Three wheels and a helluva lot of fun

Posted on July 31st, 2007 in Aerodynamics,Economy,Handling,Opinion,Power,Suspension by Julian Edgar

As I write I’m getting over a cold. I am well enough to be mobile but not well enough to work. Well, that’s what I tell myself anyway.  

As many of you will know, I am becoming more and more interested in lightweight vehicles. One of my cars is a Honda Insight – amongst the lightest of all production cars on the road – and I find the downsides of its design usually quite minor. (If I need to carry more than two people, I take Frank the Falcon.)  

Now the Honda might be light, but it still has four wheels when surely three would be enough. Using a tadpole configuration (two front wheels and one rear) would also allow the car to be nicely streamlined, something that would be helped by a front mount engine and front wheel drive. That way, the classic teardrop shape for low aero drag would be much easier to implement.  

The starting point for such a car would be a FWD half-cut, say a Mira or Suzuki 660cc 3 cylinder turbo. Use the complete driveline, subframe, steering and front suspension and brakes, add on a tube frame chassis and then run the single rear wheel and suspension from a motorbike.  

It’s an intriguing idea. So, being as I say too sick to work, today I went off to my local Japanese importing wrecker. But: “No mate, nothing like that here – and we wouldn’t get something like that small in anyway.” Hmmm. Then I took myself off to a really big wrecker, one that sources just locally crashed cars. The prices are high, but the access to the full yard of crashed cars is priceless. There I wandered at will, looking at the cars and thinking of weird hybrid three wheelers.  

And the best bet was, I thought, a Daewoo Matiz. The Matiz, which we tested here, certainly doesn’t look like a good starting point. Not with just 37.5kW from its 3 cylinder 800cc engine. But then again, this four door car weighs 780kg, giving a standard power weight ratio of 44 kW/tonne. Still pretty awful – but what if the car lost 200kg of bodywork and wheel? Then the power weight ratio would be 65kW/tonne. And what then if it got a little turbo – a neat 100 kW/tonne would be pretty easy. 130-odd horsepower per ton isn’t something radical, but it would be fun. 

The Daewoo has a major advantage over other cars – it is built as a low mass car. The suspension is small and light, for example. It’s not a downsized big car but a small car to start with. The suspension from (say) a Toyota Yaris is huge in comparison. The other advantage for us here in Australia is that it was sold locally, so all parts are available. (The Japanese 660cc turbo FWDs were never sold here.)  

And what form would such a car take? Well, I don’t think it would need a windscreen – just wear helmets. The driver’s side could be made spacious and roomy at the expense of a rarely-squeezed-in passenger. If you don’t have a windscreen you don’t legally need a demister, so (in the wonderful climate in which I live) you could get rid of the complete heater/demister unit. In fact, get ride of the whole dash, retaining just the instrument pod and switchgear. Put the fuel tank behind the seats and ahead of the single rear trailing arm suspension. I assume that as a motorbike/trike, the front and rear brakes could be separated in function, so have the foot pedal work only the front wheels and use a hand lever for the back wheel. Don’t have doors – just hop in over the high sills. The absence of doors makes legal registration easier (no need to prove door-bursting strength) and with the depth of the sills, the tubular frame structure could be made very strong but still light.  

In fact, really the only thing that would prevent the design being aero efficient, fun and fast would be the bodywork. To get low drag you’d need a pretty trick body – especially without a windscreen. (You’d need full fairings behind the heads, etc.) And the required compound curves would need complex and expensive moulds – even if done in fibreglass. The work of a really good industrial designer or car stylist would also be a necessity. (You could instead go for a minimalistic bodywork design, but then with the available power, the performance wouldn’t really cut the mustard.) 

Hmmm, now what I need to find is a Matiz that’s had a rear-end collision and is for sale for about five hundred bucks, a brilliant beginner car stylist and a fibreglass expert…

6 Responses to 'Three wheels and a helluva lot of fun'

Subscribe to comments with RSS

  1. Cary Wintle said,

    on July 31st, 2007 at 9:38 am

    Sounds like a fun project! For more fun, retain the driveline from the motorbike for twin-engined AWD (gear-shifting!) madness.

  2. John Littler said,

    on July 31st, 2007 at 3:45 pm

    You’ve got your approach back to front. Far cheaper and easier to convert a motorcycle to a hammerhead trike than carve everything off a car to make one. Would only make sense to use a car as a start point if you intend to keep the roof (and hence by extension some sort of modified monocoque shell). Trikes are registered as motorcycles in NSW (I believe that is also true of other states), so you’ll be facing all sorts of compliance fun and games to get your car based trike passed (unless you can get an engineers certificate as a 3 wheeled car – not sure if that’s possible).There are numerous low cost bikes that would be ideal starting points for a design of this type 600-1300cc 4cylinder Jap bikes from 70 to 150Hp aren’t expensive a 10 year old suzuki bandit (GSF) 1200 will set you back a couple of grand at most for example. Remove front suspension and setup up a two wheeled front end pivoting on the front of the frame. You start with a low weight frame and engine (bike engines tend to be MUCH lighter than cars) and then you can add weight. It’s a lot easier to control the rate at which you add it than it is to take it back off…

    Would be a lot cheaper and easier for achieving the outcomes you’ve described IMNSHO

  3. Andrew McKellar said,

    on August 1st, 2007 at 1:41 am

    Ah Julian, it’s almost as if you took a look at my own files on small, fun commuter cars, right down to the open cockpit and crash helmet. Needless to say, I think the idea is good.

    A few observations:

    Because the car would be so light, the occupants would comprise a significant proportion of the overall weight. I think that going all the way to a pure single seater would make it easier to get the weight distribution right without having to compromise between one vs two occupants. In a three wheeled vehicle, keeping the weight on the centre line also aids stability. I don’t think that a Formula Ford steering rack, while more expensive than a production car rack from a wrecker, is prohibitively expensive. (There is an ADR that says the driver cannot be to the left of the centreline of the vehicle but I think this can be interpreted as permitting a central seating position. At least one signatory engineer agrees).

    A Diahatsu Copen might make a good higher power, if more expensive, alternative donor, if it could be found.

    The possibility of tuning chassis balance with changes to roll stiffness distribution doesn’t exist with three wheelers.

    I think there are some issues with the suggestion of using a motorcycle as the donor:

    No reverse gear.

    A dry oil sump or at best, a deeper wet sump, would be necessary because of the different direction of ‘G’ loading between a bike and car.

    Durability might be questionable given the greater weight of the car vs ‘bike.

    A ‘bike rear suspension is not designed to take side loads, the same as for the sump.

    It would be difficult to keep the wheelbase/length reasonable because occupant(s), engine, gearbox, rear suspension/wheel must all be in line. Also the weight distribution is likely to end up towards the rear.

    Three wheelers, while generally considered as motorcycle derivatives by most (all?) state registration authorities are considered uniquely in terms of applicable ADR’s for each configuration; tadpole, one front with two rear etc. It’s been a while since I’ve read the ADR’s but from memory there are also weight breaks (> or

  4. Matt King said,

    on August 1st, 2007 at 10:41 am

    Is there really that much benefit in a 3 wheeler vs a 4 wheeler? Obviously you lose one wheel and suspension, but on the flipside you are stuck with:
    *engineering new rear suspension that must withstand large torsional loads
    *distributing those loads to the existing front chassis
    *a vehicle with thee “tracks” instead of two (since front and rear wheel pairs are aligned). Makes it hard to miss bumps/potholes…
    * Far less cargo space and carrying capacity
    * Inherent instability, with no ability to control weight transfer.

    Surely the rear suspension, wheels and brakes of a 4 wheeler could be pared down if you lighten it?

    I’d be curious to see what could become of a Matiz just by stripping all unnecessary weight (interior trim, carpet, rear seats, a/c, radio, etc).

    Or maybe it’s time you built a 500kg clubman car?

  5. Gary McDonald said,

    on August 1st, 2007 at 1:06 pm

    Suggest you take a look at a historic race day when the sidecars and the Morgan three wheelers put on a demonstration.
    Others have been down this track – see
    not a bad dream at all!

  6. Chris said,

    on August 1st, 2007 at 4:29 pm

    Inspiration for you:

    70mpg, 100mph. And that’s with a 2CV engine.

  7. Frank Thomas said,

    on August 1st, 2007 at 6:31 pm

    suggest you check out these guys in WA, for a CX500 based hammerhead trike. A real headturner on the road.

  8. Leon Weekes said,

    on August 1st, 2007 at 11:43 pm

    Hi Julian

    I am extremely interested in design and engineering of a high performance trike. I recently graduated from the ANU and am a performance + motor sport fanatic.

    In the simples possible terms the only significant reasons bikes don’t go faster than high level race cars is because firstly they don’t have enough grip ~ contact patch which reduces their braking and corning abilities (specifically on the limit) and they also produce very significant drag for their size.

    Imagine for a second a vehicle that brakes and turns like a (race) car yet has the power/weight and hence acceleration of a sports bike. Not only that but the driver has control, confidence and relative safety to push towards the limit. To add to all this there is huge potential to reduce drag from this vehicle. The vehicle also has ridiculously low fuel consumption as a result of low weight + drag. That (in my opinion at least) is the potency of a well engineered motorcycle engined, single seater, hammerhead ( 2 wheels forward) trike.

    I am keen to design and build such a vehicle and get it ADR approved (i have already had a look at the vehicles classification and the applicable ADR’s) though I would appreciate some advice as to if people think this is possible.

    There is potential to build in safety crash structures for the driver (simplified F1 safety cell). There is also potential to move to hybrid and/or full electric in the future. I am a performance fanatic but I well and truly understand that oil and all its benefits will not be around for much longer and if we want to have anything even resembling motor sport in the not too distant future we have to move towards sustainable vehicles.

    If the prospect of developing design for a trike as described above is more than just a passing thought for anyone, then feel free to email me at

    useful links and inspiration: (very highly reccomended) (also see the V8 from SR8!!!)

    Thats my two cents

  9. John Littler said,

    on August 2nd, 2007 at 12:32 pm

    Typical car guy comments Andrew 🙂

    – Reverse gear – Honda Goldwings ship with them and they’re available after market (electric, nice and light)

    – Dry and wet sumps in all sorts of configs are available, choose one that suits you.

    – Durability – sure if you were planning on bolting it into a commodore, but the spec was a trike circa 580Kg. That’s less than a typical motorbike with sidecar setup – and there’s plenty of those I’m aware of with several hundred thousand Kms on them.

    – rear suspension doesn’t take side loads ? That’s patently not true.
    I’m assuming you mean not loads of the same order, which may well be true, however I’d suggest a bit of investigation before dismissal out of hand. Some models here
    may be of assistance

    – weight is going to depend on what you do with the design but bikes are typically ~50/50 front rear, you add more weight to the front with your two wheel setup then you’re going to need to be adding a heap more to the back to cause it to be rearward biased.

  10. Peter said,

    on August 2nd, 2007 at 1:10 pm

    More inspiration on a motorcycle powered 3-wheeler.

    Very impressive!

  11. John Littler said,

    on August 2nd, 2007 at 3:53 pm

    Nice one Peter, how about this

    doesn’t meet your hammerhead criteria sorry Julian, hence why I haven’t bought it up before, but tilting trikes (including human powered ones) get the best of both worlds – the handling and manouverability of a 2wheeler and the grip of a 3 wheeler

    More here:

    Also there’s a new hammerhead trike about to be released on the market in Oz (I’m told ADR is “imminent”)

    A little off topic for the above I guess.


  12. Andrew McKellar said,

    on August 3rd, 2007 at 10:59 am

    John, yes i’m a car guy :).

    I was aware of the electric reverse on Goldwing etc. (and I think Grinnall use something similar). It adds to the expense and I question whether it is a true reverse, as in manoeuvring laden up a hill, rather than an assist.

    Re. dry sumps: Again $$$$.

    I disagree with the side loads. Any side load on a bike will be due to the moment generated by a difference between the CoG of a rider hanging inboard during cornering and the wheel vertical axis. Everything else goes straight down the wheel axis. The result is much less loading than in a non-banking three wheeler, which is also heavier.

    Maybe re. weight distribution. The extra wheel base achieved by moving the wheels forward would tend to counter the extra mass 4ward though. And the wheel base does end up very long, as some of the pictures in the links show.

    This might be one way around. Expensive and specific to RHD though.

  13. Andrew McKellar said,

    on August 3rd, 2007 at 11:01 am

    I mean LHD.

  14. John Littler said,

    on August 3rd, 2007 at 2:39 pm

    Never Mind Andrew, I’ll forgive you

    – Sure, the electric reverse adds to the cost, but then you’re starting at a lower cost point and it’ll be a less expensive conversion, so – swings and roundabouts. As for the functionality, errm I thought we were talking about a barebones ultralight (sports?)car – I didn’t think we were building a builders ute ! Fit for purpose design !!

    – Dry sumps = $’s ? Rooobish ! Like I said you can get bike engines in a myriad configs, if you want a dry sump then buy the one with a dry sump. There’s plenty of them. Also plenty with heavily baffled and highly managed oil flows (don’t forget a bike engine has to cope with being tilted 50 degrees either way on one axis and 90 degrees on the other. I’d be extremely surprised if most weren’t able to cope with the comparatively trivial oil flow movements created by the increase in cornering G forces )

    – Sideloads – I hope you’ve mentioned to gravity about how it’s supposed to behave – get back to me when you’ve whipped into line :-p
    (it’s quite possible it’s massively different but you’re not convincing me you’ve thought it through and calculated it yet – the above comment makes that quite clear). Sidecars and existing trikes seem to utilise bike swingarms successfully without catastrophic failures. Besides make it a tilting trike like I posted above and it becomes a moot point 🙂

    – Lastly – yep that’s exactly the type of solution that I think would be the most cost effective implementation of the above parameters. Baffled why you think doing something like that is going to be less cost effective and weight effective than taking an Oxy to a half cut and turning it into a trike. Particularly as by the end of the article Julian’s already talking about a tube frame (in which case a space frame with a light engine and a GRP or fibreglass body like your URL is the simplest and most cost effective way to achieve the outcome.


  15. Andrew McKellar said,

    on August 3rd, 2007 at 7:18 pm

    As I understand it, the electric reverse is either on or off with no other control (actually the starter motor engaged via the gearbox). For a toy, that may not be an issue but using it everyday? I’m picturing a reverse park up any sort of hill. How much leg assistance is required for them to work on the bikes?

    OK, the sump issue looks like it depends on which engine.


    (That is with the engine Nth/Sth).

    Maybe I’ve missed something with the side loads. On a bike, viewed from front or rear, I’m looking at it like this: the vector sum of the vertical reaction (up) at the ground and the lateral force generated by the tyre (into the centre of a turn), at the contact patch, will be equal to the weight of the bike/rider (down) plus the inertial force due to centripetal acceleration (out of a turn) at the CoG. The net result is resolved straight down the centre of the bike regardless of bank angle; no side load. (I know that is complicated by the contact patch moving laterally with bank angle and tyre distortion, or the rider moving the combined CoG off the centreline of the bike).

    The SUB solution to shortening the wheelbase is expensive because it requires an extra machined shaft, complete with splines, bearings and mounting without misalignment. That is also why it is specific to LHD; the output sprocket from the gearbox is on the wrong side for RHD.

    I don’t think that Julian is talking about anything other than a tube frame.

  16. Doug Foskey said,

    on August 22nd, 2007 at 6:21 pm

    Wonderful idea! I am thinking seriously of building a FWD Guzzi powered cyclecar. The easiest gearbox to use is the Alfasud, with inboard brakes. (The only issue is fitting the starter motor).
    The advantage of this layout is the motor is slightly forward, so reducing the length.
    I am also considering central driver, & passenger behind. This will greatly reduce the width of the bodywork.
    The other advantage of sitting ‘on’ the chassis, is the compliance is slightly different (so perhaps a little easier).
    The FWD arrangement is the best, because the inner wheel will tend to spin when cornering hard, so reducing the tendency to roll. (There is a treatise on this on the web, but I cannot remember exactly where).
    Look forward to hearing more.

  17. Ron said,

    on December 28th, 2007 at 10:14 am

    Have you thought of using hub motor & using the motorcycle engined to generate electricity
    Electric Car Hub Motor
    V: 48V-96V
    P: 1500W-4000W
    Rpm: 350-850
    Efficiency: >85%

    the rim is 10” .
    With disc brake

  18. Julian Edgar said,

    on December 28th, 2007 at 8:00 pm

    I haven’t seen hub motors as powerful as that. Got a link?

  19. Ron said,

    on December 30th, 2007 at 3:31 pm