Are auto driveline losses a set % or a set HP ?

[bbiswar]

I've always believed that there was a '% rule of thumb' for driveline losses, yet I've been challenged lately and wanted to seee what others had to say.

For example, most V8 Mustangs are said to lose 15% of engine bhp through a manual transmission and up to 25% through an automatic at most power levels. Others think that the driveline losses 35-45 HP at ANY power level. 35-45hp may be 15% at a certain power level, but what happens if it's a 700hp race motor?

 

[Philrock]

Reality is somewhere in between. An approximation would be a constant amount plus a percentage of power being transmitted.

 

[evelrod]

A good friend owns and operates a chassis dyno service, "Evanspeed", in the Socal area. Most of the British car owners and racers use his service and this subject comes up regularly I suppose because we all want to know 'the big numbers'.
What we seem to agree on is that a figure of 19% seems to work pretty much across the board. In my case with our most recent run last year of 110.48 hp at the wheels translates to about 137 at the flywheel...That's pretty close to what my competitors claim, if they are being truthfull and close to what I see claimed by many of the Mini engine builders in the UK.

Rod

 

[GregLocock]

I'll concentrate on the diff. A similar discussion applies to a manual box.

At low torques the dominant loss is friction - from oil churning and the bearings.

There is also a fair amount of sliding velocity in the gears, but as the torque is low, this is a relatively small absolute loss, but it is speed dependent. The efficiency of a diff at say 5% of its nominal capacity might be as low as 30%

At about half of its capacity the teeeth mesh properly, so the sliding velocities are less. Typically the efficiency peaks at 95%, but it could be as low as 90%.

At the full rated capacity of the diff the distortion in everything forces the teeth out of perfect alignment again, so the sliding velocities are greater, and the efficiency drops, by say 5%

Now, the manufacturer can change the point at which optimal mesh occurs, but he's got two things to manage. Big losses at high torques will cook the oil, but big losses at low torques will stuff the cars fuel consumption in the EPA tests.

Neither is a good look.



Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[Philrock]

Greg,

How can sliding velocities depend on load? How can sliding velocities depend on anything other than rpm and gear geometry?

 

[GregLocock]

distortion in the casing etc.

Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[drwebb]

If the driveline losses were a straight number (say 45 HP) then wouldn't the engine have to produce >45 HP at idle to idle-roll without stalling? That seems like a lot.

 

[peterhill]

Real live but very old data. Overall trans efficiency for a 1L side valve, 33.7bhp@4600rpm, 7cwt van, direct top 4 speed, wide open throttle. As it's direct top, there is no meshing loss in the gearbox, only the diff. Both diff and gearbox have oil fling. Chance of gears being superfinished NIL.

10mph 95.25%
20mph 94.75%
30mph 94.05%
40mph 93.25%
50mph 92.15%
60mph 91.15%
70mph 90.1%
80mph 89% (won't go this fast on road only on trans dyno test)

Between 40mph and 80mph Efficiency % = 97.45 - (0.1055 x mph).

I have no data for what happens at speeds above 80mph, it could do anything (well this 1L van won't - it could only do 61.5mph when laden and 70mph unladen). It's unlikely but maybe the line can be extrapolated to infinity (well 0% effy).

Analyse the hell out it all you want.

I invite people to add similar data for modern vehicles to this discussion it will give bbiswar a basis for a better empirical formula than most motorsport/tuning dyno firms currently use. With a 1/2way decent formula they may even be able use the wheel power to cook up the same flywheel power graph in 3rd and 4th.