Building the Seagull came about as a convalescence project following an operation. One of the main reasons for choosing Seagull was the number of parts shared with other engines designed by Edgar T Westbury. Things like the timing cover, the cam profiles, carburettor, timing tears and ignition parts. These are all shared with the Seal and a single cylinder version of the Seagull that I believe was called the Seamew.

The photograph showing work in progress shows the extra cylinder and timing cover for the single cylinder version, along with spare conrods. While making the cams for Seagull I also made additional cams for Seamew and also for the Seal, which will also benefit from another timing cover made at the same time.

Fig 1 Engine parts early into the project

I made several changes to the engine during construction, mostly to aid construction and alignment. The first of these were two hollow dowels to locate the sump casting, relying on the two end covers is adequate but I felt it could be done in a more positive manner. The same hollow dowels were also used to locate the centre main bearing and a smaller size was used to locate the big end caps on the conrod, photographs are included showing these details.

Fig 2 Sump showing hollow dowels

Fig 3 Conrod showing hollow big end dowels

Fig 4 Hollow dowels on centre main bearing and oil scallops

Further refinements came in the form of scallops to catch and feed oil to the main bearings. A photograph of the rear of the front cover and interior of the main casting as well as the rear bearing cover illustrates the technique used, in practice they work extremely well.

From the outset I had decided to make Seagull air-cooled, the fining detail shown in the plans for Seamew were probably intended to be cast on. Not knowing if in the original text ETW was going to rely on a cooling fan or the air stream produced by that of a boat underway, I did feel that some additional material in the size of these fins would not be a bad thing.

Further, inline twins always have a cooling problem with the rear cylinder running hotter, any additional cooling surface would I thought only help things along. I, therefore, made the outside of the cylinder more rectangular in profile, by using a slitting saw and rotary table, the bottom of the fin groove is D-shaped and is cut quite close to the cylinder liner. The cylinder head follows the same rectangular theme and is made somewhat taller than the original design, mimicking to a large extent the R/C car engines of today.

Fig 5 “D-shape on cylinder and fin detail

The combustion chamber design is based on a ‘Ricardo’ type cylinder head, published in an article Piston Rings and Compression Ratios by Prof. Chaddock, (page 665 & 666,Model Engineer 2 July 1976). As machined the compression ratio is 6.5:1 as per the above article and I have situated the spark plug over the exhaust valve, rather than in the original position on the side of the head, as in this position it would foul my intended position of the carburettor. I have supplied two photographs of the Cylinder heads and the milling jig used to produce the combustion chamber.

The locating holes for the various ‘centres’ for the chamber shape are clearly seen in the second photograph. The D-shape profile that forms a cooling channel the same as that on the cylinder also being milled at the same time using this jig. All work being carried out on the rotary table, those with CNC’s will find this a much easier task. Spark Plugs, too, are homemade. (ED: see the excellent article at:

Fig 6 Cylinder head combustion chamber and fin detail

Fig 7 Rotary table chamber milling fixture

Engine cooling had always been a problem in that I could not seem to decide upon which option to take, by this I mean whether or not to use an engine driven fan or a 12-volt electric fan. I had always admired the large air-cooled Diesel engines produced by Deutz, it is sad that they are no longer in production, the cooling fans on these engines gave them a distinctive profile.

Fig 8 Finished engine showing cooling fan

Talking it over with one of my electronics colleagues at work one day brought an unexpected surprise in that it would be possible if I followed the latter proposal to have the fan thermostatically controlled by using a very cheap electronic RS Thermostat, (usual disclaimer), that had a T0220 standard semi-conductor profile. This would mean the fan would only be running when the engine went above its optimum operating temperature. There is quite a temperature range of devices available so there was no need to compromise and I opted for a 90°C version.

A month or two later our department went through an upgrading of the computers in readiness for some new radiotherapy treatment machines, one of the units which apparently is now obsolete was to be scrapped. It contained two small cooling fans of the right voltage so it seemed the decision had already been made for me. One of the items in question can be seen on the front of the engine in the accompanying photograph.

I must admit I had my reservations that the fan would be adequate, but it is very pleasing to see the fan cut in and out while the engine is running. The shroud that forms the mounting for the Fan was originally the bottom of a biscuit tin and was formed over a plastic former, it was the third attempt, having tried to bend the earlier attempts free-hand with varying degrees of success.

Fig 9 Internal Piston profile detail

The piston rings are also to the proportions recommended in the above article by Prof Chaddock, the result is, I feel, a most realistic interior profile. I heat treat my rings individually using a small Butane gas torch. A slender wedge is used to spread the gap; the shank of a suitably sized drill is used to gauge the gap and heat is applied working from one end of the ring to the other. The ring starts off at right angles to the wedge and finally rotates through 90° to touch the side of the wedge due to ‘the set’ imparted to the ring from the heat and under the influence of gravity. There is very little oxidisation of the ring and I have found minimal change in circular form when inserted into the cylinder bore and held up to the light.

While many have used the jig method to do the same function, I feel that attaching what is a rather large lump of steel by comparison to the ring section not a good idea. My reason for saying this is that all steels whether they are round or rectangular in section have an enormous amount of stresses in them especially if they are of the bright drawn variety. Heat is used to relieve these stresses in industry and in putting your nicely machined rings in between two steel discs means any movement of the disc’s due to these stresses coming out during heating is going to be transferred to the rings. If this jig is used again, then the surfaces that were once geometrically perfect when they left the lathe are no longer so, the next batch of rings therefore are distorted the same as the first, as are any further batches.

There has also been much written about machining the rings back to circular form after heat treatment, again I feel this depends on the size of the ring, for the average engine like Seagull I consider it an overkill. A piston ring heat-treated using my method when inserted into the cylinder bore on the piston and worked up and down a few times does show high spots. After the first run most of these high spots have merged into one. By the third or forth running of the engine the ring has taken on a bright shiny lustre all round. 

A separate photograph shows the wear on a commercial ring fitted to a 1 inch bore petrol engine after about 10 hours running. You will see on the lower ring just to the left of the wrist pin hole an area that has not quite bedded in.

Fig 10 Commercial piston rings (see text)

If some thought is given to what is actually happening during the initial running-in process then all becomes clear. The compression pressure tries to escape down the side of the piston, it meets the top ring, there are minute gaps every so often around the bore where the ring does not quite touch the bore due to circular error.

These gaps are filled with lubricating oil to a large extent, there is however a much larger gap by comparison by which to escape, that is the clearance between the ring and the side of the ring groove, the compressed gases rush through here more readily to occupy the space between the inside of the ring and the bottom of the ring groove in the piston. The gases now look for a new exit. The ring is in firm contact with the piston groove on its bottom face due to the upward movement of the piston. The only exit left is the gap at the two ends of the ring.

Pressure starts to build, while this pressure is building uniform pressure is exerted on the inside of the ring forcing it into contact with the cylinder wall and any high spots begin to wear away, quite rapidly. Finally, the pressure is high enough to expel the oil that was blocking the ring gap and the gasses escape only to have to go through the same thing all over again with the second piston ring. As the piston goes over top dead centre both the rings stop momentarily as the piston starts to move down the bore. The ring is then brought into contact with the top of the piston groove; the only exit now for the gases is the ring gap and any very tiny gaps around the bore, remember these are reducing due to wear, those that are remaining are still full of oil.

Just bear in mind also that the temperature of the materials is changing, the piston ring is growing in length and the ring gap is, therefore, closing down, possibly non-existent. The piston is getting larger in diameter and its initial cold clearance is very much smaller now perhaps only a few molecules of oil wide. Keep in mind that all this is happening many times a second.

Believe me there just is not enough time available for an non-circular ring to have any effect, double the cross section of the ring and a whole new set of conditions exist. For one thing the ring is not so flexible, it will need greater pressure to force it against the cylinder wall. Any high spots will tend to be at the ring gap and diametrically opposite, these high spots will take much longer to bed–in, but the engine will still run.

I get called upon to repair the odd lawn mower. One such case was an engine that could literally be turned over with finger and thumb compression was non-existent.

Stripping the engine down revealed the rings to be a hand full of ring segments and luckily the bore was un-damaged. The only piece of cast iron that I had was only just big enough to make the rings. Tongue-in-cheek I made the rings and heat-treated them using the above method and re-assembling the engine I found a good compression.

It is has now been seven years since I did that engine and it is still going well. Perhaps I was lucky but sometimes one has to ‘suck it and see’ and not get too bogged down in the theory.

Making a jig to heat-treat a ring of this size would be very time consuming and require a large amount of heat to accomplish. I hasten to add if the reader has always done his or her piston rings to the traditionally accepted method I am in no way saying that my method is superior, merely easier. Incidentally Barrington (Barry) Hares who built that magnificent Rolls-Royce Eagle 22 engine used a similar individual heat treatment technique for the contracting rings of the sleeve valves of this engine.

The Cylinder liners were made from Silver Steel and left soft, both were rough turned,being finally ground to size inside and externally using my toolpost grinder. After bench running for about an hour in total all grinding marks have disappeared and the sheen on each bore is quite pleasing.

The crankshaft was made from EN8 and from the outset I had decided to make two, if you end up scrapping one during manufacture then you have the other to fall back on. Chances are you will not need it, but I guarantee if you have not produced a spare you will.

The crankshaft was roughed out as regards the main journals and overall web dimensions. The counter-balance web shape was then milled to and at the same time material from between the web was removed until a square section remained that would ultimately form the crank pin.

A jig was then made to turn the crankpins, this jig utilises the cheeks of the web to positively locate the crank pin, for positive driving and radial position. Using this jig in the dividing head I first of all milled the square section crankpin to a circular section. I had decided to make the crank pins 8mm diameter; the original size was 5/16” or 7.9mm. The crankpins were therefore milled to 8mm diameter plus a small grinding allowance.

Fig 11 Crankshaft Jig

Fig 12 Crankshaft in jig, note undercuts

Two special turning tool holders were then needed to machine each side of the web and produce a radius and undercut at the web crankpin junction. The undercut being only 0.05mm deep it would not weaken the pin as it was still at the original designed dimension of 7.9mm diameter, but it greatly simplified the grinding of the crankpin. When machining the crankpins I only ever machined the pin closest to the jig, this way maximum rigidity in the set-up prevails. I would never contemplate machining the second pin with that amount overhang and with the wasted portion of the centre main bearing in between, it is just asking for trouble

The idler gear spigot in the timing case was changed to a simple pressed in pin and circlip to retain the gear on the pin, a large central hole in the pin provides an oil reservoir filled by oil trickling down the timing case inner wall. Oil from this reservoir gets to the idler gear bearing via a radial drilling.

The timing gears were all cut with reference to the keyway positions, in that at piston TDC the keyways in the crank and camshaft are both at TDC also. Checking the valve opening times against a disc protractor confirmed the correct valve events.

The manifold in the original design always seemed to me to be a rather bulky affair and I did feel that it would interfere with the airflow in this air-cooled version. Having toyed with the idea of bending copper pipe to fabricate the manifold as suggested by ETW in the text of his article on the Seagull, I thought better of it and fabricated the manifold out of brass, a photograph shows the component parts before Silver Soldering.

Despite being made of brass the finished manifold is lightweight, as the wall sections were kept to an absolute minimum. The two large brass discs are for holding the outer ‘bosses’ of the manifold at the correct centres. When the silver soldering was finished these were removed with a pin punch through the centre stud hole. All that remained to be done after Silver soldering was to take a light clean up cut over the manifold mounting face after giving the assembly a quick clean up in the bead blasting cabinet. A second photograph shows the finished manifold, with the use of a mirror both sides can be seen at once.

Fig 13 Manifold components before Silver Soldering

Fig 14 Finished manifold assembly,( mirror shows mounting face)

The finished result is I feel a more compact fitment and allows air to exit from the centre section of fins between the cylinders.  You may notice that the manifold is now nearer the engine than in the original design. With the use of an up-draught carburettor as designed additional space for the mixture adjustment needle is required. Thus anyone intending to use this method of manifold construction please keep this in mind if you intend to adhere to ETW’s original carburettor arrangement.

With the manifold sorted the next item on the agenda was the contact breaker and distributor. While there is absolutely nothing wrong with ETW’s original design, it did not appeal to me. I wanted the contact breaker to look more like those that I used to put in my Mini, commercial points sets were out of the question as they were larger than the contact breaker housing. Many months passed and I continually tried different configurations, but none seemed to please me. Whilst looking through Model Petrol Engines by ETW for inspiration on a different matter I came across a drawing on page 139 of contact breaker assembly for the 1831 engine, it used the same Mini (Lucas to be precise) points that I wanted to replicate. With this newfound inspiration I managed to get the points designed just how I wanted, I think they look great, but then “Beauty is in the eye of the beholder”.

The points are salvaged from old Morse keys and come with a 10BA thread that makes the attachment to the points quite straight forward. I have included a GA of the assembly and a photograph of the finished items, as well as the parts that make up the assembly.

The distributor follows again the 1831 engine but modified as regards construction to again mimic the Lucas type distributor caps with the same method of attaching the leads. That is a very sharp pointed screw down the centre of the cap electrode that punctures the insulation and passes through the cooper core. Dowels are provided to locate the distributor cap and the rotor arm. Slots being provided in the distributor body and ignition cam to locate each part, should they need to be removed then they can be refitted in exactly the same place with little effort. The ignition leads themselves were sourced again from RS and are in actual fact a spare meter test lead.

Fig 15 Contact breaker and distributor parts

The final item to be made was the carburettor; I was torn between fabricating the original design, machining it out of the solid or using a more modern type of carburettor as used on R/C engines, but again manufactured ‘in house’. I eventually decided to stick with ETW’s original concept but individually tailored to suit.

It has now logged up several hours of running and compression is extremely good, but starting from cold has been problematical, that I feel is more ‘Operator Error’ than anything else. I did, however, move away from ETW’s knurled mixture adjuster as I felt it was just too close to the spark plug lead for comfort. I have experienced electricity ‘leaking out’ before at work, my Electronics’ colleagues will often laugh telling me that there is no real power, and that I am over sensitive. However, recently the tables have turned and they have experienced this leaky invisible substance, with no real power behind it. I had better reassure readers that the radiotherapy treatment machine was “off” when this happened, and we were merely changing a “Field Light” lamp.



By Graham Meek