nortonshrike.com

Zen and the Art of Three-wheeler Design

Discussions of design philosophy

People write in with great questions about the Shrike, its design, the building, and philosophies for clean living. Here is a "best of" digest of those great questions and my responses.

 (latest, 6/6/16)

Dave,
Do reverse (2F1R) trikes oversteer or understeer?
 
My response: 
Thank you for your question.  Unfortunately there is no simple answer.  In short, it all depends.  Here are my comments.  Please bear in mind that these are only my opinions based on my own personal experience, and should not be considered to be absolute truths.
 
Reverse trikes may either understeer or oversteer, depending on circumstances.  There are a large number of variables.  Wet or loose surfaces tend to produce oversteer.  Heavy braking into a turn favors understeer.  If the rear wheel locks before the fronts under hard braking oversteer is likely.  Loss of rear wheel traction for any reason will produce oversteer.  Tire tread depth and design and rubber compound will be factors, as will pressure.
 
Before I built the Shrike I assumed it would oversteer based on the very small contact area between the relatively narrow and rounded motorcycle tire tread and the ground, compared to the wide tread of the conventional front tires.  I did not find this to be true.  I am surprised how much cornering force the standard motorcycle rear tire can produce.  On smooth clean dry surfaces cornering speeds are quite impressive.  
 
The real question is what happens under steady-state lateral G (cornering) near or at the limit of adhesion.  Getting a useful answer to the question requires some testing on a flat level "skid pad", something I have wanted to do for 30 years now, and have never had the opportunity to do.  The best I can tell you is that I have not found either understeer or oversteer to be limiting factors in aggressive driving on clean, dry, smooth street roads.  I have had several times when the Shrike left the road out of my control, both nose end and tail end first.  None of these resulted in personal injury or serious damage to the Shrike.  Two of these incidents were due to a stuck throttle or a locked rear brake.  A few were due to gravel in the middle of the lane, a common problem on our winding mountain roads.
 
I also find it interesting that I can produce hydroplaning (the tire losing contact with the road due to a layer of water) at either the front or rear wheels at 60 mph in really heavy rain.  I assume that running a wide flat-section car tire on the rear would be a real concern at high speed in heavy rain.

Dave,
On Front Wheel Drive:
The biggest problem I see with a FWD 2F1R is one of aesthetics.  Unless it's done "just so" it's going to look really... well, bad.     Under braking I can see how handling could get really tricky, with the rear wheel potentially lifting off the ground.  I know a really low CG could help/prevent it, but there's only so much that can be done to lower it given it's a road car.     The design cries out for a boxer engine (e.g. Subaru flat-four), though with the engine way out in front of the tranny, which is already out in front of the axle centerline, yikes, styling just got much more difficult.  How the heck to get it to look good is the question. 
And finally, what would its handling be like on the skid pad?  When you’re on the gas at or near the limit of adhesion the wheel torque reaction would push the rear wheel down, improving grip, but let off the gas just a bit, or worse, touch the brakes, and I can see it spinning like a top.  Maybe skinny front tires and a large very sticky rear one would help fix that, but I still fear what it would do, slinging around a mountain road and having to come to a fast stop on a curve...     I would like to roll this idea around with you more though...
Kurt

Kurt,
Braking: 
I shared your concern over braking stability when building the Shrike: wide front track. grippy front tires, little weight on the rear under hard braking, skinny rear tire, add just a bit of differential friction coefficient at the front and it's just GOT to pull hard to one side...wrong.  The wheelbase is so long that even the minimal contact at the rear acting on a nice long moment arm resists being hauled sideways, producing dead straight hands-off braking at over 1G.  Amazing.  Unless the front weight bias is such that you're literally approaching a rear wheelie, braking will be straight and true.

Rear wheel traction loss on braking in a turn: 
Good point there.  The BMW Paralever rear suspension negates squat under acceleration (and rise under braking), and the front suspension geometry has about 75% anti-dive to it, so there is very little change in vehicle pitch going from power to brake, but there is still the front weight transfer due to the acceleration vector acting through the CG.  In any case, I don't think limiting front tire width is the answer, at least for a vehicle that is intended to be driven aggressively by a competent and experienced driver.  Is it better to exit the road sideways with all three sliding, or loop it as the rear comes around?  My approach is to use the driver's skill as the limiting factor.  Makes it more fun that way.

Added comment on anti-dive geometry:
Double A-arm front (or rear) suspension can be designed (by angling the inner pivot axes in side view) to reduce or eliminate front end dive / rear lift under braking and rear squat under acceleration. There are good reasons not to overdo it, however.  The degree of dive under braking serves as feedback to the driver of impending wheel lock.  The more the nose dives, the closer you are to locking a wheel.  With 100% anti-dive, the nose stays level and you lose that feedback signal.  Also, the greater the anti-dive, the harsher the road feel over bumps and irregularities, and the greater the tendency to lock a wheel on bumps.  Shrike has 75% anti-dive at the front, with essentially no nose dive, and I like it that way.

Dear Dr. Science,
[On layout]
 Rolling some ideas around in my head, I'm wondering how this packages.  A side-by-side two-seater, but with the passenger-side taken up by a bike engine, where the chain goes forward to a center diff, then half-shafts out to the front wheels.  This helps styling some, keeping the nose from being too God awful ugly, like if a FWD car engine were put up there.  Of course now it's fat side-to-side, it's always something.  I must admit driving the rear wheel with the engine behind you packages so much more elegantly and simply.... Just no traction.

Have you ever driven a FWD 3-wheeler? I'm wondering, after all the work and expense to drive the front wheels is done, might it handle so nasty it is deemed a big mistake?
Still thinking about it.
Kurt

Kurt,
Nope, haven't driven a FWD 3W.  My driving experience in 3Ws in limited to
mine and an HM Freeway: 2F1R, RWD, 12hp, rubberband tranny, about a 4 ft
track, way tall, way way tippy.

Road & Track (and others) tested the TriHawk, generally found it to be quite
acceptable.  Google it, there's lots out there on TriHawk.  Yes, of course
it's possible that after all that trouble you're left with the off-throttle
oversteer situation you described.  So was Porsche.  People learned to live
within its limits, and Porsche learned how to mitigate the problem.

I like your layout, but I'd be inclined to also consider a shaftie, sitting
backwards, with a shaft forward.  I hate chains: messy, high maintenance,
short life (when highly stressed), and a continuing expense (this from a
person who pops for a $150 Metzler rear tire every 3000 miles, so what do I
know...).  Downside on shafts is limited gearing choices.  A hot bike motor
geared for 160 mph on a 17" rear wheel might be close to right for a 100 mph
autocrosser on 13 or 14" rims.

[November 2012: On your side-by-side driver/engine idea: since your initial comments our friend Dennis Palatov has taken the concept to its limits.  Readers see dpcars.net.]

FWIW,
[Layout, continued]
As I continue thinking about 3-wheelers the more afraid I grow of cars, SUVs especially.  While I want to build a 3-wheeler looking like an F1 car, [See: http://www.indycycle.net/main.htm]

I'm pretty sure I'll be virtually invisible to any truck or SUV when I'm closer then maybe 40ft.  It's making me better appreciate your complete roll-cage structure...

Sadly the vehicle needs to be higher up, with a light at the top like yours.

Darn you Norton!

Kurt

Kurt,
I think that's probably about the second nicest thing anybody's ever said about the Shrike.  The nicest was "You make simple look good."  Yeah, physics and the Active Safety/Passive Safety/Appearance/Performance tradeoffs are a bitch, ain't they?  Ya pays yer money and ya takes yer choice (or chance, in this case).
 
Well, here's some more thoughts to stir into the mix:
Check the front suspension on the SportCycle (see: http://www.sportcycle.co.uk/Press.HTM).  [as of late 2012, the company and website are long gone] Note particularly the photos and text on the sixth page.  The front suspension pushrods both push on the single pivoting link acting on the single coil-over damper (analogous to the Hydrolastic suspension used on later early Minis, only front to rear rather than side to side).  It appears to have a nearly infinite roll rate.  When the right wheel hits a bump, it pulls the left wheel up the same deflection.  When I first saw this a couple of years ago as the IndyCycle, before Jim Musser was forced to dump the INDY name,  I assumed this was an ill-advised attempt at cost savings.  Now I'm reconsidering.  How could an infinite roll rate be a good thing?  The article sidesteps this issue, saying perhaps it only works on 3Ws but not realizing why. 
 
Consider the Shrike: I have a very high roll angle, around 6° (simply because I haven't added a sway bar yet).  Why not?  Because the handling is adequate and the ride is very comfortable.  A major weakness in the Shrike is the use of the bike tire at the rear, but not for the limitation on lateral G, as one (I) might expect.  The rear end (rather surprisingly, I think) stays glued past 1G, where with the sticky autocross tires I begin to lift the inside wheel.  The limitations are due to wheelspin in lower gears and on low-adhesion surfaces, and rapid tire wear.  The BEST I've seen is a tick over 3,000 miles on a rear tire.  At $180 a pop that's 6 cents/mile, or about a third of what fuel costs.  For me, that's about $2/day.
 
Now, what would happen if I put a flat-section car tire on the Shrike?  Much improved forward traction and mileage, but with the chassis rolling 6° the car tire now corners at 6+° positive camber.  Not good, not good at all.  Gee, if there were only a way to keep the rear tire perpendicular to the road.  But that would take some real geo-mechanical trickery ( I've bumped my head on that knotty problem a few times without anything reasonable coming out) or, wonder of wonders, a zero-roll front suspension!  I'm considering trying something like that, actually.  Of concern is the added side loads into the rear wheel bearings and the swing arm and bearings.  But I've already shown that the bike tire already loads up those components to the limit of the rest of the chassis.
 
Now, if one wanted to push the cornering limits past those of the Shrike, with a lighter rig, more power, lower CG, more rubber, less frontal area, slicker (read: F1-ish) body, then the above might be the way to get there cheaply.  Alas, the SCCA in their wisdom sees 3-wheelers for what they are, inherently unstable and not to be allowed to compete with "real" sports cars on Solo II events.  Ah well, I can dream, can't I?
 
Dave

Wedge principle of crash energy management:

Dave-

I hope you are well.  The Shrike is looking fabulous!  I've been thinking a lot lately about your "Deceleration wedge" concept.  I'm thinking that extending the same type wedge all the way around the vehicle would provide protection for the driver.  Low vehicles are at a  severe disadvantage in collisions with larger (SUV) vehicles.  I think that designing as low as possible with wedges on all sides would use the SUV's only vulnerability, a high COG, against it.  Instead of absorbing impact, deflect it!

There was a tragic head-on collision a while back in Maryland where an SUV was traveling in the wrong lane.  I was appalled that the news coverage was subtly blaming the deceased driver of the '88 Chevy Nova because her car was small.  I really resent that line of popular thought - I drive a '97 Ford Aspire, and people regularly tell me that I should drive something more substansial for safety reasons.  Anyway, I'm not trying to rant here - I'm seeking your thoughts on the 360 deg ramp design.

 

Also, how did you solve the issue of reverse with the Shrike?

 

Thanks!

 

Hi Fritz,

Sorry it took me so long to get back to you, but thanks for your thoughtful comments on the wedge principle.  The wedge carries its own limitations.  It only works with ramp angles that are fairly small, say 30° or less.  That works in side view, as you have enough distance between the low nose and the driver (nearly 6 feet) to have the driver under the ramp.  Not so useful in side impact, since there isn't enough distance between the low point at the side and the occupant (perhaps 2 feet).  A 360° ramp with the outer edge below bumper height but high enough in the middle to cover the driver would be cone 12 feet in diameter.  So I beileve the usefullness of the wedge concept for crash energy dissipation works only for front impacts, only in side view. 

The wedge works in a different way in plan view (top view): being tapered from both ends to a wide cabin encourages crash energy management by deflection rather than by deformation.  Striking a tree straight on the nose would be bad.  The narrow nose makes it more likely to hit the tree a glancing blow, shearing off the front suspension without crumpling the coll cage and without high rates of deceleration, much safer for the driver.  The same is true of rear-end collisions: if the striking car misses my rear wheel, chances are good that the Shrike will be deflected, pushed aside, rather than crumpled.

There are some weaknesses in the overall Shrike crash energy management system, of course, but I definitely feel much safer there than on a bike.

Cheers!

Dave Norton

Thanks Dave.  So, did you resolve the reverse issue?

Yes.  I park carefully and don't get caught halfway into an intersection.  So far, it works. All of which is to say: no.

 

 

Engine choice:

Dave,

I am interested in building my own Shrike but I would like to base it on R1200RT twin oil cooled. How far is that from your design?

Actually, I think the R1200 boxer motor would be a fine choice.  It should allow a slightly shorter wheelbase and lower CG (both good things), and I expect the low-end grunt of the R bike will make starting from rest a bit easier.  Good choice.

Dave

 

Conversation about mileage, CG, etc., March 2006.

Dave,

Great job on Shrike!! I am interested in three wheelers. I would like to build a gas or hybrid SOA three wheeler for highest MPG possible. I like the general layout of Shrike and might evolve from it or at least learn from it. Questions: What highway Milage (MPG)do you get with the BMW based Shrike? Better than the 2 wheeler? What is your frontal area? Any idea on coeeficient of drag? Is a Miata front end the right place to start now? How close to an optimum cg front-to-rear is Shrike?

Regards,
John
 

John,

Excellent questions!  It's a joy to work with someone who has traveled enough of the same ground to know the right questions, if not the right answers.  I'll tell you my opinions.

First, the Shrike was not designed with fuel efficiency in mind.  In its first iteration, no body panels and the Yamaha Maxim 650 motor (60 hp), I averaged 32-35 mpg.  I hoped that the added power (and potential performance) of the K1200 (130 hp) would be offset by the improved aerodynamics.  Turns out, it was a push.  I get from 30 to 40 mpg now, averaging 32-34.  If I really baby it, drive very conservatively, I can hit 40.  If I were to seriously attempt the highest possible mileage on the street, I'd go bananas.  Can't do it! But someone else might squeeze 50 mpg.  I have no real data, but I'm sure the bike, with its rather slippery fairing and much less frontal area, would do at least 30% better at the same level of performance.

I just did a rough frontal area calculation: approx. 12 sq. ft.  Cd has to be terrible, perhaps 0.8? For front suspension, I see three possibilities: Miata, Mustang II and early Fiero.  The Mustang and Fiero are both popular with the kit car people, so there are numerous companies offering their versions of components (brakes, arms, dropped spindle uprights etc.)and complete systems.  All expensive, and all heavier than we need, alas.

Optimum CG: sorry, there ain't none.  No such critter.  Only compromises leaning (you should pardon the expression) rearward, toward improved drive traction, or forward, toward increased overturning resistance.

The first Shrike, with a much lighter drivetrain and 16" shorter wheelbase, had approx. equal distribution (67 front / 33 rear).  Handling was superb, traction was adequate for the power.  The car was eminently "tossable", throw it around with gay abandon, stick your boot well and truly in it at just the wrong time, and still recover.  Much less demanding.  An absolute hoot on smooth dirt fire-roads in the local hills, with long (and, in truth, quite slow) power slides kicking up great rooster tails from the sliding rear tire. BMW K-power changed all that.  Another 150 lbs or more at the rear, way longer wheelbase, and 2X+ the power.  Application of power through the loud pedal becomes a serious endeavor indeed.  The combination of high revs and significant throttle had darn well be used must judiciously indeed.  The result of proper application of this much power produces a rush leaving me grinning constantly, and laughing outright at the absurdity of having this much fun sitting down.

But I digress.  With limited power, you won't go far wrong by keeping 30-35% of the loaded weight on the rear.  Do not fail to take into account that front/rear distribution is only 1/3 of the critical trio for 2F1R 3W stability: CG height must be kept at an absolute minimum, and front track width must be wide enough to provide adequate overturning resistance in combination with CG location and the intended use of the vehicle.

Cheers!
Nort

A thread of conversation (edited for privacy), from March of 2006.

Dave,

I have been tinkering with a three-wheel design for about 15 years now. I first started thinking about it after discovering the TriMagnum by R. Q. Riley.  In all the research I've done, your Shrike comes closest to my vision.  Even my design pillars match up similarly to yours.  Although I have talents in many areas (electronic and computer engineering, mechanical design, etc), I have no experience in chassis/suspension design.  From what I've seen, your front-end design is elegant and simple, much as I pictured it in my mind (I keep leaning towards formula 1 type suspensions).  I may have put a higher priority on the aesthetics, but your general shape is almost exactly like mine.  One thing I want to achieve is a nice looking body without a great deal of fabricating.  To get away from the mess of fiberglass, I had the idea of using sheet lexan with just a few hot-folds to fit a chassis similar to yours.  The idea is that the paint job can be done on the inside that allows a very scratch resistant surface on the outside. The body portion is actually painted on the lexan, leaving the windshield and headlight opening unpainted.  A parallelogram of link arms can be used to raise the entire body off, like a massive forward lifting door (much like your engine cover, except tilts forward).  This takes the simplicity of your design and adds an attractive seamless one piece skin/windshield (two pieces if you count the engine cover) that is just as simple.

My question to you is related to the front suspension.  Where can I acquire information regarding the front-end design and where can I acquire materials?  If you need, I would be happy to provide assistance with electrical/electronic design and implementation in exchange for some help with the mechanical stuff.

As fare as current projects, I have an old Bradley GT on a VW chassis that I am converting to an electric vehicle.  Being a 4 wheeled car, I have all the legal constraints to deal with.  I like the idea of the 3 wheeled vehicle, particularly because of the reduction in constraints.  I would like to build my dream car, and perhaps even convert it to electric someday.

I would enjoy seeing your machine in person.  I am very impressed with what you have accomplished.  I also think you should go to the next step and offer your plans for sale.

Thanks for taking the time to read this and for any help you may offer.

Geoff S.

Hey Geoff!

Thanks so much for your kind words.  I'd be happy to work with you in making your 3W a reality.  We should meet.  Shrike and I'll be at the Knott's kit car show on April 22-23.  [Note: the annual Association of Hand-Crafted Automobiles car show has moved to the Pomona Fairplex, site of the L.A. County Fair] There are many interesting vehicles and talented designer/builders there. I was fortunate enough to win a trophy (2nd place, Specialty Street Class) last year.  Spectator admission to the show is free, though parking is $8 (I think).

I do get up to Orange County occasionally.  Are you aware of the Crystal Cove gathering on Saturday mornings in Newport Beach?  I do that one now and again.

I've learned much from the 22 years and 93,000 miles with the Shrike. Some of my concepts worked very well, others need improvement.  One area needing improvement is, yup, the front suspension, for several reasons.  I'm actually considering buying aftermarket components and redoing the Shrike IFS, using Mustang II or Fiero bits (expensive and too heavy but well proven and adaptable).

As explained on my website, I do sell an info package for $70.  One fellow built a Shrike clone in Canada using my package as a basis, two years ago now.

I'd like to see sketches or drawings on what you have planned so far, your ideas sound quite novel and very interesting.  I had considered doing Lexan panels, left clear, on one side of the Shrike just for car shows.  Still might.

Let's chat more.

Cheers,
Dave Norton

Hi Dave,

I will try to make it out to the car show at Knott's.  It will be a great place to pick up ideas.  I have included a rough drawing of my body idea and the basic vehicle shape I had been thinking of.  I traced your Shrike for reference so you can see the principle on your own configuration.  I plan on building a scale model to represent the idea.  For the most part, I can purchase model kits for various motorcycles and I intend to kit bash something like a Ferrari kit to get the front end pieces. I also have a BST 3d printer at my disposal (it prints in ABS plastic), so I can make custom parts, if necessary.

As far as seating arrangements, I would like a two seater.  I played around with the tandem seating, but was not happy with the look of a stretched trike.  It became quite difficult to come up with a cool body design using just sheet lexan.  The proportions are much easier to work with on a single or side by side design.  I agree with you about the distribution of weight. I had an idea of putting the driver's seat, steering wheel and pedals on a frame that rides on a lateral track.  When the vehicle is being used for a single occupant, the passenger seat can be removed and the driver's seat relocated to the center.  Depending on the body design, the passenger seat can be stowed in back with the engine.  I don't mind the extra width, but I don't like the extra length.

Let me know what you think.

Geoff S.

Geoff,

Thanks for the sketch and added info.  The body concept looks sweet, and would eliminate the major impediment to many folks considering the Shrike: difficulty of entry/exit.  Over the years I've gotten several inquiries from ex-'cyclists who are to varying degrees limited in mobility, and would love to regain some of the freedom they lost when they gave up 'cycles.  The price to pay for easy acces comes in decreased crash protection and increased weight and expense.  But the seamless body would sure be a showstopper!

Make it out to Knott's for sure.  You'll see some very inovative stuff there, and manufacturer displays of some useful suspension and other hardware.

Dave

Hi Dave,

I've been thinking about this three wheeler quite a bit lately.  But first, I have a comment about your last reply:  I have considered cool vehicle designs for the handicapped and it may apply to your requests for three-wheeled motorcycles.  I thought it would be cool to design a three-wheeler such that a person in a wheel chair could roll right into it and lock the chair into the vehicle.  The chair would then complete the cockpit and become integrated with the vehicle.  This could even be an anti-theft mechanism because the vehicle wouldn't function without the chair.

I have several ideas to run by you.  If you could, I'd like your expert opinion on them:

1. I've been looking at front suspensions.  I've found a lot of Mustang II suspensions, but the one that looks the most promising is the Heidts Superide (http://www.heidts.com/heip18.htm or http://www.heidts.com/heip22.htm) Without the crossbar, it looks fairly light weight.  Also, it is available as a complete package without the crossbar for around $1500 (includes the disk brake assembly). [2012: much more costly now]

2. I still like the side-by-side seating.  I am thinking of moving the front wheels back to a position about even with the driver’s knees.  Because of the reclined seating, there is a significant channel that goes under the knees. This could provide a reasonable place for the rack and pinion actuator.
In addition, this moves the weight of the driver forward to reduce the affect of an off center balance when the driver is solo.  I figure the motorcycle will weigh plenty enough to give it reasonable traction.  Do you see any issues with this arrangement?  I believe the footprint will have a fairly symmetrical spacing between all three wheels (like a tripod).

3. I don't know how much your steering components cost, but I can see additional complexity if I move the front wheels back.  It will likely require a gear box to reverse the direction of the steering shaft back to the rack and pinion.  I had an idea to use hydraulic steering ( http://ww2.seastarsteering.com/) and would cost around $700.  This is used frequently in marine environments.  It involves a simple hydraulic pump that the steering wheel attaches to and a hydraulic cylinder that replaces the rack and pinion.  They typically allow 4.5 turns of the wheel to traverse the full distance of the actuator.  This would allow the link to be routed any convenient way, since it is flexible hydraulic hoses.  Will this be a safety issue?  Is the motorcycle requirements lax enough to allow this? What is your take?

I also have solutions for ventilation, heating and air-conditioning (necessary inside a fishbowl with windows that don't open).

I'm looking forward to your opinions.

Geoff S.

Geoff,

Not much time ritght now, but here goes:

Integrated wheelchair/3wheeler:
intriguing idea.  I saw something similar many years ago, more along the lines of a motorcycle with sidecar.  The bike stayed fairly stock, the 'chair rode up onto the platform by the side wheel and clamped into the structure, with controls across to the bike.  I think that has merit, and would appeal to enough people to support a small 2-3 people production company.  Risky though.  Litigation rears its ugly head once again.

Heidts Superide:
Been looking there myself.  It's nice to buy things in one lump, all developed. Heavy though. There just aren't enough applications for an engineered IFS system designed for sub-1500 pound cars to support the cost of developing and marketing same until somebody breaks through and starts production on something useable.

Side-by-side seating:
'Tis all a tradeoff.  TriMagnum, Scorpion, T-Rex, TriHawk, FireAero and others make
(or made) perfectly useable side-bys.  You pays your money and you makes your choice.  And choosing tandem seating cuts off most of your market (if market is what you're after) in one swell foop.

Wheelbase/Track width/Front-rear weight distribution:
My approach: Select a drivetrain. Locate the people. Select the front suspension components. Do a rough chassis structure. Throw in most of the other heavy components where you guess they might end up. (critical step follows here, seldom followed, pay attention...) Do a two-dimensional CG analysis (plot the weight and CG of anything weighing more than 10 pounds, note distande from (arbitrarily) the rear axle and height above ground.  Don’t forget glazing, sound deadening,fluids, and the driver only (at this point).  It helps if the passenger CG ends up fairly close to the vehicle CG.

NOW move the fore-aft location of the front axle to a position that gives the desired percentage of weight on the rear wheel.  This is a tradeoff (surprise!) between rear wheel traction (you want more rearward bias with more power) and overturning resistance (favors more weight on front, which is less harmfull to traction the less power you have). With moderate power (perhaps 70 hp in a 1200 lb car), equal distribution on each wheel is a good target. Then, move the front wheels out to the minimum track that gives the overturning resistance you want.  Tradeoff time again!

The worst (and most common) mistake here is to simply layout components where they all fit nicely and are conveniently supported by a simple chassis structure.  If that happens to coincide with a close-to-ideal overall CG location, wake up, you're dreaming.

Steering:
Approaches to a steering wheel placed well forward relative to the front axle: Switch to forward-steer hubs (the steering arm on the front side of the uprights).  This can usually be done by just swapping sides. Use the double U-joint arrangement used on Shrike (must support the
mid-shaft with a bearing). Consider SteerClear (http://wizardsteerclear.com/).

 It's a chain-drive offset steering box in two lengths.  Costly. I'd shy away from hydraulics in the steering.  No problem with the mechanical design requirements for motorcycle steering in California (none).

Enough I say!
Dave

Exchange about FWD vs RWD and a little about, March 2007.

Dave,

Hey does your car use a electric shifter as seen in magazines to shift and drive this three wheeler. I'm toying with a three wheeler using a front wheel drive and trailing rear wheel.

-Timothy
 

Hi Timothy!

Thanks for the holler.  No electric shifter, I'm more into simplicity.  In this case, a push-pull cable from the shift lever to a bellcrank at the tranny.

Regarding a FWD 3W, that's the next simplest approach (the only thing simpler and still workable is a bike drivetrain at the rear).  That gives you some significant advantages: improved low-speed acceleration (you can put all the power to the ground, I can't), potentially improved overturn resistance (but you've gotta work at it), improved cooling (no need for giant ducts hanging out the sides to get air to the motor), available heating system (for cabin heat, defroster), and greatly reduced tire cost (I use a rear tire every 3K miles, at about $160 each).  Disadvantages: heavier, more work (you really need to widen the front track for adequate overturn resistance), perhaps more cost, and styling becomes a problem with that big lump up forward while trying to keep the driver way low for a low CG.  It can be done, however, witness the Ale (see: www.fuelvaporcar.com).

 Stay with it, and please keep me posted on progress: photos, sketches, drawings...

Cheers!

 

 

 [The following is a continuing conversation with Dan Wyke in Pennsylvania, posted with his permission. 
 Dan can be reached at danwyke@yahoo.com.]

 

Hello Mr. Norton,

I'm sure you get all sorts of people sending you different designs asking for advice so I apologize in advance.  I've been toying with the idea of building my own car or 3-wheeler for a couple years now (I'm 6'9" so I don't fit in any of my ideal sports cars so I figured I'd build my own) and finally bit the bullet and bought a used bike and some other parts to get started.  Attached are 2 pictures of my design(s), keep in mind they are works in progress.  I'm certainly open to critiquing but the main question I have is the value of starting with a full motorcycle back half. 

 

I started with that design as I figured it might be easier to convince the DMV and insurance

companies that it's still a motorcycle as well as get it on the road faster.  However I'm starting to worry about traction as my bike ('84 Kawasaki GPZ1100) only has a 130 series back tire and 120Hp. I'm afraid I'll just get frustrated and wish I had pushed ahead with the larger car tire in the rear.  Any advice/criticism you might have is appreciated.

Dan,

You really ought not to locate the front axle centerline until you have laid out the rest of the components in their preliminary locations and done a first-cut at the CG location in two axes (three if you have any significant side-to-side asymmetry).  As a starting point, assume equal weight on all three wheels (33% on the rear) and see what wheelbase that gives you. Do a CG cone based on the angle of a line from CG to the half-track line, see Riley's website for a discussion on this).  The angle of the CG cone will give you a rough idea of the lateral G at which the vehicle will lift the inside wheel in a turn. 

If you're looking to do some serious canyon-carving or track day driving, this number should be over about 1.4.  More is better.  You'll find that's not easy to accomplish.  It takes serious effort at
lowering the CG and a pretty wide (front) track dimension.  You'll have the same challenge I had getting from the steering wheel to the steering pinion, due to the driver sitting so far forward relative to the front axle.  Two simple U-joints back-to-back works (note the orientation, it's important).  Provide a good sturdy support for the lower shaft as close to the U-joint as possible.  Loads here are high.  Failure here is unacceptable.  Spend the money, get good roller bearing joints. Ideally, you'll want chassis tubes to intersect very near the front control arm pivot points, in order to reduce bending loads in the chassis tubes. 

A truly triangulated space frame will have no bending loads at all, only compression and tension loads.  This is a practical impossibility, but a worthy goal.  Evan a passing glance at the Shrike chassis will show that I chose simplicity over theory in this area, and later paid the price (note the afterthought tubes from the upper control arm pivots to the bottom chassis tube).

I must scale my soapbox and expound on one other, more practical area.  If you are going to spend any time driving on the street, please consider a high center-mounted headlight ala Shrike.  In 106,000 street miles so far, I'm yet to have anyone pull out in front of me because they didn't see me coming.  A bright light 4-1/2 feet high will be noticed.  Yes, it's odd (did someone say ugly?), yes it interferes with visibility in fog.  No it doesn't help much in forward visibility (due to compromises in how it has to be aimed), but yes it works, and yes it's legal (so long as the center of the lamp is no higher that 54" off the ground (last time I checked Calif Vehicle Code). 

Actually, with your high roll bar, it might look cool to snuggle the light below the radiused hoop rather than on top.  FYI I was concerned that light from the headlamp would light up my dash and instruments to the point of interfering with my night vision.  This is true, but not all that bothersome.  You will need to compensate by providing two really effective headlights conventionally mounted, and by not over-driving your lights at night.  An alternative is to kill the overhead light at night, but I think the high light will be seen (and noticed) earlier by oncoming traffic on a winding road. 

The persistent nightmare of enthusiastic and aggressive sporting drivers on the road is the possibility of encountering an oncoming driver who has just made as bad an error in judgment as you have, at the same time and place.  Generally, if this were to happen you would die.  It is only by Grace (and probability) that we survive.  Consider this fact in designing your lighting, driver restraint, roll cage, and overall crash energy management systems.

Good Morning Dave,

I had a thought last night that I was wondering if I could run by you?  I've been trying to design a clean rear axle to adapt a car wheel to a swingarm suspension and use the bike's original sprocket and rotor and haven’t come up with a design I like yet.  Last night I was flipping through the channels and came across one of the "documentaries" on choppers with Jesse James.  I noticed the rear tires they are running and had the thought that running a wide bike rim & tire could be an easier solution.  Some thing you said came back to me when you mention that your bike tire has pretty good lateral abilities and I was wondering if that is due to the convex shape helping to maintain a consistent contact patch with body roll?

Costs aside (which is hard to ignore when a 10" wide motorcycle rims seems to go for $1200+ and the 330mm Avon tires I looked up were around $350) I was wondering if this would be a win-win solution for running a single rear tire?

Take Care

Dan

Dave,

Thank you for all your comments and input.  It's very helpful to get advice from someone how has already been down this road already.  Most people just think I'm nutz. The front suspension geometry and mounting locations are not finalized yet.  I keep tweaking them as the frame, steering, CG etc. all evolve, as I spend more time working on it and reading more about suspension and chassis design.  I probably did the wrong thing in starting the front design with ergonomics for me but it is a limiting factor and you have to start your compromises somewhere, right?  I have done the analysis from the CG and half track line to determine the roll over factor.  The beauty of using software like Solidworks is I'm able to keep pretty good track of my CG as I add components.  I've attached a jpeg of that analysis based on the current design.  If I'm looking at it right I'm coming out with 1.6? 

 

I was concerned about the steering as well and have been playing with that constantly to keep better angles on the u-joints.  That’s why it looks like the tie rod is hanging down in the pictures because I gave up keeping that aspect linked in the CAD file until I sort out the final rack and suspension locations. The frame is constantly evolving due to exactly the points you made. I've moved the second main hoop further forward to try and brace the front suspension and line up with the axle centerline.  Once I figure out the final suspension locations I'll probably go through yet another round of adjusting the frame tubes to align with as much as possible.

 

The safety aspects are something I've been struggling with the entire time.  Part of me has the mind set that regardless this will be safer than a motorcycle, which I was able to justify riding until my son came along then it started to seem too risky.  But on the other side I want it to be as safe as possible for exactly the same reason; I want to be around for awhile.  So yes I'm trying to make it as safe as possible.  I haven't come up with an elegant means of energy dissipation in the event of a crash.

 

I like your suggestion about the high center light.  I hadn't thought of it that way but it would act like a buggy flag in the dunes? I live in Pennsylvania and my understanding is that the laws for motorcycles are very similar so I'm sure I can argue the letter of the law with the DMV but they can certainly make things very difficult for you if they want to.  I guess I'm just paranoid I'm going to invest all the time and money and then end up with something I can not get registered.  but I guess as long as I keep within their guidelines then I should be able to register it eventually regardless of the number of hoops I have to jump through.

 

Thanks again for your time and input.  It's always nice to receive knowledgeable constructive criticism.

Thanks, Dan

Dan

Yes, your numbers are correct, assuming your CG height is 13.5" and assuming that the tires will lose traction at a lateral load equal to the vertical load.  That's what the 45 degree cone is based on.  If your tires would maintain traction up to 1.6 G before sliding, then your factor of safety against roll-over would be zero.  At 1.6 G needed to lift the inside wheel, you're in good shape.

Note the CG cone / half-track tool involves some fairly broad simplifications, all of which will reduce the actual capabilities on the road.  These simplifications include zero suspension deflection under extreme lateral G, optimal suspension geometry under chassis roll, linear tire response to lateral load, constant friction factor, not to mention road irregularities and transient inertial loads.  Also, the breakaway lateral G will be the lower of the rear or (outside) front tire.  The tool is not very useful in calculating actual performance, but it is quite valuable in estimating comparative behavior between alternative configurations.

Dave,
 
I attached the most recent rendering of the project.  I've been playing with the rendering aspects of the software while I try and sort out my front and rear suspensions, it looks a lot less cartoonish rendered which turned up the burner on my motivation!  The frame side structures were added for side impact protection and also to duct air back to the engine. 

Take Care, Dan