Seagull Engine Construction Diary

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

Cooling System

I am intending a passive convection cooling system with an open topped tank piped to the cylinders. I want to achieve neat pipework with swept Y joints, and two-bolt flanges throughout.

2014-07-24 - Making sample flanges

As a break from fettling the cylinders today, I turned a couple of discs for sample water flanges, to help decide on the right thickness to make them. I made one about 0.078″ thick and the other 0.086″. This involved making a narrowed facing tool to clear a step in the soft chuck jaws, so that I could face them to thickness. (1 hour)

2014-08-04 - Making a sacrificial milling plate

After drilling the flange blanks I started work on an aluminium plate to hold the flanges for profile milling. (2¼ hours)

2014-08-05 - Profiling a flange

With the plate drilled and tapped, and the central hole reamed, I set up the the plate and the thicker blank on the small rotary table. In cutting the sides I managed to get a jump in the table position and milled too much off one side, and probably did the same again milling the angle flanks. (1½ hours)

2014-08-07 - Finishing the flanges

I achieved a better result with the thinner (564″) flange by turning it 180° on the jig and machining from the one side to ensure both sides were identical. This is tedious because of the small screws but works well, it came out with a very small allowance for finishing.

Next, I made a pair of 316″ filing buttons for the bolting lug radii. Using the buttons on the flanges showed that the flanks were not milled to a very accurate angle, but they are good enough for evaluation.

Using a bit of pipe, bent in the newly made pipe bender, I did a trial assembly with dummy studs and nuts. I think the thinner flange looks about right.

2014-08-08 - Silver soldering

I tried silver soldering one of the flanges. I will clearly need a jig for this as the parts cannot be guaranteed to stay put by themselves, especially as the flux dries. Soldering from the back if possible should help to ensure a neat solder fillet on the visible side. Cleaning up after soldering will best be done in the jig too, as the least vice pressure on the tube soon distorts it.

2015-06-06 – Turning flanges

Today, as well as making a start on the silencer, I made 13 bronze discs for water pipe flanges. I parted them off, then faced both sides using soft jaws in the chuck. I left 11 of them 0.002″ over thickness to allow for lapping flat after silver soldering. (2¼ hours on these)

2015-06-13 – Drilling flanges

I drilled and reamed all the flanges to 732″. (¾ hour)

2015-07-14 – Pipe bending

With the silencer finished this afternoon, I had a little time to make a start on the water pipes. I found the edges of the bender die are too sharp and cut into the tube. (1 hour)

2015-07-15 – Easing the Pipe Bender

I took a shave off the edges of the bending die, and discovered that a fine cut-off wheel in the Dremel is a good way to cut the soft copper pipe without crushing it. I also discovered that setting up the pipes in the soldering jig is going to be tricky. (1¼ hour)

2015-07-24 – Preparing to drill the flanges

I am planning to use the small rotary table for milling the flanges, so I might as well use it for drilling them too. I already have a 732″ locating pin for the table, but this needed a polish and some of the flange holes had to be eased a little to fit the pin.

First I centred the rotary table on the milling spindle centre and set the dials to zero. An aluminium jig plate already has a milled reference edge which Ilined up. Returning to the zero position I, drilled & reamed a 732″ hole in the plate and drilled a 2.2mm clearance hole to one side and on the other a 1.7mm tapping hole which I tapped for an 8-BA stud.

I drilled a flange 2.2mm on one side (without a stud in the jig) but found the stainless studs made for them would not fit. Even drilled out to 2.3mm, it would not fit over some. I found an ancient, worn, broken and resharpened 332″ drill, measuring 0.0925″ would produce a hole that goes over all the studs, but not necessarily a pair of holes that will fit on every pair of studs. These can be adjusted on assembly if need be. I gave the drill,a new 4-facet sharpening. It is short enough to use without centre drilling.

I now have a method. The first hole in each flange is drilled with the flange centred by the locating pin. Then, with a stud fitted in the jig to locate the first hole, the second ones can be drilled, all without moving anything except the parts. The rotary table serves merely as a convenient way to position the locating pin and jig plate. (3¼ hour)

2015-07-25 – Drilling and turning the flanges

Before drilling the flanges I decided to check the mating studs on the engine for accuracy, by measuring over the fitted studs and subtracting the stud diameter. Gratifyingly, they ranged from -0.0005″ to 0.0020″. It is surprising that a 0.0035″ diametral clearance seems to be not quite enough.

After making sure all the flanges pass right over an 732″ reamer,and making sure the locating pin would fit in the first hole for drilling the second, I drilled 24 holes to complete the drilling stage. I then made a thumb nut and stub mandrel to turn the flange diameters, in batches, to a shade over final size, and picked and rubbed off the burrs. (6¼ hours)

2015-07-27 – Scrapping some

To make flat sides on the flanges I gripped them, four at a time in the milling vice. On the second side of the first batch, with a cut over 18″ deep, the flanges lifted out of the vice and went in the scrap bin. I had no suitable bar left, but I did have some 1″ diameter offcut discs in thicknesses between 18″ and ½″. I started roughing them out. (1 hour)

2015-07-29 – Making more

The thickest offcut was large enough to make two flanges, so I parted that, and then faced one side of each flange.

Both of my sets of soft jaws had been used enough times that I could no-longer find a way to seat the flanges for second operation turning. This was not helped by the jaws not having a 120° point. I pressed the cheap Chinese tilting vice into metal working service for the first time in the years I have had it, and milled bevels on all six jaws. I still had to machine quite a lot off one set to get a useable seating for the flanges without the facing tool fouling the jaws. The soft jaws seem to be mostly used for thin pieces like flanges, so I think one day I will fit them with disposable 'shoes'.

Finally, I faced the backs of two blanks, leaving them over size. (3¼ hours)

2015-07-31 – Facing the second side

I drilled and reamed two of the flange blanks, and prepared the other set of soft jaws, with an undercut internal corner. Even so, the flanges showed a thickness variation of up to 0.003″, and I don't know why. (1 hour)

2015-08-01 – Facing again

First I refaced the soft jaws with a short slug clamped near the tips. Then I refaced the discs. They were a lot better, but still not perfect, and two of them have very little allowance for cleaning up after soldering. If they are drilled for the studs in the right place, the thinnest part can be cut away when making them oval. I drilled and reamed the remaining spare flanges, and turned them all to 0.635″ diameter, using the stub mandrel as before. (1½ hours)

2015-08-05 – Drilling and milling

Firstly I set up and drilled the fixing holes in the five replacement flanges. Moving on to the next stage, I rough milled one to about 0.360″ wide, in one pass each side, turning the flange on the jig for the second pass to ensure symmetry. I was unsing an 18mm cutter, and was not getting a good finish, with a heavy burr. (1¾ hours)

2015-08-07 – Milling flange sides

I milled another one, using the self-act. The burr was still ajob to remove. Using a 12mm cutter at 2200rpm, and still using a fairly slow auto feed, the flank finish is much better. I did two in an hour, and the next one in 10 minutes, finishing with 10 done and 4 to go. (2½ hours)

2015-08-08 – Button making

I finished milling the remaining flanges, and then made two sets of filing buttons 532″ and 332″ diameter, one of each having an integral locating pin and the other being annular. Because of the closeness of the holes, I milled flats on both sets of buttons, to allow them to be used together. Next I hardened them, and tempered them a little to avoid snapping the delicate bits.

I cleaned the flange holes with a 332″ reamer and then filed up one flange. I found that the outer edge of the centre buttun had marked the flange a little.(4 hours)

2015-08-11 – Filing

I filed a second flange to the buttons, which took about half an hour. The button shoulder is still cutting into the surface, in spite of being polished up. I tried milling the flanks of the next closer to size, but misread the rotary table protractor, which is marked in four quadrants of 0-45-0 degrees, and which can be confusing. Another for the scrap bin, but I was able to file it up partially, and it proved much easier to finish. So it is probably worth trying to mill the flanks closer to size. I milled another, and dropped it while filing. The small ring button came off and I heard it bounce and roll. Eventually I found it, and finished filing the flange, but I am not very happy with the finish quality.

I milled the flanks of the ten remaining flanges near to final shape, and after deburring them, filed one up. Using a needle file and sitting at a small vice I was able to make a better job. It is close work and quite tiring. (4¼ hours)

2015-08-22 – Much more filing

Over the next couple of weeks I spent most of my workshop time on another project, but filing a flange or two each session. It would be too tedious to report each day. I had them all done by 30 Aug. (4¼ hours in all)

2015-08-30 – Preparing pipes

Starting work on the pipes, I cut two pieces for the top, outlet pipe. The long piece I clouted with the Dremel while cutting it off, and the short one I bent so that the down leg was too short. Oh, well - I can use them for working out how to make the Y-joint. I made a trial, cutting with the Dremel and filing to shape, but it was not very good. I annealed the two remaining lengths of pipe. (3¾ hours)

2015-08-31 – Making an experimental joint

I made some collars to hold the pipes in the soldering jig, and they worked well. I soldered the joint and cleaned it up. The result was not bad considering the scrappy joint faces. The assembly is very difficult to hold for cleaning up and requres patience.

2015-09-26 – Making pipe grips

To avoid distorting and bruising the pipes I made a pair of semi-circular shoes for holding them in the vice (1½ hour)

2015-09-29 – Making a joint profile jig

I decided to make a jig for profiling the convex joint face. I have not gone into the geometry of the joint line mathematically, but setting it out with CAD, it appears in practice to be very close to a circular arc.

The jig is a sandwich in which the filler positions the tube and the side plates define the shape of the joint surface. I cut the side plates out of gauge plate and milled them rectangular, milled the filler piece to thickness and drilled and reamed the locating dowel holes, drilled the bolt holes, and tapped one plate. I made the dowels, and machined the guide radius on the side plates and then hardened them. (4½ hours)

2015-10-03 – Bending pipes

The final job on the jig was to mill one corner of the filler to the inside radius of the pipe bend. I started on the wrong corner, but the job finished up OK.

I made a sample convex half of the joint. The jig certainly makes this a great deal easier.

Next, I started bending the pipes, making three good ones but bruising a third. I had to shorten the straight, outside, forming die to allow for the closeness of the two bends on the inlet pipes.

This afternoon, after bending the final piece, I started working on the Y-joints, completing the convex curve on the inlet pipe for No.1 cylinder. The distance between the two bends turned out to be 18″ too great, but I was able to correct it by adjusting the top bend. (6¼ hours)

2015-10-06 – Fitting the joint

I tweaked the No 1 inlet pipe a little more so that it now sits in place with a little play. I carried on fitting the joint and adjusting the bends to fit the jig comfortably, getting the parts all ready for soldering. (3¼ hours)

2015-10-07 – Soldering the inlet

I painted some correcting fluid round the flanges to try to keep the solder off them, and cleaned and fluxed the joints. It felt as though I made a mess of soldering (my notes express it more forcefully.) On starting to clean up it did not look so bad. After lunch I carried on with the clean up. The flanges will need more work. The joint is not perfect but looks OK where it can be seen. (4¼ hours)

2015-10-17 – Preparing the outlet pipes

I filed a satisfactory external radius on the short No 1 pipe, and removed the bulk of the corresponding notch in the long No 2 pipe. The inlet and outlet pipe joints are arranged differently so that the joint is out of view on the underside of each. After more work in the afternoon the joint is now close but sits slightly skewed in the jig. With the short pipe trimmed to project less from the back of the flange, it will be possible to put a clamp over it to hold it in place. (3½ hours)

2015-10-23 – Soldering the outlet

First I tapped a hole in the jig baseplate for a clamp to hold the short tube in place. I cleaned, fluxed and assembled everything in the jig, soldered, and stripped down. This one seems a better job, although I seem to have soldered one of the nuts to a flange. Most of the nut could be ground off quite easily using the Dremel, with which I also cut back the pipe tails. I found that in filing excess solder off the joint, it is very easy to cut into the soft copper.

In the afternoon I removed another nut that was stuck on but not soldered, and continued cleaning up. That file mark will not rub out easily. I also managed to cut into the bolting face of the flange for No 2 cylinder. It still allows a seal, but is noticeable. I am not sure what to do about it.

I rubbed the inlet flanges flat on fine paper on the surface plate and did a trial assembly. (5 hours)

2015-10-24 – More finishing

Today I carried on with the outlet pipe, and flattened the outlet flange face. (2 hours)

Work on the engine now stops for quite some time, as I set about making a camshaft grinder.

2020-05-01 –

Since last writing about the cooling, it has become more complicated. I have acquired a little radiator and fan, intended for cooling souped-up computers. The plan is to use a microcontroller for some limited engine management. The fan will be switched on and off, and its speed controlled according to a temperature sensor on the engine.

This evening I wired up the fan to a 12V battery for the first time. It is very quiet and does not push as much air through the radiator as I thought it might, but I think it will be enough. It did not seem to be running very fast, but a check with the hand-held tacho indicated 3100 rpm.