BUILDING A CYCLEKART
BUILDING A CYCLEKART
BUILD SERIES no. 3: Front axle
We have used three types of front suspension on our cars: dropped axles with semi-elliptic springs, straight axles with quarter-elliptic springs and a straight axle with a single transverse spring.
Most cars in our inspiration period utilized dropped front axles and semi-elliptic springs.
The Panhard, Fiat and Austro-Daimler have been done this way (matching their inspiration cars). These three are very much the same in concept with subtle variations in detail.
All three use 24” buggy seat springs. These are sold as elliptic springs. One half of the pair will have loop ends and the other will have tabs. To get two matching semi-elliptic halves you will need to by two pairs of springs. Typically, two builders have shared so each gets a matched set. These springs have just the right amount of “give” for our cars. Normal automobile springs or even light trailer springs are far too stiff and heavy. Our favorite source for buggy seat springs is Justin Carriage Works.
All three use drop axles. The Panhard’s drop was bent on a pipe bender. The Fiat and Austro-Daimler have been welded up from cut parts. All three cars use Azusa spindle sets with the stub axles already welded by Azusa. The spindle yokes are welded by the builder onto the ends of the axle. There are three important angles to be aware of when laying out your front axle. Camber is the angle toward the car or away of the wheel relative to a vertical line from the ground. Virtually all cars from our inspiration period had positive camber, that is the wheel leans outward at the top, sometimes quite significantly. Contrast this to a modern F1 car which has obvious negative camber (the wheel leans in at the top). Negative camber adds tremendously to the cornering performance of modern, fat, tires. We, however, are using skinny motorcyle tires. The angle of contact is not critical for these tires. The reasons for positive camber in our inspiration era were improved driveability and control with the very heavy, un-aided, steering of the era. Our cars are supremely light and it doesn’t really matter (within reason) what the camber is. What is important is the correct look! About 3 degrees. Setting the camber is as easy as changing the slant of the cut on the end of your axle. If one is adventurous enough to weld his own spindle sets, the stub axle can be welded at less than a 90° angle (about 7° less) to the king pin and the spindles inclined to offset this. This will change the camber as the wheel is turned, generally starting neutral and becoming more extreme. Watch a modern Mercedes parallel park to see this in action. All this work is fancy but flying over a lumpy dirt track, you will never feel it! Our advice is keep it simple. Use the easy jig-welded Azusa parts, at least that will be one weld you won’t have to worry about breaking...
The second angle is caster. This is the angle the king-pin leans back from that same vertical line from the ground. Caster has a more apparent effect on how the car feels. Tipping the angle of the king pin back helps the wheel track (stay straight). Think of the front forks on a bicycle. Too much slant back however and the steering will get heavy and sluggish when you want to turn off straight ahead.
The third angle is toe-in/ toe-out. This is whether the wheels run parallel, slightly toward each other or slightly away from each other. We will touch on toe angle when we get to steering in a future post. Toe is set with the steering tie-rods and is not determined by how you weld the spindles.
You can see that the depth of the drop on each axle is different determined by how high the chassis is relative to the ground, where the springs are mounted on the chassis, how much suspension travel is allowed and whether the axle is on top of or below the springs.
Springs are attached to the chassis by bolts at their ends. One end must have a shackle (usually the rear end) to allow the spring to get longer as it compresses. If both ends are fixed the spring will not compress or “give” - you will just be making a very jumpy almost rigid assembly. Shackles can be as simple as two short plates with bolts joining them at one end through the spring and at the other through the chassis (or a tab). Springs are typically bound to the axle using u-bolts. It is important to incorporate some sort of stop on the axle at this point to keep the axle from rotating on the springs. The deeper the drop of the axle the greater this effect. If the axle does move relative to the springs, ride height, caster and toe will all suddenly change. The results are never good! If the springs are not parallel but rather are slightly closer together at the front than the back you will have a touch more “bite” when you corner. But again, you’ll probably never feel it on a bumpy dirt track. The Austro-Daimler springs are set up in this way. The Fiat and Panhard are parallel.
The Panhard axle flexes the most of the three and the Fiat the least. A flexing axle will cause the critical angles to change while driving making it feel “loose”. The Austro-Daimler builder is threatening to shorten his drop and move his axle to the top of the springs.
Of all that parts you will make for your car the front axle deserves the most care and close attention.
Panhard axle material: A513 DOM 1.25” diameter x 0.65” wall steel tube. Front wheels: Honda CT90 fronts 17x140.
Fiat axle material: A513 DOM 1.25 diameter x 0.120” wall steel tube.
Front wheels: picked out of a local used parts yard, provenance uncertain. Best guess on one is something Suzuki. The other is a mystery. Suzuki wheel boss shortened to look more like the mystery wheel. Both 17x140.
Austro-Daimler axle material: A513 DOM 1” diameter x 0.095” wall steel tube. Front wheels: Honda Passport rear wheels 17x120.
Azusa spindle sets are available with 5/8” or 3/4” diameter stub axles in various lengths, see page 31in the Azusa catalog. The diameter and length of the these will be determined by which wheels you find and use. If you are lucky enough to find Honda Passport front wheels (17x120) standard inch/inch go-kart bearings drop right in. Only the Peugeot was that lucky for us. Usually you will be faced with a metric bearing seat OD and getting the right combination with the inch axle ID is going to lead to compromise and shims. NB: we have had more (durable) success shimming the OD of the bearing (inch OD) with lengths of split thin wall brass tube than using shim stock on the ID (metric ID).
The GN, Blood Mary and Miller use straight front axles on quarter elliptic springs. The inspiration Bloody Mary was in fact built using the front suspension off a GN. On the GN and Mary, the springs are attached to tabs on the top of the axle. The springs are 36” buggy springs cut in half and leaves removed. These springs are not sold in elliptic pairs like the seat springs. They also came from Justin Carriage Works. Radius rods are attached to the bottom face of the axle. On Bloody Mary there are two AZ 1849 “tubular” tie-rods as radius rods. On the GN there are four AZ 1842 “deluxe” solid rods. In both cases they are angled out slightly to improve lateral location of the axle. Adjusting the radius rods changes the caster. The GN has strong positive camber matching the era of its inspiration car. You might notice (in a less wide angle shot) that the camber does not match side to side. This axle was built for the Peugeot, our first car. The axle cuts were very carefully figured out and the yokes very carefully welded on. Only after was it discovered that the king pin holes on the yokes did not match. Neither one was correct and unfortunately they were slightly off in opposite directions. Welder beware! With this experience, Bloody Mary matches side to side because the cuts were made at subtly different angles. Mary does not have positive camber for more of a ‘30‘s race car look (tires were getting wider). As a note of interest, Bloody Mary is set up almost neutral in caster, camber, toe and bump steer. The GN is opposite to that.
The Miller uses junk yard golf cart springs bolted directly to the underside of the axle. Golf cart springs have a subtle “s” bend to them. The are mounted to the chassis above the line of the axle but with the curve come in under the axle. On this car the radius rods take off from the top of the axle. A stop has been incorporated into the chassis to restrict the axle’s range of motion. This works well to keep the front end calm. Again, the GN is opposite to that! The GN springs and bounds like a wild animal. (There are pros and cons to that but, most certainly, it takes some getting used to whereas the Miller drives in a more expected manner) Bloody Mary was even worse. The inspiration car was notorious for its “difficult” handling and the cyclekart followed suit. On the cyclekart the combination of the 1/4 elliptic cantilever, extreme light weight and overwhelming rear bias conspired to keep the front wheels off the ground much of time. Steering was more a matter of luck than intention. This ALL CHANGED however with the recent addition of friction dampers. These are intended for use on vintage hot rods, T-buckets in particular, and they are fairly heavy. Mary though had weight to give away and getting more of it up front was actually a strong positive. It is the builder’s opinion that the dampened Bloody Mary front suspension is the best to date. This confirms our suspicion that making Andre-Hartford type dampers for the cars is a really good idea. Watch for more.
GN axle material: A513 1” x 1” x 0.083” wall square steel tube.
Front wheels: Honda CT90 fronts 17x140.
Bloody Mary axle material: A513 DOM 1.25” diameter x 0.083” wall steel tube. Front wheels: Suzuki TS50 fronts 17x120.
Miller axle material: A513 1” x 1” x 0.095” wall square steel tube.
Front wheels: Honda CT90 fronts 17x140.
The Peugeot, as mentioned above, started with a 1/4 elliptic scheme. This was not because it was right for the car but rather because the builder didn’t have the confidence to start off blind with the suspension type the inspiration car had. In due course, the “starter” suspension was removed and farmed off to the later GN and a new transverse spring arrangement concocted. This suspension is extremely effective with massive, very soft travel. Unlike the GN the long, soft travel on the Peugeot stays under control. It only needs the eventual addition of dampers to get its leading place back from Bloody Mary. However, it is very complicated with extremely critical geometry. It took a many hours of testing and fussing. I won’t describe it in detail because it will not be generally useful. The simple description is a straight axle with a 24” transverse spring and a Panhard rod. The more complicated description would get in to how the rubber “pots” at the top of the spring are doing much the real work and that it just looks like a rudimentary early ‘20’s economy car suspension. The inspiration was a humble car not a racer.
Peugeot axle material: A513 1” x 1” x 0.083” wall square steel tube.
Front wheels: Honda Passport fronts 17x120.
Thursday, January 20, 2011