5 ms·
Aerodynamic shapes may be well known, but can still be very dissimilar, and counter to what we perceive as "aerodynamic." Point in case: the Citroen AX has a d
by dognotdog 11y ago
Aerodynamic shapes may be well known, but can still be very dissimilar, and counter to what we perceive as "aerodynamic."
Point in case: the Citroen AX has a drag coefficient of 0.31, which apparently is the same as for a Porsche 997 Turbo. The AX looks like a box on wheels.
We aren't the slaves of the wind tunnel, as in that case we'd probably have some pretty mad shapes driving around, but slaves of wanting to look, well, the same as everyone else, to avoid any risk. But this is more a feature of entrenchment in the car industry than anything else. There is really no incentive for the big OEMs to do anything risky.
- sliverstorm 11y agoOk, to be more accurate a number of variables come together with drag. If you are buying a crossover, you're probably interested in a roomy interior, safety, a higher vantage point for visibility, a short hood for easy maneuvering and again visibility. Put all these things together, and the basic shape is that of an enlarged classic mini cooper. Then put this somewhat large, somewhat tall vehicle in a wind tunnel and sculpt it for efficiency, and you arrive at today's crossover. You're right that aerodynamics does not leave us with only one possible body. We still have sports cars, vans, sedans, etc, each with a basic shape dictated by basic goals. But drag sculpts the shapes that go into the wind tunnels in generally uniform ways, because the wind treats a Dodge the same as a Ford.
- tadfisher 11y ago> But drag sculpts the shapes that go into the wind tunnels in generally uniform ways, because the wind treats a Dodge the same as a Ford. As do safety regulations and testing bodies. Hence the same high belt and hoodline, same lack or rearward visibility, same super-thick A-pillars that trade visibility for roof strength (and in my opinion, this final design point ends more lives than it saves).