The Development Shop and Clubmans TT
The primary aim of my development shop at BSA apart from improving standard machines was to win the Clubmans TT. We were not supposed to be interested in International racing, other than in connection with the MC1 project. The scrambler motor was considered powerful enough. For the Clubmans models we were always searching for a little more speed and of course reliability. This went hand in hand with American class ‘C’ AMA racing, since what was good for the Clubmans was also good for class ‘C’.
BSA did not dominate the early Clubmans 500cc TT races which were won mainly by Triumph and Norton. The 1949 race for 350cc was BSA's first victory. Many BSA 350's were entered and few finished but that year the winner was a BSA.
They had plunger rear suspension, leaked oil, and tended to blow oil out of the engine breather, to the extent of running low on oil, causing seizures etc. Nevertheless, the two or three best BSAs on that occasion did very well, competing against Velocette KSS models and 350 Norton Internationals. In 1952 BSA had hoped to provide a challenge in the 500cc Clubmans with an all alloy twin engine which would be promoted as Gold Star Twin if successful.
However the regulations prevented their use as it was not yet a production model. A few standard A7 500cc twins were built to Clubmans specifications, they pushed out 36-38bhp which was about the same as the production Manx Norton, but in fact were unreliable and had a peaky performance, also did not handle as well as the single. In 1953 it was the same story, the old CB350 did quite well as did the 500cc singles of which there were two or three entered.
The 500cc BSA twins were getting pretty unpopular by that time because when the crankshaft broke, they often destroyed the crankcases and fetched the rider off. If this happens at 100mph it’s pretty scary. Triumphs were still going strong with their 500cc twins although they had their share of blown engines too. They won by the use of specially prepared hand picked machines and some good riders. By 1954 the Gold Star had new heads with big area finning, eccentric rocker spindles, shorter connecting rods, tubular push rods (instead of solid alloy) and Nimonic exhaust valves, a timed crankcase breather (which controlled the tendency to blow oil out on to the road and rear tires) and needle roller bearings on the gearbox layshaft. So far as factory involvement in preparing machines for the Clubmans went there was none, however we used to maintain a workshop in the Island for the benefit of all riders of BSAs. The riders who showed the best in practice might get the benefit of someone like Cyril Halliburn working on their motors. Triumph and Norton did the same thing.
Triumph had Ernie Nott an ex- TT winner working on some of their customers’ bikes. I was in the Island workshop keeping an eye on technical problems and helping Charlie Salt with his TT machines. Before coming to BSA Hopwood had designed the Norton twin, its outstanding feature was the way the cylinder heads were splayed, cut at the front giving better cooling and more fins where needed. This feature was patented, so of course could not be used by BSA.
Nevertheless the new A7 he designed for 1952 was a good rugged engine compared with the earlier BSA twins. It had a weak point in the timing side main bearing and rather small tappets, they wore more rapidly than they should, and also a tendency to vibrate and break crankshafts. We carried out a lot of tests to improve the timing side main bearing trying an all white metal bearing, a copper-lead bearing and a needle roller bearing. This last was far and away the best but required a different oil feed to the crankshaft to lubricate the rod bearings but as the factory did not want to make a change, we continued with the short life main bearing. The crankshaft breakage problem was really due to two things, the diameter of the crankshaft rod bearing journals gave a natural frequency of vibration that coincided with 6,600 & 3,300 RPM, any prolonged full throttle running at 6,600 rpm soon broke the crank adjacent to the rod bearing journal.
A member in the Group Research dept. gave this matter some thought and came up with the answer - a better finish in the fillet radius each side of the big end journal. In production the journal radius was done by the same grinding wheel that ground the journal, this meant that tiny microscopic scratches were left and this is where the fatigue fractures invariably started. His solution was to roll under high pressure a ball around each journal, leaving a perfectly smooth finish.
The layout of the Test shop.There were five Heenan &Froude hydraulic dynamometers and one electric dyno, an elaborate set up ideal for research work. The electric dyno could be reversed to drive the engine you were testing. Everyone was half scared of it, except Arthur Bridgewood, he used it a lot. He insisted that there was always an apprentice standing with his hand on the big knife switch to cut it off dead, if an engine blew up whilst running, current would continue to flow rotating the already damaged engine to destruction. The operator had to be wide awake and act quickly if something broke or seized. We eventually traded this dyno to Research for a nearly new DPX1. The other dynos were two DPX2, one DPX0 and two DPX1. When the MC1 came into the shop we had a new high speed dyno- a DPX3 .
Originally all the dynos were driven by chain from the engine sprocket, a very inefficient method, but simple. We used a 2:1 ratio to keep things simple. Anyone who favours primary chain drive should attend a test using chain drive. At about 6000 RPM take a reading, then get a squirt oil can and oil the chain, the engine will pick up an apparent 2bhp.
Heenan used to recommend a shaft and two universal joints like in a car, but we found them unreliable, the universal joints used to break up and fly all over the shop like shrapnel. We eventually went to a bonded rubber coupling which if lined up carefully gave good results. All the dynamometers were arranged in a line with their axis parallel, very neat, but very impractical, as the blast of cooling air and exhaust blew from one to the next, making it impractical to use more than one engine at a time, due to blast of air and if testing a race engine the noise and fumes, to say nothing of oil.
To try to get away from these difficulties and to control the blowing oil we had some metal shields on castors made which could be rolled around behind the engines. When cooling air blowers were turned on the shields tended to move around, they soon got coated in oil and one had to avoid brushing against the oily surface. This system remained in used for some years until the shop was demolished. The development shop was once a stable, it was quite old, with over 30 foot ceiling and heated by steam pipes. In summer the only cooling was by opening ventilators.
Originally there was no provision for letting air and fumes out, so a rather expensive sheet metal ducting system was built with a 6 hp fan drawing air from just behind each engine. Cooling air was blown over each engine by a powerful 'squirrel cage' blower, giving an 80mph blast. This cooling air was ducted in from the roof and it was possible to select cold outside air or warm inside shop air by operating a flap valve in a ??? WY' joint by means of a cord. When you run an engine under full power on a dyno you have to be wide-awake, if it starts to tighten up or the exhaust note changes you shut it off immediately. If there are two engines running, it is difficult to tell which is about to falter.
There was no objection to running an engine on the shop exhaust system for 'running in' purposes, this running in was an important part of testing a new engine, it will not deliver its best power until it is 'broken in'. Once a month we would take an engine at random from the production line and after a couple of hours of gentle running, take a full power curve, if it was down for power we would strip it to find out why. Should it be badly down on power we would take another engine from the same batch and test it, if it also was bad, we had to get that particular assembly line stopped until the fault could be located and corrected. Needless to say this was an unpopular action, but was necessary to preserve (the factory's reputation.
The only time I can remember actually calling for a halt to production was on the A10 line and later on the C12. Usually what happened was a cumulative effect due to tightness, maybe a worn tool, making one part slightly oversize and a corresponding part slightly undersize. Sometimes an extra hour running on the dyno would be enough to restore the power.
All Gold Stars and Road Rockets were tested by the engine test dept after a period of running in. The cylinder heads and cylinders were removed for inspection and individual power curves issued with each engine. Some BSA production engines were always consistent and actually gave more than average power. The B31 for example always gave at least 17bhp often more. The B33 500cc standard engine was another consistent performer giving 22 - 23bhp when new. The twins were not so good, the A7 500 was better than the A10. The B 31 was a good little engine, a 350cc single, with cast iron head and cylinder. By fitting a few Gold Star parts they were easy to improve. By merely fitting Gold Star cams (touring version) and a Gold Star silencer they would put out 23bhp. By going a little further, fitting eccentric rocker spindles, lighter push rods and tappets, shorter connecting rod, shorten cylinder to suit, Gold Star piston giving 8:1 compression in place of 6.5:1 they would give 27bhp and almost 100mph at no extra cost since these were all production parts. The B33 500 was much the same and could easily give over 30bhp.
Ariel used a BSA A10 in their Huntmaster and one day they complained of lack of power from these engines. I requested them to send me a 'bad' engine and I would test it on our dyno. Sure enough it gave only 27bhp. So we stripped it to see if there were any obvious faults, then gave it another 1 hours running and it was a little better. I noticed that they were using Ariel exhaust pipes and silencers which were different from the A10, larger diameter and slightly shorter and different silencers. We fitted standard BSA pipes and silencers and immediately got 31bhp by fitting one size larger main jet the power was just under 33bhp. We informed Ariel with what we had done but they did not seem keen to change things. Perhaps they liked the appearance of the larger diameter pipes.