Words by Phil Price
With the exception of a few gems from the BBC, free to air television is pretty much rubbish everywhere. Certainly this is true in Australia and New Zealand where I spend the bulk of time. I sadly admit to catching a few episodes of Australian Ninja Warrior, a kind of sports game show where pumped individuals test themselves against a kind of confidence course carefully designed to have them falling unceremoniously into the drink. The contestants get to have support from a friend or family member who runs along beside them offering advice on strategy, motivation, and often condolences should they fail. The failures are of course the best part, the crowd with hands over mouths, tears, the compères offering insincere condolences. It’s a fairly pathetic formula that apparently makes good tele. What becomes quickly apparent is the undeniable truth that there is a certain body type that is capable of scaling the obstacles and arriving in good time at the big red button. Its all proportional power to weight. The cute little one no matter how determined is not going to cut it with those short arms, neither is the over proteined upper body builder. Unsurprisingly it’s the wirey around 70kg — 1.7 metre types who without too much fuss and motivational pumping just get on with the business and get to the end quickly and efficiently.
Now then, if you are not interested in vintage and classic racing Velocettes best you sign off here because its only the power to weight analogy that’s of interest from above.
So, you are going to build the best ever 350cc Velocette road race machine capable of and specifically built to tackle the Isle of Man mountain circuit, the oldest, most respected and challenging motorcycle road race. You must stay faithful to the Velocette factory efforts and simply carry on where they left off. You can’t and don’t want to use many (if any) ready to go bolt on bits largely because that’s just too easy and waters down far too much the idea of this being a genuine effort. Plus you’re a kiwi so you’ll want to do it yourself the hard way.
Just about everything we get in or on these days has a double overhead cam engine, that is unless you are a clean green modern thinker committed to batteries not petrol, in which case you should also sign off here.
The very little yet capable racing department at Velocette’s designed and built two versions of a twin cam single cylinder junior engine for racing, nearly two decades apart. Neither of these efforts it could be argued were optimized for different reasons.
The first debuted in 1936. Once the KTT engine achieved mechanical reliability during the twenties, valve control would have been the next big thing on the mind of Percy Goodman who would have been on his straps with business and highly committed to racing. He had already not only conceived, put into production and tested with great success in the preceding ten years the single overhead cam K engine but was already at the mk5 version of it.
Harold Willis was in charge of the race department at this time and his fertile mind and practical abilities both in the workshop and on the track would have been a key ingredient, and let’s not forget Charles Udall who I always pictured as something of an engineering oracle that no hair brained idea was going to get past. Add to this hotpot rider feedback from the likes of Stanley Woods who had become the first modern professional, controversially working with the Italians and riding the red devil Moto Guzzi to senior victory in 1935. He knew his mind and on it was the clear advantage of rear suspension. So in 1936 Velocettes come up with a brace of three what could best be described as superbikes of the era, double over head cam 350 singles with 4 speed positive stop gearboxes, as many magnesium castings as possible, gearbox, crankcase, wheel hubs and brake plates, this was fairly new and experimental even though Harold Willis called the stuff ‘electrified dirt’. Full loop frames, a rear swinging arm with air and oil suspension units the design of the former pretty much unchanged in your Thruxton.
We’d have to agree that the 1930’s was the most exciting for Velocette road racing both for design development, new materials, and a real spirit of progress. All fuelled and motivated by commercial successes from the factory and marketing efforts.
Possibly with war coming but for whatever reason the dohc was not pursued beyond 1936 even though Ted Mellors on the only rigid of the three built had had great success on the continent that year. The three engines were put back under the bench and eventually given to Frank Musset to take back to Australia, one of which formed the basis of the Sid Willis 250 twin cam on which he gained a creditable fourth? place in the 1953 TT. The factory pursued the single cam for 1937 which was in Ivan Rhodes’ words “as dull as dish water”. They got it wrong there is no doubt about that: the early ‘dog kennel’ and twin cam had steeper valve angles, the general consensus of the day was to go for deeper combustion chambers wider angles bigger valves and finning projecting out into the cool air, hot and heavy stuff, no red button for that ninja.
Fast forward through the war to around 1950. Velocette won the junior world championship in ‘48 and ‘49 largely on machines of prewar design similar to what took Stanley Woods to TT victory in the 1939 junior. Bertrum Goodman son of Percy is now in charge of the racing department following the unfortunate loss of Harold Willis. Percy himself is an old man and the racing budget is very tight with the board changing direction toward mass production of scooters. Just five special twin cam engines are built at 250 and 350cc using the MK8 as the basis and extensive use of magnesium. The other big change was a modified frame, using a MK8 frame as the starting point, removing the cast steering head and replacing it with a machined version from the LE project, lowered and with additional support from two near horizontal tubes a la McCandless and another angled brace all the way from the top of the steering head, through the tank to the seat tube. And for the first time a special 5 speed gearbox which no doubt helped keep the little engine on the boil.
Its worth making note of the climate of the two dohc efforts. Prewar the factory was relatively rich working with the “win on Sunday sell on Monday” philosophy, they steadily progressed technology and design year by year since the mid twenties on successes, the minds and hands that developed the K more or less from the beginning directly involved. Post war there were other factories throwing big money and progressing the racing motorcycle largely on the back of what Velocette and others had achieved. The Mk8 and factory pimped racing Velocettes were a highly developed machine that was advanced in the 30’s but not anymore. Money was very tight and Bertie while having some good ideas and encouragement from dad may not have been able to optimize and develop the thing properly.
So, you are back in the Hallgreen nae Ngaio kitchen choosing your ingredients. You have at hand one Nick Thomson, for this effort the most essential component as this cake won’t bake itself. Nick is a lifelong Velocette engineer extraordinaire so we’re in good hands whatever the ingredients.
We know that a MK8 can deliver as an excellent vintage racer and if you read about our “hotrod” which is essentially a standard Mk8 with roller followers on methanol and a 6-speed Nova box nestled neatly into the original housing, oh and one Chris Swallow aboard. Unheard of and embarrassing for others (sorry JC) results came out of this combo here in NZ. My own unproven theory of its capable handling being that this bike had had a front on stove somewhere in its history in Victoria thus providing a few precious extra millimeters of trail. We had to ask the question what if this hulking 325lb motorbike could shed a few kilograms and have the girders reworked to optimize not only trail but essentially have actual hydraulic damping not just the normal tightening of friction washers so that nothing much moves up front? Would we then have a real ninja warrior vintage racer with power to match weight and good handling to boot? A combo capable of playing at the sharp end of any classic 350 field?
As you know the 73x81mm bore and stroke of the K never changed throughout its 25 year run, although the Rhodes’ believe there was at least one special engine near the end that had a short stroke. And history tells us when any British 350 was downsized to a 250, the short stroke version was always superior. Really to be at the high end of power output a short stroke has to be on the agenda, so after much musing about the highly successful Guzzi 350s at 75x79mm, a bore and stroke of 80x69mm was settled upon.
Now in making all these special castings I must send a very big thank you to Colin Quartly of Christchurch who followed our wish to have the patterns made very much along the factory lines, that is with the feeling of the handmade over something spat off a CNC mill. Colin must be the most experienced and capable patternmaker in NZ and the foundries all respect him which is quite a plus. Back to the cases, with the exception of a bigger mouth capable of the venom stud pattern and our desired bore we changed nothing, that goes for the gearbox too. At every turn we took the approach only to change a thing if it truly needed it, and one thing you can say about the early design on a Velo, it was mostly thorough and correct. On a KTT engine the oil pump bore which is a slight interference fit and is very close to the main bearing on the timing side, quite an achilles heel for the engine as most main bearing bores end up not being round even if they start out that way.
Nick had the good idea instead of attempting to strengthen the bore with an insert or even more extreme changing the centres, rather to reduce the OD of the bearing by choosing a needle roller but also and importantly to shrink in a slug of steel the same size as the pump prior to boring for the main bearing, so the hole is not round without the pump in but is when it is. This example of the depth of consideration is throughout this engine where every part has been handmade. The conrod follows the tested method of being one piece not having a pressed in big end ring rather a hardened surface of the actual rod itself. Remember the ninja, its all power to weight and if you can reduce weight anywhere you should and if you are building from scratch the opportunities are literally everywhere and for the taking.
So without every detail, the top knot much like the Eldee has been hewn from a solid billet of Alloy, this simply because he can and because the level of accuracy and consistency of quality material is guaranteed. Here again although it is one of the few parts not cast I’d like to mention and thank one Steve Howell of CH tooling in Melbourne, another enthusiastic old school totally capable guy on the magnesium, he says ‘I don’t know whats in the water over there in New Zealand but I make more maggy parts for kiwis than anything.’
Getting back to the cambox which is in itself a wonderful piece of form and function. Here I must explain our choice of engine. We went with the 36 thinking more than anything about weight, this design has the cambox bolted onto the head where as the later is one piece and really its superior to the later version not only because its much lighter, the path of thinking from the late 30s and into the 50s was to increase the mass of the head and extend the fins out into clear air, all very nice aesthetically as Velos demonstrated with the famous ‘Huntly and Palmer’ heads of the late 30s, but it weighs far too much and there are better ways to keep the thing cool. Not least of which being a ventilated exhaust port à la late two valve Ducati engines. Plus we wanted to stand the inlet tract and valves up more, we’ve got a good quarter of an inch skull cap around the combustion chamber but other than that the vintage exterior is bang on.
On the delivery of all this urge we’d have to go with what has been proven through years of short circuit racing in NZ. The synchroflex belt has been faultless with Nick milling both the engine and clutch wheels, the engine pulley in one piece with shoulders complete. If you could hold one of these parts in your hand for a close look its quite a thing. If it’s possible to have less parts to achieve the result then just do it, it’s less to fall off or go wrong. Once set these belts never and that means never need adjusting. The clutch similarly a tested item with aluminium pulley wheel and off the shelf Suzuki friction plates.
On this bike Nick has made a tool plate aluminium backing plate as well, with that 100kg dry weight in site. The Nova six speed with low first has been a success and again because we can and it’s better we’ve gone for the extended mainshaft. Put simply because on a Velo the drive sprocket to the rear wheel is outboard of the clutch the mainshaft gets a hammering, and quite often operates in a state of bending. With a new shaft extended we can have an inline reamed bore and needle roller outboard and kiss goodbye to any problems pulling the gearbox forward and nothing is under stress through the box casting itself, Nick learned this lesson on his quick Venom racer and while we’ve not bothered on the smaller bikes, but for this ‘little un’ we have high hopes.
What else, oh yes the frame is a nice piece of work. Ivan Rhodes loaned us a proper works example of this semi duplex item and we had a measure and a think. Now you’ve heard the stories and possibly even seen the photos of Stanley Woods testing weight distribution on the rigid dog kennel bikes by fixing a lump of lead to the downtube. The outcome of his involvement was twofold, the swinging arm and a steeper front downtube to get the engine more forward, and remember here that the action of the girder is largely forward in a kind of ‘S’ line so the front wheel can be very close to the down tube.
Both these features were employed as standard on the MK8. On the late works bikes a 19 not 21 inch front wheel was used so there’s a bit more room to increase trail, plus we are not convinced about the advantages of the engine forward so choose a standard RS frame as our starting point. Chris and Nick drew up and modeled on a sheet of MDF the rake, trail and suspension travel of the girder with Chris coming up with the type of thinking we like which says we want the wheel to have constant trail everywhere on the suspension spectrum which should be in the order of 100mm so why not move the wheel through that trajectory and see if the arms and pivot points of the parallelogram either side of the steering head can accommodate this? The outcome interestingly can be achieved almost perfectly with the bottom pivot in the same place as standard on the centreline of the steering head and the top point moved forward and with a short arm. On a previous burst of enthusiasm on girders we had had punched by Southwards Tube Mills some sets of tapered tubes in 4143 grade steel.
These in hand and with more patterns made for the lost wax steel casting process for lugs we were in business. Plus Nick went ahead and actually made what may have been a myth and that is ground pins and needle roller bearings for every junction on the girder. Everything is straight and true and with minimum friction, and did I mention rigid, wow this front end is impressively rigid compared to most teles. I’m compressing the level of effort required for all this, in particular the making of frame lugs offered reasonable challenges.
For example the swinging arm which we wanted to be pretty faithful to the MK8 which has quite a specific look and advantage over the RS frame version, that being the arms are connected with a strut between and just behind the pivot. The special lug for the frame down tube we had milled in two halves from solid steel with the goal of achieving even wall thickness everywhere. The special swing arm lug cut and milled from solid. Good friend and fast becoming Velocette royalty Pat Clancy spent a solid week fettling the latter just to be right, thanks Pat. Front wheel is standard late MK8 with our own brake plate with internal double leading shoe linkage and the rear hub in alloy.
There’s the ingredients boys, a blend of the two double overhead cam factory efforts, engine from the earlier and frame from the latter versions. A lot more than able to be described here has gone into this of course but how much time do you have? Outcome, a 100kg dry, cobby looking works replica. Lets go racing.
Signs from testing to date look positive. A pleasing outcome starting from the ground up making all new choices and combinations, not the usual compromised starting point.
Built specifically for Chris to ride at in the Classic TT and as testament to the excellence Velocette had achieved many decades prior to the type of machines that are currently in use in these modern classic events. We are in no doubt there is unfinished business in terms of the development and optimization of the single cylinder junior twin cam Velocette racer and were keen to see what we can do with this package.
A slightly more complex competition than the ninja warriors on the tele, nonetheless with many a greasy pole to slip off into the metaphoric drink ahead of us.