Introducing ... the "Steel Webster"

[awake]

Part 2 of the build log is very brief, but it turned out to be one of the better design decisions I made. Attachments 1 and 2 show the design in question: a pair of bearing holders that mount in the frame.

The primary motivation for these bearing holders was to have enough "meat" to be able to mount two flange bearings on each side of the frame. The secondary motivation was that this approach would make it easy if I decided to change the diameter of the crankshaft - instead of reboring the frame, I could just make a couple of simple parts to hold a different set of bearings. As it turned out, I never did the latter; I used F6800 bearings, 10mm bore, with a 10mm (.394") crankshaft, and it worked out well.

Making these was quite simple - hardly worth taking pictures, so I only took a few (attachments 3 and 4).

I said this proved to be a valuable design decision - even though I never needed to change to a different size of bearings. Here's why: The relatively large bore in which these bearing flanges sit greatly eased the assembly of the crankshaft and the components that ride on it. I could partially assemble the crank shaft with the bearing flange, the crank gear, and the keys, and then slide the whole assembly in through the bore. Given the way I made the crank, with interrupted keyways, I don't know if I could have successfully assembled everything otherwise.

[awake]

Part 3 of the build log is the crankshaft - plans shown in attachments 1 & 2 below.

A few notes about the design - as you can see, I chose to go with a built-up crankshaft for my first time. I also chose to go with an interrupted keyway, but I'm not sure that was my most brilliant decision - as I alluded to in a previous post, this caused problems with assembly, since I couldn't slide the key along the keyway as I put the shaft through the bearings. One other note is the pin end, sized at a hair under 6mm so that the pin can ride in 6mm ID bearings used in the big end of the connecting rods. The length of the pin is a few thousands longer than the thickness of the big end with the bearings installed - this allows a small custom washer and a 3mm screw to fasten tightly to the end to hold the big end in place without binding up the bearings. (On final assembly, I added a dab of blue Loctite to this screw, but in the initial runs it did not seem inclined to loosen, somewhat to my surprise.)

I'm afraid I didn't take many pictures. Attachment 3 shows the main shaft underway, and attachment 4 shows the pin in progress. Attachments 5 and 6 show the cutting of the keyway. A couple of things to note here. One is that I was, for the first time, using a 2mm endmill, smallest I've used to date. Why not use a 1/8" endmill? Part of the reason is that, originally, I was planning to make the keyway 3mm rather than 1/8", since the shaft itself is 10mm - keeping it in the metric family, so to speak. But I didn't have a ready source of 3mm keys, and I did have several 1/8" keys, so I switched to 1/8". Meanwhile, past experience suggests that endmills can cut just a bit larger than their diameter, so I prefer to use a smaller endmill and then shave the sides to size.

Something else that sharp-eyed readers might note in attachment 6 - you can see that the shaft is held in a 3/8" 5C collet. The use of the collet in the spin indexer was just a convenient way to grip this round part ... but wait; didn't I say the shaft was 10mm, .393" rather than .375"? No, 5C collets cannot expand that much! If you look back at the plans, you'll see that the end of the shaft, the part that goes into / connects with the web, is at 3/8" diameter, so the shaft is being held by that little stub in the 5C collet - but it is quite secure because I put a center in the other end, and as you can see, I am supporting that end.

Finally, attachment 7 shows a poor shot of the web end of the crankshaft after I assembled it - actually in place on the assembled engine, because for some reason* I neglected to take a picture of the finished crankshaft by itself.

*for some reason ... maybe embarrassment over the poor execution? In keeping with the spirit already established in the making of the frame, I decided to TIG weld the parts together. How did I do? Well ... the picture shows the results after cleaning it up as much as I could ... still not so great. But it works, so I'm calling that good for a first attempt!

[awake]

Part 4 of the build log gets into something that I think might be more interesting than some of the other parts - the flywheel. To be sure, the plans are quite simple, as you can see in attachments 1 & 2. But the execution was - perhaps - unconventional.

I was looking for a flywheel with diameter of 3.75", at least .75" thick at the rim. I had on hand some pieces of 1" thick steel, 4" wide, but not a piece that was 4" long. So, having already established the pattern of using welding as a primary tool in this build, I prepared two pieces, 4 x 2 x 1" in size. Part of the preparation was setting each piece at a 60° angle in the mill vise and cutting bevels to facilitate a full-penetration weld (attachments 3 & 4).

To weld these together into what I hoped would be a solid 4 x 4 x 1" blank, I began by spacing the two pieces with a 1/8" gap to help with penetration (attachment 5), then tacked (attachment 6) and ran a root pass on each side using TIG welding (attachment 7). Not the best-looking root pass, but good penetration.

I probably should have finished welding it up with TIG, but I was a bit concerned about the rods and argon that would be consumed, so I switched to stick welding with 7018 to provide quicker and less expensive build up - at least, in theory. As attachment 8 shows, the results were so-so -- the 7018 is old, and not nearly as dry as it should be, so I struggled with porosity and inclusions here and there. When that occurred, I tried to grind it out and build up again. Maybe it would have been less time and trouble to have stayed with TIG throughout!

To be continued ...

[awake]

Continuing part 4 of the build log, the flywheel:

Once the blank was all welded up, I began preparing the blank, starting with a carbide-insert face mill in the Bridgeport (attachment 1). In general, 7018 weld on mild steel machines reasonably nicely, but I also filled in some gaps or defects with TIG. Normally, TIG welded mild steel machines beautifully, but if someone happens to dip the tungsten into the weld, it forms a small spot that is hard as a rock. I'm not saying that I did such a thing ... but I did feel the need to start out with carbide.

Once the surface was brought down nearly to level, I further prepared the blank using the shaper, getting the two faces flat and parallel (attachment 2). Then I cut the corners off on the bandsaw to get a rough octagon (no pictures of that), mounted the blank in the 4-jaw chuck on the lathe with roughly half of the blank above the jaws, and begin the roughing passes to get an oversized round blank (attachment 3; you can see the centering mark that I put in blank to help get it roughly centered in the 4-jaw chuck).

Once one side was roughed in, I turned the piece over, switched back to the three jaw chuck, and mounted it up to the octagonal rim that remained on the other half of the blank. Now I could rough this other side, again leaving it oversized at this point (and also leaving just a tiny "rim" of the former octagon that had to wait for a later step). At this point, I began to do the finish turning on the exposed face (attachment 4).

Once the face was finished, I bored it to a nice, tight, but sliding fit on the 10mm crankshaft. As shown in attachment 5, the rim area is still oversized in diameter, but the inner hub and web are to final size. Just after this picture was taken, I finalized the face portion of the rim as well, facing the rim down to the proper height above the web.

I removed the flywheel, switched to the 3-jaw chuck, and mounted some 1.125" diameter mild steel stock (scavenged long ago from some now-forgotten source). I turned a .9" long stub mandrel down to a nice sliding fit in the bore of the flywheel, with a flat face behind it (neglected to take a picture of that). Leaving this mandrel mounted so that it stayed perfectly true, I used Loctite to attach the flywheel to the mandrel and let it set up overnight. The first go round, I tried using the "blue" version that would be easier to remove, and I got as far as roughing out the inner hub and web, but ultimately it just couldn't take the torque, so I had to clean it up and do it again, this time using the "red" version. I held it firmly in place using the drill chuck in the tailstock while it set up overnight (attachment 6).

With the flywheel Loctited (is that a word??) to the mandrel, I could turn the rim to final diameter, including cutting off the little bit of octagonal "flange" that was left in previous steps (attachment 7).

To be further continued ...

[awake]

Further continuation of part 4 of the build log, the flywheel:

At this point, what is now the backside of the flywheel, up against the face of the mandrel, is fully finished, including the height of the both the inner hub and the rim relative to the web. Careful measurements were taken of the OD of the inner hub and the ID of the outer rim. Also, in the last bit of the previous post, we had finalized the overall OD of the flywheel.

Now to finish the exposed face to size. The rim area can be faced to achieve the desired .75" total thickness (attachment 1). But the web is another matter - how to measure it? The movable anvil of the micrometer can go up into the inset area of the web, but not so for the fixed anvil.

The answer, of course, is to use a gauge block in the back side of the web, against which the fixed anvil can set; after measuring up into the inset area with the movable anvil, just subtract the size of the gauge block from the measurement, and hey presto! We get the current thickness of the web.

One small problem: I don't have any gauge blocks. (On my wish list, in case anyone was wondering what to get me for my next birthday ... ) But the problem is not insurmountable; I just need to turn a piece of scrap so that it has parallel faces, measure it carefully, and use it as my gauge block. But another problem, not so small: my whole procedure depended on leaving the mandrel mounted in the 3-jaw so that it would stay true, and thus also the flywheel would stay true as I cut it. If I removed the mandrel to turn a piece of scrap, I'd lose my carefully planned precision.

This is where having a second lathe is invaluable. (Note to spouse: See? This really was a good purchase!) I turned a bit of scrap in my rebuilt 7x14 lathe (attachment 2), measured it carefully (attachment 3), and then used it to take the measurements of the web of the flywheel (attachment 4).

This allowed me to cut the web to final size, and once that was done, to finalize the ID of the rim and the OD of the inner hub. Voila! The turning is complete on the flywheel (attachment 5). Now it is just a matter of cutting the keyway.

To be continued one more time ...

[Roger B]

That's some serious work on the flywheel  I do have to wonder if it would have been cheaper/easier to buy an appropriate piece of material rather than the welding consumables    A 124mm (5") flywheel from RC is EUR 24.

https://www.rcm-machines.com/en/flying-wheels/flywheel-cast-iron--124-mm,-partially-machined/rcsr124

[awake]

The final continuation of part 4 of the build log, the flywheel. If you have stayed with this all the way through, you deserve a special treat; please help yourself to your beverage of choice and settle in for the final installment!

I cut the key way in the flywheel (and later on, in the cam gear and the ignition cad/starter spud) using the shaper. First, of course, I had to make a cutter. I tried a couple of approaches, but settled on a cutter ground from the remains of a 1/2" end mill (attachments 1 & 2). Never, ever, ever would I throw away an end mill just because it is no longer usable as an end mill - heaven forbid that I waste all that HSS that is just waiting to be ground into something useful! (This is also meant as a note to my spouse, who seems to think I am a pack rat!)

I also had to make something that would let me mount this cutter in the shaper. You can see what I came up with in attachment 3: I cut a scrap piece of 1" thick mild steel to give me a "leg" that can go into the tool holder on the shaper, a body with a round hole to accept the cutter, and two cross holes that hold split-cotter clamps. (Only one of the latter is in use in this picture, because I had to extend the cutter out further than I had first anticipated - but just the one held securely, with no fuss or muss.)

Attachment 4 shows the cutter in action on the flywheel. The cutter was ground to around .100" wide, rather than the full .125" needed for the key way. Part of that had to do with what was usable from the old endmill that formed the cutter blank, and part of it had to do with the thought that a narrower cutter would chatter less, and allow me to "sneak up" on the proper size, shaving down the flanks until the key just fit (attachment 5). After cutting 4 key ways this way, though, I am wondering if I should have gone ahead and ground it to the desired final size - something different to try next time, I suppose. I have lots more broken 1/2" endmills.

Finally, the flywheel is done (attachment 6). Well, almost done - I still had to hold the flywheel at a bit of an angle to drill and tap the hub for the set screws that screw down onto the key - one set screw on each side, placed over the key way. Fortunately unfortunately, I neglected to take a picture of my rather sketchy set up, but I did accomplish the task in the end. After breaking a tap. Which fortunately I was able to persuade out. (I did NOT want to have to go through all of that again to make a new flywheel!)


The good news is that the finished flywheel runs absolutely true. The even better news for you, patient reader, is that this part of the build log is finally complete!

[awake]

That's some serious work on the flywheel  I do have to wonder if it would have been cheaper/easier to buy an appropriate piece of material rather than the welding consumables    A 124mm (5") flywheel from RC is EUR 24.

https://www.rcm-machines.com/en/flying-wheels/flywheel-cast-iron--124-mm,-partially-machined/rcsr124

Now what would be the fun in that?

Actually, that is a very nice looking flywheel, and I would never have imagined a blank like that was available so reasonably priced. If I could find a US supplier, and/or if the shipping were not too exorbitant, that might well tempt me to a more sensible alternative to pursue this alternative on a future build.

But in fact, as long as you don't count my time - or rather, count the time as enjoyment spent in pursuit of the hobby - I actually spent considerably less. The metal was free, literally - it came out of a scrap bin to which I have been granted access. Despite my stinginess Scrooge-like parsimony sensible frugality with my TIG rods and argon, the amount used really doesn't add up to much, maybe a couple of dollars at the very most, and even less for the 7018 rods that I used to complete the weldment. Even allowing some for wear on the bandsaw blade and other tooling, I'd guesstimate that the total cost invested in the flywheel was less than $5.

[awake]

Part 5 of the build log is the rod (attachments 1 & 2). Once again this will have to be a multi-part post, as I took quite a few pictures - despite being a relatively small part, the rod took quite a lot of setup and machining, including figuring out how to use my Christmas gift - a 6" rotary table.

The process began with machining a 5/16" thick blank from which the rod would be cut - once again using my trusty shaper to get it flat, parallel, and to size (attachment 3). Next was boring the hole for the big end, sized to accept a pair of F686 flanged bearings with a light press fit (attachment 4). I also drilled the hole to accept the plain bearings for the small end - nominally .250", but the exact size didn't matter, since I later made the bearings to fit.

The next operation may look rather strange in the picture, but hopefully will make sense when you look at the plans - I drilled 4 holes, carefully located using the DRO and started using a center drill, which will define the transitions between the rounded ends and the slightly tapered body of the rod (attachment 5).

The next step was to make a mounting plate with "buttons," each tapped and drilled at the top (attachment 6). The buttons were sized and positioned precisely to accept the rod blank (attachment 7), and a screw and washer in each button held it securely in place. The plate is enough larger than the rod blank to allow clamping onto the rotary table, and it also acts as sacrificial material for the milling process.

Finally, I began trying to figure out the best way to center the new 6" rotary table under the spindle of the mill. I put an MT2 dead center in the center hole of the RT, and started out trying to use a DTI to center (attachment 8 ).

Part 5 continues below ...

[awake]

Part 5 of the build log, the rod, continued:

Continuing to try to get the RT centered: I was not getting entirely consistent results using the DTI mounted on the spindle of the mill, so I tried switching to an edge finder, and compared the results (attachment 1). They did not exactly agree, but I got it within .003" or so and got frustrated and quit decided it was good enough. With the RT centered on the mill spindle as best I could, zeroed the DRO X & Y. Then I clamped the mounting block in place, centering the big end "button" under the spindle using the edge finder - thus it should also be centered to the RT. I also used the edge finder to line up the small end button so that it was exactly 0° along the X-axis (attachment 2), and I zeroed the dial on the RT.

Getting the RT centered AND the part centered on the RT was by far the hardest and most time consuming part of making the rod - is there an easier way to go about it? (I was badly lusting after Craig De Shong's Volstro rotary attachment by the time I got done with this!)

Once centered and aligned, however, the actual machining proceeded pretty much according to plan. First I moved the X-axis to position the 0.250" diameter cutter on the OD radius of the big end. I raised the knee until I just grazed the blank, zeroed it, then raised it .031" and rotated the RT 360° to machine the boss around the big end (attachment 3). I continued to rotate the RT until the cutter was aligned with one of the .250" holes previously drilled to define the transitions, raised the knee another 50 thou or so, and rotated the RT to cut around to the other transition hole on the other side; rinse and repeat until the OD of the boss is completely cut out.

Next I reset the RT to 0°, lowered the knee back to where I was taking a .031" deep cut, and machined away the inset section of the body of the rod (attachment 4). Once again I reset the RT to 0°, moved the X- and Y- axes to position the cutter at the location of one of the transition holes, rotated the RT 2°,  lowered the knee 50 thou or so, and began cutting the tapered flank of the body section, stopping when I reached the transition hole at the small end (attachment 5). Rinse and repeat until the flank is completely cut away.

I lowered the knee again, rotated the RT to 2° on the other side of 0, positioned the X and Y axes to the other transition hole at the big end, and again raised the knee and began cutting the other flank of the rod by 50 thou or so at a time. Finally, the sides of the rod, along with the OD of the big end, were completely machined (attachment 6).

I removed the screws holding the blank in place, flipped the piece over, and took the .031" deep cut around the boss and down the length of the inset section of the body of the rod. Then I reset the X and Y axes back to 0,0 (centered over the RT), removed the blank from the mounting plate, unclamped the mounting plate, and reversed it to center the small end on the RT / under the spindle. This time I tried using the cone side of the edge finder, and I decided that was way better than trying to center the part using the edges (attachment 7).

I followed a similar process to the one described above to cut the boss section on each side of the small end, as well as the OD of the small end. At last, the rod is complete (attachment 8 ). Well, almost - we still need bearings!

Continued one more time ...

[awake]

Final continuation of part 5 of the build log, the rod:

The next step was to make the plain bearings for the small end. Unfortunately I did not have any 660 bronze smaller than .75" in diameter, so I had to turn that down - but these plain bearings are very short, so not too much waste. I turned down a length to the diameter of the flange and drilled out a .125" hole through the middle. Using a dial indicator to get the exact length, I turned down the body of the bearing to .251" diameter and .150" long, then moved over to leave a .020" thick flange and parted it off (attachment 1). Repeat the last few steps and now I have two plain bearings. Note that the length of these bearings will leave a .013" gap in the middle of the small end of the rod. I drilled through the top side of the small end using my smallest center drill, giving me a "dimple" in the top and a tiny hole through, so that oil can drip through from the cylinder to the piston, from the piston to the small end, and make its way around the .013" gap to lubricate the wrist pin. At least, that's my theory!

You may have noted that these plain bearings were sized at .251", to go into a hole in the small end that is a nominal .250" - in other words, a press (or rather, a shrink) fit. .001" is actually more interference than I thought I really needed, so I actually sized it for about .0005" of interference. I heated up the small end of the rod using my trusty heat gun (attachment 2) - which was a bit of an impulse buy on an extra-cheap sale, but has proved to be invaluable over the few years I've had it - and only had to press very lightly, really just a slip fit, to seat the bearings on either side (attachments 3 & 4). Of course, once the rod cooled down, the bearings were going nowhere.

There is nothing much to the wrist pin (plans for which are actually together with the piston, coming as part 6 of the build log); the main consideration is achieving a rather precise OD of .187" (attachment 5). This precise sizing was intended to fit into an equally rather precise hole of .188" diameter through the small end bushing. Two problems: the first is, how to hold the rod at this point, with all of its rounded and angled surfaces? The answer was to use the mounting block again, this time in the vice; I mounted the big end on the mounting block and screwed it down with the small end hanging over the edge so that I could drill through (attachment 6). This worked way better than it seemed that it should - I was able to drill without movement or chatter.

The other problem was how to get the precise .188" diameter hole when I didn't have a reamer of that size. The answer was an old-timer's trick  - I "snuck up" on it by drilling it first with a 5/32" drill (through the undersized hole that I had predrilled in the bearings); then I drilled again with an 11/64" drill; then one last time with a 3/16" drill. The idea is that the 5/32" drill will drill a bit oversize, as is typical for the average twist drill; the 11/64" will mostly just clean that up; and the 3/16" drill will basically act like a reamer. It worked; the result was a very nice sliding fit on my wrist pin (attachment 7).

The final step was the addition of the F686 bearings in the big end (attachment 8 ). These were a light press fit, and I used them that way throughout the first few runs of the engine, but once I took it apart, finalized everything, and reassembled, they were secured with "red" Loctite.

The rod is now complete; on to the piston!

[awake]

Whew! It has been an insanely busy week. Apparently "work at home" has been translated into, "it's really easy to schedule a virtual meeting, and I'm sure you have nothing else to do, so let's schedule 10 or 20 of them!" Sheesh - glad it's finally the weekend so I can focus on something important!

Here is part 6 of the Steel Webster build log, the piston (attachments 1 & 2).

As you will see, I did not take many pictures - this went more easily than I thought it might. Attachment 3 shows the OD of the piston already completed, but not yet parted to size. Here's the most interesting bit of this part of the build log - if you look below the piston in the chuck, you will see a rusty chunk of cast iron, cut off of an old pump housing. This is another piece of the stock from which I made the piston that is mounted in the lathe! I cut a piece down on the bandsaw as best I could, then used the shaper to shave down the lumpy outside to get something round enough to hold in the chuck. Then I turned it down ... and as you can see, it turned beautifully!

Since the stock that I reclaimed in this way was more than long enough, I decided to try making some cast-iron rings. Attachments 4-6 show the process - what you are seeing is the piston, now reversed in the chuck and protected by my highly sophisticated aluminum shim (AKA cut-up Coke can). First I turned it to the OD I wanted for the rings, then bored it, then parted out 5 rings using a narrow parting tool. The round rod that you see, on which the parted-off rings are resting, is just a bit of cold-rolled steel that I put in a drill chuck in the tail stock, there just to catch the small parts as they come off the blank.

Confession: I haven't actually done anything with these ring blanks. Since this was my first engine build, I decided to try going with a viton o-ring, and that seems to be working well - just a single o-ring, but you will note in attachment 3 that I cut two grooves for rings - this in case I decide to experiment with the CI rings down the road.

I failed to take any pictures of the next couple of steps: After getting as many rings as I could from the blank, I made the final facing cuts to bring the piston to the desired length. I then mounted it in the mill vise "crosswise," i.e., with the axis of the piston set along the Y-axis, resting on parallels. After finding the center and the end of the pistion, I drilled the wrist-pin hole, using my "sneak-up-on-it" approach described in the build log for the rod.

I could then put the piston in the mill vise as shown in attachments 7-8, using the drill bit through the wrist pin hole to line up the wrist pin parallel to the vise; this let me then mill the oval slot where the small end of the rod will reside at a right angle to the wrist pin. You may be wondering about the threaded hole that shows up in these pictures - that was part of my original reclaimed cast-iron blank. I had carefully planned how to cut the blank out of the chunk of CI so that I would be roughly centered on that hole - otherwise, I would not have been able to get a big enough blank to work with!

Not shown, but easy to do, was the final step, remounting the piston "crosswise" as described above, and once again using a drill bit through the wrist pin holes to align the wrist pin parallel to the table. Then I found the center and the end, and drilled out the 0.094" hole that allows oil to drip through to the matching hole in the small end of the rod. In Webster's original plans, this should have a brass tube pressed in to carry the oil right up to the small end ... I haven't done that, at least not yet - maybe a later improvement.

One bit about which I am quite uncertain - my decision on how to keep the wrist-pin from floating out and scoring the cylinder. As you can see in the plans, I chamfered the outside of the wrist-pin holes. For the final installation of the wrist pin, I cleaned it all well and used a bit of red Loctite as the first line of defense. Then after that was set, I came back and lightly staked the ends of the wrist pin so that they protrude a bit into the chamfer. Had I done a better job, that would have been the end of it ... but I didn't get the wrist pin positioned quite right when Loctiting in place, so I had to do some strategic filing to bring the end of the wrist pin back down to where it would not scrape the cylinder. The engine runs nicely, so it did work, but clearly I need to refine my technique ...

[Admiral_dk]

Nice progress .... well we have seen the video, so we know it runs - but nice documentation - thank you

[awake]

Thanks!

[awake]

Part 7 of the Steel Webster build log - the cylinder (attachments 1 & 2).

I lucked out in having a scrap piece of DOM tubing with an .875" ID; the exterior was a little rusty, but the interior was pristine. Using this meant having to make the complete cylinder in two parts, with the fins cut from a separate piece of aluminum. That mostly went well ... mostly.

First step was to face it (attachment 3), cut it just over the desired length, and face the other side to the target length of 2.5".

At first I was thinking I would turn this short length "between centers," and I experimented a bit with using the live center in the tailstock as I began to rough it out (attachment 4).

But I just wasn't confident that I could guarantee that the tube would ride concentrically that way, so I changed gears (figuratively, not literally) and made an arbor. I began by facing and center drilling each end of a 4.5" long piece of cold rolled steel. Then I put my "machine-in-place" dead center in the three-jaw, set the compound to cut the 60° included angle, and skimmed it so that I had a perfect dead center to work with (attachments 5 & 6).

I mounted the arbor blank between centers (attachment 7) and machined it with .75" diameter ends, a little under 1" long, and the center section just over 2.5" long at just under .875" for a nice sliding fit in the nascent cylinder. I installed the arbor in the cylinder with Loctite; after it set up overnight, I could return to machining the OD of the cylinder (attachment . The arbor worked great, letting me flip end for end as needed to machine the details.

Part 7 continues below ...