1952 Sunbeam S8 – Resurrection

Part One of Four – Preparation

This project is unusual to the extent that it was delivered to me as a collection of parts. Most bikes come to my workshop looking at least partially like a motorcycle. This one arrived in mid-December 2016 in the form of six large crates, four smaller boxes, two built wheels, two mudguards and one frame. This has only happened to me once before and, as in this case, it was because the owner had started a full restoration project and had, for a variety of reasons, found the prospect of completing it daunting.

In this case, the project had progressed quite a long way. Many of the parts had been refurbished and a number of additional parts had been acquired. The wheels had already been rebuilt, chrome items had been re-plated and many alloy parts had been polished. There was a complete set of additional bare cases for the engine, gearbox and rear drive, all of which had been vapour blasted. Most of the tin-ware had been powder coated.

As in the previous case, it would have been a mistake to assume that all of the parts had come from a single motorbike or even that they had all come from the same model. In this case, a few of the parts remained unidentified at the end of the project. They were returned to the owner when the build was complete, together with a considerable number of other superfluous parts.

The activities that characterise a resurrection project of this kind are:

  • the identification of parts;

  • the selection of the most suitable parts;

  • remedial work as necessary;

  • Re-finishing (powder coating, painting or plating);

  • acquiring any missing parts;

  • assembly;

  • testing, and setup for the road.

I find it hard to get a grip on these activities unless I have a thorough understanding of what is in the crates, what condition each item is in and what is missing. So the first major undertaking is to create a detailed inventory. Every container is emptied and every single item catalogued – description, location, status and condition. Once complete, the inventory becomes a tool to help manage progress throughout the project. It saves a massive amount of time by helping to find parts quickly and avoiding looking for things that don't exist.

The crates contained many duplicated parts and they didn't contain quite a few missing parts. Most of the external engine parts had already been vapour blasted or highly polished and most of the tin-ware had been powder coated, including the frame, mudguards, stays, wheel hubs, battery box, tool box, electrics box, headlamp shell and fuel tank.

Generally, I like dry powder coating. It is seldom beautiful but it is durable and smart, being highly resistant to damage by fuel, oil and heat. But, just like paint, dry powder is not just one product. There are many grades of powder, with different characteristics, and there are varying levels of skill employed in preparation of the metal and application of the coating. This particular work had been executed by somebody who was having a bad day. The metal hadn't been properly prepared and corrosion and contamination were clearly evident beneath most of the treated surfaces. The fuel tank and the mudguards looked particularly sad. I took photographs and shared my initial impressions with the owner.

There are two other salient features of powder coat that should be considered by anybody who plans to use it. It looks good when new and is very robust but its surface is surprisingly soft. Any attempt to polish it will leave thousands of microscopic scratches on the surface, which eventually appear as a dull haze rather than the deep shine you may have hoped for. Powder coat is very difficult to remove. If you change your mind, the only practical way to get rid of it is to soak it in a bath of very aggressive chemicals. It is possible to paint on top of powder coat but only if it has been applied properly. Trying to fix a badly applied powder coat by painting over it will give disappointing results. Only powder coat onto a well prepared surface and only when you are quite sure that it is what you want.

It was at this point that I made the first of two poor tactical decision, which, like many bad decisions, was based on an ill-founded assumption. The frame had been powder coated and, unlike a number of the other components, the finish didn't look too bad. It was wrapped in protective foam material and was stored safely out of harms way, so I decided to leave it there until I had tackled the major mechanical components: engine, gearbox and rear drive. I took the precaution of locating the fork stanchions and checking that they were straight, to confirm that the bike hadn't been crashed but I should have stripped the frame of its cocoon and examined it thoroughly as soon as it arrived. It was mid-January 2017 before I had made sufficient progress with the big mechanical bits to think about bringing the frame out of hibernation.

The instant I removed the protective foam from the top tube, I could see that the frame was bent in that very characteristic way that tells of a head-on collision – a gentle upward curve in the top-tube and a gentle rearward curve in each down-tube. In this case, not quite symmetrical, so the steering head was angled slightly off vertical. Fairly subtle but obvious enough if you are looking for it. The owner at the time (not the current owner) must have replaced the bent stanchions, assuming that they had absorbed all of the force of the impact. Oh dear. The rule is: if the forks are bent, the frame is bent – always. It may be slight but it will often be enough to ruin the handling.

A few years ago, I was working at Army HQ in Andover. The Colonel in charge would bark at anybody hapless enough to use the word “assume” within his hearing. He would say “don't assume, check”. Annoyingly, he was right.

I only know of one place where I can get a Sunbeam frame properly straightened and that is Maidstone Motoliner, in Kent. Fortunately, they are fine craftsmen who specialise in classic bikes but it is impossible even for them to perform such a miracle without applying considerable heat to the metal, to make it malleable. The new powder coating would probably be incinerated in the process and so it was.

The engine, gearbox and rear drive all arrived fully dismantled, with their internal components packaged separately. Since there were duplicate casings, that left me with a bit of a puzzle working out which casings the components came from. That is less important with the engine but the gearbox and rear-drive internals are shimmed to suit the castings in which they left the factory. So I was resigned to starting reassembly from scratch and re-shimming them.

The castings were fairly good but none of them was perfect. It was a case of picking the ones that needed the least amount of remedial work and proceeding with those. One of the engine blocks had been modified by a previous owner in an interesting but destructive way. It looked like some extra drillings had been made to facilitate re-routing of the oil flow outside of the engine, possibly in order to fit an external oil filter. In effect, it rendered the whole block scrap, which made the decision about which of the two blocks to use quite simple. It also had another consequence, which will become clear later.

There was only one crank-shaft and, although there were two rear main bearing carriers, one was so badly corroded that it couldn't be regarded as a serious candidate. Now that the key engine components had been selected, it was time to carry out a thorough assessment of their condition so that the scale of any remedial work could be assessed and planned.

The cylinder liners had been bored to +0.020” and there were no signs of significant wear or damage in either of them. Measuring them in several orientations with an accurate bore gauge confirmed that they were straight and round and the correct diameter. Lightly honing them facilitated close examination of the surfaces of the bores, to check for any cracks or pits caused by corrosion. Testing the top surface of the block showed up some distortion and the fact that the liners were not completely flush. That would require skimming. Close examination of the whole block revealed that the threads all looked sound, there was no serious damage to the cooling fins, and there was copious evidence of the vapour blasting that had been carried out.

That leads me into an aside. Vapour blasting leaves a clean surface that has a texture very like the original factory finish. It is also quick and fairly cheap. However, it employs a blast medium that comprises tiny particles of aluminium oxide. Aluminium is soft but aluminium oxide is very hard indeed, consists of tiny particles that are exactly the same colour as the freshly cleaned casting and will destroy a newly rebuilt engine within a few hundred miles if any of it gets into the oil. The cylinder head, gearbox and rear-drive casings had also been vapour blasted.

Any form of blasting terrifies me. I would much rather clean aluminium castings chemically. However, these castings had already been done and a wipe with a finger confirmed that they were still contaminated. The only option was to clean them as thoroughly as possible to try to make sure that all of the blast medium was eradicated. That is a time-consuming and unpleasant process involving multiple stages. Firstly, any residue that is packed into threads, cracks and crevices has to be fragmented and loosened with a probe. Secondly, the bulk of the residue must be washed away. I use a pressure washer and force water into every corner and through every passageway, paying particular attention to the oil ways. Thirdly, I scrub the castings in the solvent tank, using a stiff brush, to destroy the surface tension and get into every pore of the metal. Then I wash the casting with a pressurised jet of filtered solvent. Finally, the solvent must be discarded, the tank thoroughly cleaned and the filters thrown away. Even more finally, the first fill of new engine oil has to be drained and discarded after the first test ride, as the last line of defence. All of that requires a lot of labour and discarding a tank full of solvent and a sump full of new engine oil represents a not insignificant cost. Perhaps vapour blasting isn't as cheap as it seems. End of aside.

Once the pistons were cleaned, including the ring grooves and the oil drain holes, the working diameters of the skirts were measured with a micrometer and the clearance inside the cylinder bores was calculated. It turned out to be 0.0045” in both cylinders, which is perfect for a Sunbeam. The big-end shells were scrap, owing to some galling of the white metal but measurement of the big-end eyes in the connecting rods revealed no evidence of ovality, so they could be re-used. The crank-shaft had been re-ground to -0.010” and all three journals were near perfect with no detectable ovality.

The front and rear main bearings were also in excellent condition but the sludge trap in the main oil gallery was packed with hard, black oil sludge, impregnated with metal particles. It seems that the engine had been rebuilt only a very few miles before the bike was taken off the road for restoration. That is fortunate because the sludge would soon have constricted the oil feed to the big-ends and destroyed the bearings. Cleaning out the sludge trap is a vital part of any engine rebuild if you want it to last more than a few thousand miles but it is surprising how often I find that it hasn't been done.

Two cylinder heads were in the inventory. Neither of them was in perfect condition but checking the mating surface, the integrity of the threads, the wear in the camshaft bearings, the condition of the cooling fins, the wear in the valve guides and the condition of the valve seats led to the selection of one of them as the best starting point. Performing a “dry fit” of the chosen cylinder head onto the engine block confirmed that the spacing of the studs aligned with the appropriate holes. It often pays not to take anything for granted.

There was some minor damage to the inlet and exhaust flanges and the remnant of a broken was stud stuck in the head. Fixing the damage using the milling machine was straightforward but all of the studs needed to be removed to facilitate access to the damaged areas. To avoid damaging the threads in a 60+ year old casting, the head was heated to 200ºC before the studs were unscrewed. It was heated once more to allow the worn valve guides to be pressed out. The camshaft was worn beyond recovery so a new one was required but the cam followers could be re-ground to restore their original profile.

After heating the cylinder head once more, four new valve guides were pressed in and reamed to give the exact clearance needed for the valve stems. The valve seats were re-cut using an indexable carbide cutting tool (the steel ones that you can buy on eBay are virtually useless, particularly on exhaust valve seats). That was to ensure that the seats are perfectly perpendicular to the axis of the valve stems and to remove any blemishes in the faces of the seats. The valves were then lapped into their new seats before the head was reassembled with a new camshaft and the refurbished rocker shaft assembly. Finally, I replaced all of the studs that had been removed from the head, plus a new one to replace the one that was broken.

The rear main bearing carrier had a good main bearing bush fitted, which showed very little wear, but with slight damage to its thrust face. The damage was repaired by skimming it on the milling machine. The oil pump, which is integral to the bearing carrier, had very little lateral play and turned smoothly so it was a good one. They are not normally prone to wear. The bearing pin for the half-time gear was heavily worn, as was the bronze bush in the gear itself. Both were replaced with new ones.

When the engine block came back from the engineer who had skimmed it, the rebuild could start as soon as any remaining swarf has been cleaned away, particularly from the oil ways. It was not known for sure whether the crank-shaft was originally paired with the selected engine block so I needed to proceed on the assumption that it was not. That meant that the end-float, the longitudinal free movement of the crank between the front and rear main bearings, had to be set to the correct value.

The procedure for that is to assemble the crank-shaft into the engine casing and bolt the bearing carrier firmly in place before measuring the end-float with a feeler gauge. If there is no end-float at all then the carrier is removed and 0.008” skimmed from the thrust face of the rear main bearing, using the milling machine. If there is between 0.006” and 0.010” of end-float then nothing more needs to be done. If there is more than 0.010” of end float then a steel shim needs to be made and inserted behind the rear main bearing bush. To facilitate that, the bearing must be pressed out of the carrier and then pressed back in with the shim fitted behind it. The carrier then needs to be fastened in place again and a new measurement taken. If necessary, a further correction can then be made but, in this case, none was necessary.

Although the pistons and bores were in good shape and the measured clearance was within the permissible range, the piston rings all had excessively large gaps. Whether that was because of wear or because they were not fitted properly is unknown. A new set of rings was inserted into their bores, one at a time, and the gaps adjusted to be exactly 0.008” for the top rings and 0.006” for all of the others. The new big-end shells were then fitted into the connecting rods, which were then attached to the respective pistons. The new rings were fitted to the pistons which, with the connecting rods now attached, were inserted into the bores and the big-end caps fitted. The caps were tightened to the recommended torque and new split pins fitted.

Fitting the big-end caps is seldom quite as straightforward as it sounds. The crenelated nuts must all be tightened to the same torque, which must be in the range 25-28ft.lbs. Too little and the end-caps can move relative to the connecting rods, which can cause damage to the rod and even failure of a bolt. Too much and it can crush the aluminium of the rod and/or the cap, which can tilt the bolt head, causing it to part company. I have, on occasion, had to discard a connecting rod because of this kind of damage. Broken big-end bolts can cause a lot of mess.

Those people who have done this job know that it is unlikely that all four nuts will align with the holes for the split pins at the correct torque. Increasing or decreasing the torque to achieve alignment is not the answer. You can try swapping the nuts between the four bolts, to see if you can get some of them to align but you will nearly always be left with at least one that won't. The solution is to remove a small amount (a few thou) of metal from the working face of the nut. It is important that the face of the nut remains dead perpendicular to the centre-line of the thread so the best way to achieve it is to use a lathe. The exact amount of metal that needs to be removed will depend on the pitch of the thread and the angular misalignment for each nut.

This engine was fitted with a sump extension, to provide additional oil capacity. The extra depth of the sump requires all twelve of the studs to be replaced with a set that is about an inch longer than the original ones. But, to extract the maximum benefit from the extra capacity, the oil scavenge pipe also needs to be extended. The best solution would be to replace the whole pipe with a longer one but I have only once ever been able to extract the original pipe without using excessive heat or excessive force. They are fitted in the factory with a lot of interference, to avoid the possibility of it coming adrift in flight, which would result in very rapid engine failure. So the solution that I adopt is to make an extension piece, usually from brass, that fits into the existing pipe with interference. Why brass? Because it expands with heat more than steel, so there is no risk that it will become loose as the engine heats up, and because it is strong enough to allow it to be made very thin, to maximise the internal diameter. It is important that the lower end of the extension is slash-cut or notched, so that it will not become constricted if it touches the sump pan.

I usually fit the studs with a dab of thread-lock because it can be annoying if the studs unwind when the sump nuts are loosened.

In Part Two, I write about the rebuilding of the engine, gearbox and rear drive.