Would it be possible that it takes less power to keep an object such as transmission gears spinning at 10k rpm than 1k, seems fishy, should be opposite.

And if you have a 10,000hp engine running at 1k, and a 100hp engine running at 1k they should have the same amount of powerloss, not the same %

Some people use a fixed number
"my new tranny only takes 15hp to run - but my old one took 30hp"

They are talking about the losses only at peak power (say 5000RPM) They dont mean it takes the same power to turn it at 1500RPM.

The losses would also vary with vehicle speed and in different gears so usually the % number that people quote are for the 1:1 ratio in the transmission (where losses are lowest). In 1st gear things are obviously worse.

Torque convertors will rob power (but also build torque at low RPM!) and then the rear end and tires rob more power.

I think for a C4 tranny (stock converter) like mine people assume about 20% loss - although a 9 inch rear end is a little worse than an 8 inch its pretty close. For a stick (no torque convertor) I think most people assume about 15-20%

how can a transmission rob more power at 1k than 10k? increased speed should have increased friction, not decreased.

and it can't possibly be a percent, cause as stated a 1000hp @ 1k engine will have to do the same work as a 100hp @ 1k engine to spin the drivetrain. Therefore there is the same amount of loss, since the same amount of friction and rotational inertia is needed by the trans.

my argument is that neither can be correct, as they both relate to hp which is subjective to rpm, which the losses also must be, but are not by either theory.

the friction of the trans must be at least constant, so if you increase the rpm and keep the frictional torque loss the same hp will increase of course

quote:

---

Originally posted by Rustang:
Sorry Dubz, no disrespect meant. This string got my head spinning and look what happened.

Anyhow, here where I work we build diesel powered mechanically driven locomotives. We usually figure horsepower loss thru the drivetrain of 15-20%. This includes torque converter, transmission, and axles. It can be as bad as 25%-30%(using worm gears and/or chain drive). At 5% eff'y, if you drive a transmission with a 100hp motor, you would get 5hp loss thru the trans. A 1000hp motor should lose 50hp thru the trans. This could be verified if you had a way of measuring heat coming off the transmission, which should be 5hp or 50hp, respectively.

---

none taken, but at the same rpm there is the same amount of resistance to movement, so therefore there should be the same amound of power lost. May have to get my hands on a couple tranmissions and to some testing, but physics says it shouldn't matter what is driving it, it should loose the same power at the same rpm. My thought pattern says so anyway.

assuming 10% loss, say 2 engines make their peak hp at the same rpm of 10k, the ford engine makes 1000hp and the chev makes 10hp, at that instant in time, when both motors hit 10k, the 1000hp engine will be loosing 100hp or 52.52ftlbs to resistance on the movement of the engine, where as the chev will be having .5252 ft-lbs with the same trans. (and for the purpose of this writeup, the chev guy uses a quality trans, a ford unit)

my guess is that a concrete tq # would more acurately represnet the actual losses, as % tq does not work either, but a constant friction loss would create a constant tq loss.

Now once you start applying horsepower to it (remember, you have a motor driving thru the trans to a RESISTANCE at the other end of the transmission) you are putting more pressure and friction against the internal gear teeth sliding across each other, bearings, etc. proportional to the horsepower your motor's putting out.

Understand? Good because I just confused the sh** out of me

if there is increased pressure should there not be an increase in the frictional force (a tq force) that resists the movement?

concrete hp numbers are false

as a % loss of hp, tq loss is the same % loss. So it follows the torque curve of the motor infront of it...not sure this is at all acurate, but my guess is it would be closer, but still not correct.

a 500fwph engine would be 400whp by that theory, independant of rpm

Alex (or others), would you be able to get a dyno of an engine, and then a dyno of the same engine with the trans connected?

------------------
Alex Denysenko
Co-Administrator and Moderator

NHRA/IHRA/SRA member and licensed Superstock driver
MCA member# 53321
NHRA and IHRA SS/LA & SS/MA National Record Holder '00,'01,'02,'03,& '04
First SS/MA in the TENS!
IHRA division 5 Superstock Champion
Fleet of FoMoCo products including 88 ASC McLaren Mustang #28

The Barry of BarrysGrrl

quote:

---

Originally posted by Moneymaker:
E=Mc 2

---

Anyway...

To calculate friction, you need to know the nature of the two surfaces that are being rubbed together, and you need to know how hard they are being pressed together. (rub your hands together while pressing easy, then do it pressing hard)

The harder two things are pressed together, the more friction there will be. So, the more power you push through a transmission or rear end, the more of a frictional loss there will be. It's not a 'set' amount.

Good Luck!

------------------
Mike Burch
66 mustang real street
302 4-speed 289 heads
10.63 @ 129.3
http://www.geocities.com/carbedstangs/cmml_mburch.html
http://www.fortunecity.com/silverstone/healey/367
http://www.mustangworks.com/cgi-bin/moi-display.cgi?220

------------------
SCOOP

oddly obsessed with big scoops on little Mustangs

2.26 60'S
14.9 @ 90.86MPH

65 coupe,351w,C4,Big Boss 429 hood scoop,8" 3.40 TracLoc.

quote:

---

Originally posted by n2oMike:
To calculate friction, you need to know the nature of the two surfaces that are being rubbed together, and you need to know how hard they are being pressed together. (rub your hands together while pressing easy, then do it pressing hard)

The harder two things are pressed together, the more friction there will be. So, the more power you push through a transmission or rear end, the more of a frictional loss there will be. It's not a 'set' amount.

---

You are right at a given RPM its torque that is lost. why do we therefore express it in % hp terms? Because its easier. Take my car for example.

Makes about 200hp @ 4500RPM (expressed in another way it made 235ft-lbs of torque at 4500RPM)

we say it loses 20% through the tranny
so 160 makes it to the road. Fine.

But what just REALLY happened. The tranny really took away 50ftlbs of TORQUE at 4500 RPM
Regardless of what engine I have - my C4 tranny and 8inch rear end will ALWAYS rob 50ft-lbs at 4500RPM because thats how much effort (in torque) it takes to turn the tranmsission and rear end at 4500RPM

So why use %? Well what If I tune my engine to make 300hp at 5000RPM:

It will be making 315ft-lbs at 5000RPM.
Say the losses in the transmission are roughly linear so - instead of the 50ftlbs at 4500 we lose about 56ftlbs at 5000. To the road the tuned car puts about 315-55 = 260ftlbs expressed in power terms as 247 hp. It just lost 21% of its power.

The example shows that as you tune an engine you also increase its speed, since both relationships are roughly linear the 20% works well (within a % or two)

It WONT work If I add a much bigger engine with more cubes AND keep the little transmission.

In the case of a big block conversion that makes 350hp at 4500RPM I still lose 50ftlbs of torque at that speed with my little C4 tranny. the result (if you do the math) is about 306hp to the road. 14% loss. which is why we only say that the % system gives you a VERY ROUGH estimate because it really depends on the relative sizes of your engine and tranmission!

Blowers and nitrous etc.. will also add more torque at a given speed so they will also make transmission losses look lower in % terms unless you use a bigger transmission.

For factory cars (including modern little 4 cylinders that rev to the moon) the rough 15-20% number works well.

You might think that a modern revvy honda engine that goes to 9000RPM might lose a lot in transmission losses because of the high engine speed. You would be right if they were using a C4 tranmission and a big rear end, but with the sewing machine torque those engines put out the engineers are able to use a sewing machine gearbox with little gears and lower losses. So even those rice burners use 15-20% loss numbers. It works well so long as you dont go wild with a Motor thats too big for your transmission!

You cant expect a blown big block to run on a little transmission.

Your example of a 1000hp ford and 10hp chevy is not something that you will see in reality because the 10hp chevy would go into production with a tiny gearbox to handle the tiny torque, and the 1000hp ford would need a monster gearbox.

The rule of thumb works well so long as you dont try to look to deep, when you look deep you realize that its not scientific but relies upon rough approximations and assumptions.

The scientific answer to describe transmission losses would express lost TORQUE due to friction based on engine speed.

Each gear would have a curve, the 1:1 gear is the one we are interested in and would be the lowest curve (least losses).

I work in this stuff all day everyday and we rarely talk in HP terms. its nearly all Torque because the torque curve of a machine is its real output. the HP curve is a product of the torque and engine speed and is just for show. Take the torque curve and the gear ratios and you get a very useful plot which shows vehicle "thrust" Vs Vehicle ground speed. This is the plot used (in conjuction with vehicle weight) that determines acceleration, shift points and all kinds of useful stuff.

Torque is King, but dont be fooled into thinking high rev engines with low torque figures are useless. More rear wheel torque is simply a lower gear ratio away. Its REAR WHEEL TORQUE that moves your car not engine torque.

quote:

---

Originally posted by indyphil:
Dubz,
But what just REALLY happened. The tranny really took away 50ftlbs of TORQUE at 4500 RPM
Regardless of what engine I have - my C4 tranny and 8inch rear end will ALWAYS rob 50ft-lbs at 4500RPM because thats how much effort (in torque) it takes to turn the tranmsission and rear end at 4500RPM

---

Not so.

The trans are rear end will absorb different amounts of torque depending upon their load. In other words, the parasitic loss will be less with the tires 'free wheeling' than with a load.

When under load, the gears and bearings are meshed together with a LOT more force, so the friction is much greater... The parasitic torque loss is more than simply RPM dependent. It will not always be 50ft-lbs at 4500rpm.

A 5hp Briggs engine could spin a tractor trailer transmission and rearend with the wheels off the ground.... but the parasitic drag of those units under real operating conditions (under a load) would be far greater than than the output of the Briggs engine.

Good Luck!

------------------
Mike Burch
66 mustang real street
302 4-speed 289 heads
10.63 @ 129.3
http://www.geocities.com/carbedstangs/cmml_mburch.html
http://www.fortunecity.com/silverstone/healey/367
http://www.mustangworks.com/cgi-bin/moi-display.cgi?220

quote:

---

Originally posted by n2oMike:


A 5hp Briggs engine could spin a tractor trailer transmission and rearend with the wheels off the ground.... but the parasitic drag of those units under real operating conditions (under a load) would be far greater than than the output of the Briggs engine.

---

Another good example, Mike! This is a very good (spell that 'educational') thread!

quote:

---

Originally posted by n2oMike:
Not so.

The trans are rear end will absorb different amounts of torque depending upon their load. In other words, the parasitic loss will be less with the tires 'free wheeling' than with a load.

---

I get that. So its not the same freewheeling, but its damn close when under a variety of loads. The reason is that the oil film on the gears and bearings should be supporting the load. Friction in this case has more to do with oil viscosity and realtive velocity of the moving parts and less to do with load. If it was very sensitive to load it would wear out too fast. Thats what makes bearings and oil so great, the real load is in the heating of the oil between surfaces not the grinding away of metal to metal contact.

You could put a spectacular load on most of the gears and the friction wont go up much until the oil film breaks down and the metal touched metal. So maybe under load my 50ftlbs will rise or fall a little but is most closely related to tranny oil viscosity. Cold tranny oil will add some losses but otherwise it will be about constant.

Engine bearings are very similar in this repsect. In my engine models bearing friction is modelled purely as a function of engine speed - cylinder pressure (load) is not even a function in the calculations.

quote:

---

Originally posted by n2oMike:

The parasitic torque loss is more than simply RPM dependent.

---

Thats the heart of the debate here. Im saying that for all practicle purposes the parasitic torque loss IS simply RPM dependant. I shouldnt have said ALWAYS! Load is a factor but I dont think it has a very strong effect.

Good Luck!

So here goes:
The torque converter is important, and I realized that it does not comply with my constant torque loss statement for a given RPM. Bigger engines applying more load will lose more torque through the TC at a given RPM. Its part of the design of a TC explains why the "stall speed" of a TC depends on the size of the engine. It also explains why the briggs and stratton engine wont work through a big tranny unless the wheels are in the air.

The TC is not the whole story. Heres where it gets hairy:

Most people think of friction in the simple case of a block sliding along a smooth surface (or hands rubbing together). This can be shown mathematically where Friction = N x C where N is the normal contact force (the perpendicular LOAD) and C is the friction coefficient. Clearly here friction is proportional to load. The harder the two surfaces push together the more the friction.

But something fancy happens when we lubricate the surfaces. Friction now becomes a fluid problem. Friction is now a function of the relative velocity of the objects, the lubricants viscosity, and the FILM THICKNESS. I cant remember the exact equation but it does not include load or normal contact force directly. interestingly if things are not moving the object is "suspended" and can be moved easily (like blowing on an air hockey puck to make it move) this is because the realtive velocity component of the equation is zero and therefore at rest the friction is zero. In practise it is however hard to maintain film thickness without the said relative velocity but im getting ahead of myself

The transition between the two regimes of friction can be sudden - like when you car starts "aquaplaning" on water. Water makes a poor lubricant generally because although it has low viscosity, it does not generally support a good film. Which takes me to FILM THICKNESS. Film thickness is itself a function of many things, including relative velocity (why slowing down prevents aquaplaning) and also LOAD and viscosity. The equation for friction is now super complex and involves fluid boundary layers and so on. But you have to trust me that although perpendicular load is important in determining film thickness it is NOT a dominant factor in determining friction in cases where the two objects are separated by a fluid film.

Case in point. A thin film of cooking oil on my kitchen floor will cause a 250 lb man to slip almost as easily as 125 lb woman. Both fall on their ass. The load Vs friction relationship is no longer linear, or even close to it. This also explains why 1000's of Pounds of pressure in an engines cylinder during firing - over an area as big as the piston, can be supported by a small main bearing area with only 60psi of oil pressure in a thin film. The OIL transfers the load to the main caps.

In this way the bearings and gears in the remainder of the transmission and rear end DO exhibit ROUGHLY constant torque loss at a given RPM regardless of load, assuming the gears stay well oiled. And it means if that briggs and stratton motor had an appropriate (SMALL!) torque converter coupled to that huge tranmssion it could pull a semi - just very slowly and requiring a VERY low gear.

What it really means in the context of the %hp loss through a tranmission is that the big block example I gave earlier would probably be in the 16 or 17% range not 14% because of the torque converter. It kind of splits the difference between my first argument and Mike Burches point about load having an impact. It means that the %hp rule of thumb is still un-scientific but works pretty well.

When things get extreme the film thickness (say on the rear end gears) may get thin, and with lots of speed, and load the oil will get hot. this causes further loss of film thickness and if your not careful you lose the film altogether and the rear end will wear out quickly. This is why F350s need big rear ends - for towing. And why hot oil is bad news. When things are this extreme the friction at a given RPM IS NOT THE SAME and has risen. You are now expending energy to wear out your gears as well as heat the oil - your not "aqau-planing" any more

Many big thick books have been written about oil films and bearings and gears. The subject is called Tribology. My engineering tutor was a tribology professor. Its really very boring. But now I can go to sleep.

im such a dork.

without exacting formula it looks like a mess to me. Gears are very hard work mathematically because the fluid film between them has to deal with the reversal of direction as the teeth mesh and unmesh. its super complicated.

What you have ended up with is a rule of thumb that is not based on formulas or direct relationships, but works pretty well. Best let sleeping dogs lie.