Okay boys and girls, those of you who’ve been paying attention will remember that last month I’d sawed up two perfectly-good motorcycles, and rearranged them more than a little.  The outcome was a Whole which, to my eye at least, is more than just the sum of its parts.   Having fortuitously discovered a unique yet aesthetically-pleasing Chopper Chief concept, my next task is to lash these parts together in a fashion that will hold up under road use.

Now, frame-building is another of these jobs which contain traps for the unwary.  I have, however, come to the conclusion that frame building is easier than frame repair.

Here’s what Bruce Palmer III's "How to restore your Harley Davidson" (Motorbooks International) says on the subject.  I haven't quoted the passage in its entirety, just those parts relevant to this discussion.

Frame repair.

Serious bends or twists, broken or rotted tubes, and other major damage to the frame should be left to a highly qualified frame repair specialist.  So, too, should be the disassembly of a frame to replace a mismatched component. 

Unfortunately, these fellows are few and far between.  First, he must have a frame table. Without a frame table, the frame cannot be held true while repairs are being made.  Ten mules and a blow torch just don't make it. 

To understand why the table is a must, visualize the frame in its entirety.  If the neck has been bent, the whole frame has been pushed out of kilter to some degree or another.  When the neck is brought back into alignment, the rest of the frame also moves.  However none of it moves back to the original position unless it is guided.  Only the table can hold the frame tightly enough to keep the rest of the frame from moving and give the repairman an accurate surface from which to make measurements. 

The repairman bolts the transmission base to a fixture and starts from there to accurately set all other points on the frame.  Each point is checked and secured before going on to the next point.  It has happened more than once where the motor could no longer fit the frame after it was straightened without a table.

For frame-building, on the other hand, all you need to do is ensure the two wheels end up in the same plane and track along the same line.  This can be done independently from making the engine fit the frame with sprockets aligned, because engine mounts can be added later.  Splitting the job up into separate steps like this makes it a lot easier compared with trying to get wheels, engine mounts and sprockets to all line up simultaneously.  For one thing, I shouldn't need a proper frame table.  A simple jig will suffice, because I'll only be welding, not pressing, clouting, pulling or heaving.  I won't have to worry that pulling in this tube is going to push out that tube, and so on.

Click on pictures for full size
 
 

Pic. 4. Frame jig welded up out of steel channel, levelled, and with centre line scribed down the middle.
The lower rear Ariel frame parts and seat-post have been assembled onto the jig, and the top tubes and Kawasaki steering head await in the foreground.  Also in the foreground is the mandrel for the steering head.
 

Pic. 5. Close-up of the frame top tubes and steering head, with mandrel fitted in place.  The various turned slugs for these joints are also shown in this picture.
 
 
 
 

(1) Jig the frame pieces to ensure the wheels will be vertical, then weld the frame parts together;
(2) Weld on engine and gearbox mounts (starting with the gearbox) in such a way that engine sprocket, countershaft sprocket and rear-wheel sprocket will all align yet have sufficient chain clearances past tyre-walls, frame tubes etc; 
(3) Adjust rear wheel rim offset to get the wheels tracking along the same line.

Note that this sequence of steps for a scratch-built frame may well be different from the sequence for restoration of a stock frame, although Bruce Palmer III did not elaborate on what this sequence should be, preferring instead that we all go to a "highly qualified frame repair specialist"  (a bit hard here in the Fiji Islands!). 

I think the main difference between my method above and a typical frame restoration method might be that restorers often treat rear-wheel rim offset values as if they'd been handed down to Moses on tablets of gold.  Personally I think that rim offset is arbitrary and can be any value you want.  The only critical thing about rear wheels is that their chain sprocket should end up directly aft of the gearbox sprocket.  This is achieved by making up suitable spacer bushes for the rear axle.  Then, and only then, do you have a look to see if the rear rim is tracking along the same line as the front rim. 

With any luck, it will be.  If not, then adjust all spokes to move the rim over a bit, using standard wheel-building techniques (see, for example, Paul Bartholemew’s VI articles on the subject).  Easy enough to say, though perhaps hard for a novice to do, but I’ll have a crack at it.  My point is that it’s more critical to first ensure the rear sprocket will have room to get a chain through the space between frame tube and tyre-wall and line up with the gearbox sprocket.  The wheel offset can be dictated by that, rather than vice-versa. 

Restorers who send wheels away to a wheel builder along with an instruction regarding the rim offset will need to assume (1) that they know the correct offset (did you measure it before disassembling that wheel?!!?), and (2) that everything else about the bike is straight and in good order.    If you suspect that either of these assumptions may be violated, then you might be better off to true the wheel yourself in its actual bike.

To implement my frame-building method, the first thing I needed to make was a jig.  I did this using 3" steel channel, and had fun trying out my recently-acquired garage-sale arc welder.  The results are illustrated in Pic. 4, and it hasn’t collapsed yet so I must have done something right.

Next step was to put in place the lower rear frame components, axle, and wheel with its tyre, to figure out the height and location of the rear axle.  The frame components were held by hose clamps to adjustable stands.  Each stand used a piece of steel channel for a base, with a long vee-headed bolt threaded through each end held by locknuts to adjust their length (Pic. 4, on the left).  I adjusted these stands to give a satisfactory frame ground clearance, then clamped them into place with G-clamps.  Later, when I was sure their position and height was what I wanted, I tacked them with a bit of weld.

One-inch steel strip was bent to make axle stands and welded in position with enough length to roughly suit the height of the axle when a tyre-shod wheel was put on the “table” between them.  Onto these stands I mounted chain-tensioners taken from a Triumph swing-arm.  My axle sat inside the loops of these tensioners, and each axle end could be moved upwards or downwards by turning the tensioner nuts until the axle was (1) the correct height off the table surface to suit my wheel, and (2) level.  After sighting with a square to ensure the levelled axle was also perpendicular to my scribed center line, braces were welded in to stop the axle stands from moving out of alignment.

Strengthening the 741 rear frame

Narrowing the 741 forks

The general idea was that, once all the rear frame components had been trued to the center line and clamped in place, using a plumb line to ensure that everything was centered and reasonably vertical, the steering head could be slipped on and the pointer used to “point” to the center line.  Once this was done, I could be satisfied that the two wheels will be in the same vertical plane. 

Kevin’s machinist turned up mild-steel slugs for me to suit each of the joints in the frame tubes that needed welding.   After cutting up the Ariel frame to stretch it out bigger (as described in Part 6), welded joints were now required in the main top-tube, the seat down-tube, the two bottom tubes beneath the gearbox, and the two rear top-tubes on either side of the rear wheel. 

My technique to weld these joints follows that in Mike Arman’s Custom Harley Cookbook, whereby a solid mild-steel slug extends at least 1.5” into the abutting tubes.  I decided to drill a 5/16-in. hole down the centers of these slugs rather than leave them solid, just to allow some space for heat to escape while welding.  Four 5/16” holes were drilled through the tube ends within the 1.5” end zone, in an arrangement that placed holes at 12, 3, 6 and 9 o’clock around the tube, with the 12-&-6 and 3-&-9 pairings of holes offset from each other by about ¾”.  Lastly, the cut tube ends were bevelled off at 45 degrees to provide a vee in the butt joint. 

Using the various adjusters in the frame stands, I made sure it was all jigged so that the things which should be vertical were vertical, and the things which should be horizontal were horizontal.   I found, for example, that when the seat tube was vertical, the lower engine tubes were not horizontal when a level was laid across them.  This was the Ariel factory's fault, not my fault.  However I decided it was irrelevant, and could be allowed for later when I came to make engine mounts.  What mattered now was that the steering head should be vertical, and the rear axle should be horizontal.  The tubes in between can do loop-de-loops for all I care.

Once I'd got steering head and axle where I wanted them, the distance between cut tube-ends on each "stretched" joint was carefully measured, and a matching-diameter tube was sawed to this length.  Taking it all apart again, I slid these tubes over each slug to sleeve them out to the same outside diameter as the frame tubes on either side of the joint.  Then I reassembled and checked all verticals and levels again.

Pic. 8.  Just checking to see if the engine will fit.  It would be a bit upsetting if I finished welding and found that it didn’t!   You can also see the bright metal where slugs are bridging the gap between tube ends in the “stretched” portions of the frame.