I have studied all the free tests on this topic and one of the latests tests still say totally smooth is the best on that currenct tested body.
http://publications.lib.chalmers.se/...512/250512.pdf
- I say less you have the wheel open the better.
- Also some of those chalmers study tests use as a test wheel normal aluminum wheel, which is then covered with add on test covers. If you just put cover on top of normal wheel it will still create turbulence through the inner side of the wheel. Plus it will require force to push the air out of each pole on each revolution. --> not good.
- None of the new study test have tested to find out how the convex wheel cover effect.
- Best results is with steel wheel. Pole thickness minimal and also inner side of the wheel is almosts smooth.
- Aerodynamic convex wheelcover with minimal cooling holes on the outer edge of the cover. Like the XL1 covers.
My version looks like this:
That is best in most cases. Of course the results are not some in all vehicles and thats why its better to test it yourself. I have done also lot of road testing and if its windy I would say better not test at all as you cannot estimate wind direction effect to your results.
Only reliable way to test changes is ABA testing on the same day and route with as fast changes as possible. Car can take up to 100km to fully warm up in tire testing! So when you get the car warm changes should take the exact same time to do. If you let it cool 10 minutes more to other direction the temperature of the warmed parts is not anymore same.
I have custom ecu program which limits injection to certain level and you can measure the test distance and after that measure the avg speed of the test. This makes it possible to test also on hilly roads.