Episode 9 || Crankshaft
Making the crankshaft for my Radial 5 brought me to all corners of my shop: hogging off a bunch of 4140 Pre-Hard on the lathe, turning between centers, dividing head work, mill work, a bit of CNC, some Dremel tool-post grinder work, and some time at the arbor press. This is gonna be a bit of a long one…
The drawings call for using 4340 steel for the crankshaft and heat treating to Rockwell C35. I’ve never really done much heat treatment before and don’t really want to have to send it out to a heat treater. It was quite easy to find annealed 4340 steel that would suit my application, but I couldn’t find any prehard. So instead of going with the drawings, after some research* I opted to go with a chunk of 2”-diameter 4140 prehard with a nominal hardness around C30 instead, which I got off eBay for a reasonable price. Here are some of the data that influenced my choice (I found data other than the plot below, but this is the best I could find showing data for both 4140 and 4340 in one plot). Let me know if you think I’ll regret this choice. I’m second guessing myself a little as I write this post.
Anyway, I slabbed off a 4-5/8”-long chunk of 2”-diameter 4140 prehard, chucked it up in my lathe, faced off the end, and put a center in there.
I had to start with 2” diameter stock because most of the crank web is integral with the shaft portion, the largest diameter of which was about 0.47”. That meant something like 2 or 3 hours’ worth of roughing to get the bulk of the material off. This stuff turns nicely, but is tough and hard, so I could only take about 0.025” DOC per pass without bogging down my 3/4hp single-phase lathe motor. That meant lots and lots of blue chips. ASIDE: I don’t know why everyone complains about long, continuous chips. I love’em. Sure, they take up more space in the trash can but it makes cleanup easy (of course, I could see why a job shop wouldn’t like them).
I roughed it all down to 1/2” diameter and pulled it out of the 4-jaw chuck. Reversed it and held it in a 1/2” collet, put in a center on the other end so I could turn it between centers, and machined it to final length.
I bolted on a lathe dog, blued it up, and marked out the locations of feature transitions. None of these length dimensions are critical so within a few thou is plenty accurate. Then I got to work machining everything to final diameter. I was getting some pretty poor surface finishes (I don’t think my lathe can spin it fast enough?) with my carbide insert tooling, so I left the diameters about a thou heavy and took it down the rest of the way with emery paper, though I did leave the diameters a few tenths larger than the high end of the specification. I can always take a bit of emery paper to fine tune bearing fits later on --- particularly because it’s got centers in both ends. I ground up an HSS threading tool and knocked out the 5/16”-24 thread without a problem. And that was about that for lathe work.
Over to the dividing head to flatten out the web of the crankshaft. Did that between centers as well; easy peasy. Used the same setup to drill/tap a couple holes in the web for bolting on the counterweight, and another hole in the shaft to receive a shear pin.
Then I had to knock out the offset hole in the web to receive a 5/16” dowel pin that acts as a crankpin for the master rod. This was a bit of a tricky setup. I had to make sure I was perfectly parallel to the crankshaft (in two directions), and at the correct offset. The setup would have been a little bit easier if I had a 5C or ER collet block, but the best I could come up with was clamping an MT2 ER collet chuck by a couple flats on the body. Not ideal, but it worked. Drilled and reamed it to spec.
I also did a little bit of CNC action in the setup described above to make the 10-thou shoulder around the crankpin. Okay, bulk of the crankshaft done!
Alright, next component to be made was the counterweight that bolts up to the web. Simple part made from 1018 CR. I rounded the sides with the basic CNC capabilities available to me on my Bridgeport, but this is all noncritical, so could be done with a saw and a file, or a million other ways. Not pictured is an adjustable parallel on the other side of the vise to balance out the clamping forces.
Once I milled out the slot and drilled and counterbored it, it was ready to be bolted to its buddy. Really nice looking part, if I do say so myself.
Alright, last bit to be made is the 5/16” crankpin. The most logical way I saw to do this was to start with a 5/16” hardened and ground dowel, the one I got is spec’d at Rockwell C60. The problem, of course, is how do you machine in the snap-ring groove into a hardened dowel?
Well, I made up a little bracket to mount my Dremel on my tool post to use as a cheap toolpost grinder. Dremel conveniently makes cut-off discs with the correct kerf width.
The only problem I encountered while ‘toolpost grinding’ is that the Dremel wheel wore away quite quickly. So I couldn’t rely on my cross-feed dial for determining depth of cut. Easy solution: I just checked with an actual snap ring every once in a while until I got a fit I liked. Then I cut off the dowel pin to length using the same process. And BINGO, perfect.
Last step was to press the pin into the reamed hole of my crankshaft. I had nominal interference fit of 0.002”; see
this other thread for a bit of a longer discussion on that topic. It took much more effort on my 3-ton arbor press than it took in my test go. I don’t know why. But I had to really crank on it with everything I had, and bang and bounce it home a bunch of times to get it all the way seated. That pin is never coming out.
Lots of fun to make this crankshaft, and it turned out great! Only thing that remains is to take a little emery cloth to the bearing surfaces during final bearing fitment.
……That is…unless someone convinces me to scrap it and use 4340 instead…….
Next installment on making the cylinder barrels may be a while. I'm mostly finished with two of them but am waiting on some more material to make the rest.
TIME ON CRANKSHAFT: 12 hours
CUMULATIVE TIME: 120 hours
*There’s a lot of scatter in the available data regarding strength vs. hardness for 4140 and 4340, and complete data sets seem to be pretty hard to come by, at least from my searches. But best I can tell, the strengths ought to be similar given similar heat treatments, perhaps with 4340 being a little stronger. 4340 ought to be tougher (in a technical engineering sense) across the board. But because the stress concentrations aren’t too bad on the crankshaft, and the loads imparted on the crankshaft will be reasonably smooth (unless I install it in a plane and it crashes), I think that 4140 will have plenty of toughness to prevent cracking or crack propagation.
To be a bit more thorough, I ran some simple calculations. Based on an estimated 6000rpm and an estimated 6hp (which almost certainly won’t be occurring simultaneously, so this is conservative), I calculate a peak shear stress in the crankshaft---at the root of the threads---of about 25ksi; maybe add as much as 50% more if we include factors like stress concentrations, small inertial/impacts, out of balance, etc. The shear yield stress of 4140 prehard is probably no less than about 60ksi, leaving me a good bit of fat in the design. I frankly have no idea what an appropriate factor of safety is for a crankshaft, but 1.6 seems reasonable enough to me. We shall see.