Test The Limits

CAFE: Pressure and Results

By Steve Remondi

November 18 2014

In a recent editorial titled “Get Real About MPG,” autonews.com (subscription required) suggests that

"...automakers should ratchet down the pressure on product teams to meet gaudy fuel economy targets. It's counterproductive if the effort results in trouble with the feds, costly reimbursement programs and unhappy customers.

We agree that OEMs’ mileage claims should better reflect performance in the real-world. But we disagree that the pressure should be reduced on engineering and design teams to reach lofty MPG targets – after all these targets are mandated by the government and demanded by consumers.

The problems have arisen due to antiquated certification procedures that don’t reflect the real world. This is then exacerbated by the fact that not all vehicle models or configurations are required to be tested but instead their fuel economy can be extrapolated from base model testing.  Not a very robust process for consumers that are making buying decisions based on the  fuel economy written on the sticker.

That being said, more accurate testing is not an easy to do.  One could imagine that the OEMs should just drive their cars on the road and measure fuel economy. However, this is not a repeatable process as each test is subject to varying weather (particularly wind)  and thus lots of tests would need to be done over many seasons.   Not a practical solution.

By leveraging simulation, computer modeling of detailed vehicles can be virtually driven in very realistic conditions.  This can be repeated and verified.  In the U.S., the EPA already allows heavy truck OEMs certify fuel economy based on accurate simulations. Why not passenger cars?

Most OEMs already rely on accurate simulations to make most design decisions about the car and how it performs.  Tying this digital environment to the certification process will help the industry more robustly and more cost effectively predict the fuel economy of their vehicles for their customers.