Hey guys!
I'm working on something a little weird...
I see many aftermarket companies are making lightweight flywheels. Why?
Here are the questions I can think of to get a discussion going.
1. How do we mathematically figure out the perfect flywheel weight?
2. What is the advantage of having a light or heavy flywheel?
3. What happens when you go too light or too heavy?
4. Do we apply the same principals from a traditional manual transmission to a new style dual clutch transmission?
In reply to darkostoj :
I'm going to answer your questions out of order:
2. A lighter flywheel will allow the engine to rev quicker due to less rotating mass. However, this can also make it easier to stall the engine when taking off from a stop. A heavier flywheel also smooths out some vibration.
3. Too light = too easy to stall the engine. Too heavy = engine doesn't rev quickly enough
4. Good question. I'd say yes & no - weight still affects the ability to rev, but the dual-clutch would prevent the revs from dropping too low between shifts.
1. I'm no engineer so I can't help with the math, but you need to determine the characteristics you want from #2 before you work out the correct weight.
I have always enjoyed the revving characteristics of a lightweight flywheel, and have frequently swapped for the lightest I could find- no science.
I recently did it to an RX8 and regretted it. Yes, the engine revved nicely. But that car is pretty heavy, and the only way to get it rolling was to ride the clutch. Burned up a new clutch pretty quickly.
So, I'm checking in because I want to learn a few answers to the same questions you are asking!
This has a lot of math so I haven't tried it at this point.
https://www.engineersedge.com/mechanics_machines/flywheel_mass_size_design_13981.htm
darkostoj said:Hey guys!
I'm working on something a little weird...
I see many aftermarket companies are making lightweight flywheels. Why?
Here are the questions I can think of to get a discussion going.
1. How do we mathematically figure out the perfect flywheel weight?
2. What is the advantage of having a light or heavy flywheel?
3. What happens when you go too light or too heavy?
4. Do we apply the same principals from a traditional manual transmission to a new style dual clutch transmission?
1. Math sucks
2. The real answer here is to have the correct flywheel weight for the application. See #3 below
3. The flywheel stores kinetic energy. Let's say you go really heavy. LIke 500 lbs for the sake of demonstration. There is so much inertia stored in that flywheel that you could pop the clutch while towing 10,000 lbs and the vehicle would just lurch forward and not stall (and probably snap an input shaft or U-joint, but that's a different story). Some European buses have gone to using this idea to save some fuel. They utilize a very small motor (say, 20hp) that spins a massive flywheel that might be 1000 lbs and 4' in diameter. Most of the forward motion comes from utilizing the stored energy in the spinning flywheel and the 20hp motor simply replenishes it. Instead of building a 200 hp motor that only uses that 200 hp for very brief moments, it is able to "average" the hp it needs and store it in a spinning flywheel. In a regular automotive application, going too heavy costs acceleration. You're not only using engine power to overcome the inertia of getting the car moving, you're overcoming the inertia of the flywheel... much like using bigger/heavier wheels and tires. The other downside is that it can become difficult to match RPMs during upshifts. Anyone with a Tacoma 4-cylinder is aware of this. They use a small 4-cylinder engine in a 4000 lb truck. In order to get the truck moving without stalling, they use a somewhat mismatched heavy flywheel. Upshifting can be difficult to do smoothly because the extra inertia in the flywheel keeps the little engine spinning faster for longer. When you let off the accelerator during a shift, the RPMs don't fall at a rate that complements the speed of the shift and the ratio change. It is a necessary trade-off for the mismatch of low torque production and heavy weight. Going too light means that the rotating inertia is lower, which means engaging the clutch requires more slipping and more throttle than a heavier flywheel. On the opposite end of the spectrum from those buses are F1 cars. They have tiny lightweight flywheels. Getting started from a stop is incredibly difficult, but once they're going, the lightweight flywheel doesn't stand in the way of the super-fast shifts and the acceleration needed to get 2" ahead of the racer beside you.
4. I would say that the same basic principles apply, but likely skewed toward the lighter end of things. The vehicle still needs to apply the clutch to get moving from a stop, but since there is little or no shifting time, it needs to be able to change speed quickly. I'm not that well-versed in the finer points of flappy paddles, though.
TL;DR... The right weight for the application is key. One person might use a 12lb flywheel in their Miata because they only Autocross it, and another might use a 15lb flywheel in theirs because it is a daily driver.
I don't have any first hand experience with this but I strongly suspect that changing flywheel mass on an existing DCT transmission application could cause significant issues with low speed drivability. I'm basing that on a couple of factors. One is that different flywheel masses with conventional transmissions require different throttle and clutch inputs and that's all software in a DCT. The other is that it's apparently difficult to get that software working well or at least it's done badly in some otherwise fabulous automobiles (I'm looking at you Porsche GT3).
I have to agree with apeowner on DCT's. With electronic controlled clutches, changing the flywheel could lead to odd issues that might be difficult to correct unless you have specialized knowledge. A few lbs either would likely be fine, but if you did like i dod with my old car, and replaced a 24lb clutch/plate/flywheel with a 11lb one, you will need transmission and possible engine tuning as well to see even normal performance. To what extent are modern DCT's hackable (by the mere mortal with a day job) I wonder?
The beauty of manual transmissions is the ease of reprogramming the shift logic. And oh boy did I need recalibrating.
For the record, anything under 10lbs (in a 2400lb car, espescially with a grabbier clutch) is annoying, and should only be considered by those who require a workout and racers (like me!) For a sane daily, get a slightly lighter single-mass chromoly and have it checked and/or machined for runout (in relation to the crank face!!!).
I have heard that some mfg's use the stock flywheel as part of their damper system, but I have no idea how they work. Be weary of messing with them, as I personally have seen many 'internally balanced' i4's self destruct with aftermarket crank pulleys. In the hondas, the extra vibration would grenade the oil pump and predictable carnage would shortly follow as the debris from the pump worked through the galleys, oil pressure would then drop, and the spinny hardened steel bits make intimate and unwelcome contact with their otherwise loving journals, now loaded with grit. In a word, heartbreak.
But now I am curious to look up dct hacking, as I might want a 'new' car one day, and so far have only owned manuals.
Edit: I wonder if a well-hacked dct could use amazingly light flywheels with the right programming... some sort of advanced 'clutch traction control' just to manage enguagement? That might even benefit the few manuals still being produced if the feel doesn't mimic 1990's ABS systems...
Knowing what darko does for a living I am really excited about where this may be going.
In reply to buzzboy :
Crankshafts, like every other material, are flexible. (they ring when struck or dropped)
Flywheel rotational inertia influences the torsional resonant frequency of the crank assembly.
If that resonant frequency or a harmonic matches the operating RPM, flexure will build up and can break the crank.
It's why some engines run crankshaft torsional dampers.
Usually, but not always, a lighter flywheel raises the resonant frequency, and is thus 'safer'.
But not always.
I think the core things have been covered, but maybe there's some value in summarizing:
Because it's possible to have a flywheel which is simultaneously lighter than ideal for pulling away/smoothness and heavier than ideal for acceleration, we can conclude that there is no single ideal mass, but rather that selection is a compromise. That compromise will be directed by requirements and preferences.
I'm now curious about the upshot, and I hope the summary isn't regarded as intended to stop conversation. I think perhaps seeking a "perfect" flywheel mass suggests the need to discuss the intent of the question as much as the answer?
For instance, if we're trying to come up with a mathematical answer, how do we quantify the difficulty a driver (or drivetrain control computer) has with pulling away from a stop with a very light flywheel? The effect of flywheel mass on acceleration is more straightforward, though weighting its importance in the "ideal mass" equation seems subjective.
Okay, I did have a follow-up thought about a flywheel minimum. Since you can substitute, to some extent, throttle and RPM for flywheel mass in rolling away from a stop, at the expenses of clutch slip and/or violence of the start, can you define a maximum rate of acceleration that's okay for a controlled start, and from there work out how much flywheel mass you need to avoid needing excessive revs/throttle/clutch slip?
E.g. if you need a lot of revs/throttle and you don't want to slip the clutch, you have to engage fairly abruptly, in a manner not conducive to comfort/etc. On the flip side, with the same flywheel, you could pull away more gently at the expense of a bunch of clutch slippage. Neither is "nice".
I suspect that near this minimum value it would be a very unforgiving car to drive, but then perhaps a "user-friendliness" factor could also be applied.
I've often wonder why some clever person hasn't come up with a variable flywheel for performance cars or have they?
In reply to Tom1200 :
How would you vary the mass (or moment of inertia) of a flywheel?
This brings up the dual mass flywheel. How and why ? I think the FiST has one.
Matthew Kennedy said:In reply to Tom1200 :
How would you vary the mass (or moment of inertia) of a flywheel?
Not vary the mass, but maybe something more like distributor weights that swing in & out?
Pete Gossett (Forum Supporter) said:Matthew Kennedy said:In reply to Tom1200 :
How would you vary the mass (or moment of inertia) of a flywheel?
Not vary the mass, but maybe something more like distributor weights that swing in & out?
Like this?
http://bestride.com/news/video-the-real-meaning-of-balls-to-the-wall-balls-out-and-big-brass-balls
My ‘05 Legacy GT came from the factory with a heavy (26lb) dual mass flywheel. When I did the clutch at 100k miles, I agonized over what to do about the flywheel. I decided to switch it out with a conventional lightweight 14lb flywheel and WRX clutch combo. Conventional wisdom was that it was too light for a 3500lb AWD turbo 4 cylinder wagon. It wouldn’t be as smooth, I’d need more revs from a stop. The car was an otherwise stock daily driver. It turned out to be a terrific choice. The driveability issues that the internet blamed on the drive by wire throttle went away. Shifting was much improved. It was even smoother from a stop. I attribute that to improved throttle response. The heavy flywheel induced a lag between throttle input and the corresponding increase in revs that was no longer noticeable with the lighter flywheel. Interestingly, Subaru went away from the dual mass flywheel in ‘07 on that car, with a lighter (22lb?) conventional unit. One thing to keep in mind. Pressure plates are pretty heavy, and effectively act as part of the flywheel. So going from a 30lb flywheel to a 20lb is not really a 33% decrease. If the pressure plate is 10lbs, the effective change is 40lbs-30lbs, or 25%.
That was my thought, moving the weights in or out but I'm not an engineer.
In reply to Boost_Crazy :
From what I understand, many dual mass act similar to a lighter weight flywheel shock mounted to a heavy one, supposedly giving the instant torque change of the lightwehight one, but the extra heft of the sprung one when you ham-foot the clutch or otherwise need more oomph. The ommph seems to come a bit late while still slowing down ungeared rpm changes like a big flywheel to me.
Also, someone has a 'variable weight' flywheel that worked by shifting heavy chunks of metal outward in radially, thereby increasing the effective inertia as if changing the weight. I'll search in a bit for a link.
Edit: 'variable inertia flywheel' yeilded more results than I can post. However if this effect was pronounced enough, you could, instead of making the flywheel smaller for faster shifts, make the center of mass smaller throughout the rpm and expand it suddenly during dct shifts to slow the engine while setting flywheel at full 'weight' for extra low rpm power during hard acceleration. Does that make sense or has the beer kicked in?
Double edit: Then would pulling the mass back in at high rpm give you extra acceleration?
There is so much good info in this thread, but almost none of it will help Darko choose a flywheel. It's all kinda theory.
In reply to SVreX (Forum Supporter) :
I thought his question was more about exploring the theory how to think about the general case than actually picking one flywheel.
Nugi said:In reply to Boost_Crazy :
Edit: 'variable inertia flywheel' yeilded more results than I can post. However if this effect was pronounced enough, you could, instead of making the flywheel smaller for faster shifts, make the center of mass smaller throughout the rpm and expand it suddenly during dct shifts to slow the engine while setting flywheel at full 'weight' for extra low rpm power during hard acceleration. Does that make sense or has the beer kicked in?
I don't think faster shifts is the central concern; a flywheel with more inertia is able to provide more stored energy to the system during momentary high load (releasing the clutch from a stop), while one with less inertia requires more energy to accelerate at a given rate (or accelerates more quickly for a given energy input). My instinct would have been to leave a variable flywheel in "low inertia" mode during shifts to minimize the "thump" of the next gear hitting, but doing as you suggest might be smoother still in that it would presumably slow the crank just a bit while the weights were transferring outward (think figure skater slowing a spin by extending limbs) making the engagement of the next gear less violent. That said, you'd want them sucked back in ASAP to avoid spending energy accelerating the high-inertia mode.
Double edit: Then would pulling the mass back in at high rpm give you extra acceleration?
Yes, for exactly the reasons mentioned above. Well, you wouldn't want to pull it back in AT high RPM. You'd want to pull it back in as soon as the next gear is engaged to avoid having to accelerate the extra inertia. It's the change in RPM that the extra inertia is fighting.
SVreX (Forum Supporter) said:I have always enjoyed the revving characteristics of a lightweight flywheel, and have frequently swapped for the lightest I could find- no science.
I recently did it to an RX8 and regretted it. Yes, the engine revved nicely. But that car is pretty heavy, and the only way to get it rolling was to ride the clutch. Burned up a new clutch pretty quickly.
So, I'm checking in because I want to learn a few answers to the same questions you are asking!
You just saved me about $1000. This was numero uno on my RX8's hit list as I was SURE this would be money well spent. Thank you. And if we ever cross paths the bar tab is on me.
In reply to Hungary Bill (Forum Supporter) :
You are welcome.
I love the RX8, but the Renesis simply doesn't have enough torque with the weight of that car and a lightweight flywheel. I wouldn't do it.
SVreX (Forum Supporter) said:There is so much good info in this thread, but almost none of it will help Darko choose a flywheel. It's all kinda theory.
Once he bolts the flywheel on, it will become practice. 
You'll need to log in to post.
