The Rabbit Sipster: Project Sipster part 8: All the Drag-Reducing Details

Project Sipster Part 8: All the Drag-Reducing Details

by Dave Coleman

Photography mostly by Jared Holstein

When we set out to save the world, we promised three 7's; so far we've only delivered two. With 0-60 in 7 seconds and 70 mpg in the bag, we still owe you an explanation of how to reproduce this fine driving machine for $7,000.

First the how-to part. A few weeks ago we explained, in excruciatingly greasy detail (maybe by TopGear standards, but it was pretty shallow by MotoIQ standards. Sorry 'bout that...), how to put a modern TDI engine in a crusty old Rabbit. Since then we figured out how to reduce the amount of work that engine has to do to motivate that sexy little box.

Weight Savings

If we make the car lighter, it will take less power to accelerate, and less power to climb hills, two of the most fuel-intensive things a car does. Over the course of this project, our readers kept suggesting we strip the interior to save weight. Clearly they've never been in a Rabbit before. The accommodations are so bare bones, there's nothing left to remove. No, our weight savings was accomplished up front, when we removed the heavy 2003 Jetta from around our engine and replaced it with a bantamweight Rabbit. We've done all we can do here.

Mechanical Friction

Friction is our next enemy. All of a car's moving bits take power to move, so we need to make them move more easily. The faster you spin the engine, for example, the more power you waste on friction between all the slidey bits inside. The tall gearing we installed lets the engine turn only 1800 rpm at 60 mph.

Disc brakes can drag, too, if the rotors are warped or the calipers are in anything but top shape, so the Sipster got nice, straight new rotors from a GTI, and we made sure the caliper pins were nice and greasy so the caliper could release without undue resistance.

Our rear drum brakes are adjustable, so we adjusted them slightly loose.

We had plans to get fancy with lubrication as well. RedLine oil sent us a few cases of gear oil and LSD friction modifier. Start with MTL, they suggested, and add a little LSD additive at a time. LSD additive is super-slippery, which is great for reducing drag, but sadly, synchros require friction to do their thing. By adding just a bit of friction modifier at a time, we could work our way up to the point where the synchros start grinding, then we could add a bit more straight MTL to dilute the oil back to the point where the synchros work again. This seemed like a great idea, but we never actually had time to do it.

Tire Rolling Resistance

Tires are a huge source of rolling friction, but good luck figuring out what to do about it. Tire companies are surprisingly reluctant to divulge any rolling resistance data, even if you beg. The only data we've been able to find is vague, old, and buried in a dry government study. Consumer Reports claims to be the only independent body testing rolling resistance, and they're so proud of that fact that they never actually publish the results. They just factor that data into their overall recommendation and keep the good stuff secret.

There is a movement afoot to standardize and publish rolling resistance data, but until then, the frugal-minded will have to rely on speculation and marketing, which is exactly what we did. Observe: when we were building this car, Yokohama was about to release their new dB Super E-Spec tire that they say has 20% lower rolling resistance than the original equipment Prius tire.

Marketing: "Yokohama says."

Speculation: "Prius tires must have low rolling resistance."

The study did give us something to get excited about, though. A 10% reduction in rolling reisitance, it said, was good for 1-2% improvement in fuel economy. 20%, then should give us as much as 3 mpg by itself. And that's compared to Prius tires, which are surely better than the ancient BFG Comp T/A's we found in the snow behind CWS Tuning. Excited, we called Yokohama and asked for a set.

"Sure!" they said, "just hop on a boat and go get them." The first set due in the country was halfway across the Pacific and wouldn't land unitil we were done. Instead, they offered a set of Avid Touring tires. They're not specifically low rolling resistance tires, but they are a brand-new design which should make them better than what we had. Or so we thought. Our testing actually showed them to be exactly the same.

That testing, by the way, consisted of cruise-controlling across a marked stretch of flat, mostly-windless road and using an aftermarket fuel economy gauge to calculate the average fuel economy over that run. A few back and forth runs with each configuration made the numbers seem believable.

Back in that dry, old government study, though, there was a bit about tires breaking in and how their rolling resistance could drop significantly over the first 4,000 miles. So maybe we just have to drive on the Yokos longer.

Our tests did show that pumping the tires up from 32 psi to the max pressure on the sidewall (44 psi on our old tires, 51 on the new ones) was good for just over 1 mpg on either tire. So at least we had that...

Our biggest challenge was to make the painfully square body slip through the air like a Prius. We had all kinds of good intentions to do some science here. Coastdown testing on a windless day could reveal some interesting truths about how much any of this nonsense did, but, as with so much else on this project, there simply wasn't time. Instead, with some consultation from John McNulty, our favorite aero professor, we did everything we could think of that could actually be executed by a couple of hacks working after hours with no budget and no skill.

Aero Wheels

One of the most aerodynamically nasty bits of any car is the wheels. They have a lumpy, complex shape full of holes, and they're spinning through the air stirring things up in a most unpleasant way. We first planned to solve this problem with spun-aluminum MOON discs, until we saw the price. Then we decided to try Pizza pans, but we found it surprisingly hard to find 15-inch pans to fit our wheels. Finally, we stumbled into a set of wheels from a 1984 Honda CRX. With the exception of a few vent holes, they're as flat and smooth as a MOON disc, and they're so horrifically ugly that they're free.

The Honda wheels bolt right up to a Rabbit, but only if you get longer lug bolts to account for the thicker aluminum wheel (H&R), and use bolts with a Japanese-style tapered seat, rather than the ball seat preferred by the Germans. Oh, and the Honda center bore is slightly undersized to clear the Volkswagen hub. Find someone who makes jewelry (Jared), hand them a die grinder, and tell them to make the hole a little bigger. After all that, they bolt right up...

For the mileage run, we used aluminum tape to cover the vent holes and lug holes, making the wheels perfectly flat-faced. Same theory as the grille here. We'll pull the tape when chasing Porsches and need some brake cooling.

Aero Front Bumper

The air's first encounter with a Rabbit is the blunt aluminum bumper, hung 6 inches out in front of the car. Air smacks into the bumper, swirls around into the huge gap between the bumper and the car, before finally smacking into the car itself. To modernize the shape, we simply grabbed a front bumper from a MK2 golf and stuck it over the original. The word "simply" ignores the fact that we had to push the stock bumper back 4 inches to make it fit (we just drilled the bumper shocks to drain the oil, then literally drove the car into a brick wall to compress them), and had to invent mounting points with a drill, sheet metal screws and chunks of pine. Yes, the pine that grows on trees.

The MK2 bumper presented new aerodynamic challenges. The MK1 rabbit takes its cooling air through the grille and through two air inlets curiously hidden behind the bumper. The MK2 Golf, though, took the majority of its cooling air from a large vent built into the lower half of the bumper. Stick a Rabbit behind that vent and suddenly its an aerodynamic dead end.

To smoothly push air around our new bumper, we covered over the vent with Monokote, a plastic film used by dorks to build model airplanes. The thin, lightweight film has a heat-activated adhesive built in, so if you simply stretch it across a big gap, like our vent, and run an iron around the periphery, it sticks. If you heat it a little more, the plastic also shrinks, so a heat gun magically pulls out any wrinkles and makes a drum-tight, surprisingly rigid membrane. Since it still probably isn't strong enough to withstand a 70-mph pebble, we filled the airspace between the Monokote and the bumper with expanding aerosol foam insulation from the hardware store. The aerosol foam expands roughly 50 percent, then air dries to a reasonable level of rigidity, which should back up our new skin with a little muscle to shrug off small road hazards.

This really seemed like a brilliant idea for about 4 hours. We filled the back of the bumper's new plastic skin, spray painted it to match the car, and in less than an hour the Sipster had a smooth new face. We went to bed with the smug satisfaction of a job well done.

The next morning, at the crack of noon, we awoke to find the Sipster glaring at us with a lumpy, wrinkly, still squishy face. The foam's need to air dry meant the backside of the foam hardened quickly, but the foam trapped up against the Monokote was sealed off, so it stayed wet and kept expanding. A 2-part epoxy foam that doesn't need air to cure would be a much better idea.

Aero Cardboard Grille

Above the bumper, the rabbit is still a jagged, square mess. The grille is straight upright. Any air that goes through it is forced through a torture chamber of radiators and greasy engine bits, and the air that goes around has to change directions 3 or 4 times just to find its way to the hood. None of this is good.

Our plan was to build an air deflector that would sit in front of the grille and headlights, smoothly shoving air up onto the hood. The center would be shaped from a block of rigid foam and covered with Monokote. The headlight covers would be made from 2-liter bottles, de-labeled and unwrapped to match the radius of our foam bit. It was beautiful in our minds.

The reality ended up being cardboard and duct tape, and the headlights just sat there exposed like little parachutes.

But isn't that radiator there for a reason? Indeed. That's why both grill blocks, the fantasy one and the cardboard one, were removable. Cruising around town or down the freeway, it takes very little power to move the Sipster (roughly 15 hp to cruise at 60 mph) so very little cooling air is required. Chasing down Porsches at the track or doing powerslides in the desert? Just toss the deflectror in the back seat.

Even at cruise, though, the radiator and intercooler need some air. The most efficient way to feed them is from the very leading edge of the car, so whatever air enters the radiator can go there directly, without changing direction on its way there. Two holes, then, in the front of the bumper. The driver's side blows on the radiator, the passenger's side blows on the intercooler. (If we were smart, we could have built a small gap into the bottom of our cardboard deflector and skipped the bumper drilling, but as you can see, there wasn't much smart going on here.)

Aero Tire Deflectors

If you're air, the only thing nastier than a grille is a spinning tire. The MK2 bumper was supposed to push air out around the front tires, but by the time we were done screwing it to the front fenders, it didn't stick out enough. Duct tape, some coroplast and a strip of aluminum tape solved this problem in about 5 minutes.

Aero Undertray

Ok, if there's one thing aerodynamically nastier than a spinning tire and nastier than a wheel. It's the entire underside of the car. If you're less ambitious or much smarter than us, you can make a good-sized dent in undercar drag by lowering the car and using an air dam to reduce the amount of air that goes under the car. It takes a lot of energy to push the air dam through the air, but not as much as it takes to drag that air across the bottom of the car.

By far the better solution, though, is to cover all that nastiness with a big, smooth sheet of aluminum. Only one problem: Have you looked at aluminum prices lately?

Ok, a big, smooth sheet of plywood. Plus some urethane deck sealer, just in case we hit a puddle some day. 

Making a big sheet of wood stick to the lumpy underside of the car without falling off, snagging on something or catching on fire is surprisingly hard. Here's how we did it:

Up front, we traced the shape of the bumper onto our sheet of wood, cut it out, and bolted the wood to the thin plastic lip around the bottom of the bumper. Lots of bolts…

The TDI oil pan hangs too low to cover with wood. Instead, we bought some rubber sheet, screwed it to the bumper wood, stretched it across the oil pan, the Eurosport lower control arm brace, and back to the underfloor tray. It's not perfectly flat, but it's a lot smoother than an exposed engine.

Under the floor, we had to space the tray down far enough to clear the exhaust system, brake lines, fuel lines and the floor's stiffening ribs. Once it was down far enough, there was nothing to attach it to. To solve this problem, we made long, L-shaped steel brackets that screwed to the pinch rail, giving us a flat surface about 3-inches below the rail that we could screw the wood to. To hold the center up, we wedged wood blocks between the tray and the floor's reinforcing rails, then screwed right through the wood and into the ribs.

Dropping the undertray low enough to clear the hangy down bits required us to sketch up some side rails that would screw to the pinch rails and hold the wood 3 inches below the rocker panel.

To make these, we simply went to a metal supply house that had an on-site shear and bending brake. It's only a few more bucks to cut your metal to size and throw a bend in it.

You can see the spacers in action here, as Jared spends his midnight screwing to the wood spacer blocks we have scattered around the floorpan.

This still put wood perilously close to the exhaust system, so we sandwiched some pink attic insulation under a sheet of aluminum and screwed it to the top of the wood, just under the exhaust. Even doing powerslides in the desert, nothing caught fire, so we're going to call this successful. Remember, though, the exhaust on gasoline engines is a few hundred degrees hotter, so if you try this on a gasser, don't blame us when it burns to the ground.

We may be dumb, but we're not stupid. That tube down the middle is where the hot stuff goes. Hot makes wood burn...

The hardware store version of undertray safety is an a aluminum heat shield (stainless steel works better to minimize heat transfer, but it cost too much) and some fiberglass insulation.

The end result actually worked (at least with our relatively-cool diesel exhaust) and the car has still failed to burst into flames.

At the back, John McNulty, our aerodynamic consultant advised us to bend the undertray upward at precisely 5 degrees. Unable to bend plywood, we splurged and bought aluminum for this part, letting the metal supply shop put the bend in for us. Since the thin aluminum bends more easily than the plywood, we attached some reinforcing ribs to the top of the sheet.

In a rare instance of measure twice, cut once, Jared cut this elaborate sheet to dodge all the suspension and tailpipe bits that simply couldn't fit above the tray.

This may well have been the best fitting car part Jared has ever fabricated.

Aero Hood Vent

Brilliant as it is, the undertray poses a problem. It may be too good. In sealing off the bottom of the car, we've sealed air into the engine bay. If we let air come in through the radiator, but don't let it out, the Sipster might inflate. Just to be safe, we cut an exit vent in the hood, just behind the radiator. This will upset airflow across the hood slightly, but that same air is about to be upset by the windshield anyway.

Its amazing to me that we so utterly failed to get a properly exposed, in-focus picture of our hood vents, but I think we cut these out literally minutes before our 80-mpg test run, and mere hours before Jared had to fly back to New York, so we were pretty frantic. Stand about 10 feet from your computer and squint and you might get the idea. Basically, we cut two H-shaped slits in the hood, grabbed onto the sheet metal with our bare fingers, and bent the back of the steel down into the engine bay, and the front of it up into a little Gurney flap that should ensure a low-pressure area right over the opening to vent engine heat.

Aero Protrusion Patrol

'70s cars are riddled with lumps, bumps, warts and protrusions (yes, it's a 1981 Rabbit, but it's still very much a '70s design). The marker lights on the rear fenders, for example, stick out almost an inch. We removed them and taped over the hole.

The mirrors are dead flat, like little air brakes. There's no law requiring a passenger's side mirror, so we removed ours. And there's no law requiring you to aim your mirrors properly, so we folded the driver's side mirror flush against the body and just turned our heads more when we drove.

The rain gutters on the A-pillar are deeply offensive to airflow, but also quite tricky to remove. The gutter isn't just there for rain, it's the seam where the roof panel is welded to the sides of the car. We planned to cut it off, a few inches at a time, weld up the resulting gap, grind the weld flat, smooth it out with bondo, sand it, primer it, and paint it to match the car. Does that seriously sound like something we could pull off? We hit the gutter with a hammer until it was as flush as possible, then smeared some silicone around to smooth over the bump. Good enough is just right.

Finally, the windshield wipers. Modern cars hide their wipers under the trailing edge of the hood. Ours just sat out there in the airflow mocking us. To minimize their impact, we removed the driver's side wiper and reset the wiper mechanism so the passenger's wiper stands straight up when parked and wipes enough of the driver's side to keep us from running over too many children.

Aero Body Kit
The Volkswagen Cabriolet used the Rabbit's shell well into the aero-conscious '80s, so Volkswagen saw fit to equip it with a body kit, consisting of smooth bumper covers, fender flares and side skirts that discourage air from slipping under the car and kick the air out around the tires. We used it to questionable effect.
Here you see John McNulty, an actual UCLA Aerodynamics professor, trying to make our car more aerodynamic by sticking parts of some sorority girl's old car to it.

Aero Rear Wheel Covers
Keeping the wheels hidden from the airflow worked on the original Honda Insight, not to mention Boss Hogg's Eldorado, so we fashioned wheel covers from Coroplast and screwed them to a steel support bar that was welded to a screen door hinge for easy tire changes.

This worked great in my mind, but in reality, the bottom of the Coroplast hung much farther away from the body than I anticipated, past even the Cabriolet body kit. Desperate for something to gently push air off the body and out to the wheel covers, we finally duct taped rusty paint roller tins in front of the rear wheels.
Now, About That $7,000 Part...
Claiming that a car as quick, economical and downright charming as the Sipster cost less than $7,000 has some people baffled. The cries of incredulity were many. What about labor costs? It's called doing things yourself, kids. We did lots of this work ourselves, and whatever work we farmed out, we only did because of our arbitrary and rediculous deadlines. Properly motivated, you CAN do this on budget.
Without labor costs, the accounting is simply a matter of adding up prices for everything necessary to hit the project's goals. So in parts only, our Sipster's triple-7 performance could be reproduced for $6993.21. Don't believe it? If not, then reading the chart below probably won't help.