Using science to invalidate a scientific test.

Everyone makes mistakes. Sometimes multiple people make mistakes, and that seems to be the case with a recent article by Motor Trend claiming that the pre-production 2020 Chevrolet Corvette Z51 the magazine dyno’d put 558 horsepower and 515 pound-feet of torque to the wheels. This, the magazine estimated, translates into a naturally aspirated 6.2-liter V8 making 656 hp and 606 lb-ft of torque at the crankshaft. Eyebrows have been raised considering how much higher those power figures are than Chevy’s own SAE-certified numbers of 495 hp and 470 lb-ft of torque.

My best guess is that the dyno operator made a mistake – then Motor Trend made another by taking the numbers at face value.

Without even knowing what went on behind the scenes, we know the dyno numbers are invalid for two reasons: first, the magazine’s own testing numbers delivered performance exactly to be expected from the engine’s SAE-certified 495 hp and 470 lb-ft of torque; and second, getting 650 hp and 600 lb-ft from this 6.2-liter V-8 is effectively impossible.

We’ll get to the latter part in a second.

2020 Chevrolet Corvette Stingray

First, let’s talk about Motor Trend’s testing results: 0-60 miles per hour in 2.8 seconds and a quarter mile time of 11.1 seconds at 123 mph. These are almost identical to Road and Track’s results of 2.8 seconds and 11.1 seconds at 121 mph (Editor’s Note: Camissa wrote Road and Track’s first drive of the C8 Corvette).

That 0-60 performance, especially, is spectacular – it’s nearly a second quicker than the old Stingray, which had a similar power-to-weight ratio to the new C8. The new car’s additional thrust comes from two things: gearing and grip. Second gear on the 2020 Corvette’s new eight-speed dual-clutch automatic is only a hair longer than first was on the old seven-speed manual, and the extra mass over the rear wheels gives the C8 the ability to more easily put its extra grunt to the ground.

The C8’s 2.8-second sprint to 60 mph, or 0.9 second quicker than that of its predecessor, was enough to trigger my skepticism, so I used my trusty, nerdy performance simulator to see what’s what. This software, which I wrote initially in 1993, takes a dyno plot and a bunch of other known metrics and calculates out theoretical acceleration and top-speed numbers.

2020 Chevrolet Corvette Stingray

Since I had no access to dyno plots of the new 2020 Corvette’s LT2 engine, I used the last-generation C7’s LT1. I found multiple DynoJet plots online and chose one that was an average performer at 417 hp and 415 lb-ft to the wheels. (The LT1 was rated at 455 hp and 460 lb-ft at the crankshaft.)

(A side note: Yes, using an LT1’s dyno plot instead of an LT2 does have an effect on the outcome, but if the real-world tests match the simulator’s estimated acceleration with a slightly-less-powerful engine, then a massively more powerful one would have produced far quicker results. If I adjust the LT1’s torque curve with an estimate to reflect the LT2’s additional peak torque and horsepower, the 60-mph time drops by just four hundredths of a second, from 2.96 seconds to 2.90 and the quarter-mile drops by a similarly small amount. So instead of 11.2 seconds at 122 mph, it’s 11.1 at 124.)

The C8’s 2.8-second sprint to 60 mph, or 0.9 second quicker than that of its predecessor, was enough to trigger my skepticism.

For the uninitiated, a dynamometer – dyno for short – is a device that measures an engine’s output. The type of dyno we’re discussing here is a chassis dyno, meaning it’s measuring the amount of power and torque delivered to the drive wheels. This will always be lower than what the engine produces because of parasitic losses that occur while power travels through the driveline. Typically, you can expect 10 to 20 percent less power delivered the wheels than is produced at the engine’s crank.

Plugging the LT1’s proven power output into the simulator using the C8’s gear ratios as supplied by Chevrolet (and verified in the real world using the car’s speedometer and tachometer), the mid-engine Corvette’s actual curb weight and tire size, and estimated drag coefficient and frontal area, resulted in the simulator spitting out acceleration numbers within a rounding error of both Motor Trend’s and Road and Track’s actual test numbers, all the way from 0-to-30 mph to 150 mph.

Acceleration Jason's Simulator Road and Track Motor Trend
0-30 MPH 1.1 seconds 1.1 seconds 1.0 seconds
0-40 MPH 1.7 seconds 1.5 seconds 1.5 seconds
0-50 MPH 2.2 seconds 2.1 seconds 2.1 seconds
0-60 MPH 3.0 seconds 2.8 seconds 2.8 seconds
0-70 MPH 3.8 seconds 3.7 seconds 3.7 seconds
0-80 MPH 4.7 seconds 4.7 seconds 4.6 seconds
0-90 MPH 6.0 seconds 5.9 seconds 5.8 seconds
0-100 MPH 7.3 seconds 7.2 seconds 7.1 seconds
0-110 MPH 8.8 seconds 8.8 seconds N/A
0-120 MPH 10.8 seconds 10.7 seconds N/A
0-130 MPH 13.1 seconds 13.2 seconds N/A
0-140 MPH 15.7 seconds 15.9 seconds N/A
0-150 MPH 19.0 seconds 19.4 seconds N/A
       
1/4-mile time 11.2 seconds @ 122 mph 11.2 seconds @ 122 mph 11.1 seconds @ 123 mph

The simulator also estimated a top speed of 182 mph. That’s just shy of Chevrolet’s 184-mph figure, and is exactly what’s to be expected with the LT1’s small horsepower deficit relative to the LT2.

In other words, the C8 Corvette’s real-world acceleration and top-speed numbers match those of my simulator’s almost perfectly. Running the simulator with Motor Trend’s result of 558 wheel-hp and 515 lb-ft of torque results in a quarter-mile performance of 10.3 seconds at 134 mph. This hard-to-believe estimate is the first reason to question the validity of Motor Trend’s dyno results.

And now the nail in the coffin: something called BMEP, or Brake Mean Effective Pressure. This is a measure of torque per unit displacement – think of it as the torque equivalent of specific power (also known as horsepower per liter). An engine’s specific power relies on two things: its ability to spin quickly and its ability to produce lots of torque at those high revs. Specific power output varies tremendously between engines.

BMEP, on the other hand, doesn’t change dramatically from one engine to another, at least where computer-optimized modern engines are concerned. It’s a pressure measurement. Here are a number of random examples of BMEP for naturally aspirated engines in the metric unit, bar:

Car Displacement Peak Torque BMEP
Toyota Corolla 1.8-liter I4 1,798 cc 126 lb-ft 12.0 bar
Dodge SRT 6.4-liter V8 6,417 cc 475 lb-ft 12.6 bar
Mazda MX-5 2.0-liter I4 1,998 cc 151 lb-ft 12.9 bar
Corvette C8 6.2-liter V8 6,162 cc 470 lb-ft 13.0 bar
Toyota Camry 3.5-liter V6 3,456 cc 267 lb-ft 13.2 bar
Ferrari 812 Superfast 6.5-liter V12 6,496 cc 529 lb-ft 13.9 bar

As you can see, the Corvette’s peak BMEP is right in the middle of a small window that ranges between 12.0 for the base Toyota Corolla and 13.9 for the most powerful naturally aspirated engine in the history of the automobile.

Motor Trend’s highest crank estimate of 640 lb-ft would put the Corvette’s peak BMEP at 17.7. That’s just not within the realm of possibilities for a naturally aspirated engine.

Turbochargers actually add BMEP. For example, the Mercedes-AMG C63’s 4.0-liter twin-turbocharged engine produces a peak BMEP of 20.5.

The dyno article goes to great lengths to speculate why and how the numbers were so high, but never considered the simplest explanation: a mistake may have been made.

Right off the bat, then, we know that anything approaching 600 lb-ft from a naturally aspirated 6.2-liter isn’t technically possible. Even if the single-cam, pushrod-actuated, two-valve-per-cylinder, 11.5:1-compression-ratio LT2 V8 could match the Ferrari 812’s DOHC, four-valve, 13.6:1 compression-ratio V12 for BMEP, the V8’s peak torque at the crank would be no more than 500 lb-ft.

Speculation aside, the dyno results are invalid on their face.

Motor Trend concludes its article with, “One thing we know for certain: the 2020 Chevrolet Corvette C8 provided for all our testing produces more power than what Chevy claims.” Actually, the only thing we know for certain is that the C8 could not, no matter what GM did to it, produce the power that MotorTrend’s dyno runs said it did. And Motor Trend’s own testing results prove it. The dyno article goes to great lengths to speculate why and how the numbers were so high, but never considered the simplest explanation: a mistake may have been made.

2020 Chevrolet Corvette Stingray

As for General Motors, Motor1.com asked the automaker for comment about this controversy and received the following reply: “We have spoken with Motor Trend about its dyno testing of the 2020 Corvette Stingray. We cannot be sure what procedures were followed during its testing, as we were not present during nor informed about the dyno test until after it was performed. The vehicle we provided was completely representative of the cars we will be providing to customers. GM follows a strict protocol on certifying engine power ratings, and we stand by our Society of Automotive Engineers’ certification of the output of the Stingray’s LT2 engine.”

Science is hard, and it’s easy to get excited about amazing numbers. The 2020 Corvette Stingray’s acceleration numbers are already amazing – there’s no reason to call into question Chevy’s official SAE-certified power ratings, at least not based on science.

 

Jason Cammisa is a freelance journalist and TV host with a monthly column in Road and Track Magazine. The former writer and host of Motor Trend’s YouTube shows is currently working on a video project with ISSIMI, a San Francisco Bay Area startup, and his Instagram is chock full of cars, everything from his short-form #InstaReviews to his adventures owning and working on a fleet of classic cars. Follow him at @jasoncammisa.

Gallery: 2020 Chevrolet Corvette Stingray