Transmission parasitic loss discussion...

[Sinister Fiero]

Will and I got into a discussion over on another thread about this topic. Basically to sum it all up, Will insists that as power level goes up; parasitic loss thru a manual trans will go down in % of loss. I argue that parasitic loss % of engine power will stay the same due to the laws of physics and friction.

Well I discussed this topic with my good friend who works at International who recently graduated from IPFW and has a degree in Mechanical Engineering. He said that parasitic loss (% of engine power) will NOT decrease as engine power increases.

Example of what he is saying:

-140hp 2.8 loses ~ 10% of its power going thru a getrag manual trans.

-300hp N* loses ~ 10% of its power going thru a getrag manual trans.

I explained to him Will's side of the arguement and he said the reason why it can't work that way is because for Will's theory to be true, (that % power loss goes down with more power); at some point % power loss would approach 0% with even higher power levels, which is not possible according to the laws of mechanical power transfer, friction, and physics.

[Series8217]

So what you and your friend are saying is that if you have, say, a 10,000hp motor, 1000 hp is going to be lost through a Getrag manual trans (considering that it holds up), right? So thats 745kilowatts lost into the transmission. Even if it could physically hold up to the torque, thats a lot of power that has to go somewhere besides into melting the transmission into one giant piece of steel-aluminum alloy. I guess that explains why Top Fuel dragsters don't use gearboxes?
It sure makes sense to me.

[Series8217]

Ok I was just getting curious about the implications of this so I did some calculations to test my knowledge and see what I come up with. So yeah I was bored.

10,000 hp (7,456 kW) motor running through a transmission of 10% drivetrain loss. Assume that the gears do not shear and we're not generating electricity. Lets imagine we produce the full 7,456kW for the whole 4.5 second 1/4mile run, because I dont want to calculate an integral right now and I dont know what a top fuel dragster power curve looks like.
4.5 sec = 4.5/3600 hr
7,456kw * (4.5/3600) hr = 9.31 kWh
9.31 kWh * 3,600,000 J/kWh = 33,552,000 Joules.

The Getrag weighs roughly 100 lbs = 45 kg.
The specific heat capacity of aluminum is 900 J/kgK. The specific heat capacity of steel is 450 J/kgK. Assuming the Getrag consists of 50% steel and 50% aluminum by mass, the overall specific heat of the transmission is 675 J/kgK.
Assuming the transmission is isothermal at 71 deg Farenheit = 295 K, over the course of the run we heat it up by:
kgK/675J * 33,552,000 J * 1/45kg = 1104K
So now the transmission is up to 1399K = 2058 deg F (the melting point of steel is 2500deg, aluminum is 1220 deg F)
In reality, the steel will get more of the heat because essentially all of the moving surfaces in the transmission where the heat is generated are made of steel. The specific heat capacity of steel is also lower. So the steel will actually get hotter than this. Also the transmission is going to have heat transferred into it by the engine. Also its likely not going to start at 71 degrees. However, the engine isn't producing its max power for the whole run.
Lets imagine we only produce half the energy. We go to 847K. 1064 deg F. Ouch.

Lets compare now look at a situation we're likely to encounter in real life (who here owns a Top Fuel dragster??). Say someone is making an average 500 hp (372kW) over a 10 second 1/4 mile run with a transmission of 10% parasitic loss and materials similar to the Getrag, though weighing twice as much (hey, it has to hold the power!), they get a 55K increase in temp per run. Thats a little over 90 degrees farenheit in just 10 seconds. A manual transmission without an active cooling system (who has active cooling on a manual??) is going to warp its aluminum housing after a run if it starts at any temperature greater than the temp of the air outside, and isnt 10% loss as good as it gets? Now I'm starting to see how Will may be right.

Does any of that make sense?


EDIT to fix math, I accidentally did it for 1,000 hp instead of 10,000 hp in the first example. Also added second example.

[stimpy]

Math is hard. :lamer:

[bryson]

I don't think that Will meant that the parasitic loss decreased linearly (as you assume by saying it will eventually reach zero). I think that what he means (and I could be wrong) is that it will require a certain amount of power to turn the drivetrain, and since that amount won't change a whole lot with the addition of extra horsepower, the percentage will go down. If it takes 30hp to turn the drivetrain on a 120hp engine, then there isn't reason for it to take more than 30hp to turn the same drivetrain if the engine is producing more power. We are studying gear trains rigt now -- I will go over my Foundations of Mechanical Systems book and if there's anything significant in there, I'll post back.

[S8n]

At a constant RPM, it should take the same amount of power to turn the trans no matter how much HP the engine has. During acceleration though, if the rate of acceleration is more, then the trans is gonna soak up more power because it takes more energy to accelerate those parts faster. In other words, if the car accelerates faster, it takes more energy to make the trans internals accelerate faster. Make sense?

[Sinister Fiero]

So what you and your friend are saying is that if you have, say, a 10,000hp motor, 1000 hp is going to be lost through a Getrag manual trans (considering that it holds up), right? So thats 745kilowatts lost into the transmission. Even if it could physically hold up to the torque, thats a lot of power that has to go somewhere besides into melting the transmission into one giant peice of steel-aluminum alloy. I guess that explains why Top Fuel dragsters don't use gearboxes?
It sure makes sense to me.

Top Fuel Dragsters DO use gearboxes. What did you think they used? Automatics? They use Jerco sectional transmissions which are of the manual variety.

Back on topic: the power loss thru a transmission is due to friction which results in heat production. The energy can't be created nor destroyed so it has to go somewhere. The more power you put thru a gear train basically increases its load. The higher loads result in higher friction levels.

Let me put this another way: For Will's theory to be true, friction within a gear train would not increase as load increases. Due to the laws of physics, we know this is impossible. Furthermore it was also explained to me by my friend there is another parasitic loss within the transmission that basically robs more power with higher RPM levels. This has to do with the lubricant's fluid sheer forces; IE: the faster you turn the gears the more drag you are going to get.

[The Dark Side of Will]

Top Fuel Dragsters DO use gearboxes. What did you think they used? Automatics? They use Jerco sectional transmissions which are of the manual variety.

You mean like Lencos? Top Fuelers don't use those, AFAIK... just a mighty stout clutch and a diff.
Funny cars use Lencos... Do they use Powerglides as well? How much HP can a PG handle?

I'm driving down to Charlotte today, so I'll have 5.5 hours to think about this.

[Series8217]

Top Fuel Dragsters DO use gearboxes. What did you think they used? Automatics? They use Jerco sectional transmissions which are of the manual variety.

They don't use gearboxes!!

http://www.popularmechanics.com/automot ... page=2&c=y

"The car has no transmission, so the five-disc clutch is set up to slip just enough to keep the tires from breaking traction. It's all controlled by centrifugal levers (adjustable by adding or removing weights) and air pressure (regulated by mechanical timers). "

[Sinister Fiero]

Sorry about that guys, you are right the funny cars use the manuals and the top-fuel dragsters use direct clutch-diff setups. My bad (I was confused).

As far as the powerglides go, I am pretty sure the fastest cars that use them run in the 7's or maybe 6's; but nobody running in the 4's is using anything resembling an auto.

[aaron88]

The transmission is a stand alone unit, it’s efficiency is based on a number of things and will change slightly depending on the characteristics pre-input (engine) and post output (axles, wheels, tires), meaning that with an engine transplant some things will change, but the overall efficiency will stay about the same. Why, because you didn’t change the gears, weight of said gears, oil, bearings, et-cetera, et-cetera, or anything else that transfers energy, within the transmission.

So, once you figure out how much energy is lost in the transmission (about 28hp in this case) that will stay (approximately) the same, assuming the same range of use (RPM doesn’t increase). If you increase the rpm on the input shaft (new engine has a higher redline) then you can expect the transmission (transmission really means transaxle including wheels and tires) to produce more heat (loss), the curve will be parabolic so I don’t know off hand by how much.

P.S. keep in mind that some people graduate at the top of their class and some at the bottom. It takes a wise man to know who to listen to. But that you have to figure out for yourself. Good luck.


Aaron

.

[The Dark Side of Will]

As gears turn, the teeth slide across each other a certain distance. This means that turning the gear from one tooth to the next takes a certain amount of work. As the trans is turned faster, the amount of work per second increases, which means that the transmission consumes more power at higher RPM.

Does the amount of work required to turn from one tooth to the next go up with torque transmitted? Yes. Does it go up linearly? In the absence of lubrication, it would. In the presence of lubrication from sophomore physics, it would as well. However, how does synchromesh actually behave? How to extreme pressure additives affect coefficient of friction relative to contact pressure?

Viscosity loss from the oil adds to high RPM parasitic loss.

I don't think it would be all that hard to adapt a V6/60 oil pump to actively lubricate a 282... it could even be driven off the outer end of the input shaft with only a new end cover being necessary to make (and a few minor oil system mods inside the trans).

I'm thinking of adding a transmission oil temp sender to my box when I get the Fiero back on the road. That way I can work Sinister's math backwards and estimate how much power my transmission is consuming by how much the temp rises during a 1/4 mile run. Comparing this to RWHP would allow extrapolation of fly wheel output and loss %.

[Pyrthian]

every transmission has parasitic loss. it is a fixed number.
gears dont get harder to turn if you turn them harder - they get harder to turn as you turn them faster.
a TH125 is going to suck out X hp's @ 4000 rpm
if you put this on a 90 hp duke, it sucks X hp's @ 4000 rpm
if you put this on a 200 hp dohc, it still sucks X hp's @ 4000 rpm
this would make for different percentages - a lower % for the higher HP motor.
now, for some more confusion - the parasitic loss increases with RPM's. the loss at 2000 rpm is lower than the loss at 5000 rpm. the 200 hp dohc can rev higher than the 90 hp duke. so, at 7000 rpm, which the dukie cant reach, the overall parasitic loss is pretty high. and, may even make the percentage the same, or maybe even higher.
so, to avoid confusion - dont use a percentage - use a fixed number for that trans at the RPM in question.

[The Dark Side of Will]

That's the discussion. I don't think it's a fixed loss amount OR a fixed percentage.
Also, the fixed loss magnitude idea does not match observed chassis dyno results for different engines.

Gear teeth slide across each other as the gear turns. The distance they slide is the same for EVERY tooth. The normal force at the contact area is proportional to the torque the gears are transmitting.

We thus have a F=Nu situation, where N is the normal force, u is the coefficient of friction, and F is the force required to slide the gear teeth across each other. The normal force increases with transmitted torque, so if u is constant (not a function of N), then F increases with transmitted torque.

Since we know that each pair of meshing teeth slides a specific distance d with frictional force F, we know that the work consumed by friction in this gear mesh is F*d for each meshing pair. That is work/energy on a tooth-by-tooth basis. Multiply by the tooth rate (number of teeth per second) and you'll get the power consumed by the gear mesh. Again, since F is related to transmitted torque, power consumed is related to transmitted torque.

Every variable in the above math is both known and calculable EXCEPT the coefficient of friction between the gear teeth. Even if it is vanishingly small, it is still there. The only way for loss not to increase with torque transmitted is for the lubricant to become more slippery as the contact pressure increases.

[Blue Shift]

I do know that on those high power blown big block multi engine powerboats, they use drive showers to keep the outdrives (which are essentially the transmission and work the same) cool under such a punishing load - 1000+HP per engine sometimes. The big block boats running a few hundred HP don't usually have them, though the outdrives are generally the same.

On an empirical level, I have reason to think that power in vs power lost isn't linear (is anything truly linear in the real world?) though I don't think it levels off at any point either - the more in, the more loss you get, though it changes as you overcome the limits of lubrication, fluid shear, differences in oiling at different RPM and load, changing tolerances under load, etc.