Seagull Engine Construction Diary

Making a 10cc twin 4-stroke to the E T Westbury design

Exhaust Manifold

The draw on the casting means that one side will not clean up to the full ¾″ width. This side can be put at the back, against the cylinder ports, as the full depth is not needed this side. The top face of the casting could probably be squared up to remove the draw, which would look better. Although the end flanges are generously sized, their positions mean that it will be quite tricky to ensure they will clean up to size. They will probably want cleaning up by hand to match a template.


2010-03-13 – Trying out the castings

I was a bit worried about what Ron Chernich had to say about the castings so I decided to try a relatively simple one to learn on. The front face was rubbed on a coarse diamond whetstone to sit firmly on the top step of the 4-jaw chuck, packers made for the ends, and 30 thou taken off the back face. Should I use some cutting fluid or not? The result of periodic applications of paraffin diluted neat cutting oil from the washing-up liquid bottle was a series of bright rings on a smooth but more mat face. Generally the finish seemed quite fine.

The job was reversed and the same amount taken off the front, this time working dry and taking a more aggressive cut. The surface was still perfectly acceptable. There are a few very small voids in the casting, but nothing to worry about. There is a tiny fragment of brass in the surface. It is more likely my dirty habits than an inclusion. The scratches at the left hand end are from packing strips; this is only roughed out, and the final face will be lapped flat.

Next, the top was machined. I set it up with parallel packers under the bottom of the body, as this surface will not be machined, and protection strips for the machined faces. Two cuts, amounting to about 25 thou, were taken, the back edge not quite cleaning up. This time I tried a squeeze from the bottle of Rocol RTD spread out on the surface. This worked pretty well, remaining on the surface throughout the cut and producing a better finish than the dry cut. As this is not a mating surface it was rubbed on a piece of 800 grit to remove the machining marks and see how it comes up. Again the large photo shows tiny voids in the casting, not any where near enough to cause porosity. The first of the flanges was then tackled, again in the 4-jaw, with protecting strips on each of the machined faces, and a shorter piece on the underside, unmachined surface to ensure even pressure against the opposite jaw. About 25 thou was taken off.

Experimentally, I lapped the back face on a fine DMT diamond whetstone borrowed from the woodworking department. A good surface soon appears, but it is difficult to avoid scratches from particles trapped in the surface pockets of the stone. (1½ hours)

2010-03-14 – Flange check

After facing the second flange, I checked the profile with a paper template cut from a copy drawing. It looks as though there will be just enough metal to clean these flanges up. Good. Let's have a go at the crankcase.

The front and back faces will just clean up to the 2716″ length shown in the drawing, but I was hoping for a bit more than this. It only allows a 332″ radius round the outer studs, or 0.037″ metal thickness round the clearance hole, not enough for a proper seating of a 6BA nut. I don't understand why Westbury made this so tight. (½ hour)

2014-09-29 – Setting up

My last job today was to check the metal to be removed from each side and set it up for facing the back. It needs 0.025″ off the back and 0.020″ from the front to leave a lapping allowance on the final 716″ thickness. (¾ hour)

2014-09-30 – Facing the back

In a very short session today, I took a roughing cut of 20 thou and a finishing one of 4 thou. The manifold was obviously not properly bedded against the chuck as, although there was only 0.0005″ in it from end to end, there was a variation of 0.0015″ across the ¾″ from to to bottom. I need a more careful set-up for finishing the front. It looks as though the machined back face is unlikely to fully cover the cylinder port faces at each end. (¼ hour)

2014-10-01 – Facing the front

I have decided to add 164″ to the thickness. This will make it easier to differentiate between the machined front and back faces and the 'cast' surfaces of the flanges at each end. This means I want to machine it to 0.455″ thickness. The witness marks (in the photo) from setting the tool show I have got the back face better seated this time. It turns out within ¼ thou across the face this time, but a bit less than 1 thou out end to end, with each face dished by about ¼ thou.

The bolting face for the inlet manifold measures 2.430″, some 0.040″ short of what would be needed for 732″ diameter bosses on the inlet manifold. The back is also 0.025″ less than needed. If there is any discrepancy between the positions of the two faces, that will make matters worse. There is about 0.020″ available to machine the end flange faces, and this looks to be distributed evenly except for the lugs underneath which differ in thickness by about 10 thou.

This casting would, with great care, just machine to the original drawing. However I would really have liked more metal round the stud holes. I am not sure how to proceed, but the answer seems to be to press on. If I don't like it, I can start again. I may be able to reduce the boss diameters a bit, but that would depend on the Inlet Manifold, so back to that next. (1¾ hours)

2014-10-10 – Starting again

Work on the Inlet Manifold has persuaded me that I can have a better job by starting again. I have redrawn the manifold to allow just a bit more room. It is 116″ longer and 132″ wider and I have enlarged the end flanges a little to allow the ¼″ bore to be extended through the exhaust pipe by accommodating a 516″ OD pipe. The 'cast' flange will be 10 thou larger all round than the nominal profile, as I have done with the water flanges on the cylinders.

I have some suitable ¾″ thick alloy bar, so I cut out a piece 3″ × 1516″. I milled the top face and then half the excess thickness off one side in two light roughing cuts and a finishing cut of 0.005″, and similarly on the second side, all the while measuring carefully for size and parallelism, leaving it at 1532″ + 0.0025/0037. Next I milled the bottom face to a height of 0.920″, which leaves a few thou for hand finishing the end flanges. Transferring to the 4-jaw chuck in the lathe, I rough faced the ends, leaving about 10 thou each end for finishing. (4¼ hours)

2014-10-11 – Making holes

To set up the correct centre for drilling the hole that runs the length of the manifold I could mark out, centre-pop and use a wobbler and DTI, I could use a DTI directly to measure on the flats, or I could use a toolmaker's button set with slip gauges. I used the first method to get it roughly true, and then corrected it with the second. The centre pop turns out to have been about 4 thou out side to side and 1 thou from top to bottom. By setting the dial to zero on the sides, I had worked out I needed a reading of 0.297″ on the top face.

I took 0.010″ off the face, turned the flange to 0.775″ for 18″, deeply centred and drilled 5.8 mm to 1½ depth. Next I took a 4 thou cut with a boring tool to 1″ deep to ensure a true running hole. This was a nice fit on a 6 mm drill which I took to the bottom and followed with a 6.2 mm one. This was all done at 1000 rpm. I then dropped to 290 rpm and reamed ¼″. On removal, it measured 2.761″ long, so to allow 1 thou each end for lapping, I will need to face 9 thou off. I set it up again the other way round and repeated the procedure.

Next I set the manifold up in the milling machine for the bolt holes and ports. I drilled the top row of six mounting holes 2.8 mm, and then drilled and reamed the four ports ¼″. I drilled the bottom two holes to 2.2 mm for tapping, with little counterbores, breaking into the longitudinal hole a short distance. Finally I milled the connecting exhaust slots with a ¼″ ball-nose cutter.

After cleaning up and deburring, apart from tapped holes in the end flanges, the manifold is now functionally complete. I just have to make it look like a casting. (5½ hours)

2014-10-25 – Making buttons for templates

I turned silver steel buttons for the end flange template 0.238″ & 0.458″ diameter, and 38″ ones for the carburettor flange on the inlet manifold, then hardened all the buttons.

Next, I cut two pieces of 332″ gauge plate for the flange templates, milled each one rectangular, and drilled them. (7½ hours)

2014-10-27 – Drilling & tapping

Today I drilled the stud holes in the end flanges, and then tapped the eight holes in total on both manifolds, using some new 6 BA taps, as the 40 year old ones seemed to be getting a bit tired. (2 hours)

2014-10-28 – Milling the underside

With the manifold bolted to a clamping block, I milled between the flanges on the underside, initially with a 516″ slot drill, and finishing with a 3.5mm long series cutter to get the internal radius. (2¼ hours)

2014-10-31 – Making templates

The filing buttons for the exhaust flange template need flats so they will fit together. Rather than milling before hardening, I decided to grind these flats, which I have partially done, and then let them cool down while I worked on shaping one side of the carb. flange template. (1 hour)

2014-11-01 – Carving begins

This morning, I finished grinding the buttons to fit, finished filing the two templates, and hardened them.

In a first session working on the exhaust flanges, the right hand (flywheel) end is well on the way.

The original design is a rather crude block. I want a treatment at the back and between the cylinders that will look more like an authentic manifold casting, with less unnecessary metal. (5¼ hours)

2014-11-04 – Filing and milling

After some more filing, I decided the back needs to be relieved before I go any further. I milled 164″ off the lower part of the back face and a groove up the middle to leave two rectangular pads for the port bolting faces. I then filed a 732″ radius along the bottom back edge. I worked on defining the lower corners of the inlet manifold, as this will provide a template for the blending the front face of the exhaust. Still working at the right hand end of the exhaust manifold, on the fillets where the port face and exhaust flange shapes intersect. (4 hours)

2014-11-08 – More filing

The filing today moved to the left hand (timing) end of the exhaust manifold. By the end of the day both ends were finished. (5¾ hours)

2014-11-11 – Sculpting the middle

Having decided to cut away at the back, between the two port faces, I made a start with a coarse round file, but did most of the metal removal with a carbide burr in the Dremel. I then refined the shape with fine files and round riffler, and then gave the whole job a final titivating with fine wet-or-dry paper. All it needs now is bead blasting. (3½ hours)


2015-01-24 – Bead blasting

My new bead blasting cabinet arrived earlier this week. Time to have a go. First I made a pair of thick ¼″ washers to cover the end flanges.

Then I ground a 0.025″ wide parting tool and used it to make ten special, temporary 6BA washers 0.215′ diameter to cover the spot-faces, trying out different thicknesses intended for the final assembly washers, and, having decided, one with a chamfer to see how that would look.

I cut screwdriver slots in a couple of left-over experimental 6BA studs and used these to fill the tapped through holes in the bottom lugs of the end flanges.

The template for the carburettor mounting flange can be used, but new plates were need to cover the port faces, as the one hardened template was lacking a hole for the central bolt (necessary to protect the spotface on the inlet manifold). New plates were cut from some 1.5mm thick mild steel strip, and milled to a suitable rectangle. After drilling the holes I filed the plates one one side almost to the template profile to ensure the relief area at the back of the manifold would be properly exposed to the blast.

The photo above shows everything bolted up ready for blasting, which of course only took a few minutes. I dusted it off, stripped it all down an wiped it over. It looks good. It is a bit bright, but that may tone down with time and grime. (5¼ hours)


I have been playing with a silencer design for some time, and decided it would be another job I could do to postpone having to make the valves. It has a tangential inlet to a cylindrical chamber with radial holes to an inner, axial, outlet tube. The offset entry is intended to produce a swirling flow in the silencer, with the idea that it ought to more effectively dampen the pulses, it also lifts the silencer well clear of the timing case. I don't know how effective it will be. I can stuff it with wire wool or something if necessary, but I would prefer not to.

2015-06-06 – Turning flanges

I started with a brass turning session, making the silencer inlet pipe and flange, two silencer end flanges and turning the blanking plate for the other end of the exhaust manifold. (1½ hours)

2015-06-12 – Making end caps

I made two inlet end caps for the silencer, making the central boss too small on the first and doing exactly the same thing on the second. (1½ hours)

2015-06-13 – Making end caps

If I use a one-under-size nut to hold the silencer together, the second of the flanges will be OK. After mangling an outer end cap while trying to hold it for second operation turning, I made a good one from the first of the inlet end flanges. (1 hour)

2015-06-16 – Drilling the silencer tube

Using a short stub mandrel turned in situ, I rounded the corner on each silencer flange. I then made the offset entry hole in the silencer tube, badly. I decided a longer tube would be better anyway, and cut and faced a new piece to 2½″ long. While chewing over how to make a better job of the side entry hole, I cut the inlet pipe to the correct length. (4 hours)

2015-06-18 – Jigging the silencer tube

After realising that the water pipe jig would be a good basis for a silencer jig, I cut a piece of steel for the job, but then decided I would do better to get the design on paper first.

I had intended to finish the oval flange profile and mounting holes after soldering, but if the job is jigged it would be easier to make the flange before soldering. With this in mind I set up and drilled the mounting holes in the flange, seting the holes from the circumference and forgetting that that circumference is currently left 0.009″ over size, so the hole positions ended up offset. Scrap.

I turned up a new inlet flange & tube using up a piece of gunmetal. As I intend to paint the silencer black, this should not matter. I also drilled the bolt holes in the blanking plate that goes at the flywheel end of the exhaust manifold.

I set up the small rotary table in the milling machine and centred it in line with the spindle. With a sacrificial aluminium jig plate fixed to the table, I drilled & tapped mounting holes for the flanges. The parts are tight fit on the jig, but are OK.

In milling the flanks of the oval flange profile, the Autolock chuck came apart, gouging a bit out of the little rotary table that I had put a great deal of care into making, and doing slight damage to the chuck threads. The rest of the day's work was spent failing to get the chuck to assemble nicely. (4½ hours)

2015-06-19 – Finishing the blanking plate

Using a 1″ cut-off disc in the Dremel, I was able to dress the bruised internal thread in the Autolock chuck body. The chuck now assembles without binding. This device may well automatically tighten the collet on the cutter shank, but the sleeve definitely needs a biff on the wrench to ensure it is tight against the body. Which is what I forgot to do. I think it unscrewed when I turned the spindle motor off.

Incidentally, most people insist on an incorrect procedure for using an Autolock chuck. They are told to leave the sleeve backed off half a turn, tighten the cutter against the centre and then tighten the sleeve with the wrench, leaving a small gap. This is not what the instructions say, is a complete misunderstanding of the mechanics of the chuck, is likely to cause the damage one often finds to the important centre up inside the chuck, and relies on the thread for the true running of the cutter.

To mill the silencer inlet flange profile I first milled straight across on each side, turning the part on the jig to ensure symmetry, and producing a width of 0.4975″. This means there is exactly 0.030″ to come off each side. I set the angle stops on the table and milled both flanks. One angle was not quite right, but I had left sufficient for file finishing with buttons.

After tweaking the rotary table stops I milled the blanking plate. Again I filed the radii to buttons, then lapped the faces and then gave the outside face a straight line rub on 1200 grit for a fine textured finish, and fitted the piece on the manifold. (3 hours)

2015-07-03 – Resuming the silencer jig

Moving back to the jig, I milled the ends of the mounting plates square and to length, and cut the hole for the silencer tube using an end-mill. This hole is perhaps a bit tight. (2¼ hours)

2015-07-04 – Continuing the silencer jig

Today I spent most of my shop time on a storage rack, but also drilled holes in the jig baseplate and one of the existing posts.(¾ hour)

2015-07-05 – Continuing the silencer jig

Between being repeatedly pestered by the junior dog to play ball, I managed to finish drilling and tapping all the remaining holes in the jig. (1¼ hours)

2015-07-07 – Using the silencer jig

Once the new posts were slit on one side for clamp screws, the jig was finished. I used it first to make the offset hole in the silencer tube, and then to silver solder the tube and inlet together. It soldered and cleaned up nicely with a good joint line and little spread. However, the jig posts were difficult to get off and left the tube squashed a bit. This will have to be rubbed out. I had forgotten the 6-BA cap-head would be trapped after soldering, and had to sacrifice it. (3 hours)

In retrospect, I suspect that heating the extruded silencer tube caused it to grow in diameter (and shorten in length). This might well be explained by relaxing internal stresses from the extrusion process.

2015-07-08 – Finishing the tube

Today I bored out the unwanted part of the inlet tube projecting inside the silencer tube, leaving a small lip all round. I also sanded and polished out the blemishes on the outside of the tube. (1½ hours)

2015-07-10 – Soldering the exhaust tube

I made a spacer to position the outer flange on the exhaust pipe for soldering. I turned the tube to length and drilled the cross holes with a deliberately uneven pattern to mitigate possible resonance. I made a plug to fit the inlet end of the exhaust tube, deciding at the last minute to make it with a locating shoulder, but forgetting to shorten the tube to correct the length.

In soldering the inner assembly, I melted the plug slightly, and the flange is not quite square either. After cleaning up, I turned the flange to fit the outer tube. I think it will be OK. I turned the end of the plug away, restoring the correct length of the inner tube. It seems to still be well soldered. (4½ hours)

2015-07-14 – Finishing the silencer

Because of the considerable amount of dressing I had to do on the outer tube, the outer flange was standing proud of the tube surface when fitted, so I turned five or six thou. off the diameter to remove the ridge.

With the tube chucked for this operation it was running true, so I set the top-slide round to 3° and bored a flare in the outlet end, out to about 0.275″ diameter at the edge. This thinned the pipe to a more scale appearance. In theory this would also reduce the exhaust back-pressure, but I tried applying Bernoulli to some big assumptions and concluded that the effect would be minuscule.

I drilled and tapped the plugged end, made a stud to fit, and assembled the finished silencer. (2 hours)