Project Sipster: Frugal Sipster (is what I was gonna call it)
Project Sipster: Sipster Indeed (is how it ended up)
by 2009 Dave Coleman with commentary from 2011 Dave Coleman
On February 20 (2009), TopGear.com America declared war on oil, poverty and slow cars, all at the same time. We naively proposed that we could build the car that no car company could manage to build themselves. Specifically, we proposed to marry timeless Italian style with thoroughly adequate performance (0-60 in 7 seconds) shocking fuel economy (70 mpg), and humble frugality (you can duplicate it for $7,000), and we decided to build it in 55 days.
We didn't exactly decide on 55 days. Word had come from the home office in New York that the projectwould finish, and Jared would be back in the office by April 1, no negotiations. Office politics trump science once again.
In retrospect, both symmetry and our own sanity would have been better served if we gave ourselves 77 days to pull this off, but no matter, one cannot change the challenge once one has made a declaration. It just isn't done.
In a perfect world, we would have started with a 1974 Volkswagen Scirocco. This paragon of low-rent sex appeal would easily swallow the turbodiesel Volkswagen engine we planned to use, and it's far more aerodynamic than the brick we finally settled on. We rejected that plan because we feared the bureaucratic hurdles involved in registering a car that had been converted from gas to Diesel. We have since been informed by countless readers who have easily registered Diesel-converted cars that we're complete idiots. That would explain a lot of other things too…
Having settled on a Volkswagen Rabbit instead, we really should have shopped for one that was cheaper (we paid $1,700, but should have paid closer to $0), hadn't been crashed, and that didn't have a fuel tank full of rust and french fry bits from its former life as a grease-powered hippy car. See idiot comment, above.
Maybe you have to own wrenches to know just how ridiculous this idea of putting a modern (2003) engine in an old (1981) car really was. Putting a Jetta TDI engine in a Rabbit is an egregious abuse of the word "put". You can put your hat on a rack, and you can put a head of lettuce in your grocery bag, but try putting an elephant in your trunk. Just as easy to say, but its a fair bit more complicated to do, and odds are pretty good something will go wrong when you actually try it.
Unconstrained by realism, we drove the car to Canada, where Cam Waugh, owner of CWS tuning, skillfully coaxed the engine under the hood, along with its turbocharger, computer, and hundreds of wires. Surprisingly things didn't really go wrong at first.
It ran, in fact, for a full 40 kilometers before we figured out the old fuel tank and new fuel gauge couldn't agree on what was in the tank. The gauge said half, but the dead Sipster on the side of the road clearly said empty. As soon as we made it to California, the car died again when our new fuel filter again filled up with rusty old french fries. Two weeks later, black smoke started belching out the tailpipe, accompanied by a distinct lack of quickness. Another fuel filter--this time a German one--and a new Diesel injection pump fixed it for a while, until, a few days later, it just refused to turn over.
It took most of a day to figure out that if you reached under the steering column, grabbed a certain wiring harness in just the right spot and gave it a squeeze, the car would start right up. At least for a while. Eventually even that little caress stopped working and we had to start the car by reaching under the hood and shorting the starter solenoid with an old scrap of speaker wire.
Each of these problems took precious hours--sometimes days--to figure out, so by the time we had nailed our 0-60 in 7 seconds goal and were ready to tackle the 70 mpg part, we had less than a week left.
There are two sides to improving fuel economy. You can make the engine more efficient, so it takes less fuel to make each horsepower, and you can make the car itself more efficient, so it takes less horsepower to make it move in the first place. We've already done everything we can to make our horses efficiently--our new TDI engine is incredibly frugal and the miraculous program we uploaded last week gained both power and fuel economy--now it's time to tune the car.
And that just about brings us here. El Mirage dry lake, Middle Of Nowhere, California. Over the past week we've been semi-blindly improving our Rabbit's aerodynamics, hoping to make the barn-shaped Rabbit slip more smoothly through the air. We've smoothed out the nose with a front bumper from a mid-'90s Golf, slapped on the side skirts and fender flares from a late '80s Volkswagen Cabriolet, installed smooth, flat wheels from an old Honda CR-X, built covers to completely hide the rear wheels from the air, covered the grille with aerodynamic cardboard (we were getting desperate), and made the entire underside of the car perfectly smooth with an undertray built from 1/4-inch plywood, sheet aluminum, and rubber floor mats. We recruited the help of John McNulty, our friendly local aerodynamicist, to help with the modifications, but even to an aerodynamicist, air is a tricky thing. When we made the smooth underbelly, for example, John could tell us that the bottom of the car is the single biggest opportunity for reduced drag, and that for optimum effect, our smooth belly should be between 5 and 6 inches off the ground, and the back of it should angle upward at precisely 5 degrees. What he couldn't tell us was if all that work was worth 1 mpg or 10.
The lakebed will tell us that. Miles of uninterrupted flatness in every direction is the best aerodynamic laboratory this side of a wind tunnel. On a calm day, we should be able to run back and forth across the lakebed measuring fuel economy without worrying about traffic, hills, or gusty winds. We can install our new aero bits one-by-one, test them all, and know exactly what we need to reach 70 mpg.
Unless, of course, it isn't a calm day. Wind makes it impossible to gather aerodynamic data, and as our frantic week of fabrication dragged on, the storm forecast to hit the desert Saturday afternoon loomed larger and larger with every setback. By the time we had our aero bits finished, it was Friday night, and the storm was set to hit Saturday afternoon.
Undaunted, we headed to the lakebed at midnight, slept on the ground, and here we are at dawn, ready for the science. Crisscrossing the lake to check for any unexpected bumps or hazards, the Sipster's massive torque and nimble handling prove irresistible. Just a squeeze of the right pedal and a flick of the steering wheel and the little blue box is sliding sideways across the desert at 80 mph. Fuel economy be damned, this is what we live for! Those H&R coil-overs we installed a few weeks back, yeah, they let us put our undertray exactly 5.5 inches off the ground, but really they make our ex hippy-hauler turn like a race car. And that efficient turbodiesel? Just feel the torque!
Uh… feel the torque? As suddenly as the fun began, it's over. Squeezing the go pedal isn't making any more go. Rolling to a stop, the engine dies, again, and a strange hissing noise is coming from the engine bay. All ears under the hood and the hissing is tracked to the brake master cylinder. Whaaa?
A phone call to Cam explains everything. The hissing is either a leaky brake booster, or a bad seal on the master cylinder letting air leak into the booster. That won't kill the car, but, since the booster and turbo share the same source of vacuum, it will make the car run erratically. Making it just die mid powerslide? That would be the g-forces stirring up the old french fries again.
Stupid hippies.
What's that, wind? Just a breeze at first, but within minutes its gusting to 25 mph. Our science just blew away, the Sipster is dead, and we need this baby running for tomorrow's mileage test. Then the driver's door blows open and slams into the front fender, bending the hinges so much the door won't close any more. Is this the part where the car catches fire and we get to go home?
Sadly, no. Abandoning the Sipster on the lakebed, we drive into Adelanto, the nearest town-shaped armpit, to get another cheap Chinese fuel filter and swing by the junkyard in search of a less-broken brake booster. We want one from a Cabriolet or a VW Pickup, according to Cam. When was the last time you saw a VW Pickup? Well, if you're looking for one, There are FOUR in Ecology Auto Wrecking in Adelanto, and SEVEN cabriolets.
One sleepless night later and the Sipster is purring like a kitten with tuberculosis. BBC America already booked a flight to ship their editor back to New York, leaving us about 3 hours from the time the car is fixed to the time he has to be on a plane.
Our mileage test just became pass/fail. We'll fill the tank, drive 70 miles, and fill it again. If it takes less than one gallon, we've broken the 70 mpg barrier.
Starting at our favorite gas station in Seal Beach, California (our favorite because they sell Diesel cheaper than regular unleaded) we'll drive down the freeway to Dana Point, California, exactly 35 miles away, turn around and come back. Getting good mileage will mean driving slowly, like 55, which is tough to do on a Southern California freeway where the average is more like 80. Trucks here still have to go 55, though, so we can duck in behind one and let them both push the air out of our way and take the blame for screwing with traffic.
About 5 miles into the drive we remember an important detail about truck drivers. They apparently don't work on Sunday.
It's ironic how squeezing the most out of every last drop of fuel--a selfless bit of environmentalism if there ever was one--means driving like an erratic, self-important douchebag. Our apologies to anyone within a mile of the Sipster who actually has somewhere to go. Keeping our fuel economy up doesn't only mean clogging the slow lane, it means engine braking at the first sign of a red light (even if it's half a mile away), accelerating like a grandma, coasting up hills, and--this is the only fun part--carrying absolutely as much speed as possible through corners so you don't have to accelerate on the other side.
As fuel economy tests go, this one was pretty chickenshit. We had planned to do a full-tank drive. We had planned to get a Prius and go nose-to-tail with it somewhere far away. We had all kinds of ideas for how to get a realistic picture of just how frugal the car was. But when the boss says the story is done, the story is done, so we did the shortest test we possibly could. Jared literally left for the airport directly from the gas station, then fell asleep in the terminal and missed his flight to New York. Hell of a way to celebrate our accomplishment.
Exactly 70 miles after starting, we clatter back to he same Diesel pump, swipe the card, and, with held breath, start pumping. The pump shuts off almost immediately. 0.14 gallons! No that's not right, that would be 500 mpg.
Diesel is like beer; pour it hastily and it's all foamy head. The foam was just working its way up the filler neck and shutting off the pump early. To get an accurate measurement, we have to turn away from the tantilizing numbers on the pump and stare down the Sipster's filler neck, dribbling fuel down the side of the neck to quash the froth. It's a painfully slow process, but finally the fuel reaches the top. One more drop, and it will start dribbling down the fender.
This is it. If the number is bigger than 1, all the bloody knuckles, all the sleepless nights, all the stress and favors and promises would be for nothing. Tempting as it is, procrastination is pointless. We turn around and there it is: Zero point eight three three gallons. Covering 70 miles with that little fuel works out to 84 mpg. We didn't just meet our spectacularly arbitrary goal, we crushed it. 84 mpg is nearly double the highway rating of a Prius. Ok, yes, a Prius would probably do pretty well if you drove it with all the hypermiling tricks we just tried over the last 70 miles, but let's be realistic. Prius drivers never drive like that.
Sometime next week we'll actually explain how to build a clattery, unreliable, silly looking, fast, nimble and incredibly fuel efficient Sipster of your own for only $7,000. And after that? You decide. Aside from replacing the fuel tank, what should the Sipster do next? Should we try to set a land speed record with it? Should we try to do the 12 Hours of Sebring without refueling? Should we fill the tank, drive as far as it will take us and then try to start a new life wherever we land? Hit the TopGear.com America forums and tell us. (Yeah, don't bother, the forums are gone.) If you have any really great ideas for it, though, Jared still owns the car, and he has proven to be rather suggestible...)
Behind the scenes:
The only plane I had that was capable of handling the extra weight and flying slowly enough for these shots was a Multiplex Easy Star, a really basic foam begginner's plane designed to be dead simple to fly and tough enough to withstand non-stop crashing. I had cobbled it together with a wildly overpowered no-name Chinese brushless motor, speed controller and LiPo battery, and had been flying it on an obsolete radio given to me by Doug Kott at Road & Track. The radio was so old and hopeless, it was going either to me or the garbage can.
Perfect, then, for our project.
The only way to get a clear shot of the ground without giving away the fact that we were filming from a toy plane was to mount the camera upside-down on the bottom of the plane. We were headed to El Mirage on a friday night, so that same day, on my lunch break, I took a razor blade to the bottom of the plane and cut a camera-sized chunk of foam out of the bottom. Then I glued big chunks of balsa to the sides of the plane to reinforce the area I had weakened with the hole. Since the nose had already broken off and been glued back on several times, reinforcement was critical.
Next problem, the plane had no landing gear, and the lowest point on the fuselage was the lens of an HD camera. Not cool. The quickie solution I came up with was to bend up some aluminum welding rod and some actual RC plane wheels and rubber band the contraption to the bottom of the plane.
The plane was completely untested when, at dawn the next day, I unfolded myself from the back of the Mazda5 I had used for a tent, plugged it in, and threw it.
The extra weight was noticeable, but not too bad, and the plane was still aerodynamically stable, but that was the extent of the good news. As you can hear in the raw video, below, the radio was glitching like crazy. Every time I touched the throttle the motor would surge spastically, and the servos kept twiching and throwing random inputs into the plane. I don't know if it was the cheap Chinese electronics, the dumpster diver radio, or the Predator drones circling silently overhead (they test them out of a nearby airport), but something was seriously trying to knock us out of the sky.
Even with a perfect plane, though, our plan was doomed. It turns out to be much more difficult than I imagined to take a camera in the sky and point it at a car when you can't see what's on the screen, the guy in the car is driving around at random with no particular plan, and your hands are frozen solid and too busy with the plane to talk to the driver on the walkie talkie that you forgot to bring anyway.
Watch the video below and you'll be shocked that any of this crap was useful.
The whole time we were trying to build this ridiculous car on an impossible schedule, we were also trying to film enough video footage to make these 5-minute clips for every installment. Thankfully, the editing for these videos was being farmed out to William Barber, a Canadian college student at the time, who took our hours of spastic footage and somehow stitched them together into something professional looking in a series of weekly all nighters.
OK, so, Jared had been filming all the video for this project on a cheap Chinese Aiptek camera. For 2009, it was an amazingly small, cheap HD camera. For our grand finale, we came up with the brilliant idea of strapping it to an RC plane of mine and getting some aerial footage of the car racing across the desert. The results, we figured, would be indistinguishable from anything we could have done in a rented helicopter, and it wouldn't cost us a penny.
The only plane I had that was capable of handling the extra weight and flying slowly enough for these shots was a Multiplex Easy Star, a really basic foam begginner's plane designed to be dead simple to fly and tough enough to withstand non-stop crashing. I had cobbled it together with a wildly overpowered no-name Chinese brushless motor, speed controller and LiPo battery, and had been flying it on an obsolete radio given to me by Doug Kott at Road & Track. The radio was so old and hopeless, it was going either to me or the garbage can. Perfect, then, for our project.
The only way to get a clear shot of the ground without giving away the fact that we were filming from a toy plane was to mount the camera upside-down on the bottom of the plane. We were headed to El Mirage on a friday night, so that same day, on my lunch break, I took a razor blade to the bottom of the plane and cut a camera-sized chunk of foam out of the bottom. Then I glued big chunks of balsa to the sides of the plane to reinforce the area I had weakened with the hole. Since the nose had already broken off and been glued back on several times, reinforcement was critical.
Next problem, the plane had no landing gear, and the lowest point on the fuselage was the lens of an HD camera. Not cool. The quickie solution I came up with was to bend up some aluminum welding rod and some actual RC plane wheels and rubber band the contraption to the bottom of the plane.
The extra weight was noticeable, but not too bad, and the plane was still aerodynamically stable, but that was the extent of the good news. As you can hear in the raw video, below, the radio was glitching like crazy. Every time I touched the throttle the motor would surge spastically, and the servos kept twiching and throwing random inputs into the plane. I don't know if it was the cheap Chinese electronics, the dumpster diver radio, or the Predator drones circling silently overhead (they test them out of a nearby airport), but something was seriously trying to knock us out of the sky.
Even with a perfect plane, though, our plan was doomed. It turns out to be much more difficult than I imagined to take a camera in the sky and point it at a car when you can't see what's on the screen, the guy in the car is driving around at random with no particular plan, and your hands are frozen solid and too busy with the plane to talk to the driver on the walkie talkie that you forgot to bring anyway.
Watch the video below and you'll be shocked that any of this crap was useful.
Not as easy as it looks, this sliding front-drive diesels across the desert. Getting it to slide is easy, but when you're trying to intersect a 50 mph camera car at a 90 degree angle and pitch it sideways just right so you match speeds, land in the frame, and don't crash into your own minivan, well, then it gets kinda tricky...
See that Mazda5? It was 3 days old when we took it to the desert, slept in it, and then drove around with the doors open shooting video of the Sipster doing big diesely powerslides in a sandstorm. A year later there was still a puff of dust every time you closed a door.
The exit from the lakebed is at the downwind side, and by the time we had the car running after the latest fuel filter failure and simultaneous brake booster leak, the air was completely opaque with dust picked up on the long blow across the desert. It took 30 minutes of wandering around in a tornado of swirling oatmeal before we found our way back to pavement.
After that crappy raw footage, the final result Wil managed to churn out between classes is shocking:
The original post:
No comments:
Post a Comment