1 ... 8 9 10 11
Keith Tanner
Keith Tanner GRM+ Memberand MegaDork
10/23/19 3:41 p.m.

nderwatersaid:

Vigo said:

Just the thought that the losses scale linearly with added power is ridiculous. 

Totally agree. If between dynoruns I take off my air filter and muffler, I might see a 5% power bump. Has the friction in my driveline suddenly increased as well?

Drivetrain loss is part inertial, part friction. If you're doing power tests at a fixed wheel speed, you can take the inertial out of the equation. But if you're doing a sweep (ie, anyone who's not actually doing hard-core tuning), it's definitely a factor.

One of our dynos allows us to adjust the sweep time as it has a closed-loop resistance setting. We did a demonstration one time where we changed the sweep time between two runs and saw a significant change in measured power. The only difference was how fast we were spinning up the wheels and the mass of the driveline. Basically, we changed the inertial losses, and inertial losses are proportional to the power level of the car if you're running a fixed resistance (drum, eddy brake). It's possible that they're proportional to an exponent or something more complex, I've never really looked into it.

But there's also the fixed losses. So, awkwardly, drivetrain loss is a combination of a fixed amount and a percentage on most dynos. 

chaparral
chaparral Dork
10/23/19 4:17 p.m.

If you're losing 100 horsepower in a compact MR drivetrain, you'll need an transmission oil cooler the size of a midsize passenger car radiator to get rid of it. I don't think the C8 has that.

Dave M
Dave M Reader
10/23/19 5:49 p.m.

So to summarize:
Either GM snuck MT a ringer

OR

The MT Dyno results are miscalculated, miscalibrated or otherwise borked

But in no way is it possible that the LT in the C8 is actually putting out monster power with nary an upgrade over the C7.

 

Knurled.
Knurled. GRM+ Memberand MegaDork
10/23/19 7:18 p.m.
chaparral said:

If you're losing 100 horsepower in a compact MR drivetrain, you'll need an transmission oil cooler the size of a midsize passenger car radiator to get rid of it. I don't think the C8 has that.

...  Don't cars that make that kind of power, that are intended to be used at a high duty cycle of that power, in fact HAVE transmission/differential oil coolers the size of radiators?  (I'm thinking of a Civic half-width radiator as sufficient for a 100hp engine assuming "100hp" of power also lost to the radiator in a 33/33/33 work/heat/exhaust efficiency ratio because it makes math easy)

 

How big are the trans and diff coolers on a Camaro or Mustang with the track package?

 

Most vehicles get away with it by having enough thermal mass in the drivetrain to handle quick bursts.

dculberson
dculberson MegaDork
10/23/19 8:59 p.m.

I watched Matt Farah's take on it and I think this is going to be an incredible car. Yes he had some hesitations about the car on the street but I think they'll be addressed over time with better transmission programming. On track this thing was an absolute beast. $85k for the z51 fully kitted model; think they'll be $35k used in a few years like the c7? That would be absolutely nuts. 

classicJackets
classicJackets Dork
10/23/19 9:37 p.m.
Dave M said:

So to summarize:
Either GM snuck MT a ringer

OR

The MT Dyno results are miscalculated, miscalibrated or otherwise borked

But in no way is it possible that the LT in the C8 is actually putting out monster power with nary an upgrade over the C7.

 

My thinking is that if GM had slipped a ringer, the 0-60 times would also be significantly faster than what GM reported, right? Instead, they got exactly the number that GM published the production car would be at, on repeated attempts (IIRC). That much extra power would have to change that time, no?

Dave M
Dave M Reader
10/23/19 9:54 p.m.

In reply to classicJackets :

A very, very good point.

I guess we'll just have to wait for some other Dyno pulls!

Keith Tanner
Keith Tanner GRM+ Memberand MegaDork
10/23/19 10:05 p.m.
classicJackets said:
Dave M said:

So to summarize:
Either GM snuck MT a ringer

OR

The MT Dyno results are miscalculated, miscalibrated or otherwise borked

But in no way is it possible that the LT in the C8 is actually putting out monster power with nary an upgrade over the C7.

 

My thinking is that if GM had slipped a ringer, the 0-60 times would also be significantly faster than what GM reported, right? Instead, they got exactly the number that GM published the production car would be at, on repeated attempts (IIRC). That much extra power would have to change that time, no?

Jason Cammisa agrees. An acceleration run is basically a dyno run with a lot of uncontrolled variables.

The C8 Corvette Stingray does NOT make 650 hp! ⁣

A certain media outlet published a story today that they dyno'ed one and got more than 500 hp at the wheels... calculating back to 650 hp at the crank.⁣

This isn't possible. 650 hp from 6.2 liters of displacement can't happen with at only 6500 rpm without forced induction. ⁣

Horsepower is a function of torque x rpm... and there's a maximum torque-per-displacement (otherwise known as BMEP, or Brake Mean Effective Pressure) that's vastly exceeded by Motor Trend's dyno runs. ⁣

MT's 630 lb-ft number suggests the LT2 makes 17.2 bar BMEP. The most efficient naturally aspirated engines are around 13. The LT2 is SAE rated at 12.9.⁣

Did it have turbos on it that no one saw? I suspect not... but they sure didn't show up in acceleration testing...⁣

Here's the real reason I know the dyno results are invalid. See the screen-shot on image 2.

This is my acceleration simulator. My nerdiness knows no boundaries, but I've been using this calculator since the early 1990s. ⁣

I inputted a DynoJet plot from a 7-speed manual Z51 C7 I found on the Internet... 417 hp and 415 lb-ft at the wheels. Then I plugged in the C8's actual weight, gearing, tire size, and I estimated drag coefficient and frontal area.⁣

The numbers (in black, on the right) line up almost exactly with the testing results @roadandtrack got (in red) - within ~0.3s all the way to 150 mph.⁣

The most damning is top speed, which calculates to 182 mph. Chevrolet claims 184 mph for the Z51 C8. I'd expect a couple mph higher with the C8's LT2 engine (495 hp instead of the C7's 460 that I used here.)⁣

If the thing really made 650 hp, top speed would be vastly higher. Like, 200+ mph.⁣

MT's dyno piece doesn't pass science muster. I'm disappointed that they published it — something was clearly broken on that dyno. They should have brought it elsewhere and re-tested. ⁣

Or at least realized that the numbers didn't line up with the reality of their tests. ⁣

Gingerbeardman
Gingerbeardman Reader
11/4/19 3:35 p.m.

In reply to Dave M :

What the fine folks at the enthusiast rags have yet to understand is the the Tremec DCT is setup like the ZF or Graziano DCT...there is a drop-gear set that feeds the ring and pinion.

The gear ratios listed aren't the actual gear ratios, those drop-gears add a multiplicative effect. So when they are testing in whatever gear, without accounting for that multiplier effect, they are interpolating a higher wheel-horsepower number.

Most of the enthusiast dynos use a weird correction factor (not talking about SAE weather/density/air temp correction factors) that is based off a 1:1 output times the final drive ratio...i.e., 4th gear. It's the reason why Porsches and Ferraris will read high or low on the dyno, depending on what gear they measured in. Do a pull in an overdrive gear, they appear anemic. Do a pull in a 1:1 + gear and they appear overly optimistic.

I'll repost with the actual gear ratios and the drop-gear ratios when I find them in my stash of data.

So take this with a grain of salt because this is a leaked document, not an official GM spec-sheet, but since GM has released 3.62 final drive ratio for the base model C8 and 3.80 final drive ratio for the Z51, that would make the drop-gear set 1.35ish for the 3.62 to net 4.89 and the drop-gear set 1.36ish for the 3.80 to net 5.17.

 

Here you can see part of the drop-gear driving the ring and pinion on the right hand side in this cutaway:

Keith Tanner
Keith Tanner GRM+ Memberand MegaDork
11/4/19 3:53 p.m.

I've never come across this correction factor. Changing the sweep time (ie, running a different gear on an inertial dyno) will have an effect because that will affect the actual torque delivered to the wheels.

Don't assume 4th is a 1:1. That's true for a lot of 5 speeds, but not for a lot of 6 speeds (the T56 being a notable exception, with two overdrive gears). Heck, the ND Miata has a 1:1 in 6th.

Gingerbeardman
Gingerbeardman Reader
11/4/19 4:22 p.m.

There is an article by a dyno engineer (the guy who designed one of the first inertial roller dynos, not engine dynos) that explain it better than I can paraphrase. But I'm not here to argue dynos, so I'm not going to search it out right now.

I was referring to the "American" state of mind regarding manual transmissions...every American design 4,5,6 and 7 speed manual transmission installed in cars, trucks, etc. has 4th gear as a 1:1 ratio because the input shaft is locked to the mainshaft in 4th gear...making it a straight-through design. Muncie, Mopar, T10, Toploader, Saginaw, etc. etc. etc.

(The one caveat was the Doug Nash/Richmond non-O.D. 5 speed where 5th gear was the input locked to the mainshaft in straight-through output).

This was also true for 3 speeds (3rd gear was straight-through) and even the 4 speed with O.D. ala truck "toploaders" and the Mopar A833 OD...3rd and 4th gear levers at the case were flipped, so "3rd" gear was actually 4th and 3rd gear was an OD.

T45, T5, T56, T6060, T6070, etc. all have 4th gear as straight-through with add-on O.D. gears in 5th, 6th and 7th.

So in this regard, I'm not comparing American to European/Asian transmission design philosophy, I'm comparing what the original (aftermarket, i.e. Mustang) roller/hub dynos were designed/programmed towards...5.0 Mustangs, 350 Camaros/Firebirds/Corvettes, etc. That was where most of the aftermarket development was geared towards. 

In any case, ignoring a drop-gear set multiplicative factor will skew the results upward or downward because it's part of the overall gear compounding affect.

Keith Tanner
Keith Tanner GRM+ Memberand MegaDork
11/4/19 4:39 p.m.

1:1 is used because it has the lowest losses. Doesn't matter what gear it is numerically or where the gearbox came from. Saying that you use the 1:1 (if possible) is accurate, saying that it's 4th is irrelevant. The dyno has no idea how many gears there are other than the one it's in.

I didn't know that all US transmissions had 4th as the 1:1 due to design legacy. That's interesting.

You don't use the 1:1 gear on an ND Miata (or, I'm assuming, a Caterham which also has a 1:1 top gear) because wheel speeds get into unrealistic (and dangerous) ranges and the car will also throw an early rev limiter. These speed-dependent rev limiters can lead people to incorrect conclusions.

Even then, if you test in a different gear, you're still going to be in the same ballpark. Ending up 100 hp off is a sign that you've got something wrong.

I suspect the reasons that Ferraris read low on dynos is due to a little bit of optimism on the part of the factory :)

Gingerbeardman
Gingerbeardman Reader
11/4/19 4:45 p.m.
Keith Tanner said:

I've never come across this correction factor. Changing the sweep time (ie, running a different gear on an inertial dyno) will have an effect because that will affect the actual torque delivered to the wheels.

I don't follow (not a dyno guru or even a higher-level maths person)...an engine develops a set amount of torque based upon airflow (intake, heads, cam), rpm, fuel, timing and programming. So when GM says an LS3 develops 460 lb/ft of torque, SAE certified, it's based on the engine, not the gearing. 
Now I do understand that the GM ECU can use a feature called torque limiting, which uses programming to derate the output to soften the load on the drivetrain, but that's an engine being electronically held back, not the engine itself outputting less than it's individual rating.

I also understand a torque curve, that not all power is available at all RPMs. But regardless of whether the transmission is in 1st gear or 6th gear, it's outputting the same amount of torque at the same RPM...i.e. 4500 rpm. So, in my mind, if the dyno is using "good" math rather than fudged formulas, you can back out that torque number using the known gear ratios (trans X final drive) and tire diameter.

So please correct me if I'm not understanding, because I really would like to understand. I'm not being a smartass, I'm just confused.

Keith Tanner
Keith Tanner GRM+ Memberand MegaDork
11/4/19 4:55 p.m.

The SAE numbers quoted by GM are at the flywheel under fairly heavily controlled conditions.

We're measuring at the wheels, which is a different thing. There are a bunch of losses in the drivetrain, and those losses aren't a constant. They vary depending on how fast we're accelerating the drivetrain (inertial losses) and how much friction there is in the drivetrain which is more of a constant. The faster you accelerate the drivetrain, the higher the inertial losses and thus the lower the torque you're actually delivering. That's why you'll see a different output for a low gear (fast acceleration) and a high gear (slow acceleration). Think about how much effort it takes to spin up a flywheel quickly versus slowly - the transmission and wheels are that flywheel. Plus, of course, the actual flywheel :)

That's assuming an inertial type dyno where you basically do an acceleration run against a fixed load like a big drum or an eddy brake.

The other way to measure is at a fixed RPM which takes inertial losses out of the equation but involves unrealistic levels of engine loading and rapid engine temperature gain.

Gingerbeardman
Gingerbeardman Reader
11/4/19 4:57 p.m.
Keith Tanner said:

1:1 is used because it has the lowest losses. Doesn't matter what gear it is numerically or where the gearbox came from. Saying that you use the 1:1 (if possible) is accurate, saying that it's 4th is irrelevant. The dyno has no idea how many gears there are other than the one it's in.

 

*This I understand and it makes sense, though it's fundamentally flawed as gear multiplication does affect torque output...300 lb.ft X 3.00 (1st gear) X 3.00 (final gear)= 2700 lb.ft @the tire...300 lb. ft X 0.50 (6th gear) X 3.00 (final drive)= 450 lb.ft @the tire*

 

I didn't know that all US transmissions had 4th as the 1:1 due to design legacy. That's interesting.

 

*Probably due to legacy manufacturing equipment...that's why the LS1 has the same bore centers as the SBC*

 

You don't use the 1:1 gear on an ND Miata (or, I'm assuming, a Caterham which also has a 1:1 top gear) because wheel speeds get into unrealistic (and dangerous) ranges and the car will also throw an early rev limiter. These speed-dependent rev limiters can lead people to incorrect conclusions.

 

*This has always surprised me about dyno runs...in 4th gear most American enthusiast cars, the wheelspeed is well over 100mph! Being held onto a platform with ratchet straps!*

 

Even then, if you test in a different gear, you're still going to be in the same ballpark. Ending up 100 hp off is a sign that you've got something wrong.

*This...so much this*

 

I suspect the reasons that Ferraris read low on dynos is due to a little bit of optimism on the part of the factory :)

*On the older ones, I agree. But with the newer ones, you can't argue with their real-world performance AND say the numbers are fudged. I think it's because of the way the formulas on the inertial roller dynos are implemented...I don't think the difference between a .89 (6th gear on F430) and 1.11 (5th gear on F430) x 4.30 (final drive F430) pull should matter if the formula is based off real-world formulas i.e. torque@RPM, divided by gear ratios.*

 

Keith Tanner
Keith Tanner GRM+ Memberand MegaDork
11/4/19 5:02 p.m.

I have to leave to go to SEMA, but here's something to look at. Same car, same gear, back to back runs. The difference? We changed the sweep time. The difference you see is the difference in inertial losses. That's why you see different dyno numbers in different gears, because your inertial losses change.

https://flyinmiata.com/tech/dyno_runs/NC_sweep_times.pdf

Gingerbeardman
Gingerbeardman Reader
11/4/19 5:13 p.m.
Keith Tanner said:

The SAE numbers quoted by GM are at the flywheel under fairly heavily controlled conditions.

We're measuring at the wheels, which is a different thing. There are a bunch of losses in the drivetrain, and those losses aren't a constant. They vary depending on how fast we're accelerating the drivetrain (inertial losses) and how much friction there is in the drivetrain which is more of a constant. The faster you accelerate the drivetrain, the higher the inertial losses and thus the lower the torque you're actually delivering. That's why you'll see a different output for a low gear (fast acceleration) and a high gear (slow acceleration). Think about how much effort it takes to spin up a flywheel quickly versus slowly - the transmission and wheels are that flywheel. Plus, of course, the actual flywheel :)

That's assuming an inertial type dyno where you basically do an acceleration run against a fixed load like a big drum or an eddy brake.

The other way to measure is at a fixed RPM which takes inertial losses out of the equation but involves unrealistic levels of engine loading and rapid engine temperature gain.

Straight cut gears are rated @99% efficient, helical gears @98% efficient. Non-hypoid gears (ring and pinion on-center like 10" QC) are rated @98% efficient (before the QC gears...so 98% X 99%= 97% efficient). Hypoid gears vary depending on off-set (amount pinion shifts from center) from 90-95% efficient. This is according to AGMA (the American Gear Manufacturers Association). None of this accounts for windage losses, frictional losses, etc. Just the "dry" ratings for the gearsets.

Most of the drivetrain loss is due to inertia of the various components...driveshafts, CV joints, axles, bearings, wheels & tires (the worst offenders). 

So I understand that there will be variability in the real-world compared to laboratory "perfect" conditions, but again (to ME) if the formulas the roller/hub dynos used were based on real data and not fudged formulas, all of this could be backed out using data at hand...gear ratios, wheel/tire weights, tire diameters @speed, etc.

There is way too much variability amongst dynos to be anything more than bragging numbers. I forget which magazine published the article, but using the same vehicle on different dynos produced wildly varying dyno sheets (the only variables that changed were the barometrics/weather...the vehicle didn't change up any parts to reduce variables).

Gingerbeardman
Gingerbeardman Reader
11/4/19 5:16 p.m.

In reply to Keith Tanner :

Have fun @SEMA! I haven't been in the last 7 years.

Any chance you can do me a favor and check out some of the vendor booths for cool stuff? I'm thinking mostly drivetrain (transaxles, transfer cases, etc.) related, Weddle Industries, Fortin, Weismann, etc.? Most of the magazines don't cover that stuff, too niche.

Dave M
Dave M HalfDork
11/4/19 6:30 p.m.

Not on the Dyno topic, but the R&T PCOT lap time for the C8 was super fast!  Can't wait to see what the Z06 can do....

Snrub
Snrub HalfDork
11/5/19 12:46 p.m.

It was a smidgen faster than a 2020 911 Carrera S. I'm not sure if that qualifies as unexpected performance.

Dave M
Dave M HalfDork
11/5/19 12:54 p.m.

In reply to Snrub :

I mean, that Carrera S is, what, $120k as tested? And faster than a GT-R as well.

infinitenexus
infinitenexus Reader
11/5/19 12:55 p.m.

Carrera S also has an engine half the size.  Or less than half I think.  

 

Has anyone posted lap times from the C8 yet?  I'm curious how the GT500 and C8 lap times will compare.

06HHR
06HHR Dork
11/5/19 1:34 p.m.

In reply to infinitenexus :

Lap times at thunderhill are in the R&T PCOTY (Perfornance Car Of The Year) article.  Engine size has nothing to do with anything, Corvettes have always had big engines (even the 54 had a pretty large Six-cylinder) relative to other sports cars.  Porsches historically have had smaller engines relative to other sports cars.  Apples to Oranges for that one.  For what it's worth, the only cars in that test that outran the C8 on track were the Lamborghini Huracan and McLaren 600, try buying one of those for 70K..

Like it or not, the C8 is still a performance bargain compared to cars it can run with.  Wonder how fast the Z06 or ZR1 versions may be (or will ZR1 continue as a distinct model?)

 

Snrub
Snrub HalfDork
11/5/19 3:03 p.m.

I'm not saying it's bad, it's just not unexpected is all. For example around VIR C&D had a 2013 C7 do 2:53.8, whereas a 2012 Carrera S did 2:58.9.

Dave M
Dave M HalfDork
11/5/19 3:10 p.m.
Snrub said:

I'm not saying it's bad, it's just not unexpected is all. For example around VIR C&D had a 2013 C7 do 2:53.8, whereas a 2012 Carrera S did 2:58.9.

I stand corrected! I should have said, both the C8 and the 911CS are really, really fast!

1 ... 8 9 10 11
Our Preferred Partners
vMmqs6iQiPZrSYLYXRG2qxdhvBYQgvOrm6HzquyR2KhYxNwEs6Zb7QPaDkzy31yV