Dyno testing and "drivetrain loss"
 

Hey I was reminded of this since there was that fake dyno chart in the leaked info...
I noticed that when people dyno cars (on youtube videos), they typically record data in the HIGHEST GEAR. Then they go, great, so drivetrain loss is 20%.
However I have never seen anyone mention the fact that testing like this maximizes rolling resistance...

I saw a Ferrari 458 (RWD) get "20% drivetrain loss" somewhere on the internet, which is really high. But a Ferrari 458 in 6th? gear hits over 200mph! The wheels are spinning much faster on the Ferrari when it is on the dyno compared to your typical car, so it's going to have a much greater amount of energy going towards rolling resistance.

AWD cars get hurt even more, because if all 4 wheels are spinning, the dyno measures the total rolling resistance of the car! If it's 2 wheel drive, only 2 wheels are spinning and only about half the rolling resistance is there. On that note, cars with heavy rear weight distribution bias and RWD get hurt even more because the drive wheels have more resistance.

Why does this matter? Well no one factors air resistance into dyno performance do they? So why should rolling resistance be? It is only a function of the speed of the vehicle, completely independent of the engine.

Anything besides a top speed run should be done in the gear closest to a 1.00 ratio, not the highest gear.

Final drive ratio...?
Wheel diameter?
Rev limit?
1.00 ratio is still a very tall gear ratio.

I understand there's limitations, but the point is that I have never seen ANYONE even mention that rolling resistance is part of the equation. The estimates that are like manual transmission 2WD 15% loss are definitely overestimated for this reason. At say 100mph, rolling resistance can be like 20hp, which is huge. The Ferrari example I gave, 200mph with rear weight bias, rolling resistance easily accounts for at least 30% of the "drivetrain loss".

Trans ratio.

Dynos are good for tuning, that's it so I've never bothered to think about it.

If you want crank HP data, put the engine in a calibrated and environmentally controlled engine dyno cell.

If you're trying to tune the car, or determine engine configuration deltas, chassis dynos are a tool in the toolbox to help...but you still need a competent operator [both dyno and car] to get useful data. Garbage in, garbage out. No chassis dyno runs should be done without as-extensive-as-you-can-afford datalogging equipment to ensure the engine is operating in exactly the same conditions between configurations.

Most production engines, assuming they're broken in reasonably well [aka not the asinine "BMW way"], will result in a relatively tight percentage ratio from SAE advertised power curve. From there, you can interpolate "driveline loss" however your philosophy desires. Conversely, with extensive enough datalogging, you can deduce a crankshaft torque curve from a known-quantity chassis dyno. This is tougher as there are many out-of-calibration chassis dynos out there, with some claiming to be "real" numbers when in fact they're wildly erroneous.

In the end, it's better to do all dyno-queen comparisons on the exact same dyno, on the same day, with the same operational parameters. If people want to compare **** sizes with cell phone pictures of screenshots from a random dyno day, they're not serious, so let them make up whatever they want for driveline loss to make their ego feel better.

They don't dyno in the highest gear for the whp #s. They only do that when they are doing a high speed run as mentioned. For a 5spd car it's usually in 4th gear on the dyno for the whp. 6spd cars usually are 5th gear dyno run. RWD cars on average are 20% loss through the drivetrain. FWD is like 15% and AWD is like 35-40%.

For the record, I use a Dyno Dynamics to get my numbers from.