or holes/transitions concentrating stress, except of course around the bolt holds in the centre.

Yes, stress has nowhere to concentrate as such, as you say, but my concern is that the *strength* of the material may not be uniform due to localised heating, and that localised stresses may occur due to differential thermal expansion (which may well cause warping or a non-catastrophic failure).

Having said that, I guess that the material would be less brittle/more ductile, if anything, at the hot spots, thus minimising the chance of failure starting here, as the stress would be forced to flow around the more ductile/weaker areas if stress was close to yield strength in these spots.

Yeah, I think you have convinced me. But my concerns partly stem from a lot of thermal FE analysis I have performed on round clamped hot plates (which don't spin), where the main issue is a combination of bending due to clamping and differential thermal expansion (hot on one side, not on the other).

In regard to the aluminium flywheel idea, it certainly would dissipate heat better than steel, but the fatigue life would mean I would never run one. But what about a composite aluminium/steel flywheel, or even a reasonable amount of finning on a lightweight steel version to help with the heat dissipation?

Perhaps the answer is simply a strong enough pressure plate to minimise clutch slip.

[madjak]

A flywheel shouldn't get that hot as the clutch shouldn't be slipping. The only time a flywheel should take heat is at launch. During hard shifting there should be very little slip and if there is slip the driver is doing it wrong. The lightweight flywheel should be fine in a turbo car as long as the pressure plate and friction material of the clutch are suitable to hold the engine torque.

I run an aluminium flywheel. I'll update this thread if ever it explodes.

[97 MXV]

I run an aluminium flywheel. I'll update this thread if ever it explodes.

aka Sticking with Probability.

I guess since Rick's flywheel is not holding on for all its dear life to a flat earth dodgy unbalanced crank in a 351 moonshine runner, I might die of old age waiting for the explosion.
So really unless Probablity is sorted for a BP engine application, that fatigued out Consequence (explosion) might coincide with the sun running out of hydrogen.

I imagine in a BP the fatigue life would be heavily dominated by relatively small numbers of high stress ranges from high energy events rather than high cycle fatigue from normal operation.
Maximum shift energy would be related to engine power/torque so I imagine a turbo engine can't help but heat cycle a flywheel harder, irrespective of steady state clamping force.
In other words, a relationship between engine power and flywheel weight to me still seems more plausible than myth.

So where do fatigue inducing stress ranges come from?
1. Diaphragming response of the disc flywheel with on/off of the clutch.(heavy clutch worse)
2. Changing reaction forces with engine speed change.
3. Changing reaction forces with engine torque change.
4. Starting torque
5. Rise of steep temperature gradients
6. Differential material expansion/contraction

Being so firmly attached to a heavily weighted BP crank, there are unlikely to be the horrible high cycle imbalance loads and cases of the wobbles the high powered moonshine runner crank/flywheel combos might get.
So the determining factor in respect of Probability is not going to be high cycles (ie number of km) and not so much the number of cycles of 1 to 6 but rather the extent of the stress range achieved in those cycles.
That points to the type of use and points to the Manufacturers Instructions all leading to high horsepower, high energy launches, low speed high torque competition (rather than high speed) as critical determining life factors.

I guess that the material would be less brittle/more ductile, if anything, at the hot spots, thus minimising the chance of failure starting here, as the stress would be forced to flow around the more ductile/weaker areas if stress was close to yield strength in these spots.

I guess the hot spots are harder due to martensitic transformation of the alloy steel during air quench of the thin section, evidenced here.
Stress flow in the elastic region of steel is determined by stiffness not hardness or strength. Given Elastic Modulus, Shear Modulus and Poisson’s Ratio do not vary with hardness, there will be no stress changes or gradients resulting from hot spots.
Furthermore hot spot fatigue is possibly more related to contact surface defect/crack size than of the reduced fracture toughness of the hard spot overall.

But my concerns partly stem from a lot of thermal FE analysis I have performed on round clamped hot plates (which don't spin)

That sounds a lot like "steel clutch face clamped to the aluminium flywheel" technology.
The complex temperature and stress gradients and heat transfer goings on in such an interface boggles my mind, but StillIC may have the enabling technology.
Not important it is spinning at the time as that is stress constant, not stress range. So this technology of StillIC sounds promising in respect of Probability.


But the CrMo is all one piece, no separable clutch face and no separable ring gear to plug into StillIC's FEA model.
No worries, I have already nailed that Probability free of FEA....I just calculated the sun will run out of hydrogen before these parts of a CrMo separate.
So Probability says no need to keep tabs of a CrMo's age

[madjak]

Well I'm revving it to 8300 with hard shifts on r-comp tyres... if there is a failure I'm almost certain it won't be in the spinning balanced lightweight flywheel rather the oscillating parts in the rest of the engine.

Even if a flywheel fails, the energy is all rotational anyway. So damage would really only be caused if it shattered and parts flung outwards. Even then the flywheel doesn't have that much radius or weight so the energy imparted to exterior shards wouldn't be that great, even at 8000 rpm. Given a such light flywheel in the first place, and a good thickness of transmission housing, the floor of the body, even the pressure plate holding stuff together etc I can't see it causing damage to the driver unless it's a freak event.

My brother's GT6 race car has 4 M12 bolts holding a massive flywheel to a triumph 6 cylinder tractor crank. Those bolts have failed multiple times at very high revs on the race track and all you get is a clattering and loss of drive. The flywheel is held in place on the input shaft.

From a risk point of view, failure of brake components, wheels, tyres at speed are a much bigger concern than exploding flywheels.

[StillIC]

I guess that the material would be less brittle/more ductile, if anything, at the hot spots, thus minimising the chance of failure starting here, as the stress would be forced to flow around the more ductile/weaker areas if stress was close to yield strength in these spots

I guess the hot spots are harder due to martensitic transformation of the alloy steel during air quench of the thin section, evidenced here.
Stress flow in the elastic region of steel is determined by stiffness not hardness or strength. Given Elastic Modulus, Shear Modulus and Poisson’s Ratio do not vary with hardness, there will be no stress changes or gradients resulting from hot spots.

Yes but...
...to reiterate, my comment is only valid when the hot spotted steel is at or near yield stress, where elastic modulus is no longer the simple constant it was up to that point. And I was suggesting that the steel might be weaker, i.e. have a lower yield strength than surrounding steel, therefore change its stiffness compared to surrounding material *when stress is equal to localised yield strength*. This comment is specifically in regard to flywheel failure, which has to start at some point where yield strength is exceeded. This is all pure speculation of course, as the steel may not be weaker at the hot spots, for all I know.

For the flywheel in the video, my guess is it was not made from 4140, or if it was it hadn't been heat treated before use. This is where I was hoping Dr. Guran could add value to the discussion, as I am quite rusty (no pun...) on heat treatment, but well heat treated 4140 should already contain a martensitic phase, IIRC???

I run an aluminium flywheel. I'll update this thread if ever it explodes.

Well, I guess we all have aluminium road wheels. If they can last until the sun runs out of hydrogen, then a flywheel could be designed to do the same, if kept within designed operating parameters. I rescind my previous comment regarding fearing fatigue of a well designed Al flywheel. My excuse is that I was swept up in a wave of overwhelming persuasiveness foisted upon me by Keith!

[97 MXV]

Join the club StillIC, I too was swept up in a wave of overwhelming persuasiveness foisted upon me by the manufacturer in their official User Instructions.

Extract from Manufacturer's Instructions.
FOR ANY PERFORMANCE/RACE USE VEHICLE: Inspect flywheel whenever possible for fatigue, cracks,
damage or adverse wear. Some of the most critical areas to inspect are: (1) The crankshaft register (2)
Flywheel to crank mounting holes (3) Ring Gear. Extreme heat can adversely affect the dowels and ring gear.
Extreme heat can, as with any flywheel, affect the ring gear causing it to grow and not return to its static
diameter, possibly causing eventual ring gear failure. Precautions must be taken in performance use
vehicles to avoid this dangerous situation. The use of a scatter shield or safety blanket for the clutch and
flywheel area is a MUST in all performance/race use vehicles.

So again at the risk of being too obvious....Probability not Consequence must be the overwhelming persuasiveness that drives the engineering decision making process.

Well I'm revving it to 8300 with hard shifts on r-comp tyres...
SIG - NA8: 200+ whp, 6 speed, 4.78 diff

That points to the type of use and points to the Manufacturers Instructions all leading to high horsepower, high energy launches, low speed high torque competition (rather than high speed) as critical determining life factors.

We have consensus ! Rick's application is not the critical type (involving big heat) being discussed.
Just not possible with 200+ whp, 6 speed, and 4.78 diff on twisty bits.

So in Rick's case a Probability discussion is meaningful and Consequence discussion is meaningless because Probability says Consequence can't be achieved.

So where does this leave the Manufacturer's Instructions. ? It leaves them as a legal disclaimer (or arse cover) which funnily enough proves the Consequence was known and expected to be managed by some bizarre Probability that the customer would act on "the fine print" and design and manufacture a guard (or no guard) having a 100% Probability of being effective .
Such ludicrous "hospital passes" takes one into the big wide world of Duty of Care Pandora's Box USA style trolled by lawyers sadly, not engineers.

On the subject of alloy wheels, again it is Probability that decides. Examination of the steel thickness and bolted joint design verses alloy thickness where critical failure is possible tells the story.
Examination of steel thickness verses alloy thickness at the tyre rim joint tells another story.
We all know the Probability that alloy wheels will fail pot hole encounters before the sun runs out of hydrogen so accept the Consequences at the outer rim of the Bling it Brings.

[project.r.racing]

A flywheel shouldn't get that hot as the clutch shouldn't be slipping. The only time a flywheel should take heat is at launch. During hard shifting there should be very little slip and if there is slip the driver is doing it wrong. The lightweight flywheel should be fine in a turbo car as long as the pressure plate and friction material of the clutch are suitable to hold the engine torque.

I run an aluminium flywheel. I'll update this thread if ever it explodes.