Swing arms and related components

The rear suspension arm is also known as a rear pivoted fork, but I think I'll use swing(ing) arm as this member on some bikes (RC30 etc) is not a forked (two armed) structure.

I've been asked to go into more detail of the construction of the swing arm pivot area of the frame, so I'll also cover that in this article.

Swing arm pivot shaft:

As with the other spindles (front/rear axles, steering stem), a large diameter hollow tube is a lighter, stiffer and stronger structure than a small diameter solid shaft. The pivot shaft can be problematic on those engines that incorporate the swing arm pivot into the back of the crankcase. I built an SOS bike around an air-cooled 510 Husqvarna four stroke, and as I recall the pivot shaft was only 12mm in diameter. The small shaft size was one reason why I used a standard-style swing arm pivoting behind the engine. A Rotax thumper (pivot in the case) uses a 16mm/.625" shaft (somewhat better). I generally use a .75" tubular pivot shaft. I believe one of the new supersports 750/1000cc bikes is using a 20/25mm shaft. The trend is obvious. A nice thing about using the .75" shaft is you can get light weight aluminum aircraft nuts in this size that will take about 50 pound/ft of torque.

Swing arm pivot bearings:

The main types of bearing in use are plain bearings (bushings), needle roller bearings and taper roller bearings. The bearings must accept both radial loads and axial loads.

Some lightweight bikes use plastic bushings, but these suffer from rapid wear and limited rigidity (they tend to squish under load). A good quality bearing bronze, such as SAE 660, gives a much superior part. You should probably avoid an "oil-lite" oil-impregnated bushing as they are not as strong as the 660 bearing bronze. The bushing can use a separate thrust washer to absorb axial loads, or can be made with an integral flange to provide the axial bearing material. A bronze bushing running on a hard-chromed large diameter pivot shaft (as in Ducati bevel-drive singles/twins - except Ducati omitted the hard-chrome) has a large load capacity, and can have a very long service life if kept properly lubricated. Accurate sizing of the bushing will provide minimal slop in the bearings.

Needle roller bearings, especially the crowded needle types (no cage separating the needles) can also handle high radial loads. They have no axial load capacity, and will therefore require a separate thrust bearing, either a bronze washer or a needle roller thrust bearing.

Both of the above types of bearing have an advantage in that their diameters can be fairly small, making for a smaller swing arm pivot tube (not the pivot shaft - the actual cross tube at the front of the swing arm). This can give extra clearance for the chain, and lets the pivot be slightly closer to the back of the engine case.

They do have some disadvantages. They can be a pig to remove from the swing arm. They require that the bearing bores at each end of the pivot tube be absolutely concentric. BSA and Honda worked around this by using a metalastic bushing, where there is a layer of rubber between two steel bushings. This lets the inner bushing move in the rubber a bit to align with the pivot shaft. Of course, it also lets the inner bushing move under load. I've tried machining the bearing recesses in the pivot tube from each end and then pressing in finished bushings. I found that with a tight clearance between the pivot shaft and bushing I generally couldn't get the pivot tube bored sufficiently straight to be able to push the pivot shaft all the way through the bushings. Ducati bushings come undersized on the ID, and have to be align reamed/honed after installation. This operation usually has to be farmed out. Both types of bearing need a hard surface to run on. Having a spindle ground, hard chromed and reground isn't cheap, but is needed if the bushing rides on the spindle and you hope to have a good service life. Needle roller bearings have a wide selection of hardened inner races available, but I find that the needle roller bearings seem to have more "shake" in them than I care to have. Plus, the needle bearing gets bigger in OD since it has to accommodate the pivot shaft and an inner race. The inner race/pivot shaft interface is another area where tolerances have to be tightly controlled to avoid adding more slop to the assembly. As a somewhat related aside, I prefer to use bronze bushings (oil-lite is fine here) on shift/brake foot levers instead of needle roller bearings, just because of the extra slop in the needle roller bearings.

The third bearing type is the tapered roller bearing. They accept both radial and axial loads, and generally have a high load capacity. You need to make a spacer to go between the cones of the bearings to set the amount of preload on them (just .001-.002" is plenty). With the spacer you can really tighten the spindle nut without worrying over excessive preload on the bearings. I use a Timken cup/cone set - A6075(cone), A6157(cup). This is a .75" bore with a 1.574" OD (approx 19 x 40mm). The pivot shaft ends up being a good size, and the OD of the bearing is small enough to allow you to use a 2" OD pivot tube on the swing arm. Since there will be no movement between the bearing cup and the frame when installed, the bearing can be shimmed with readily available steel washers. With a bushing or needle thrust bearing a hardened thrust washer should be used. The tapered roller bearings also seem to be tolerant of slight amounts of misalignment, which makes things easier on the home frame builder.

Swing arm pivot tube:

As I mentioned, I use a 2" OD pivot tube, normally of .065" (16g) wall thickness. I make a flanged cup to accept the bearing and press the cup into each end of the cross tube. If the ends of the tube are faced in the lathe, pressing the cup flanges solidly against the tube will do wonders for ensuring things stay in alignment. I tack weld the flange to the tube in 3 or 4 spots. I don't make the pivot tube until after I've completed the frame. This lets me measure the inside of the swing arm pivot area in the frame, and I cut the pivot tube to fit so I can minimize shimming the bearings. I don't have troubles with chain interference with this size tube, but then I tend to use the largest countershaft sprockets I can get.

Swing arm pivot area of the frame:

As requested, I'll go into a bit more detail on how I build this section of the frame.

The pivot is mounted in a sheet metal box structure. This is light and strong. Some frames use a single piece of steel plate instead - this is heavy and weak. This box can be made out of .049" or lighter sheet if it is properly designed. I use the washers mentioned below to spread the load from the swing arm pivot into the light weight sheet metal.

With a .75" pivot shaft being used, I make a .75" ID steel bushing with a 1" OD to be welded into the pivot area. I bought some large diameter washers (about 3" OD) that are about .090" thick that have a hole a bit less than 1". I bore these to be a good fit on the bushing. I clamp a washer on the welding table, insert the bushing and weld them together. If the washer warps a bit I'll lightly face it on both sides in the lathe. These washers will butt against the bearing cones, and need to have a parallel face to the bearing. I'll then run the .75" ream through the assembly to clean up any warping of the bore from the welding. This is a good time to mention that I try to touch up the bushing bore after each welding operation. If you have the parts on the fixture, mounted on a .75" mandrel, and then weld all four ends of the two bushings without cleaning up the bores after each weld, you may never be able to get the mandrel out. It is amazing how little warpage it takes to really clamp the mandrel tightly in a close tolerance hole. I then weld a washer to the end of the other bushing and clean up the bore.

At this time I'll make the sheet metal boxes for the pivot, offer them up to the frame, and mark where the holes need to be for the bushings. I make the holes slightly oversize just in case there is a little warpage when the boxes are welded to the frame tubes. It is a bit of a chore, but I like to make a thin wall, large OD tubular spacer to go inside the metal box, concentric with the pivot shaft. This spaces the sheet metal walls apart, and will let you clamp the pivot washers solidly against the sheet metal when welding. If you can, try to put a couple of tack welds between the sheet metal and the spacer so it doesn't move around. I've also tried welding a washer with a 1" ID hole in the middle of the spacer to locate it on the bushing. Both methods work OK. Do your best to ensure the sides of the sheet metal boxes are perpendicular to the pivot shaft when they are welded into the frame.

Now put the bush/washer assemblies into the boxes, with the washers to the inside of the frame, and reinsert the fixture mandrel. Tack the periphery of the washers to the sheet metal, taking care to limit the amount of heat going into the assembly to minimize warpage. Continue making short welds with plenty of cooling time between welds (try alternating between the two sides of the frame) until the washer is completely welded to the sheet metal box. Now slip the mandrel out and install the remaining washers on the outside ends of the pivot bushings, clamping the washer/box/bushing assembly firmly in place. Repeat the procedure of small tack welds until the outer edges of these washers are fully welded. By tacking the outer edges you can minimize distortion, and avoid pulling the bushes out of line. When the washers are welded you can then weld ONE bushing end to a washer. Now remove the mandrel (wait until everything cools down) and ream the freshly welded bushing. Reinsert the mandrel and weld the other bushing, and repeat the reaming process. If you were careful, you should find that you have two very close tolerance holes that are in line so that the swing arm pivot pin will be a light push fit in the frame. You now have a light weight, strong, and no slop pivot area.

The swing arm(s):

As I've stated earlier, I'm not wild about aluminum as a frame material, and that goes for swing arms too. Some, if not most, of the alloy swingarms that appeared on production bikes in the 1980s were both heavy and weak when compared to a steel swing arm. Such a deal. At the last Sears Point race there was an RZ350 engined bike on display that was equipped with a Parker/GTS style front suspension. The builders had gone to a lot of trouble to make hollow aluminum castings for the front swing arm and upright, and had the raw castings and patterns on display. The swing arm was sooo heavy. I spoke with one of the people involved, and got him to admit that they might have been able to make the parts a lot lighter if they had done a steel sheet metal fabrication. The builders have hopes of selling the conversion, so it looks like the aluminum castings are another marketing ploy, plus there might be some advantage in time/money savings if production numbers were to be large.

There are two basic swing arm cross sections that can be used, round or rectangular. A round section tube will be stronger in torsion and bending than a square section tube of the same gauge and weight. Joe Bolger, a New England frame builder/inventor and scrambles champion in the 1960/70s built a light weight Bultaco MXer about 1970, and said that he used round tubing in the swing arm because he tested both round and rectangular tubing and the rectangular tubing was worse. Unfortunately, I don't have any specs on the size of tubing he was using, but I presume he was trying to maximize strength for a given weight. Foale and Willoughby present some calculations that show that a 2"x1" tube is a close match in strength for the ratio of bending loads put into the swing arm by a sideways load on the wheel. They then point out the advantage round tubing has over square tubing. They also point out that these sideways forces place the greatest load at the swing arm pivot, so a cross section that tapers from the pivot to the axle (as used by Greeves (rectangular tube) and Velocette (round tube)) is more efficient. I've built swing arms with both 2"x1" tubing (fits nicely to the 2" OD pivot tube mentioned above) and 1.625" round tubing with good results. A drawback with round tubing is that as the diameters get bigger, the tubing starts to crowd the frame, wheel and foot pegs in the pivot area.

I think that the best option is to go with a "deltabox" style swing arm, fabricated from sheet metal. Properly designed, you can basically fill the entire area between the chain runs, maximizing the strength of the swing arm for the least weight. You can also taper the arm, reducing the unsprung weight a bit. This type of construction does require a lot of welding when compared to using tubing.

If an external brace is installed you must remember to triangulate it. The U shaped braces that have a couple of tubes at the base of the U running to the pivot tube are largely useless. Triangles, as with frames, are the way to go to design a light weight rigid structure. I think that for most of our projects, a deltabox type swing arm will be plenty strong/stiff without an external brace.

Axle area of the swing arm and axle adjusters:

There are several ways of building the swing arm at the axle. You could use a thick piece of steel plate welded to the tube, as used by Triumph etc. This is heavy and not very strong. You can put a sliding axle carrier inside the tube as did Seeley (round tube) or many modern dirt bikes (rectangular tube). This is nice, but requires some careful machining if a no-slop fit is to be made between the sliding block and the tubing. If you use a good sized axle (something we all try to do, don't we?), the round tube ends up without much metal left in it next to the axle slots, and needs reinforcement. A rectangular tube doesn't have this problem. Nico Bakker used round tube for the swing arm, and welded rectangular tube sections on to carry the axle/adjusters. Instead of a sliding internal block, you can weld steel strap in the slot in the tube (Maico, Can Am) and use an external adjuster. This would be similar to the current 500GP bikes which use a large slotted aluminum block welded to the end of the swing arm. The tube/strap combo could possibly end up lighter in the aluminum swing arms than the milled from solid part.

Adjusters can be a pull/push bolt on one side, a "stirrup" arrangement where the adjuster engages the axle on both sides of the swing arm tube and uses a central bolt at the end of the swing arm, an external cam as used on Bultacos and B50 BSAs, and an eccentric axle carrier clamped in the swing arm (some Kawasaki road bikes, Bimotas). Some bikes (Rickman, 860 Ducati) mounted the adjuster at the swing arm pivot. This does let you clamp the axle very solidly, and saves some unsprung weight, but ends up moving the pivot point further away from the countershaft sprocket and makes the swing arm pivot area of the frame more complex to build. Which ever method you decide upon, make sure it is strong and easy to use.

Damper mounts:

If you are shrewd enough to have your damper(s) near the axle, the load on the pivot area of the swing arm is greatly reduced. Also, the swing arm is subjected to much lower vertical bending loads, allowing it to be lighter. In addition, if you design it right you can make your damper mount serve double duty as a brace for the back of the swingarm.

Standard swing arms vs single-sided swing arms:

The single-sided swing arm (Elf, RC30) is a great idea for endurance racing, as it can allow for very quick wheel changes. I don't think it is too great in a typical high performance situation. The swingarm must be stronger due to the high torsional loads it experiences, and the axle and related components must be must stronger as well. Generally, any one who doesn't have a marketing axe to grind uses a standard swing arm in a high performance application (exclusive of endurance racing).

As always, let me know if anything needs clarification. I'm not sure what subject the next installment will cover, possibly front ends.


Michael Moore

Euro Spares, SF CA


Back to the top frame page
Back to the home page
© 1996-2009 Michael Moore, last update for this page 04 February 2009

For more information contact us by using