My Douglas T35
Frame Repair - November 2006
Account & Photographs

If you have arrived at this page via the discussion forum no doubt you have read the initial post about this subject and have also read the reponses and advice proffered by those who have had similar experiences. If not and you have happened upon or have been referred to this page by others, as a precursor it would help if you read the forum discussion before continuing to read this narrative. The discussion can be found here.

On the operating table with the lifting sling still attached for security, and yes! - it's not obvious in the photograph but it does have three-point slinging - there's a rope sling at the head piece in addition to the webbing sling looped to the foot pegs.

In stripping down, the petrol tank was first removed, then the rear wheel, chain guard and sprocket cover. The rear sub-frame was disconnected, the saddle nose bolt removed and the whole rear guard assembly removed as a unit and moved to storage.

The wiring terminals were carefully tagged as the wiring loom was removed and coiled. It was then wrapped in cloth together with the voltage control unit which had been detached from the tool box and the distributor cap and taped to the front r/h front fork leg. The throttle cables were similarly coiled and wrapped together with the carburettor slides and also taped to the front forks. The carburettors were then removed. The clutch cable was detached from the control lever, coiled and attached temporarily to the motor. Removal of the tool box, battery carrier and generator followed, the latter being removed to establish why it has failed to generate and to rectify whatever is at fault, then the torsion bars and linkages to the rear fork were removed and the rear fork frame itself detached which as anyone who has been through the process already knows, is a bit of pain involving removing the retaining pins and drawing the bushes.

Taking the weight of the partly dissembled bike on the chain hoist the centre stand was removed, the front bench table taken away, the rear one repositioned forward and the bike lowered on to wooden blocking on the table with the bike bearing on the two bottom frame tubes. The three point slinging was maintained for safety with the front rope sling being removed from the steering head and re-attached to the frame top tubes. The top fork yoke was then removed and the head released surprisingly without loss of any balls, not even from the top bearing! The complete front end with the wheel still attached was then carefully jacked down to the floor and wheeled into storage.

The main webbing sling was removed from the foot pegs and re-attached to the frame behind the rear frame lugs to allow the removal of the foot pegs and their connecting through bolt which doubles as support for the motor. The front engine mounting through bolt was then removed and the motor lifted from the frame. Disassembly was complete - the frame was bare! 

The frame was mounted on one of the tables in use as shown in the photograph at the right. A 10 mm plate was fabricated and bolted to the table and to the frame using the torsion bar anchor bracket fixing points. A spacer tube was bolted between the two engine mounts using the engine through bolt and the broken frame tube was stabilised with a vertical support 'post' bolted to the table and to the exhaust pipe fixing bracket. The rear end of the frame was supprted on wooden blocking at the correct height off the table and securely clamped in position.

Prior to dismantling the bike an inspection hole was drilled through the 1" (25.4 mm) o/d frame tube next to the fracture (the hole is visible in the photograph) in order to determine the tube wall thickness and hence its internal diameter. As near as could be determined without a vernier it measured 0.125" or nominally 3.0 mm, indicating an internal diameter of .75" (19.05 mm) so a piece of tube of these dimensions was assiduously sought and ultimately found. However, the initial measurement proved incorrect  for when the existing tube was cut it was found to have a bore diameter of 20.6 mm indicating a wall thickness of 2.4 mm compared with the expected 3.0 mm. Upon later inspection it was revealed that the error was due to burred edges about the hole. Moreover and surprisingly, the tube was found to be welded seam tube and not drawn tube as expected. The replacement tube was turned out to match the bore dimension of the existing to avoid having an internal step at the weld.

The next task was to bore out the broken frame tube from within the cast lug that also provides the connection to the front down-tubes.

As can be seen in the photograph, with the table turned on its side, the mounting plate doubled as an ideal base for the drill press. Incidently the press and the two speed Sher drill was picked up in a garage sale just a day before a decision was needed on how to mount a drill  for just $A30.00 the lot! The bargain drill coincidently is an identical twin to one already available and originally intended to be used for the job. It proved to be a winner.

Two cutting tools were used to drill out the broken tube from the lug. Both are seen in the photograph at the left. The first used was the non-fluted 4-edged piloted end mill shown lying on the mounting plate. The brass pilot was turned to fit within the bore of the tube with minimal clearance. Due to the lack of fluting the mill tended to jam frequently on the swarf - this meant progress was slow as frequent withdrawal from the lug to clear the swarf was necessary. The mill was used only until the arbour interfered with the mounting plate. This occurred at a depth of about 1.5 - 1.75" into the tube. The drilling was then completed using the broach bit shown in the drill chuck which of course self cleared and penetration was much quicker leading one to wish it had been employed sooner. Drill speed for both tools was limited to about 450 - 500 rpm by reducing the 240 volt power feed to the drill to 150 volts. initially, then to 175 volts using the transformer which can be seen in the top right corner of the photograph above and to the right.

Returning the table to its upright position, the broken end of the fractured tube was cut using a hand hacksaw. A scarfing of 60 was employed and a matching cut made to the new extension piece and fitted closely before milling it precisely to the required length.

 The sawn ends of the existing and new tubes were prepared for V-butt welding by hand file.

Welding was carried out using the MIG welder which can be seen in the photograph above and to the right of this page. About 120 to 130 of the circuferential weld was done initially and completed following removal of the frame from the its mounts on the table. The aluminium bar mandrel was successful as a backing plate for the weld and was removed after the weld was completed with just a couple of sharp blows on a steel bar passed down the tube from the rear end of the frame. The surface of the mandrel was virtually unscathed by the welding process.

The photograph at the left is of the welding of the tube in progress and on the right the finished repair after cleaning up the weld. Grinding of surplus weld material was achieved using a 10 mm high speed pneumatic belt sander. The photograph has been retouched to eliminate flash shadows and reflections. The repair will be practically imperceptible after repainting, at least seen from the top side anyway and I harbour no personal wish to see the underside ever again whatever the circumstance!

The photograph at the right shows the brazing of the replacement tube to the lug in progress. Silver brazing and normal flux was used. Full penetration was achieved quite readily.

At this time (10 November 2006) the frame is being grit blasted and is expected to be at the paint shop together with the chain guard, centre stand and a few other ancillary parts within a couple of days and hopefully it will all be repainted within the next week enabling the reassembly of the bike to begin. I can hardly wait to be aboard again, it's such great riding weather down here currently albeit as dry as hell!

Thanks to all who have proffered advice and help on this project and a special mention of my friend and 'engineer' adviser and doer of the difficult bits - thanks heaps Humby!