What is it about exhaust nuts? They seem to be one of the parts you can generally rely on to be a problem on an old motor cycle. Consequently it seems they regularly get 'bodged' at some point in an engine's life. Here are my trials and tribulations in the hope they may help someone struggling alone in their shed.

A short sorry history of a Sunbeam Lion's exhaust nut

My 1931 Lion was acquired with an exhaust pipe held in place by a nut that didn't even connect with the port's internal thread. Two holes had been drilled and tapped through the cylinder's projecting exhaust port and into the nut. An ingenious solution, erm?! Through these holes two small screws were used to hold the nut, and hence the exhaust pipe, in place ... or had done once! (photo, left). The nut rocked and wobbled, and lasted one trip around the block before the exhaust pipe blew out leaving remnants of two old screws fouling the internal thread of the port.

Luckily, a replacement was found in the form of a well-worn original item from an identical 1931 engine (photos, below). It just had enough 'bite', fouled exhaust port thread and all, to hold the pipe in place. It saw me through a few years until one day the connection between the internal and external threads became non-existent. Unsurprisingly, this occurred 20 miles from home! A few feet of garden wire fashioned into a tourniquet, of sorts, held the exhaust pipe in place long enough to get me back. (My late grand-father - a Marston's employee - advised always carry a few feet of wire as well as string in your tool kit as it won't burn if you need it to be in contact with anything hot. It turned out to be useful advice.)

A new nut was needed. I turned to The Marston Sunbeam Register as they advertise side-valve exhaust nuts as one of their available spares. Unfortunately when I enquired it turned out the batch had been made up some time ago and had now all gone. The Lion was consigned to the shed.

Sometime later a further engine of 1928 vintage arrived in the shed, from which projected a gleaming chrome exhaust nut. It was clearly home-made. It was, however, a beefed-up, more robust item than the original design (photos, above). It seemed the Lion would have its missing part and be back on the road.  Unfortunately the ungainly projection of the gleaming nut from the exhaust port was explained not by the usual explanation - the laxity of previous owners in tightening up nuts and bolts when assembling stuff for sale - but by a failure of whoever fabricated it to cut the correct thread to fit the port. The nut was no use at all. That's when the measuring started in earnest! Time for a bit of 'do-it-yourself'.

Getting serious: measuring and cleaning-up the exhaust port

It seems that the 'blind' cylinder, side-valve engines up to 1931 have a 1½ inch diameter exhaust pipe, whilst the side-valve engines introduced in 1932 with a detachable head have a 1¾ inch diameter pipe. At least engines I've had a good look at seem to be consistent with this.

Out with the thread gauge to measure the screw pitch. For the 1½ inch diameter exhaust pipe (the blind cylinder engine) the internal thread for the port measures 26 t.p.i. (threads per inch) as did the original, much-worn exhaust nut (photos, above). This explained why the shiny, home-made nut at 25 t.p.i. wouldn't fit. Thankfully it hadn't been forced so the internal thread to the port on the engine with which it came seems fine.

However, the condition of the internal thread in the exhaust port of my 1931 Lion engine was a cause for concern. There was a momentary dread that it was worn beyond any practical use and I would need to renew it, possibly with some form of insert. I spoke to people about this. It was all worryingly complicated! And, very few of the engineering companies that advertise in the old bike press wanted to touch the job.

Before going down this road, I thought I'd clean up the thread with the correct-sized tap and see what I'd got. A huge 1¾ inch diameter tap in British Standard Cycle thread pitch of 26 t.p.i. was called for. Thankfully, Tracy Tools Ltd of Dartmouth were able to supply just the thing (tel. 01803833134 or web site: www.tracytools.com). And, even more thankfully, the exhaust port thread, complete with its bodged remnants of broken-off old screws, cleaned up well (photo, left). This required a very gentle hand, and involved lining the tap up and letting it 'bite' by itself rather than forcing it into the port.

All it needed now was an exhaust nut that fit. Time to fire up the lathe. The shed had recently acquired one of those very useful combined lathe and mills from Chester UK (www.chesteruk.net). I had promised myself one after my access to a Colchester student lathe ended with the closure, a few years ago, of the model engineering evening classes at the local college.

Exhaust nut fabrication

The original Sunbeamland nut appears to be made of mild steel. Its maximum outside dimension is 2 inches with a 1½ inch hole for the exhaust pipe, which meant I ideally needed 2 inch steel tube with a ¼ inch wall. Finding mild steel tube with a ¼ inch wall proved impossible. As I didn't fancy removing all that metal from a solid bar I decided to use brass tube which was available in this size.

Mallard Metal Packs Ltd of Birmingham supplied me with a 6 inch length of brass tube at 2 inch diameter with a ¼ inch wall (web site: www.mallardmetals.co.uk or tel. 0121 624 0302).

I started the fabrication of the nut by sawing off a few ¾ inch lengths of brass tube with a hacksaw in the bench vice. These would allow me a few attempts to get it right! The attempts included a replica of the original, Sunbeamland-made nut with its more slender flange as well as a nut with a wider, more robust flange for the four slots for the C-spanner, similar to the home-made jobby that had arrived with the 1928 engine (photo, above left).

In the lathe, I faced-off back and front of each section of tube. Next, I marked off a length of 7/16th inch to be turned down. This was turned down by an ⅛ inch deep cut, leaving a reduced diameter of 1¾ inch to take the screw thread (photo, above left). N.B. I use the formula in Lawrence H Sparey's very useful 'The Amateur Lathe' (first published 1948) for setting the lathe speed for turning brass. He suggests a cutting speed of 200 ft. per min. for brass. To set the r.p.m. for the lathe, he advises take the cutting speed (200) and divide by a quarter of the diameter of the work in inches (0.5 inch for 2 inch tube). Which in this case requires I set the lathe to 400 r.p.m. Finally, once the length was turned down, the 'shoulder' between the two diameters was faced-off.

The result was a 'blank' that didn't quite fit over the exhaust pipe, so boring its internal surface to remove 1/16th of an inch was called for - i.e. a 1/32nd inch cut (photo, above right).

The resulting 'blanks' fitted over the 1½ inch diameter exhaust pipe a treat. They were bored out sufficiently to cope with the various curves in the pipe whilst not being so wide internally that they failed to contact against the exhaust pipe's terminating flange (photo, left showing three nuts on the exhaust pipe).

This leaves two operations, screw-cutting a 26 t.p.i. pitch on to the nut (requiring a 60° lathe cutting tool, required for British Standard Cycle thread) and milling of the four slots in the outer flange of the nut used for tightening it with a C-spanner.

Milling used an 8mm (9/16th inch) end mill to cut slots to the full depth of the outer flange. Advice in Sparey on cutting speed for milling brass simply advises that end mills can be run at higher speeds than other types of cutter, and brass at higher speeds than steel and cast iron. It seemed, the faster the better!

Care is needed in lining up the work, too, to ensure that the cut is square and occurs at 90° intervals around the circumference. I marked the nut out accordingly before milling. I used a length of narrow steel rod in the lathe chuck as a guide to line up the end mill and exhaust nut.

The end result is an exhaust nut missing only its screw thread. Unfortunately, my lathe arrived without the necessary gear for cutting 26 t.p.i. - this was annoying as I was advised last year, when I went to look at the lathe that I eventually bought, that it came as standard! I stressed that the ability to screw cut at 26 t.p.i. was essential for me. Thankfully, Chester UK have been able to supply the appropriately-toothed gear quite quickly and at a reasonable price.

The exhaust nut was firmly clamped in the three-jaw chuck and gears set up as per the instruction helpfully provided by Chester UK. I was ready for screw cutting. The power feed lever on the lathe proved invaluable, allowing me to turn off the power feed to the carriage with the cross-slide and the tool post. This meant I didn't have to remove the tool from the job and remained permanently engaged with the lead-screw of the lathe. The reverse button on the lathe then allowed me remain engaged with the lead-screw whilst reversing the direction of the cut. By this means I was able to re-position the tool at the start for the next cut (photo left above).

Spacey's advice on screw cutting is to take light, shallow cuts of no more than .005 inch (5 thou) for each pass. I went down to 4 thou which seemed noticeably smoother. The cutting tool has to be the right profile (60° in this case) and at the exact centre and 'square' with the work (the tool is just visible in bottom centre of the left-hand photo, above). The lathe should set at its slowest speed - in my case 160 r.p.m.

The total depth of cut was monitored on the cross-slide's dial. But how deep to cut? A letter from David Hull in 'The Classic Motorcycle' of June 1985 helpfully advises that for a 60° thread, the depth of thread is 0.5237 multiplied by the pitch, i.e. 1 divided by 26 in this case. (For 55° threads, the figure is 0.6403 multiplied by the pitch.) That gives a cut of .020 or 20 thou. I tried this on the first example I cut. It was just too big for the port, so on the second version I cut to .025 or 25 thou (photo above right). This fitted a treat!

The finished result, firmly holding the 1931 Lion's exhaust pipe in place is shown in the photo (left).

And finally, a bit of Sunbeam research

Out of interest, Marston's 1931 'Price List of Spare Parts', identifies the part as 'exhaust pipe union', reference D.423, used only on the Lion model that year. It cost 3/- (three shillings). The 'D' number is a specific reference for the 1931 spare parts list.

The factory reference number for the part is also given, it is part no. 16. Marston's 1929 spare parts list shows the same part was used on pre-Lion saddle tank Models 1 and 2 (350cc side-valve) and Models 5 and 6 (500cc side-valve). The reference number from the 1929 spare parts list is C.308.

An earlier spares part list of 1927 indicates it was also used on flat tank Models 1 and 2 (350cc side-valve) and 5 and 6 (500cc side-valve). The reference number is A.363. I have no earlier spare parts catalogues, but as 1927 saw the introduction of the cross-over exhaust pipe for the side-valve engine - to replace the earlier 'coffee pot' exhaust - I cannot say with certainty that a 1¾ inch, 26 t.p.i. union nut was used earlier than 1927.

The spare part lists all refer to an exhaust union washer with the nut. I have not seen this on the three Longstroke engines I have removed exhaust pipes from. Illustrations in Marston spare parts list show it as roughly the same size as the exhaust nut but I have read no reference to it in Pitman, etc. I am not sure whether such a washer would sit in the port prior to inserting the exhaust pipe, or be placed on the pipe itself between flange and union nut. I haven't used a washer but any thoughts welcomed.

The union nut was not used from 1932 onwards as a 1¾ inch exhaust pipe was introduced, requiring a larger union nut (factory part no. 6828 for those interested in these sort of details!).

Return to: The Workshop