How To Machine Crankshafts

There are probably more ways of holding an overhung crankshaft to machine the crankpin than there are of inducing involuntary discorporation in cats. Trying to cover them all at once would be a daunting task, so until I feel blessed with excess time and energy, this technique page will constrain itself to the most common sort of shaft a home-shop builder is likely to make: the overhung, single throw shaft, machined from a solid bar. Discussing this topic with The Motor Boys, we concluded that each of us had a favourite technique and all were different in some way, frequently being governed by the different types of equipment we each have, and the content of our respective scrap boxes. Each has merits, and all work. Naturally, I prefer the way I do it, but that has changed 4 times over the years, and may change again. So use this page as a grab-bag of ideas and see what will work for you.  

Turning the Main Journal

I've read a number of approaches to producing an overhung 'shaft. That provided by Westbury in his book, Building the Atom Minor Mk III must rank up there with the most time consuming, but "correct" way, involving meticulous marking out, center drilling on opposing blank faces, followed by (in order), turning the crankpin, then the main journal. I actually followed this on my Atom Minor and produced a useless shaft owing to a misreading of the drawing that gave it more throw than there was head room to accommodate. Never again. What this really convinced me of was that I hate marking out! I've long since evolved a process based partially on the sequence described by David Owen in his well thought-out instructions for machining the DIY Mate 2cc Diesel that obviates the need for any marking out whatsoever!

First a word on the front section of the shaft—the bit that will carry the prop. Most amateur built engines will be produced from steel and left in the soft condition. Heat-treating causes distortion, requiring that the shaft be finished oversize, then ground to final size and alignment. For 'soft' shafts, the best choice of material is "stressproof" steel, if you can get it. This machines ok with sharp tooling. Good old 12L14 makes machining the shaft a joy, but it's a bit soft, especially for larger engines.

Now since the shaft will be left soft, it is vulnerable to crash damage—generally in the form of a bend at the point where the shaft emerges from the front bearing and reduces in diameter, where the shaft is at it's most vulnerable. I've found that if the diameter here is 3/16" or above, it has acceptable strength. If it's smaller, even stressproof will be a bit on the weak side, so for smaller shafts, I strongly advocate drilling the shaft for a screw-in stud made from a cut-off cap ead screw, or even using a high-tensile bolt if you wish. Many commercial engines used this feature, and for exactly the same reason: it's easier to replace a bent stud than straighten a bent shaft!

Now, onto the markout-less shaft. Cut a blank that is about 1/32" to 1/16" longer than the overall shaft length. Pop in the 3 jaw self-centering chuck (3JSC) and face the ends with a minimal cut. Center drill the second face, then extend the shaft blank so sufficient protrudes from the chuck to form the main journal. With the chuck jaws only lightly tightened, bring up a dead center in the tailstock to align the blank, then tighten the chuck firmly. All chucks have run-out, especially 3 jaw self-centering ones. But even if the 3JSC is not 100% aligned with the axis, the final journal will still be formed correctly provided we don't disturb things as the finishing cuts are made. It's the unmachined web portion that will be out and we can fix that later.

First cuts will be deep ones to remove lots of metal, so angle the tool towards the tailstock so that if it deflects backwards, the depth of cut becomes shallower, not deeper. This will guard against a "positive feedback" situation where a dig-in becomes quickly self-perpetuating. You won't be able to get onto the corner of the web from this setup, but that can be fixed later. Reduce to within 50 thou of the journal size, then reset the tool to cleanup near the web and get ready for the finishing cuts.

I like to arrange my finishing cuts to be three or more passes at no more than .005 deep (0.010" of diameter). The final one I like to be only 2 or 3 thou deep to leave the shaft at 0.0005" to no more than 0.001" above the reamed bush size. Since the shaft actually needs to be less than this size, we have enough metal left for finishing. For this, I use a shop made hone that is a copy of an old Sunnen design. Alas, the source of the casting used for this tool is no longer in business. Alternates are to use a split donut with diamond lapping paste smeared on the inside, or even a piece of 600 grit paper backed up a flat steel ruler and lots of oil. Even though your shaft may look smooth after the finishing cut, it's really a very fine screw and the microscopic helical ridge needs to be removed. Hone and polish until it is a good running fit in the bush—experience will guide you.

If the shaft is to have a threaded, reduced diameter section, turn and finish (but don't thread) this before finishing the main journal. Threading is done last as we need to remove the center for this operation. Regarding centers, I believe "live" (revolving) centers are too inaccurate for finish turning shafts. You can use one while hogging the blank down, but change to a "1/2 dead-center" for finishing. The cut-out in this type of center lets you get the tool tip onto the work as it get smaller.

The diameter of the crank web should now be turned to be concentric with the main journal. Just how concentric depends on how you will form the crankpin. If you've left enough protruding from the jaws, good—do it at the same setup. If not, gripping the shaft on the journal in a collet will be fine and produce very accurate results, provided you have an appropriate collet. In desperation, the shaft can be lightly gripped in the 3JSC with a shim of drink-can aluminium used to protect the finish. This can produce a few thou of eccentricity depending on the state of your chuck. Purists might grip in the 4 jaw independent chuck (4JIC) and clock a short, protruding section of journal true. However if you use the offset jigs described below, a thou or two of eccentricity between the web and journal will not matter at all.