Any publication strap this car on a dyno yet?

[Yenadar]

Originally Posted by Dave-ROR

Wow, glad to see someone in here understands what's important

I love curves. In all their forms.

[fatoni]

Originally Posted by Yenadar

I'm only interested in the curve.


Not a momentary number under somewhat imprecisely logged conditions on a dyno with any possible unknown miscalibrations or correction factors that will just spark more debate anyway. Is it high? People will say the dyno reads high and others will say 'expected'. Is it low? People will say it reads low and others will say 'expected'.

I don't care one bit about that number. But when I hand it to Agile to tune, I will be comparing pre and post curves. THAT I care about.

Carrying only about the if probably as ignorant as carrying only about the numbers

[serialk11r]

Originally Posted by fatoni

Carrying only about the if probably as ignorant as carrying only about the numbers

What?

[NastyNate]

Remove the words carrying and insert caring.

[LSxJunkie]

Drivetrain loss is not a percentage, it's a fixed number. It takes a fixed amount of torque to turn the drivetrain. That does not increase as you make more power.

I can't wait to see a torque curve though.

[serialk11r]

Originally Posted by LSxJunkie

Drivetrain loss is not a percentage, it's a fixed number. It takes a fixed amount of torque to turn the drivetrain. That does not increase as you make more power.

I can't wait to see a torque curve though.

I think the friction between gears increases with higher forces. So if you increase the torque of the engine the loss increases. Not to mention the loss is different in each gear, as the differential sees higher forces in lower gear. Otherwise say a Ferrari which can lose almost 100Nm after the input shaft would always need >100Nm to turn the transmission, which I don't think is the case. Otherwise Ferraris would be getting more like 3mpg instead of 10 on the highway or whatever lol.

[Yenadar]

Drivetrain loss is neither a flat % nor a fixed number.

Rotational-only components before the transmission (such as flywheel, crankshaft) soak power, but only when RPMs change. They do not soak any power when the RPMs remain constant. Higher mass (weight) parts increase the power soak. This is one reason people go after lightweight flywheels. There is no difference between a lightweight and a heavy when you keep the RPMs the same, but the heavier one resists engine change more, which steals a bit from acceleration, but can also be of a benefit when you want to maintain momentum (typically not an issue for performance oriented driving, but a common factor for commuter cars)

Rotational-only components after the transmission (such as driveshaft, axles, wheels) soak power, but only when the vehicle speed changes. They do not soak any power when the vehicle speed remains constant. Higher mass (weight) parts increase the power soak. Carbon fiber driveshafts improve engine response, but are really tied to vehicle speed, not RPM, so they have a greater impact when the speed change desired is the greatest. Smaller vehicles of the same drivetrain configuration typically have less loss from this aspect since their driveshaft, axles, and wheels are typically smaller. It's the mass of the parts that matter though, not the size (different materials can make larger parts weigh less than smaller ones of a heavier material)

Mechanical linkage before the transmission (such as the e-shaft gear in a rotary, input gear on the transmission, alternator, pulleys) soak power at an increasing rate as RPM increases. This number can easily change for cars based on lubricating fluids and surface treatments. Race teams apply REM and WPC treatments to gear contact surfaces specifically to reduce drivetrain losses, and it works.

Mechanical linkage after the transmission (such as the output gears, diff gears, axle joints, wheel hubs) soak power at an increasing rate as vehicle speed increases. This number can easily change for cars based on lubricating fluids and surface treatments. Race teams apply REM and WPC treatments to gear contact surfaces specifically to reduce drivetrain losses, and it works. The frictional surface area matters here as well. For example larger wheel bearings generate less heat and drag than smaller ones (of the same material and velocity).

Fluid and air pumps before the transmission (oil pump, water pump, any vacuum drag from emissions components or the engine itself, power steering pump if you have a fluid system) soak power based on RPM at an exponential rate. The rate of loss increases dramatically as RPMs increase. The viscosity of the fluid and pump efficiency can make dramatic differences as well. For example, the RX-8 OEM water pump uses flat vanes on the impeller to push the water through, and a common upgrade is to go to a spiral impeller design, and this has been proven to increase measured WHP between 5-8whp. Obviously it doesn't do anything to the engine's output itself, it's just reducing some of the drivetrain loss.

Fluid pumps after the transmission (diff) soak power based on vehicle speed at an exponential rate. The rate of loss increases dramatically as the vehicle speed increases. Power losses are also impacted by fluid viscosity. The dyno difference of an RX-8 on the same dyno/same day between cold diff fluid and hot diff fluid is around 5-7whp.


Some components have more dynamic drivetrain loss explanations, such as the transmission fluid. Part of it's loss is based on RPM, part on vehicle speed. Some components also change based on ECU or triggers. For example the A/C compressor on an RX-8 soaks lots of power, until you pass a certain point in engine load, and then it internally shuts itself off on command from the ECU until engine load drops back under the threshhold.

There is also loss from the tires themselves, somewhat based on vehicle speed, but it's more dynamic than that, due to the change of shape a tire undergoes as the speeds increase, and every tire is different based on internal composition, sidewall flex, tread squirm, etc...

So in the end, it's always a moving target, and many people just try to give a % of what they think the loss rate is at, but even that isn't really that precise since it can change easily just by dyno'ing in a different gear (and thus having a different 'vehicle' (wheel) speed that changes the loss rates of the diff and tires) or with a different / non-stabilized fluid temps (repeated runs will heat up the fluid, reducing viscosity. One team just runs their RX-8 on a dyno until their transmission, oil, and diff fluid temps stabilize before bothering to record the result.)

If anyone is interested in the discussion, I can link in two dyno's from RX-8s where one is OEM components and one has REM WPC treatments on the gears, low viscosity fluids, etc... for curve comparison. The race dyno is notably higher across the board, but the curve itself doesn't drop off in high RPM at nearly the same rate as OEM. The owner of that team shares lots of his R&D, and is quite serious about reducing drivetrain loss (especially since the RX-8 has so much of it from the start)

Edit:
Added more examples and adjusting factors.

[Levi]

^Nice post!

[old greg]

Originally Posted by Yenadar

If anyone is interested in the discussion, I can link in two dyno's from RX-8s where one is OEM components and one has REM WPC treatments on the gears, low viscosity fluids, etc... for curve comparison.

Hell yes.

[Yenadar]

Before I post these, PLEASE UNDERSTAND THIS:
I am not posting for "Bragging rights" or "omg the rx-8 is slow" or whatever results. I am only posting these for comparison on what anti-friction coatings can do to reduce drivetrain loss when applied to the gears

I am also not posting them as embedded images to avoid people that ignore what I'm saying and see a dyno chart and immediate start with a knee-jerk response completely outside of the intent. So please refrain from comments that would turn this into a flame war in any direction.

This chart (http://www.rx8club.com/attachment.ph...0&d=1305560087) is one of the clearest OEM charts we have, and also a very good baseline to compare the REM/WPC curve to. The lower stock curve is in blue. The upper curve in red is with a quality rebuild and porting work. This comparison is important since the rebuild and porting work are roughly equivalent to what the race team does for their race engines, and can help you factor out what is just engine modifications.

This chart (http://www.rx8club.com/attachment.ph...7&d=1308147670) is from a race team's dyno (Owner's website: http://www.meyer-motorsports.com/Home/Home.html Sponsor driven website: http://www.xowii-racing.com/XOWii_Racing/Home/Home.html) with the REM / WPC treatments on the transmission and diff gearing.

The torque curve is significantly flatter throughout the entire range. The low torque increase is from their intake, exhaust, and tuning (mostly), but it holds that torque REALLY well, because they are able to mitigate so much drivetrain loss that normally increases as the RPM / speed increase. The team owner stated that most of the lack of drop-off from the 7k+ RPM torque curve is as a result of their REM / WPC treatment. He did not post a pre-treatment dyno chart for comparison however.

It should be noted that this team has unlimited access to a particular dyno shop (he doesn't share the name), so all of their dynos are on the same machine, and they have run literally thousands of dyno's on their car (with various engines). So they measure the hell out of any small change.

[serialk11r]

I think pump power increases as the cube of speed, not exponentially, small detail.

Also the losses to the fluids in the transmission/diff/whatever increase with speed, but the frictional losses (as a proportion of power/torque) should be torque dependent, not speed dependent. But I guess we are talking max torque curve so that's relatively constant.

I think I've seen people try just different viscosity fluids in the transmission and the differences in loss are pretty huge.

[fatoni]

Originally Posted by NastyNate

Remove the words carrying and insert caring.

yeah. stupid swype fails me again

[fatoni]

Originally Posted by serialk11r

I think pump power increases as the cube of speed, not exponentially, small detail.

Also the losses to the fluids in the transmission/diff/whatever increase with speed, but the frictional losses (as a proportion of power/torque) should be torque dependent, not speed dependent. But I guess we are talking max torque curve so that's relatively constant.

I think I've seen people try just different viscosity fluids in the transmission and the differences in loss are pretty huge.

increasing by the cube power is increasing something exponentially

[serialk11r]

Originally Posted by fatoni

increasing by the cube power is increasing something exponentially

No it's not