posted on April 03, 2010 10:00
Project Miatabusa - Part 1: What The Hell Are We Thinking
by Dave Coleman
With all due respect to the rest of the project car stable, this is destined to be the single most awesome project car in the short history of MotoIQ. The dirt cheap, bantamweight, nimble and perfectly balanced Miata needs no introduction in this crowd. Neither does the giant-slaying, brass-ball-swinging, 10,000-rpm-spinning king of the squidbikes, the Suzuki Hayabusa GSX1300R. What they both desperately need, though is an introduction to each other.
If we were to simply bolt a flywheel to the Hayabusa's crank (nothing simple about that, really) we'd end up with a situation like the dyno chart above. The Hayabusa's 1.3-liter engine is actually weaker than the Miata's 1.8 until you're past the Miata's redline.
If we take power off the Hayabusa's input shaft, though, the gear reduction between the crank and flywheel bumps up the torque and pulls down the rpm. Of course, this is flywheel rpm on the chart below, and since there's a gear between the crank and flywheel, the engine itself will still be spinning to 10,500 rpm. Oh the sweet music...
(The Hayabusa's dyno data, by the way, comes from Sport Rider Magazine's fantastic dyno chart database.)
In a way, these two machines couldn't be more different. One is terminally underpowered, lithe, nimble, cute as a button, and driven by someone so secure they don't care the rest of the world thinks they're a hairdresser. The other is only ridden by someone who does.
Somehow, though, their key elements of greatness dovetail perfectly. The Miata's racecar chassis is paired to a stout but anemic econobox engine, and the car's lightweight, "gram strategy" development mantra is seriously dented by the engine's bulky, cast-iron block. The Hayabusa engine, like most sportbikes, is like a little bundle of race engine technology in a convenient, bargain-priced, crash-recovery package. Unlike most sportbike engines, though, the Hayabusa's unusually large displacement makes it a relative torque monster. Its 99 lb-ft toruqe output is nearly a dead match for the original Miata's 100 lb-ft, but its 171 hp (or 197 if you drop the big bucks on a 2008 or newer engine) will give the Miata the longitudinal urgency it so desperately needs.
Even better, the 'Busa's little bundle of power is a bantamweight 135 lbs. That's less than half the weight of a Miata engine. With the proper application of weight weeniedom, we're confident we can drag the Miatabusa's curb weight down to 2,000 pounds. That's Lotus Elise power, Lotus Elise weight, and even a slightly Lotus Eliseish rear weight bias, all at much less than Lotus Elise prices.
The idea for this project was the inevitable result of sharing the pits with the nutjob creators of the MetroGnome and the Angry Hamster, the two bike-powered creations that dominate our Best Engineered Cars of Lemons series. The sight of the lowly, pedestrian Geo Metro, coupled with the spine quivering F-1 soundtrack of its CBR-1000R engine, is prescription for obsession. The thought of that sound coming from a Miata is hard to shake.
Alone, I would have the self control to keep the Hayabusa and Miata apart, but stick me with twisted geniuses like Alex Vendler (the Gnome's creator) and Tim Taylor (father of the Hamster) and I'm as suggestible as a sorority girl at spring break. Nerds gone wild.
Here's how it's gonna work:
The Hayabusa's monsterous 171-hp I-4 sits transversely under the rider's massive stones, sending power through a 6-speed dog box and on to a chain drive, like virtually every sport bike on the planet. This layout is far from ideal for a car. Most bikes carry tall first gears, since a motorcycle's ability to handle short first gears is limited by their tendency to flip over backward. The rest of the Hayabusa's gears are pretty tall too, since the bike is geared to go nearly 200 mph. Then there's the lack of a reverse gear, and of course, the fact that there is simply no practical way to connect a chain to the rear wheels of a Miata.
||Most FR Hayabusa installations are like this Starlet we spotted at a rally in New Zealand. The 'Busa's fragile transmission is retained (in the dirt this is probably strong enough), and the sprocket is replaced with a driveshaft. In our car, the engine will actually face the other way, with the timing chain against the firewall and the header on the right side.
The typical solution for front engine/rear-drive Hayabusa conversions is pretty simple. Replace the output sprocket on the transmission with a driveshaft flange, turn the engine north/south and you only need an extra-long driveshaft to make everything work. Unless you want to back up... Adding reverse means sticking a reversing gearbox (roughly $1500 from Quaife) in the middle of the driveshaft.
Endurance racing a bike-powered car gives Alex Vendler a uniquely experienced perspective on this peculiar form of hybrid, though, and it only took three or four destroyed gearboxes for him figure out that while their engines are surprisingly robust, sportbike gears simply aren't beefy enough for car-sized loads. Then, in the midst of one of his many mid-race transmission rebuilds, the answer hit him.
4-cylinder sportbikes all transfer power from the crank to the gearbox through a big reduction gear. That gear both reduces the rpm at the clutch by a factor of 1.596 and increases the torque by the same amount, making both much more appropriate for a car. While the Hayabusa redlines at 10,5000 rpm, its clutch never sees more than 6500. That makes the input shaft, not the crank, the ideal place pull power for a Miata transmission.
There's an amazing amount of stuff going on in a Hayabusa's crankcase. The silver can on the end (yellow arrow) of the beautiful, fully couterweighted forged crankshaft is simultanously the outer shell of the generator and the trigger wheel for the crank angle sensor. The big gear in the middle of the crankshaft (dark blue arrow) drives the engine's single balance shaft. The small gear all the way on the far end of the crank (light blue) is for the starter. The big gear we really care about, (orange arrow) is the primary drive gear, which drives the even bigger primary driven gear (the other orange arrow) that's wrapped around the clutch at a ratio of 1.596:1.
The green arrow is the transmission's input shaft. We'll be replacing this with a longer shaft (without any gears) that will extend out past where the Hayabusa's current clutch sits, and will hold a MIata flywheel and clutch. The purple arrow is the transmission's output shaft, which will just be removed.
So, if you just replace the bike's clutch and input shaft with a slightly longer shaft, you can bolt a Miata flywheel to the end of it. Then you can make an adaptor to bolt the Miata's bellhousing to the Hayabusa's clutch cover flange. Suddenly you have big, car-sized gears backing up your bike engine.
Comparing the SolidWorks model (above) with measurements we made by taping IKEA measuring tapes to the bottom of our $1500 Miata (right), revealed the Hayabusa sump and Miata steering rack really wanted to occupy the same space.
If "just" fabbing up this new shaft and bellhousing adaptor seem daunting, remember we're also sharing our pit with Tim Taylor. Go read about his Magna-powered Honda Z600 and all your doubts about this project's feasibility will vanish.
Work started with ones and zeros. Tim made a rough solid model of the adaptor in SolidWorks before we even had a Miata. This gave us enough information to start to approximate where big things, like the oil sump, were going to fit.
As soon as I saw the model, I found a clean, reasonably straight, 130,000-mile 1996 Miata with a factory Torsen LSD for sale for $1500 and pounced on it. Within a day the car was up in the air and we were comparing the model to the reality of where it would have to fit. This first approximation revealed a few very crucial issues. The Hayabusa's oil sump, for example, wanted to occupy the same space as the steering rack. The exhaust ports also dumped right into the frame rail, leaving no room for the factory header, and little space for a custom one. The offset engine ran right into the Miata's front subframe, and, finally, the lowest point on the sump, a V-shaped bulge on the bottom intended to keep the oil supply steady during wheelies, hung dangerously low.
|The oil pan is off in this photo, but you can still see how a portion of it sticks forward (to the right). That portion used to hit the steering rack, until we rotated the engne, removing the slant that it normally sits at when in the bike. The simple shaft on the right was used to align the Miata's input shaft with the Hayabusa's.
With no SoilidWorks model of the engine bay, it was hard to know how to resolve these problems. We would have to re-enter the real world. After removing the engine and front subframe, we stuffed a spare Hayabusa case into the engine bay. Tim whipped up an alignment shaft that indexed on the Miata's input shaft, and pressed into one of the Hayabusa's input shaft bearings. With that alignment set, we just jammed a half-inch spacer between the bellhousing and engine case to approximate the thickness of the as-yet-unbuilt adapter flange.
With these key dimensions set, the solution to 3/4 of our interference problems was as simple as rotating the engine until it was standing straight up. This lifted the offending portion of the pan off the steering rack, pulled the exhuast ports off the frame rail, and pulled the deep vee of the sump up flush with the bottom of the bellhousing.
Since this rotation approximates what the Hayabusa does during a mild wheelie, the pickup is positioned perfectly for this layout. And, since the oil level in the sump is limited by the the need to keep the crank out of the oil, raising the crank with this rotation also increases the engine's oil capacity, which will help keep temperatures down.
This rotation did nothing to fix the interference between the subframe and the engine case, though. There is simply no way to avoid making a new subframe. While we're in there, though, our new subframe will be lighter, stiffer, will have more negative camber adjustment range, and will use the higher caster geometry from the NB Miata. If you're gonna do it, do it right...
|That's the look of satisfaction and relief you get when a simple rotation of your Hayabusa engine makes it fit in your Miata...
One last detail you should know before the obsession sets in: we're not greedy. This concept is too good to keep for ourselves, so we're making tooling to reproduce the adaptor and subframe. Miatabusas for everyone! And I really shouldn't have to point this out, but just in case all this Miata talk is causing cranial misfire, the Hayabusa engine/Miata trans combo will bring featherweight horsepower to any front-engine, rear-drive car you can dream up. Toyota Corona, Lada Riva Estate, Borgward Isabella, the possibilities are endless. When there are eventually parts available, you'll find them at Miatabusa.com. Meanwhile, while there is only one, you'll find it here.
Part 2 is up, check it out:
Monday, March 29, 2010 11:06 PM
Very curious to see how this one turns out! I dreamed up a boosted ExigeBusa a couple years back with a 2-speed final drive. Like a low/high for 4-wheelers, to keep the revs decent while cruising on the highway. What can be more fun than a 2000lb, 500hp, turbo, 1.3L, 10k+ rpms, with 6-speed sequential gearbox?
The part I was worried about was the strength of the motorcycle tranny... so I'm very interested to see how your solution turns out!
Tuesday, March 30, 2010 4:42 AM
This is project makes me very excited. Will the lightweight Hayabusa engine negatively affect the balance of the car in any way?
Tuesday, March 30, 2010 6:59 AM
This sounds like the perfect solution for a Spitfire or a Morgan. I've also thought about doing something similar for an old CRX (1st gen), would the bellhousing trick work on an FWD transmission too?
Tuesday, March 30, 2010 7:29 AM
At a very high level, the FWD version should work, but having been very intimate with this project, I can already spot several problems. Older Honda engines (pre K-series) rotate the wrong direction so to get things to work, you'd have to stick the flywheel on the opposite side of the Hayabusa's case. The bearing on that end is too small, though, and the shaft would have to extend past the Hayabusa case more than 8 inches, so I don't think that option is mechanically feasible.
In addition, the offset between the crankshaft and the flywheel would put the engine back toward the firewall (good) where it would interfere with the axle (bad). You might be able to fix that with clocking, like we did with our steering rack issue, but its hard to say without trying.
Tuesday, March 30, 2010 7:35 AM
Miata4G63, this definitely will affect the balance of the car, and we plan to do some custom suspension tuning to make the car work right with its new, slightly rearward weight balance. We'll also probably be the first people in history to re-locate a Miata battery OUT OF the trunk! The end result should work great, though.
Tuesday, March 30, 2010 8:10 AM
Seriously bad ass project Dave. I'm looking forward to seeing how this one will turn out. What's the plan for the steering and brakes? Will they both be manual now or will you be still aiming for some form of power assist?
Tuesday, March 30, 2010 8:31 AM
This is inspiring!
Tuesday, March 30, 2010 8:36 AM
What can I say other then Dave you cease to amaze me! And although I wish Sport compact car was still around, This will do just fine!
Tuesday, March 30, 2010 9:25 AM
Hayabusa engine + my 510 = a BIG smile on my face!
Tuesday, March 30, 2010 9:25 AM
I read a while back about people making busa V-8's for cars because something about the engine not having enought torque. Seeing as your design uses the stock gear reduction makes tons more sense to me.
Tuesday, March 30, 2010 9:45 AM
Fly'n_Z, I'm on the fence about power steering in Miatas normally. On the track, its completely unnecessary, but on tight mountain roads, I prefer the quicker reflexes power steering gives you. With so much less weight on the front wheels, though, I'm confident it will be unnecessary with the Miatabusa. In fact, I've already de-powered the rack. We'll still have power brakes, but its unclear whether we'll need to do any upgrades to the brake system, or just change the proportioning to match the new weight distribution.
Tuesday, March 30, 2010 10:57 AM
This. Is. Freakin. AWESOME!
I've always thought it would be awesome to drop a motorcycle engine into a Civic or something light-weight (Mini? Original Beetle?). You could always source a K-series transmission, since K-swaps are relatively commonplace and I'm sure the transmission mounts needed are available OTS. How do you think that would work out? Better/worse than pre-K-series tranny?
Tuesday, March 30, 2010 11:35 AM
How about some other FWD trans? Since you have to fab up all the mounts and adaptor plate, you could use pretty much anything you want. How about a QR25 trans with the SE-R's LSD? Or an SR20 transmission? I would assume the lightest trans that you can get an LSD for would be ideal, the querstion remains which trans do you pick? The K20 would be a good pick because of the 6-speed, especially if you can get an Si LSD.
Tuesday, March 30, 2010 12:07 PM
Will a turbo be in the future?
Tuesday, March 30, 2010 1:45 PM
Dude. Seriously. Are you kidding? OF COURSE a turbo is in the future! Everything is better with turbo...
Tuesday, March 30, 2010 8:02 PM
Only awesomeness can come of this project!
And Dave, I totally agree. Everything is better with a turbo. A car, a moped, a golf cart, a toaster, everything.
I can't wait to see it finished!
Tuesday, March 30, 2010 8:39 PM
thanks for inspiring me Dave! i cannot wait to see this in action
Tuesday, March 30, 2010 10:10 PM
You guys should run for president!!
Great project.... Tokens to you!
Have you seen read about the guy who makes a V8 out of two of these busa engines? I heard one of those made its way in to a miata...
Tuesday, March 30, 2010 10:41 PM
Very comprehensive write up. Thanks for all the interest and comments everyone. I have given a lot of thought to the idea of mating this setup to an FWD trans. Initial investigation makes me think it can be done although not with a Honda FWD trans due to the rotation direction issue. The hard part will be finding an engine position where the passenger side half shaft will clear the engine case on it's way to the wheel. I think it can be done and it's something I want to explore since it opens up even more platforms for this kit.
Tuesday, March 30, 2010 10:44 PM
holy mutha-freaking shiznit! I cannot with words express my gratitude for you doing a 'busa swap in a miata and planning a kit for it. hmmm swap kit for a Datsun roadster with miatabusa everything makes my carnal hoondom tingle with glee.
Wednesday, March 31, 2010 7:01 AM
Uber cool project guys.
Just one question, dry sump?
Wednesday, March 31, 2010 10:38 AM
You NEED a turbo.
Thursday, April 01, 2010 5:08 PM
With that extra weight off the front end you might want more brake bias in the rear. A little trick we did on our track miata is to use the proportioning vavle from an ABS miata and it makes the bias almost 50/50 and it has worked great for us for the past 2 years.
Thursday, April 01, 2010 7:29 PM
This project is wicked! Can't wait for updates. One thing I was wondering, since a bike's oil pan is designed to keep oil at the bottom of the pan because the engine leans in corners and your motor is now stationary, do you plan on making a custom baffled oil pan or go with a dry sum system? This comes up a bit in the Formula SAE competition since we also rigidly mount the motors to the frame. Currently our team is running a dry sump on a CBR600 F4i, one pump takes the place of the factory oil cooler and the other gets driven off of the water pump drive shaft while we run a small electric water pump.
Friday, April 02, 2010 8:34 AM
ahh i miss these write ups. oh wait all i have to do is come here. how do you keep up all these projects?
Friday, April 02, 2010 8:41 AM
Good question. There are years of precedent for bike engined cars in Dwarf Car circle track racing and oiling is never an issue. Remember that in the Miatabusa the engine is in a north south configuration and the lateral g forces are in the direction of what used to be braking and acceleration on the bike. Since bikes can do wheelies and brake stands without running the oil system dry, we'll be fine. I have over 100 hours on my motorcycle powered LeMons endurance car and we've never even had a hint of oiling issues and that car has the engine east west.
As for your F SAE car. You probably didn't need to go dry sump but since it's all about learning it's great that you got to set one up. I did an informal test years ago with a GSXR by letting it idle and tipping it slowly to see when the oil light would come on. 45 deg. of tip simulates 1 g in a car etc.. By the time the fairing was on the ground the engine was still oiling fine and the test was over!!
Friday, April 09, 2010 12:32 AM
I love this project and what it means for Lotus Seven clones:)
Please let us know when you make more progress.
Sunday, April 11, 2010 9:08 AM
That makes a lot of sense going north to south. East to west probably still poses the issue of sloshing but probably not enough in a car to affect it too much. The F SAE cars can hit close to if not over 2 g on some tracks and can dart left to right pretty quick. In our previous cars we just cut the deep sump and flattened it out with baffles. The dry sump actually isn't so much for corner starvation as it is for weight balance. The honda CBR's have a greater lean than say a gsxr which lowers the center of mass, cutting the tall oil pan out completely, allows us to drop the motor way down while the negatives being slightly increased weight. But yea the learning is also a great pro to this set up.
Also about sport bike powered 7's, there is a Caterham dealer in Chemainus, BC here on the island that as far as I know is the only one licensed to sell bike powered 7's, let alone the only Caterham dealer in Canada. I'm not sure how they are set up, it looks sort of like a chain powered output shaft possibly still using the bike's transmission.
Wednesday, April 21, 2010 12:24 PM
Few things, maybe you missed, maybe you didn't.
The gearing will be wonky with nearly 10,000 rpm input, and likely the shifting at high RPM may be.. not so willing. A solution to both is to drop the retarded and expensive idea of building your own longer input shaft, and just use the stock output shaft. With a pair of 3rd gears always engaged(a further 1.526 reduction), and a simple coupler/support to the transmissions input shaft, you can just use the bikes box as a reducer box. Lastly, use the bike wet clutch(I know they make HD ones, and it's easy enough to change) arrangement and add in a large oil cooler/filter setup to keep clutch material out of the shared oil, which you'd do anyway. This also shoves your engine over ~3" and you can rotate it around that new axis. Which, when vertical, will likely give you the room you need to make everything work, maybe even with the stock subframe and oil pan. I have things all planned out to do this myself to some car. most likely a metro, but i have other projects right now.
Also Coleman, about the FWD, if there's space, and there likely is, the honda transmission would be ideal. BUT you flip it around so it's mid-front engine and the transmission is on the drivers side, diff forward. I'd have to check the bearings on the shafts to make sure they can handle the reversed thrust loading, but IIRC they're the same bearings on either end. Also you'd have LOADS of space for nice intake manifolding and turbo stuff once you got the exhaust out of the firewall. assuming it'd fit. I'd think the engine would be above the steering rack, and you raise the transaxle up a few inches to fix the standard driveshaft angularity issues with lowered hondas. Maybe even rotate it around a bit so the busa engine is leaning forward, exhausts up... if not just use a K series transaxle.
Wednesday, April 21, 2010 12:45 PM
my apologies on that! Was going off memory and a 14krpm busa build a friend of mine did. It dominates my memory of busas, and was just thinking, no, they wouldn't have messed up the math...
Your numbers are good with the 1.596:1 built in reduction with the stock 10,900 redline. I still propose the stock output shaft, clutch and doubled gearing. even if you're keeping the stock redline. sorry about that.
Tuesday, May 18, 2010 3:13 PM
This is pretty exciting, I just happened to buy a wrecked 07 hayabusa and didnt really know about putting the engine in a car. **UNTIL NOW**
You guys rock.
Monday, January 17, 2011 8:23 AM
This is such a great project idea! Heard about you guys through Mike at MD, and Jay Lamm at LeMons.
Qs: What do you think the total cost for the (I like: Hya-Mia) project will be in P&L?....
Would my mint '90 SHO motor fit in Miata or RX7? Too heavy?
What do you think the total cost in P&L might be?....
Wednesday, June 01, 2011 12:14 PM
OK, I just discovered this a couple days ago, and Ive been obsessing over it ever since. Absolutely top notch work! Couple of questions though. First; when are we going to see some more updates? And second; It would seem that with the original sprocket opening (output shaft) sitting at the front of the car now, that it would be fairly easy to re-purpose it to be an accessory drive. You could (in maybe another edition) run A/C, a bigger alternator, or a supercharger off of it. Just a thought, and Im looking forward to a kit, as I have a home for it already-so hurry up! I need a fun commuter!
Wednesday, June 01, 2011 12:34 PM
Good catch, and yes, we do have an accessory drive planned eventually. The actual sprocket won't be there any more, since we're removing the original gearbox, but the input shaft will poke through the front. Its very small at that point, though, so its not clear how strong it will be for an accessory drive. First car (or three) will have no accessory drive and just run off the Busa alternator.
As for the lack of updates recently, the next step is a big one. Lots of machining steps to make the shaft. Its about 80% done, so we should have some updates soon...
Monday, June 06, 2011 12:16 PM
OK, after further thought I has another question or two.
"4-cylinder sportbikes all transfer power from the crank to the gearbox through a big reduction gear. That gear both reduces the rpm at the clutch by a factor of 1.596 and increases the torque by the same amount, making both much more appropriate for a car. While the Hayabusa redlines at 10,5000 rpm, its clutch never sees more than 6500. That makes the input shaft, not the crank, the ideal place pull power for a Miata transmission."
I read and re-read this a bunch of times as this didnt make sense to me. If the stock transmission is seeing the 1.596 reduction, and Suzuki measures their published power ratings at the output shaft, then hooking the Mazda transmission to the same source of output potential isnt going to change the available torque any. The input shaft of the Mazda trans is seeing the exact same thing the the input shaft of the 'Busa trans.
Am I looking at this wrong? In my mind the concept of the whole adapter kit is still viable, but I just dont see the surge in torque thing happening.
Monday, June 06, 2011 1:31 PM
The stock rating is at the crank, even though the point of measurement might be somewhere else. (I'm not sure its safe to assume any of us know where Suzuki measures power)
Similar thing happens when you put a car on a chassis dyno. The actual torque measured is mutiplied by the gear and final drive, but the the number the computer spits out is still a crank torque number. This is why you have to hook up an RPM signal on the dyno. The computer needs to know engine rpm to do this calculation.
MotoIQ Proudly Presents Our Partners: