A Simple Uniflow Engine

[gary.a.ayres]

Some endmills arrived today but I reckon I'll hold out for the counterbore...

[gary.a.ayres]

G'dang!

The counterbore arrived today but the guide section on it is too big for the holes!

I thought that might happen, which is why I bought the endmills alongside. I'll just use one of them. Everything will be locked down tight so the counterbores they make *should* remain centred on the holes.

I don't want to open up the holes just to accept the counterbore guide as the screws make a neat fit in them as they are and I'd rather keep it that way...

[gary.a.ayres]

The counterbore which I didn't use as the guide was too big for the holes:



An endmill was used instead, which went ok (screws dropped in as I went along, to confirm the depth):



I used a piece of the same steel as used for the crankshaft as a gauge to locate the bore of the flywheel before taking down the setup. It's a nice bearing fit so I hoped it would be accurate enough. This was in order to keep the DRO at the same setting to make a corresponding bolt circle on the flywheel boss. To be honest I'd have been better just taking a note of the co-ordinates of the centre. Ah well...



The completed counterbores, just visible here:



The machining marks that are visible here look worse in the photo than in actuality. However, I will try to deal with these later.

The back of the flywheel, showing the M10 bolts protruding to attach the boss:



I couldn't resist another quick mockup of the unfinished flywheel on the unfinished engine frame:



The boss was then set up in the milling vice with its centre located using the same piece of bar as a gauge. Five holes were centre-drilled, drilled and tapped M10 using the same pitch circle setting on the DRO. The depth of these holes was measured in order that they would be blind but long enough to accommodate the screws. The photos of the drilling and tapping all came out blurry, so here's one of the initial centre-drilling. Nothing too exciting, I know, but all part of the process:



And then the moment of truth. I had tried to be careful so that no inaccuracies would creep in, but while I hoped this would be the case I did not expect it. Actually, I was fairly pleasantly surprised. On assembly, some of the screws were very tight and I thought I may have to widen the holes through the main flywheel to create some wiggle room. However, with a bit of perseverance they all went in, and to depth. There was a bit of inaccuracy, however. I found that in order to carry out the assembly, it helped to have a short piece of bar through the bore of both components. By the time the screws were tight, however, the bar was pretty well clamped due to a very slight misalignment between the bores. Some inaccuracy had indeed crept in. Overall, though, it's not a great deal and I suspect a pass with the reamer will sort it out. Or even better maybe take it apart, rotate the boss and reassemble a few times to see if there's one position that works best. I'm actually quite pleased with it as I was half-expecting it to be a whole lot worse. The joys of the DRO! The boss will still have to be turned down and probably also given another light facing cut. I aim to do this with the parts assembled, and I guess it can be treated a bit like a casting at this point:



Finally, the wheel was chucked in the 3-jaw, held by the shallow raised section on the front side of it:



 There is a slight wobble and at this stage (which is where I left it for the night) I think that's largely down to the workholding. I don't think the 3-jaw is the way to go with this. The ER32 collet chuck and a superglued arbour may be better, but I need to sort out that slight misalignment in the bores first.

[gary.a.ayres]

So... I tried rotating the boss step by step and bolting it to the wheel in all five different alignments to see if any position would allow me to pass a length of the shaft material through the bore of the wheel/boss assembly:

 

No joy. The shaft slips in neatly from the wheel side and the boss side but in either case it stops at the point where the wheel and boss meet. I returned it to what I feel may be the optimal position and the misalignment looks very slight - I suspect a reamer from the front of the wheel will sort out the bore. If this works it will put the bore of the boss slightly out of shape but it should leave the bore of the wheel part intact and as that will be the longer section of bore the whole thing *should* still run true on the shaft.

However, first I need to get the outside of the boss turned down to diameter in situ on the wheel. I took a short piece of shaft and superglued it into the flywheel to make an arbour:



This will be left to cure overnight and then held in the ER32 collet chuck so that the boss can be turned down to size and faced. The wheel is heavy and there will be a lot of shearing force on that glued join so I'll have to go very gently, especially as there are currently a couple of rogue notches on the outer surface of the boss which will in effect give me an interrupted cut initially. If the superglue arbour doesn't hold I'll have to try it another way. If it does, I'll be able to turn the boss down to size then reverse the assembly, hold it by the boss in the three jaw and ream the bore right through...

[gary.a.ayres]

With very careful turning I managed to true up the boss without breaking the join on the superglue arbour held in the ER32 collet chuck:



The problem (as described in the preceding post) is that there is a misalignment between the bores of the main part of the flywheel and the boss (which is on the right in the photo). The superglue arbour only reaches as far as the join between the flywheel and the boss - it will go no further. However, by turning down the boss in this setup I now have the outside diameter of the boss running concentrically with the superglue arbour and the bore of the main flywheel.

The next step is to remove the superglue arbour using heat, and hold the assembly by the boss in the three jaw. Here is a mockup of this with the arbour not yet removed:



The plan has worked so far, as the superglue arbour ran true in the mockup shown. My hope is that with the superglue arbour removed I'll be able to run the 12mm reamer through both components to create a continuous bore that will be concentric enough with the boss to function and run true. I appreciate that this may not happen - the existing misalignment could deflect the reamer and result in an irregular and unusable bore. If this happens, all is not lost - I'll open the bore up wide and put in a bush (maybe bronze) pre-bored to 12mm.

In any case, there is also currently a wobble on the widest part of the flywheel even with the bore centred. I'm not sure where that came from as I have somewhat lost track of the various steps I have taken. This is less of a concern because provided I get a decent bore through the whole assembly there is no shortage of material which can be removed in truing it all up.

[gary.a.ayres]

The short superglue arbour was removed using heat:



The assembly was then mounted in the 3-jaw and the bore reamed right through to correct the misalignment between the main flywheel and the boss. This was a moment of truth, as if the misalignment was anything other than miniscule it wouldn't have worked. Fortunately the reamer went through with little resistance, suggesting that the misalignment was indeed tiny:



A piece of 12mm PGMS slid through nicely and showed no lateral play. The bore seems fine:



In theory, the bore was at this point concentric with the circumference of the boss, but there was a wobble on the main part of the wheel, so the assembly was again mounted in the 3-jaw for an initial rough truing up:





It was then mounted on a second superglue arbour, this time one which passed through the whole assembly:



This was allowed to cure overnight. The arbour was then mounted between the ER32 collet chuck and a revolving centre in the lathe tailstock, and the wobble was turned out of the front of the flywheel:



This done, the assembly was reversed and the back of the flywheel (including the boss) was also trued up:



Unfortunately a significant amount of chatter made its presence felt:



While this isn't wonderful, it is (honestly!) not as bad as it looks in the photo. Futhermore, it is mainly on the back of the flywheel and the boss, in which places it will be relatively hidden *if* I don't turn it off. That may depend on how true the wheel runs in its final position on the crankshaft. If it runs true, I may just leave it as overall the flywheel as a whole looks pretty presentable. If it needs further truing up, that may offer an opportunity to remove the chatter marks. In retrospect, I think the problem may have been due to a carbide insert getting blunt and rubbing instead of cutting. This was probably made worse by the fact that I had to angle the tool due to the diameter of the wheel stretching the capacity of my lathe.

Another factor is weight. The finished flywheel weighs in at a whisker or two below 4 kg - in other words, almost the same weight as the lathe chuck which served as a surrogate flywheel during the test setup and had the engine running. For that reason alone I'm reluctant to remove more weight from the wheel if I don't have to. Interestingly, machining the wheel removed approximately 0.8kg in the form of swarf. It was obvious that the machining would lose some mass, and that was why I was happy to start off with blanks whose combined weight was higher than my target of of 4kg. But the amount lost in swarf surpised me. You never really feel the weight of it in the form of swarf. I had been prepared to make a pitch circle of holes to reduce the flywheel's weight, but it appears that this won't be required.

The second superglue arbour was removed:



The flywheel was ten set up in the milling vice and the centre of the boss was located using an electronic edge finder and the half function of the DRO:



These gadgets make it effortless.   

This is quite a big flywheel so instead of a grub screw I'll be using an M6 cap head screw to secure it to the crankshaft. The position of the hole was centre drilled...

 

Then drilled right through with the tapping drill:



The hole is stepped, and the section adjacent to the bore will be tapped M6. Above that, the hole has been widened to 10mm to give clearance to the head of the cap head screw. To do this I had to transfer the vice over to the floorstanding drill press as the length of the 10mm drill bit plus the height of the workpiece in the vice exceeded the capacity of the z axis of my small mill:



With the drilling done, the setup has been left in situ for tapping in the drill press tomorrow.

[gary.a.ayres]

Tapping the boss M6 for the retaining screw (is that the correct term?) in the drill press using a piloted spindle tap wrench:



The stainless cap head screw sits neatly in the hole:



Mounted once again on a piece of 12mm PGMS with the screw tightened, it runs truer than I dared hope:



I'm very pleased with this flywheel in all respects apart from the machining marks in certain areas of its surface, and I'm calling it done for now. I don't want to carve away any more metal using a lathe tool, but am wondering how big a task it would be to polish the marks out using various grades of scotch-brite and/or wet and dry paper. If anyone looking in has any advice to offer on that (or any other abrasive that might work on EN8 steel without taking forever) I'd be grateful to hear it. Meanwhile, I need to top up my supply of scotch-brite and wet and dry anyway so I'll order some.

Next, though, on to another part of the build...

[john mills]

Hi
Scotch brite wont remove much metal when i was working turning shafts the turner had 36 grit emery paper
for when he won'ted to remove metal .the finer grades will take a long time to remove much even a few thou
move on to finer grades to get a finer finish  try finer grit until you get the finish you want.
                John

[gary.a.ayres]

That's great thanks John.

I will investigate that.

 

[propforward]

Looks great to me Gary, super job!

[gary.a.ayres]

Thanks Stuart. Very kind of you.

I'm pleased with it, but once I have seen it running true on the actual crankshaft I'll have a go at it with abrasives to try to get the machining marks out of it.

[gary.a.ayres]

On to the job of securing the cylinder to the frame.

The cylinder will be supported on the underside by two columns of round aluminium bar. These have already been drilled with a 6mm clearance hole through the centre. For each, a long M6 screw will pass through the base of the frame, up through the centre of the support column and fixed into a threaded blind hole on the outside of the cylinder.

Also, the cylinder will be anchored to the sides of the frame by four side braces - two on each side. These are thinner than the support columns but will be fixed in the same way.

Drilling the clearance holes through the side braces:



To install these columns and braces it was necessary to drill and tap six blind holes in the outer wall of the cylinder. This required quite a bit of care, as the last thing I wanted was to overshoot into the cylinder bore. In fact I erred on the side of caution, keeping the holes as shallow as I felt I realistically could. This meant that I had to optimise the amount of thread available in the holes for the screws to connect with, so the tooling sequence was centre drill, tapping drill for M6, 5 mm endmill to square the bottom of the hole, the usual three M6 taps and - to finish - an M6 tap ground square at the end to get a thread right into the bottom of the holes:



Tapping the hole for one of the the support column fixing screws in the drill press (as the mill doesn't have enough vertical clearance):



A fiddly marking out operation ensued for the position of the holes in the base of the frame for the cylinder support column fixing screws. I knew the distance between the face of the cylinder and the centre line of the crankshaft, so this was easy enough to measure and mark on the base of the frame. The cylinder was then aligned with this, allowing me to scribe the position of the back of it on the base as a visual reference. I then spent ages searching my shop for pieces of stock which could be used as guides to locate the cylinder centrally on the base in a lateral direction, finally settling on a parallel plus a piece of plastic bar  (visible in the photo) on each side to create a sliding fit between the cylinder and the frame support bars. Together, these steps enabled me to place the cylinder directly on to the base but in the correct position. Transfer screws were then put in the two threaded holes on the cylinder base. These can be seen at the bottom of the picture below. Ideally they shouldn't have stuck out so far but they wouldn't go in any further due to the shallow depth of the holes. However, with care it was possible to mark out the hole positions by holding the cylinder between the guides and carefully making a rocking, side-to-side scratching motion:



Al that was needed was to then scribe the centre line of the base to intersect with the marks made by the transfer screws:



The base was then transferred to the mill and clocked in:



The holes were drilled, using the DRO to double-check that the distance between them was the same as that between the two holes in the cylinder:



So, drilling these two holes was an evening's work. To be fair, they are important holes.

Aren't they all, though...?

[crueby]

Great sequence and planning. And drilling towards the cylinder is always scary!


 

[gary.a.ayres]

Thank you Chris.

Scary indeed! The prospect of having to bore and otherwise machine a whole new piece of cast iron if things went wrong kept me on my toes though!

Doesn't bear thinking about...!   

[propforward]

Well planned, and very well executed I would say Gary. It worked out very well for you. Great progress.