RWHP to Crank HP reduction

[69znc]

I have been researching the percentage reduction for a manual transmission car between the flywheel/crank engine dyno HP vs the rwhp on a chassis dyno. I was only looking for empirically data and not "the rule of thumb". I never trust rules of thumb without some science! Knowing there lots of variables, starting with the engine dyno usually done in optimal conditions to every aspect of hp reduction points on the car. After many hours of searching the best I have seen with people trying to address this question is an 18.5% reduction. Most empirical comparisons were actually in the low 20%'s. Way more than the "rules of thumb". Any thoughts on this topic? What is your experience?

[Flash68]

My first comment would be you can't work with just one % without addressing manual or auto tranny. An automatic is going to eat up much more power than a manual.

There is no magic number I completely agree, but I have always operated under the range of 12-18% loss for manual and 20-25% for automatics.

[Che70velle]

No "real world" experience here, but I do have experience in the NASCAR world, although It won't apply to street cars, because we built everything to last one race, and it was all low to zero drag components. Numbers there are unbelievably low, but it only has to make it one race...
What drivetrain are you contemplating? Not only the gearbox, but the rear end also. The rear end makes a bigger difference than you'd think, GM vs. Ford, that is. Like you already stated, there are a lot of variables.

[69znc]

Not trying to compare auto vs manual. I was only trying to understand the urban legend of 10-15% lose for manual. Can not find any facts to back this up and I think the low 20's is more realistic. I cannot even find an empirically supported 12 - 18%. The best is 18.5 with a lot of transmission work and very light axles. All the rest are the 20's.....

[Ron Sutton]

We have had all of our race car engines on the dyno ... and then the whole car on a quality Superflow chassis/wheel dyno over the last 8 years.

This is around 35-40 cars & engines ... some with manual trans & 9" Fords & some with direct drive & quick change rear ends.

For all the "standard" drivetrains, we measured 18.1-18.8% difference (loss) from the engine dyno.

For cars we reduced the parasitic losses in ... with REM polished gears, ceramic bearings, no drag seals, lightened gears, etc ... we saw 16.4-16.8% difference (loss) from the engine dyno.

But we never lightened up parts that would have reduced the reliability, like driveshafts, u-joints, etc. More could be gained, but the risk wasn't worth it to us.

[Che70velle]

We lightened driveshafts, and lost power due to shaft flex causing harmonics.

[Flash68]

Not trying to compare auto vs manual. I was only trying to understand the urban legend of 10-15% lose for manual. Can not find any facts to back this up and I think the low 20's is more realistic. I cannot even find an empirically supported 12 - 18%. The best is 18.5 with a lot of transmission work and very light axles. All the rest are the 20's.....

Sorry, I missed that in your first post.

Just for a reference, when I was in the Porsche world for a couple years. Engine to chassis dyno conversions were routinely in the 12-15% range.

I have read about sub 10% losses for NASCAR... any truth to it, Che70velle?

[Ron in SoCal]

I have read there is also dimishing returns at higher horsepower engines. In other words, 15% (or so) at 550 - 650 hp, maybe a percent or two less over 750.

I have never seen proof of this theory though.

[Ron Sutton]

That's interesting Scott.

For our race teams, not finishing races due to parts failure was unacceptable.


So we went the other direction. Stronger driveshafts (3"), bigger u-joints (1350's not 1310's or 1330's), nothing "borderline".

[Ron Sutton]

I forgot ... we did run some powerglide automatics for 1 season & did chassis dyno test them.

These ran NO converter, nor clutch, so they were light & designed to lock the gearsets hard. With only 2 light gear sets & nothing else they used less power than typical automatics.

The numbers were 19.1-19.4% difference (loss) from the engine dyno.

But, I don't think we can compare these #'s to the automatics in street cars, even well built pro touring cars.

[Che70velle]

Sorry, I missed that in your first post.

Just for a reference, when I was in the Porsche world for a couple years. Engine to chassis dyno conversions were routinely in the 12-15% range.

I have read about sub 10% losses for NASCAR... any truth to it, Che70velle?

Actually yes, there are those types of numbers out there, but again, this technology does not apply AT ALL to anything that you'll see on here, which is why I spoke about "real world" experience. The money going into drivetrain testing is in the millions of dollars, in the NASCAR world.
I worked for a firm that built cup engines, but also did testing on things such as lightweight reciprocating assemblies vs. lap times, which would take months to complete, and cost several million dollars also...The results would surprise you. We were involved in early development of many things that most people would call BS on. Some stuff I can talk about, some stuff I can't.
The owner of this firm is a dear friend of mine. He was behind Cale Yarborough at the Daytona 500 in the late 70's, when Cale blew up. No big deal, until you find out he broke 8 rods...carbon fiber rods. This was Late 70's! Totally off topic, but true.

[Che70velle]

That's interesting Scott.

For our race teams, not finishing races due to parts failure was unacceptable.


So we went the other direction. Stronger driveshafts (3"), bigger u-joints (1350's not 1310's or 1330's), nothing "borderline".

We found this during testing on dynos, not on the track. Same philosophy here.

[Matt@BOS]

Sorry, I missed that in your first post.

Just for a reference, when I was in the Porsche world for a couple years. Engine to chassis dyno conversions were routinely in the 12-15% range.

I have read about sub 10% losses for NASCAR... any truth to it, Che70velle?

The Porsches all (almost all) have transaxles, which I think helps a little. I know my engine builder, who only deals in LS stuff always tells Corvette owners to factor out roughly 12 percent from their chassis dyno numbers, as compared to 15 percent for F-bodies. (all of that is to guestimate SAE net numbers)

Now, I'm also curious, are we trying to compare engine dyno numbers with no accessories to chassis dyno numbers?

[badmatt]

Actually yes, there are those types of numbers out there, but again, this technology does not apply AT ALL to anything that you'll see on here, which is why I spoke about "real world" experience. The money going into drivetrain testing is in the millions of dollars, in the NASCAR world.
I worked for a firm that built cup engines, but also did testing on things such as lightweight reciprocating assemblies vs. lap times, which would take months to complete, and cost several million dollars also...The results would surprise you. We were involved in early development of many things that most people would call BS on. Some stuff I can talk about, some stuff I can't.
The owner of this firm is a dear friend of mine. He was behind Cale Yarborough at the Daytona 500 in the late 70's, when Cale blew up. No big deal, until you find out he broke 8 rods...carbon fiber rods. This was Late 70's! Totally off topic, but true.

What CAN you talk about? me and Dave will blow up your PM box soon.

[BBC71Nova]

Peter, this is easy. Surely NRE did an engine dyno. So just strap your ride down to a chassis dyno and let us know the result :D .

Seriously though I've put some thought into this lately as well. Primarily because I can't figure out why the LS setups put down so much rwhp compared to some pretty strong big block combos. I would generally think flywheel hp was flywheel hp regardless of it being LS, BBC or even a Ferd FE :).

However, I then read a few examples where some solid 600+hp BBC builds were only getting high 400/low 500 rwhp numbers. That would support the closer to 20% drivetrain loss figure.

On the other hand, it seems fairly common/easy to get 500 rwhp out of an LS3. I know they are efficient and all but geez. Similar rwhp and 100 less cubes???? I also can't see a basic LS3 putting out 600 flywheel HP to start with.

That just doesn't seem to add up so I gotta think drivetrain differences maybe. Most certainly any BBC rwhp number you'll find is likely through an auto trans so that may explain some of it.

[DTM Racing]

Don't get to wrapped up on peak numbers. Average HP/TQ numbers over the operating RPM is much more important.

[V8TV]

We tested a 1970 Buick 455 on an engine dyno several years ago, and then dropped it in a 1970 GS ragtop with an M22 4-speed and chassis dyno tested it. If I recall, we lost 17.5% at the wheels.

[Ron Sutton]

We tested a 1970 Buick 455 on an engine dyno several years ago, and then dropped it in a 1970 GS ragtop with an M22 4-speed and chassis dyno tested it. If I recall, we lost 17.5% at the wheels.

Kevin, do you recall what rear end was in the car?

[Ron Sutton]

Don't get to wrapped up on peak numbers. Average HP/TQ numbers over the operating RPM is much more important.


I completely agree. We were looking for total power "under the curve" ... meaning within the usable rpm range that race car was running in. Peak numbers are just to brag about.

[69znc]

I agree the power curve is the holy grail! But I think the urban rumors about the 10% loss have a real stickiness with no empirical evidence. Ron your testing is exactly what I have found in all the research and studies I have found conducted by various magazines and by individuals posted in forums.

Specifically I was thinking of street legal cars so I think that Ron's reduction of 18% in a race car set up supports my belied 18.5 with PS, AC..... is realistic. I think the low 20's is more realistic for someone who is not overall anal about minimizing the reduction (sorry some self reflection there!)

Scott thank you for your insight. I was not trying to figure out what race teams and cars can do. Really just for the types of cars people on this site talk about and own. NASCAR, Indy, Grand Prix, NHRA are in a different world!!

[INTMD8]

It can't be a fixed percentage. It would be a fixed amount of HP loss plus a percentage variable.

For instance, double or triple the RWHP and you didn't double/triple the drivetrain loss.

What trans? Converter locked or unlocked? fluids up to temp? Same exact exhaust and induction system? What type of rearend? Do you have air in the tires? What is the gearing? Same rate of acceleration on both dyno's or is one inertia? Etc/etc...

I've seen some cars go from a manual trans to an auto with a very loose converter and "lose" 80 rwhp.

I'll just leave it as, too many variables. Depending on the car and the sum of it's parts there will be a huge spread in realized loss to the tires.

[Rybar]

I just quickly skimmed the thread, but did anyone mention what type of chassis dyno here? Mustang and Dynojets read differently.

I think theres about a 10% difference between the two. (Dynojet's reading higher that is)

[hp2]

I agree the power curve is the holy grail! But I think the urban rumors about the 10% loss have a real stickiness with no empirical evidence. Ron your testing is exactly what I have found in all the research and studies I have found conducted by various magazines and by individuals posted in forums.

Specifically I was thinking of street legal cars so I think that Ron's reduction of 18% in a race car set up supports my belied 18.5 with PS, AC..... is realistic. I think the low 20's is more realistic for someone who is not overall anal about minimizing the reduction (sorry some self reflection there!)



Depends on the crowd you hang with. I've heard 10% losses from some the same crew spouting off 3/4 race cam like its still relevant. Other guys who deal with more practical methods such as e.t. and m.p.h. have usually been in the 20% range because they see the results of their installations and are trying to figure out ways to minimize it.

[Ron Sutton]

I just quickly skimmed the thread, but did anyone mention what type of chassis dyno here? Mustang and Dynojets read differently.

I think theres about a 10% difference between the two. (Dynojet's reading higher that is)

We used a Superflow chassis dyno for all of our tests.

[Ron Sutton]

It can't be a fixed percentage. It would be a fixed amount of HP loss plus a percentage variable.

For instance, double or triple the RWHP and you didn't double/triple the drivetrain loss.


We did not find it to be a fixed amount of power. We found it to be a percentage. It varied "a little" ... but otherwise it was a similar percentage throughout the power range.

In fact, if you lay our engine dyno graph over our chassis dyno graph, in the same usable RPM range (3000-7000 for us), the graph looks the same ... just lower by 18.1-18.8%.



What trans? Converter locked or unlocked? fluids up to temp? Same exact exhaust and induction system? What type of rearend? Do you have air in the tires? What is the gearing? Same rate of acceleration on both dyno's or is one inertia? Etc/etc...

I've seen some cars go from a manual trans to an auto with a very loose converter and "lose" 80 rwhp.

I'll just leave it as, too many variables. Depending on the car and the sum of it's parts there will be a huge spread in realized loss to the tires.

I agree. It absolutely depends on what you do to the car, drivetrain & accessories.

Different transmission designs ... torque converter size, stall & weights (if running one) ... clutch size & weight (if running a clutch) and the ultra light option of running neither a clutch or torque converter in some cases ... will all play a role.

Pinion drop plays a role. 9" Fords use up more power than most rear ends due to a lower pinion, with more angle to the ring gear & more tooth contact (friction). That is also part of what makes the 9" strong for it's size.

Of course gear oils, bearings, no/low drag seals, micro-polishing gear surfaces ... or lack of ... play a role.

And lastly, but pretty obvious, is the weight of the entire rotating drivetrain assembly the engine has to turn.

We tested our cars just like we raced them & saw 18.1-18.8% less power than the engine dyno showed. We all know there are variations in dyno readings, but we still wanted to know with the best equipment we had available to us.

We started looking for ways to reduce the loses & unlock power ... first with our lower powered cars with spec, sealed engines. We spent a lot of money & time testing to free up speed in those cars. Angular contact bearings, ceramic ball bearings (EVERYWHERE), micro polished races, micro polished gears, no drag seals, lighter weight gear oils, etc ... dropped the loss by 2% +/- ... to 16.4-16.8%.

On track, the data acquisition showed the cars accelerated faster (predictable) ... but also carried more rolling speed after braking. :-)

The ceramic bearings cost a lot of $ ... and were worth every penny to us for our race team goals. Not only did they reduce rolling resistance & freed up power ... an added bonus was cooler running stuff (remember I said we put them everywhere) ... AND ... the bearings lasted 4-5x longer.

Later we did similar upgrades to our other race cars that had more power than they could use. Why? Because they carried more rolling speed after braking.

[69znc]

hp2 - yes I have "heard" many things but the point is not "heard" but empirical data. This I can believe in with many points of verification. Thus my post. No it does not matter the crowd you hang with if they are serious.



the dyno does matter thus the need for some structure in testing.

Ron you still have the most factual opinion with empirical data and test time.

Again I think for our cars featured here lose of low 20% is great!

[Ron Sutton]

69znc & hp2 ...

I think we're saying the same things ... 20% loss in a street car is to be expected ... give or take a little ... depending on auto or manual transmission & a few variables.

I think the newer cars in the last 10-15 years have accessories that draw less power & more efficient drivetrains. I suspect if we compared a newer Camaro, Mustang, Challenger or Corvette, we would see numbers under 20%.

Just my 2 cents.

[Greg from Aus]

Thanks for the great insight Ron. :thumbsup:

[Vegas69]

Since this is a pro touring website I'll chime in with my results from a pro touring car. :poke:

I had the engine dynoed and then a chassis dyno. It worked out to 16% hp loss and 15% torque loss. Big block chevy, tko 600, 3.5" aluminum driveshaft, and 12 bolt with 33 spline axles.

There is only one way to figure it out. Do as I just said...

[Ron Sutton]

Since this is a pro touring website I'll chime in with my results from a pro touring car. :poke:

I had the engine dynoed and then a chassis dyno. It worked out to 16% hp loss and 15% torque loss. Big block chevy, tko 600, 3.5" aluminum driveshaft, and 12 bolt with 33 spline axles.

There is only one way to figure it out. Do as I just said...

Hey Todd thanks for sharing.

Was the 15% torque loss at peak torque? and if yes, what rpm? Was the 16% HP loss at peak HP? and if yes, what rpm?

[hp2]

hp2 - yes I have "heard" many things but the point is not "heard" but empirical data. This I can believe in with many points of verification. Thus my post. No it does not matter the crowd you hang with if they are serious.


I guess I didn't get my point across adequetly. IMO, the guys with a fist full of timeslips or a log book of info have empirical data. These guys tend to speak from data driven results. Liek you said, these are the people who are serious about their performance and how to improve it.

The guy who swears his 396 Chevelle with dual quads put down 500 horses because it would burn rubber in all four gears is at the other end of the spectrum. In my experience, these two groups typically only tolerate each other for about 90 seconds at a time before hostilities begin. They don't hang together very often, so association with one group will never get you data. Association with the others will get you as much info as you are willing to digest.

[Vegas69]

Hey Todd thanks for sharing.

Was the 15% torque loss at peak torque? and if yes, what rpm? Was the 16% HP loss at peak HP? and if yes, what rpm?

Ron, I found the old thread. http://www.lateral-g.net/forums/showthread.php4?t=19857&highlight=payback+chassis

I was wrong, it was 17% HP loss and 15% torque loss at peak.

This was my 1st combo, not the ZL1.

[Ron Sutton]

Thanks for the update Todd.

Take care !

[69znc]

Since this is a pro touring website I'll chime in with my results from a pro touring car. :poke:

I had the engine dynoed and then a chassis dyno. It worked out to 16% hp loss and 15% torque loss. Big block chevy, tko 600, 3.5" aluminum driveshaft, and 12 bolt with 33 spline axles.

There is only one way to figure it out. Do as I just said...

Todd, that is impressive that your reduction was only 16% / 15%. Best that I have heard of and my research. Also curious to see the answers to Rons questions.