Dave Norton's

SHRIKE

last updated 5/28/2006

The story of the Shrike

Anyone who uses any of the information and advice presented here assumes the risk of doing so. Building a road-going vehicle from scratch is not a task to be taken on lightly. Following the advice here in no way guarantees a safe or successful vehicle. There are simply too many variables. This is serious business, and lives depend on its being done right. Enlist the help of experienced people in the detail design and construction, and good luck!

I'm Dave Norton, Senior Product Engineer for Flowserve Corporation, in Temecula CA. I designed and built the Norton Shrike in 1984 in San Diego CA. This document is provided to give those interested in the Shrike a more complete picture of what it is and how it came to be. It is the result of a chance combination of knowledge, need, and opportunity.

My background includes over 40 years experience in mechanical design, including flight control actuators on the Lockheed C-5A aircraft, chassis and suspension design on Indy and Baja 100 (it was a 1000 mile race, but alas only a 100 mile motor…) race cars, motorcycles, taxis, and steam-powered vehicles. I have raced Lotus open-wheeled formula cars, go-karts, sports cars, and motocross and dessert racing motorcycles. All of this with notably little success but a lifetime of enjoyment. I survived 20 years of street motorcycling, which speaks volumes of both skill and luck.

The idea for the Shrike followed a tragic accident in 1978 involving a sad mixture of motorcycle and dune buggy, resulting in the loss of my riding buddy's legs, and nearly his life. Over the next few years, ideas percolated and evolved, combining my sports car chassis and suspension knowledge with my motorcycle experiences. I've always been attracted to minimalism in my personal transportation. My first car was a Morris Mini, and my favorite sports car has always been the Lotus Seven. Form follows function, appearance be damned.

Automobile (or motorcycle) design involves a complex mix of compromises. My goal for the Shrike was to achieve a unique balance of competing factors, suitable to my own tastes with almost no concern for mass market appeal. There may or may not be others to whom my weird ideas would appeal, but even if not, I would have built what I wanted.

Here is a rough listing of my priorities, with 5 being most important, 0 least:
Crash Safety: 5. Accident Avoidance: 4. Lateral Acceleration: 5. Longitudinal Acceleration: 3. Cost to Produce: 4. Individuality: 5. Convenience: 1. Appearance: 1. Operating Economy: 2. Ease of Manufacture: 4. Simplicity: 4. Passenger(s) Accommodation: 1.

I decided early on that I wanted a 1+1 seating capacity. That is, I would be the sole occupant most of the time, but I wanted the ability to carry an occasional passenger. This raises the question of side-by-side vs tandem seating. My priorities place a much higher value on lateral acceleration than on passenger accommodation. The significance here is in weight distribution. With driver only, a side-by-side will be assymetrically loaded, resulting in a lower overturning threshold speed turning one direction than the other. This can be compensated by offsetting the battery, fuel, or other items, but that only makes it worse when carrying a passenger. Tandem seating maintains lateral balance under all load conditions.

The reasons that the Shrike has 3 wheels are two, one aesthetic and one practical:

  • I wanted something really different, unique. A 3W would fill that bill.
  • I wanted something simple, and easily reproducible if I should decide to market kits or plans. In California, and throughout most of the US, a 3W passenger vehicle weighing less than 1500 lbs is legally considered a motorcycle. That fact greatly reduces the amount of equipment required in order to be legal. Bumpers, windshield (and washers/wipers), doors, bodywork, and several other items are simply not needed on motorcycles.


When I started doing concept sketches, I started with a single front wheel and a transverse rear engine. There are advantages to that layout, and it did lend itself more readily to the monocoque (stressed structural skin, no separate frame) construction which was my preference. But as I got closer to the realization that I might actually build something, and that my own fragile structure would be at risk, the idea of a more substantial protective cell became more attractive.

The next series of sketches developed a very simple welded tubular steel chassis structure surrounding the driver, forming a cage within which the driver could be well protected even in a severe crash. The rear engine gave way to the much less expensive alternative of using most of a complete motorcycle chassis, with the drivetrain intact, and a single rear wheel. As the chassis/roll-cage evolved, I realized that the inclined plane or wedge principle could be used to advantage to reduce the magnitude of deceleration during a crash. This principle is discussed in greater depth in my SAE paper written at the time. In essence, the low nose and wedge shape encourage the chassis to slide under the vehicle it runs into, dissipating energy and decelerating the Shrike (and its driver) much more slowly than would be the case in a direct impact. Suffice to say, I never came close to testing the theory in practice.

As the sketches for the chassis evolved, I built a series of paper and straw models of it to test for rigidity and as an aid in visualizing the results. It would be difficult to overemphasize the importance of this step. You can learn so much from a series of quick-and-dirty models built of paper and balsa or straws. The last of these chassis models I covered with paper to work out the shape of the bodywork. It was at that point, with very simple wheels hung on, that I began to realize that this might really work.

The chassis structure is designed so that no tooling is required. The two sides of the chassis are welded up on a flat surface, then raised to the proper angles, and joined by straight tubes forming a triangulated structure. The aluminum floor pan and firewall form the triangulation members for those areas.

The next step was the big one: spend money. I bought a racing seat and steering wheel. That night, I sat the seat down on the floor of the garage, blocked it to an angle that felt right, held the steering wheel where it felt right, and started taking measurements. As I sat there holding the wheel, I could really see the chassis taking shape around me. I could hear the engine rev. I could feel the lateral g. I would really do it. It would work.

The front suspension became a simple matter of finding lightweight components with acceptable geometry. The easiest solution here is the VW trailing arm setup used successfully on millions of bugs and thousands of dune buggys. The system has the advantages of low cost and ease of installation. Four things remove it from consideration, however:

  • Camber control (the wheels remain parallel as the body leans, producing adverse positive camber on the heavily loaded outside wheel in a turn, limiting front end adhesion. This works great in the VW, as it tends to let the front end "wash out", lose traction, before the tail-heavy rear end. This is not what I wanted for a higher-performance sports vehicle.
  • The system is relatively heavy.
  • The track is too narrow, and widening the track requires significant modifications to the components.
  • The torsion bar springs and their supporting tubes cross the car between the front wheels, requiring that the driver's feet be behind the front wheel centerline. My initial CG (Center of Gravity) estimates indicated that I wanted the driver to be placed further forward than this.

Most small light automobiles then (more so now) used McPherson Strut front suspension, most with front wheel drive. Mac struts give neither the camber control I wanted, nor the packaging flexibility I needed. I settled on the same hardware that Colin Chapman chose when he designed the Lotus Seven, one of the most successful kit cars in history.

The Triumph Spitfire was (in 1984) still readily available in junk yards, my entire front suspension system including sway bar and wheels was $200. The parts are light, the disk brakes are oversize for the weight of the Shrike, and I could redo the cheesy stamped A-arms with proper tubular weldments to give just the geometry I wanted. The Spit uses a tricky, typically English weird lower pivot bearing using a bronze bushing with a ¾" Acme thread. I replaced this by bolting a ¾" Heim joint over the upright stud. Works like a charm. I was able to keep the Spit springs and dampers, moving them inboard, being activated by a cantilevered upper A-arm.

A note here about CG location: when starting from scratch on a 2F1R (two wheels forward, one rear) three-wheeler driven by the rear wheel, a reasonable starting point is to keep an equal amount of weight on the three wheels, fully loaded. More discussion on this area can be found in my SAE Technical Paper. The location of the front wheel centerline is determined as follows: layout the motorcycle drivetrain, firewall, rear seat (if present), and driver's seat. Layout the frame around these locations. Place the other major components (battery, fuel, people, bodywork) and estimate the mass and location of all items. Place the front wheel centerline in a position that will give roughly equal mass distribution after including the front suspension and steering rack.

The steering rack was a challenge. I wanted to use a rack and pinion, but the central driving position, with the driver well forward of the front wheel centerline, precluded a conventional offset rack. I was fortunate in locating a Porsche 911 rack, with centrally located pinion. It took two sets of U-joints and a steady bearing to get the steering shaft pointed down into the rack properly.

The body design was worked out in paper, detail cross-sections were made for the simple formed sheetmetal edge and corner pieces, but a body was never actually built. I was having too much fun driving to justify taking it off the road long enough to add weather protection. Besides, it never rains in California! As a result, I spent many a mile cold, wet, and happy!

The Shrike was completed in 1985, and was my daily driver until '89, when the motor (after 82,000 miles of WFO motoring) went BANG at the top of 3rd gear one morning. It has been garaged since, waiting for a heart transplant. Now, the transplant is finally underway. The donor bike this time is a BMW K1200RS, a liquid cooled 4 cylinder, 16 valve, fuel injected, 6 speed, 130 hp monster motor.

This will allow me to push the other end of the performance envelope, and see just how much power the Shrike concept can safely handle. Also, the BMW is such fine machinery that I must now up the level of design and fabrication to a much higher standard. The body will be finished to the original concept, using Alucobond material. Full weather protection, with defroster and heater, will be provided. Estimated completion date: 1 October, 2000.

Here is a running commentary on progress to date:

6/10/00: Bite the bullet: bought the crashed Bimmer as a salvaged bike for $4000. The bike is stripped, mounted on a dolly, and placed in position relative to the chassis. I'm currently pondering how much of the Bimmer to use. Leaving the bike chassis alone and retaining the 1+1 seating capacity of the Shrike, I get an overall length of nearly 14 feet, and a wheelbase of 105". These numbers are in increase of about 1 foot over Shrike I. I may elect to shorten that beautiful (expen$ive) cast aluminum BMW frame and relocate or replace the fuel tank, to save about 6". Much time is now being spent sittin', thinkin', sketchin', and figgerin'…

7/10/00: The layout is progressing. The bike is stripped down to what I will be using. I've dumped the ABS unit (about 20 pounds!), retained the fuel tank, rear fender and light group, and center stand. The bike is in place on the floor in it's proper location relative to the chassis. I now have a (relatively) simple subframe designed to carry the rear suspension loads into the Shrike chassis. A Recaro racing seat has been selected. Pedals are next.

8/8/00: Layout nearing completion. The critical brackets connecting the chassis tubing to the Bimmer frame are being machined. Purchased the seat (Corbeau), steering wheel (Momo), pedals (Wilwood), belts (Simpson), instruments (VDO) and other bits. Present plan for the speedometer is to use a GPS, wired into the 12V system, with backlit display. No joke, they give an extremely accurate speed readout, and are much less costly than a VDO or S-W electronic unit with toothed wheel and sender.

Space in the cockpit is really tight as well. The (35 year old!) Spit dampers (shocks) are getting tired, and will be replaced by Konis. I'm not happy yet with a rear bodywork concept, so I'm modifying the 1/8 scale model (built 14 years ago to work out the original chassis design) to the new wheelbase, and will use it to mock up the new rear panels.

Body material will be Alucobond, as originally planned. I hate fiberglas! Still undecided is whether to use the Bimmer rear fender & lighting, or dump it. I'll have to do the weight and balance numbers soon, to determine for sure whether I can leave the front wheel centerline as is or move them aft a bit. Must find a 500 lb capacity floor scale… Total cost so far for the rebuild is pushing $5000, which was what the original Shrike cost complete!

6/15/2002: The best laid plans of mice and men…
I experienced a gran mal seizure in September, '00, caused by a dormant glitch in my brain that decided to become not-so-dormant. The seizure happened to happen during a backpacking trip, 6 miles from nowhere, 9,000 feet up in the Sierras (Tuolumne Meadows) and I had to be airlifted out, but that's another story… It turns out that (so far) this has been a one-time occurrence, and life goes on. I was a bit bogged down for awhile, adjusting to my medication (Dilantin). They say I'll be on it for life. We'll see. What do they know!

The Shrike is progressing, albeit slowly, ever so slowly. The engine support structure is done, and the engine has run (always a load off, when you find that the motor really does run after sitting for so long). The shift mechanism is in place. Next is the brake lines and instrumentation. I located (thank$ to Brother Dan!) a new instrument cluster from the K1200RS, so many niggling problems are eliminated there. The support for the steering column bearings will have to move, no big deal.
Onward!

7/26/03
Much water over the bridge in two years. After lengthening the wiring harness to the new instrument cluster the engine would no longer start. After nearly a year of troubleshooting, continuity checking, translating German wiring diagrams, checking relay functions and bypassing various interlocks the engine remains mute. I decided to move on with the bodywork and trailer the vehicle to Irv Seaver Motorcycles for their diagnostic computer check.		 A check of CG with 95% of the mass now in place shows 46% of the weight on the drive wheel, without driver. This is fine for drive traction but far from optimal in overturning resistance. More calculations will follow. One possibility, should it prove necessary, is moving the front wheel centerline rearward. If I do this, I will also chuck the Spitfire parts and switch to the more modern (!) and readily available Chevette parts.

Designing and fabricating the Alucobond (actually Dibond, another product by the same company) panels has been one of the most enjoyable aspects of the project so far. What fun! I'm now installing lighting, horns, battery, and a few remaining body components. I am extremely susceptible to the Paralysis of Analysis, getting bogged down in the minutia, and remaining stuck until either the final elegant solution presents itself (as was the case with the shape of the rear body cover) or I reach the Just Do It point. The uninitiated have absolutely no concept of just how much time and energy is consumed in this stage of each design and decision point. I have the utmost respect for those who have developed the self discipline to maintain momentum through these inevitable pauses in the overall process. The beat goes on…

9/17/03
On the Road Again!!!

Monday was official On the Road Again day, with the first trip to work and back. Several challenges have arisen, chief among them being cooling and front suspension feel. My elegantly simple concept for the rear cover looks great (I think), but severely restricts air flow out of the twin radiators. Combined with being hidden behind the barn-door of a firewall, and the lack of radiator inlet ducting, cooling was woefully inadequate and had to be addressed pronto.

Opening up the engine cover just behind the radiators took care of getting the hot air out, and I have mocked up a cold air inlet duct on one side to judge effectiveness. I also added a rather large engine oil cooler (core approx. 10" square, cross-flow, 3/8" NPT in/out) with a rubber scoop hanging below the chassis to get some ram air. The A/C condenser fan from my crashed CRX pushes cold air up through the cooler. With the oil cooler and ducting to one of the two radiators, engine temp is manageable so far. I won't subject it to a long hot run until I get the second radiator ducting mocked up.

The front suspension difficulty is still a factor. It acts like zero caster. The steering is dead: no self-centering effect at all. It also "hunts", requiring constant minute corrections. None of this was present in the original Shrike: it was almost literally hands-off at 100. I raised the upper A-arm inboard pivot to the upper point to increase ground clearance, forcing more deflection in the jounce direction (the A-arms point downward considerably). The effect is increased track change on symmetrical bump, which will tend to break traction in bumpy turns. I will recheck bump steer and roll steer, as they may also be aggravating the situation.

10/14/03
Development Continues:

Never a dull moment! With about 500 miles now on the clock, I haven't had an uneventful drive yet! Notable moments include loose lug studs on the rear wheel (twice! It turns out that the seating torque required to keep them tight is quite high: 76 lb-ft per the dealer), a loose radiator catch tank cap, loose A-arm spacer shims, low tire pressure, NO tire pressure, occasional high coolant temp, various bits of temporary ductwork coming from together, and a couple of real 0h-shits. The first involved filling up with gas on the way to work and parking in the back lot. 20 minutes later I got a call saying my fuel tank was leaking. Badly. Cold fuel in a tank perched atop a hot motor on a hot day, expanding fuel was drip-dripping onto a puddle ten feet across! I suspect I had crossed the two fuel tank vents, one to the ground and the other to the carbon canister. No confirmation yet, as I've been reluctant to fill the tank full.

The other incident is comical. There is no floor panel yet, as that will be one piece and will need to be done at the sheetmetal shop. I made a grocery run to Costco and bought a bit more stuff than I expected. The load filled the shelf behind the seat with a couple of items sort of cradled by the seat belt, between the seat and the side panel. On the way home I came to the last signal out of town and braked to a stop beside one of Riverside County's finest, to the sound of a clunk as a skinpak of three quart jars of pasta sauce sluiced its way down and onto the street! Fortunately I had time to unbelt, reach down through the lack of floor, retrieve the three loose jars, stuff them in my open jacket, and proceed just as the signal changed. I never looked over (well, up, really) to see if the good officer was watching this little drama.

I finally was issued my license plate and insurance policy (on the same day!) and that night Ellie and I celebrated! 'Twas a long uphill battle, but the Good Guys won again! I was so elated I popped for the vanity plate "SHRIKE1", to be ready in 2-3 months. We chortled over the possibility that our troubled grandson may end up building the plate! One finds one's levity wherever one can!

Back to Development:
The flat front tire yesterday morn prompted a replacement of the tires. The excellent Yoko' A008 slalom tires are no longer made in my size, so I upped the ante to their A032 (at $122 each. What the hell, it's only money!). The change from 175-50/13 to 175-60/13 will increase the rolling radius by about .7", which leads me to return the front suspension upper A-arm pivot point to the higher location I used before. This will return the A-arm angles to a much better situation, lower the ground clearance and raise the roll center considerably. I expect a return to the excellent tracking and stability feel of the first version. I still have not measured and laid out the actual geometry on AutoCAD to see just how bad the geometry is now.

Front wheel bearing grease seals: I asked Dave Bean (supplier of all parts English in support of the small horde of Lotus people on the left coast…see him on the internet) if he had the felt grease seals used on these Spitfire/Lotus hubs. He recommended a rather trick all-metal replacement that seems to be working out all right.

I'm still undecided as to whether to bite the bullet and move both coolant radiators down to just behind the side air inlets, for a much more direct air flow path. That will mean adding a bit of water piping and moving the pressure cap to the top of a new pressurized coolant reservoir. Some significant work, but it may make the difference between being able to guts-ball it up to Palomar Mountain in 90° heat or not. Another probable mod is the addition of a separate coolant overflow tank and fuel overflow tank, where the BMW battery used to reside. Perhaps.

I treated myself to a birthday present: the Vector FX! (?) accelerometer. Initial trials in our Toyota Previa indicate that this will be a useful (and fun!) diagnostic tool. Properly calibrated, it will provide G data in two planes, braking distances, 0-60 and ¼ mile times and speeds, and estimated horsepower (derived from time vs. rate of acceleration). All for under $150, with literally 5 minutes of installation and calibration. Impressive.

More anon…

Dave Norton, builder and driver of the Shrike, can be emailed here.

You can get your own information package from Dave Norton.

Go back to the Shrike's main page.

 "Ask Mr. Shrike"

New section!  Search NortonShrike.com.