I saw a show on how motors are rated by SAE and it is no joke. The motors are set up on the stand EXACTLY like they will be in the car right down to all the accessories and entire exhaust systems. The SAE has a testing regiment that all motors, all manufacturers must go through before they can make claims, advertise, etc. Representatives oversee all aspects of the testing regiment. The feeling I got from the show I watched was that the SAE people were “The Man” and passing the test, getting good numbers, etc was very, very important. The engineers I the show certainly were suffering from some anxiety during the testing process.
I think in this day and age it’s fair to say that the manufacturers are not going to be able to make any outrageous claims and get away with it.
As for comparing crank hp/tq to rwhp hp/tq that will always be a very, very rough calculation. The variables in using the chassis dyno are potentially huge. I know from thousands of hours of dyno testing that it’s very, very hard to get repeatable runs. The smallest thing can affect the run and there are a lot of “smallest things” that can crop up. Then add to that all the parts from the flywheel to the tires and you have a whole new list of variables. I have seen tire pressure change hp readings 5%+/-. Another big factor is the temp of the fluids in the Trans and rear end. There is a big, big difference between hot and cold.
The claimed numbers from the manufacturers, because of oversight are very close to actual outputs. Some will be more, and some will be less due to tolerance stacking during assembly and inconsistencies of the weight of parts, etc. That’s why simply blueprinting a motor can be worth as much as 10% hp. The process of blueprinting eliminates or greatly reduces these variables.
Everyone should always take rwhp numbers with a grain of salt. There are a number of ways of skewing the results to justify claims. It’s easy. And even if someone is not trying to pull a “fast one” it’s still hard to get repeatable runs even for the best operators. Comparing two different cars on two different dyno’s is simply stupid. Comparing dyno runs with the same car and the same dyno but a couple months apart is just as stupid. For the best, and only nearly “scientific” results dyno runs need to be made back to back and if parts are being changed they need to be changed as fast as possible and all engine and drivetrain operating conditions must be identical. And lastly after a parts change a new map must be made to support the change. This is the most prevalent thing I see skipped in magazine tests and that is a big no, no.