DIY Rubber Front Wheel Skirts

[newtonsfirstlaw]

I just finished the front skirts a couple hours ago. It took four hours. The first one took 2.5 hours, the last took 1.5 hours.



My main motivations for doing it this way:
1) Ease of construction - I know how to cut aluminium (shears), use a drill (to drill steel car body, you should spray some lubricant such as wd-40 on the drill bit before and sometimes during use) use a screwdriver, and cut rubber with boxcutters. Nothing fancy and you don't have to take the wheels off the car.

2) I was dubious about the elasticity of the rubber when faced with a wheel turning at full lock, either direction. For that reason, I wanted the rubber to be as long as possible, so the amount of stretch was minimized. Two reasons for this, one that the longer the rubber is, the less it will have to stretch per length of rubber. The other reason is that for small angles, sin ~= tan, i.e. adjacent side ~= hypotenuse, when the adjacent side is large wrt opposite side, the adjacent side is roughly the same as the hypotenuse, meaning that the rubber will hardly need to stretch.

3) I didn't want the rubber rubbing at all during highway speed, so I estimated the maximum angle of the steering wheel from center position I'd steer at on the highway, and doubled it (basically 180 degrees from center position of the steering wheel). That's what the aluminium is for, to push the rubber out.

4) I could either put the rubber on the inside or the outside of the aluminium. I decided to put it on the outside, figuring that the ugliness of all the washers and screws would be counterbalanced by not having the aluminium visible. Note that the aluminium has a nice smooth curve where it could rub on the rubber, to minimize abrasion. If I put it on the underside, I'd have to curve the metal slightly outwards, which would be ugly aerodynamically and aesthetically.

I figure that the increase in frontal area will be counterbalanced by the fact that the the rear is curved in a shallow way. The front also is a smooth continuation of the front bumper until parallel with the side of the car. Although of course I'd anticipate a reduction in Cd, even using the old frontal area figure, because air is not being sprayed out to the front and side in a turbulent fashion.

I will test drive it tomorrow some time hopefully. I tried it with one side, it seemed to work ok both at speed and cornering at low speed.

Basic steps involved:
1. Make a cardboard/card template for the front and the back. (figure out ahead of time where you will screw it in (should be screwed in both sides for rigidity). Should flare out more at the bottom, very little at the top. Size it by estimating how big it needs to clear the wheel at highway type turns + safety factor.
2. Trace onto aluminium sheet (0.6mm is what I used, that's fine).
3. Attach strip weather seal to aluminium where it contacts metal. Chances are you will scratch the paint a bit anyway, this minimizes scratching over time.
3. Drill holes then screw in to attach.
4. Stretch pond sheet rubber (or inner tube rubber) over the frame you have made up, duct tape in place.
5. Drill, screw in screws (you will need washers so that the rubber doesn't tear).
6. Cut rubber with box cutters, be careful not to cut the paint.
7. Take for a test drive.

[Lazarus]

Very nice job. Can't wait to see some FE numbers. What do you think you will end up with about 3% increase?

[MetroMPG]

Cool beans. I admire your "Just do it-" ness.

Looking forward to your driving impressions. Any plans for a coastdown A-B-A?

(Let's also hope some "regular" car forum types pick up on this so we can enjoy the apoplexy! )

[SVOboy]

Planning on a coast down test?

[newtonsfirstlaw]

Originally Posted by Lazarus

Very nice job.

Thanks! To you, and everyone else who commented.

Can't wait to see some FE numbers.

Me too. Won't be happening for a couple weeks (time off for Christmas), so no commuting. As it is, it takes 6+ business days to empty a tank, where it used to take 4 or less.

What do you think you will end up with about 3% increase?

Good question.

Let's talk about drag coefficient first. If we use the figures from the original Volvo paper, which basically formed the impetus for DIY aero FE modifications in the last couple years, drag contributed by the front wheels is 13.1%. Of what? Since the best case drag coefficient for close to the ground cars is around 0.11, and the drag contributed by the "exterior" is 31.7%, I'm estimating that the car they started out with had a Cd of 0.35. It seems a fair assumption, giving that cars are often around that - mine is 0.37.

13.1% of 0.35 is 0.0455.

Whether we can eliminate all that is questionable. I'm going to shoot for 10%, or 0.035.

If I did the coast down test properly, my drag coefficient was 0.26 (rounded up). My new Cd should be somewhere in the area of 0.26 - 0.035 = 0.225.

That's a 14% improvement, significant to my eyes.

On the highway (i.e. where I don't have to stop), I will see almost all of that gain, because I pulse and glide (proper glide most of the time) and air resistance is the dominant form of load. Where there are unavoidable stops, I will see less of that gain, because drag coefficient does not help with storing kinetic energy that is shunted into brake pads.

I'm estimating on my commute I will be 5% better off for having done this.

[newtonsfirstlaw]

Originally Posted by MetroMPG

Cool beans. I admire your "Just do it-" ness.

Thanks! The thing is, I've had the materials lying around for at least 2 months... I COULD have done it earlier. I wasn't sure exactly how to do it though as I tossed things around in my head, and kept putting off going to the hardware store to get the washers and screws.

Looking forward to your driving impressions.

So do I. They do increase the looks your car gets, I can tell you that now, just from the one test drive I had with one side done.

The turns aren't very noticeable resistance wise, no noise from them rubbing. I think having one side only made some oscillations that made the undertray more noisy. Will have to try with them both like that.

Any plans for a coastdown A-B-A?

I had a feeling you'd say that. I'll do another coastdown test, but I don't have the motivation to unscrew everything and screw it on again as an ABA test requires. That's something I'll do if/when I go into commercial modification manufacturing. At this point, it's basically a proof of concept that the sum of my vehicle modifications provide a tangible benefit to drag coefficient. And a cheaper way to get to work. And the feeling that I'm not wearing out my poor engine wasting fuel just to push air out of the way for nothing (other than to look like every other lemming out there driving a car standard as produced by a factory). And the knowledge that I've taken a $3500 AUD car and achieved a better Cd and CdA than pretty much any modern production car, with about $200 worth of materials.

[newtonsfirstlaw]

Well, I just went out and the Christmas traffic on the roads was too much to do a coast down test. I also forgot to pump up my tires before I went out. The other ones are probably ok, but there is one tire with a screw embedded in it that loses 20 psi in a week, and that certainly affects my gliding. It has been a week since I've inflated my tires.

I picked up some rubbing, and that was one of the aluminium fairings rubbing on the wheels. Other than that, no marked abrasions, turns well.

I'm undecided about the under car wobbling, whether it is increased or not. The undertray just does that sometimes. It does seem somewhat quieter though.

Maybe I'm imagining it, but it does seem to glide a little better.

I really need an altimeter so that I can find a good area to do coast down tests. My local highway really isn't the best.

I'm also a bit annoyed, since I'm ruining what has been a really good tank of fuel. I've already gotten 600km out of it. Oh well.

[newtonsfirstlaw]

I've just been thinking about the effect of altitude on coast down tests.

Consider that the gravitational potential energy = mgh, therefore with only 1 metre of elevation (not noticeable when evaluating an area for altitude), the GPE = 9.8 m, where m is total mass of car. (All units of energy are Joules)

The kinetic energy at 100kph is E = .5mv^2 = 385m. The KE at 90kph is 312m. So the difference is 53m. Just a 1 metre difference in altitude between start and finish is enough to throw your results off by 18.5%.

That's not even counting the effect that headwinds or tailwinds will have on your results, which will be significant. Maybe the terminal velocity on a known hill would be a better test? At least that way you take the altitude out of the equation, athough you will of course need a dumpy level to calculate your CdA.

I'm coming to the conclusion that to measure automobile CdA properly, the minimum tools required are at least a dumpy level or highly accurate altimeter.

[MetroMPG]

If you're trying to actually measure/calculate CD or CDA, it's important.

But if all you want to do is determine if a mod is effective or not, having a flat road or access to altimeter info isn't critical - the relative differences in the coasting times will do.

I realize you're after hard data - just pointing it out for others who may only be going after improvements.

[newtonsfirstlaw]

Update: Just did some more coast down testing and a couple hours worth of driving.

I guess I'll cut to the chase - using an average of all my figures, taking into account that I was going into a 13kph headwind (as measured by the bureau of meteorology), I get a drag coefficient of 0.19. (I was on the way home). I will probably need to do more testing, but I'll probably be conservative and say that it should at most have a drag coefficient of 0.24.

Regardless, that is definitely an improvement. The lower road noise is noticeable, and coasting seems more effortless than it was. Probably the acid test will be when I do some terminal velocity glides down some hills on my commute - typically I see 80-85kph going down them.

Someone yelled out "What's the go with the car, mate!?" at a set of traffic lights. I said (before doing the testing) "The drag coefficient is about 0.22". He probably didn't know what I meant. I guess I should have said "fuel economy!". Oh well.

I still have had no issues with rubbing.

And I think my solution to the front of the rubber/car join will be industrial strength UV proof clear tape.