Another Rupnow 1 by 1

Home Model Engine Machinist Forum

Help Support Home Model Engine Machinist Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.


Well-Known Member
HMEM Supporting Member
Feb 10, 2018
Reaction score
Central Florida
Starting around the time I was first getting my Webster to run, Brian Rupnow started a build thread on a vertical 1" bore by 1" stroke engine he was designing. I watched it within interest because I was trying to decide what to build for my second engine.

I spent a while looking at hit and miss engines, and talked about scaling models to get closer to what I was thinking of building, but in the end I decided to build Brian's engine from his plans.

Brian finished in a few months. That's highly unlikely for me. First off he's a much more experienced machinist, and second off, I'm working from 2D prints in pdf files. While I have both a manual and a CNC lathe, my mills are only CNC, so my work flow or processes are different from most folks. For parts that I want to CNC that will mean turning the 2D drawings into 3D models in CAD, and then using CAM or writing manual files to get the tool paths. There are other things the prints call for that I should do; heat treating O1 (oil-cooled) steel; silver soldering some steel parts together, and so on. These are things I think I know how to do in concept, but I'll need to test the details.

My CAD program (Rhino 3D, but ver. 5 while 7 is current) is supposed to import .pdf files, and it sorta does. It doesn't import the dimensions and some things it imports (like circles) don't behave at all Rhino's native version. It just puts lines on the screen, and the commands to create dimensions don't work for many things, so it's going to be a completely manual process. Before you machine a part manually, you need to study the drawings. That doesn't go away with reading pdf files in.

For my first experiment, I thought I'd import the crankshaft counterweights. These have a rather complex outline, with only three straight lines in the whole part. To do these on the rotary table would take multiple setups. I believe that the whole part would take seven setups. If I could do it in CNC, I thought I could reduce that to three setups. First, I tried to import the outline, scale it in size and then convert it to a solid. It went better than I thought. As a test, I printed it in PLA, and it came out looking like it should.


It seemed like the best way to make this would be to not do that rectangular cutout in the bottom until the outline was cut. Then I could drill a hole where the center of the large radius (top in that view) is, and use that to bolt the workpiece to a tooling plate. Starting with a roughly 2" x 6" piece of solid steel I'll drill two holes far enough apart to cut two of these counterweights; with the tool paths based on a 3/8" end mill that was 2.000" apart. The cutout would be done using what CAM programs call "waterline" cuts: constant depth passes around the shape. Once the outline is cut, I'll stand it up in a vise, and cut out that 0.500" wide rectangle. After that, there are two holes, a through hole from top to bottom (you can see it in the center of the plastic model) and its counterbore.

This shows the two pieces right after cutting them out. I like to say it came out almost as if I knew what I was doing. It was kind of cool watching the cutter cut the one on the right out of the steel without touching the one on the left.


Look between the two of them where the piece of stock that hasn't been cut away has a point that's pointing out of the screen. See the ledge between the left and right pieces? It's an arc from that point to the right side of the left counterweight. What you're seeing is the left one isn't cut to full depth while the right one cut through the steel into my scrap aluminum backing piece in places. The backing piece is visible on the front left.

All that irregular depth of cut meant was that when I took them off that fixture they're on, I had to do some file work to make the problem child match the other one.

Now we get to the real meat of this task. How do I hold those rounded parts so that I can cut out the rectangular relief that's under the mounting screw? I want something with a concave radius that matches the convex parts, to hold it upside down in the vise. I thought about using the rest of the steel that I cut them out of, but that's the wrong radius. That radius is 3/8" more than the part because of the cutter going around them cutting out the parts. At some point, it occurred to me to use my 3D printer to print a fixture. An hour and 30 cents worth of filament later I had a fixture.


The two parts squeezed together are .620 thick, and I made the plastic thinner - like about .575 - so that the jaws wouldn't just squeeze the plastic but would lock the parts in place. I made sure the bottoms of the blanks were parallel to the mill table with an angle gauge.

Once I cut the rectangular cutout, the printed fixture was a perfect way to hold the two parts to drill the through hole. Being backed by plastic and not my vise made it risk free to the vise so a no-brainer to do. They're not technically done in this picture, there was one more step, but they're pretty close to being done.


This is after the second setup of the three needed, with the two counterweights alongside the fixture and the 3D printed test counterweight I did to make sure the dimensions were right after I translated from .pdf files to a CAD drawing. To do the final setup, all I needed to do was stand the two parts on their legs and center the drill bit that made the holes in the parts. I did the 1/4" counterbore with an end mill held in the drill chuck. The only motion is the up/down of the drill.

I'm not sure what the next part will be, but I either plunge more deeply into porting the drawings into CAD files or do something where I can work off the pdf files. I had been thinking of doing the crankshaft because I've never done one like this. I noticed that in Brian's thread he had the mating piston connecting rod done first before getting into the crankshaft, so that points me toward doing the conn rod before the crankshaft.
Go Go Gadget!!!--Good luck Bob.--If you have questions, don't be afraid to ask. Sometimes I put way too much information on one drawing, so you have to look really close to see what I am trying to tell you.---Brian
Just an update.

It was a computer-intensive week and not one single piece of metal was harmed or even inconvenienced.

I upgraded my 3D printer setup by adding Pronterface. Pronterface gives you computer control over the printer for things that the built in interface doesn't allow, including an immediate mode line where I can tell it things to go over 100mm in X or Y (G01 X100), like I'm used to from Mach 3 on the CNC tools. Well, I can do that in the built in interface, but it's much clumsier and involves turning the selector knob about 125 clicks and presses.

There's a back story there of why it took days to do that rather than a few hours, most of which I can sum up by saying, "I hate Windows 10." There's one exception to that, and maybe it's only important for the Creality Ender 3 V2, but let me tell anyone else in case they need it. The micro SD card that comes with the Ender includes drivers for Windows to convert a USB port to emulate a regular serial port. They wouldn't install. I'd point Windows Device Manager to the files and it didn't recognize them. It turns out the driver I wanted is called CH340 and one of the Ender documents has a screen shot showing a piece of software to install that driver that they didn't include on the memory card. If you get into this trap, do a web search for CH340IR.EXE, run that and it'll install the drivers for you.

The rest of the week was spent turning the pdf files of the engine drawings into solid models so I can get tool paths. Chances are I wasted about half that time by correcting the pdf import into Rhino 3D CAD. It doesn't import dimensions or distinguish hidden lines from regular lines, so the drawings are a mess. The point was to just convert them into solid models and not fuss over the drawing. As for where a hole goes and hidden lines, just keep both the pdf open in Acrobat reader and switch back and forth between them as needed. This is one of the side plates that define the crankcase.


I converted a few other parts to solid models, but they were much simpler than this one. There's more aluminum left in the chip tray than the part. I think.

Before the cutters touch this part, I'll convert the other side plate to a solid model.
Bob, something to consider is getting a Raspberry Pi and setting it up with Octoprint. It is a very easy process, not very expensive (just the cost of the Pi), and gives you wireless control of the printer with all of the manual features that you get from Pronterface, plus the ability to upload files to print, the ability to use a camera (cheap USB webcam works) for live monitoring of the print over the wifi, etc. You can also set it up to work over the internet, so that you can control your printer away from home - some security issues to attend to, but again very easy.
I have a Raspberry Pi 2, and the reason I didn't start down the road to using that is that the NUC was literally on the next workbench over, while I'd have to dig up some things for the Pi. I might well replace the NUC with the Pi if I can get it to run the only app that was on the NUC until now.

I looked at Octoprint but really don't care about the wireless control aspects which seemed to be the emphasis.
I have set up Octoprint on my printer and will never look back.

I can do a complete job of controlling my printer through a web browser -- which is immune to Windows 10/11 issues (when they issue the required update, which eventually they will).

It makes it trivial to control my printer. I can flash firmware very easily. I can do all the control I want.

I get it that you have Pronterface set up. I used to have Repetier host set up. The best investment I've made in my 3D printer was Octoprint.

I haven't updated this since the 20th, and while I don't have cut metal to show, and won't for another couple of days, I've been plodding along on the computer side of things.

Here's where things are now.


These are (big secret) side plates 1 and 2. The blanks aren't much bigger than the final part, and instead of laying out where to drill with Dykem and scratching, I printed the two drawings at 1:1 and taped them to the blanks. Then I marked the point referred to as (0,0) on the drawings using a punch, and drilled them for a "tight clearance" hole of the 5/16-18 bolts holding them to a plate. The holes in the plate were drilled and tapped the same distance apart as the drawings show for the two big holes in the side plate (1.563").

With the (0,0) located, I know the location of all the features, but I'm intending to cut the outlines in the next step, and not drill all the matching holes. I'll then bolt them together, like Brian did using a 3/16" drill rod in two places, and drill the five screw holes that hold the two halves together. I'm not sure how I'll hold the two pieces to do that. It would be good to drill, tap and counter bore all five holes first, before I cut the outlines, but don't have a cutter long enough to cut the outlines of a 2" thick plate or a good way to hold them for that.

The next step is to create the tool paths, and since they're mirror images of each other, the tool paths have to be different. The program I'm using is an old version of DeskProto, version 5. They appear to be offering it as a free version now, and you pay for more capabilities in the SW. I need to look at the free version and see if it's better than what I have. I approach a part like this by waterline machining and I have to arm wrestle DeskProto to get it to what I want. The problem is that even then don't get exactly what I want. This is a screen capture showing the tool paths before I save them to a file.


The red and green lines are tool paths; greens are rapid (G00), reds are regular (G01) movements. The problem is that the program moves the cutter across the top of the part, and that will cut off my mounting bolts if I do that.

I end up editing the tool paths in a text editor to remove those loops around the raised center area. Then I check it in GWizard editor from to ensure it's cutting right. There still appears to be wasted motions that I'd like it to not make, but since I'm not trying to save every second of machine time, I won't try to edit them out.


(I'm not sure why the green tool path lines in this turned blurry when I saved it). This is a perspective of the same side plate (2) showing that the cutter just goes around the raised area and doesn't trim it.

The next step is to correct a mistake in my original mounting of those blanks and put them on a sacrificial piece of 1/16 thick aluminum. I'll recheck the center of the (0,0) hole with my edge finder to ensure it hasn't shifted around the mounting holes. Once that's done, I'll air cut the first few loops around the part to prove to myself it's not doing something I didn't expect. If that's good, I'll cut the first one.
Looks like your making good progress. A couple of questions regarding the issue of the tool paths cutting your hold down hardware: does the software allow you to set a "Clearance Height" or a "Retract Height"? Alternately, does the software allow you to select a start Z height and a finish Z height? If so you could make your model taller to include the hold down bolts, then start the actual cutting lower so as not to cut air before you get to the part.

It looks like you plan to cut the part in one go. I like to make a roughing pass and then a finishing pass. I typically leave about .020" of material all around the part with the roughing pass and then climb mill to dimension. For the roughing pass I use an older end mill, and remove material faster than I would for a finish pass. then I use a newer end mill moving across the work piece slower for the finishing pass. This takes longer, but I get a much better finish and I am able to hold tighter tolerances.

Do you have a band saw or some way to remove most of the material before you take it to the CNC? I have an old carbide tipped table saw blade that hit a couple of nails in a board year ago. That broke several of the carbide tips so I don't use it for wood, but it does a great job hogging out aluminum on the table saw. Much faster than my band saw for large blocks.

I really hate to have to go in and edit the G-code because I end up modifying and tweaking the tool paths a lot, and if I have to edit the G-code each time I make a little modification, it can be quite tedious. As you stated, it is about tricking the software to give you what you want. I usually make several version of the 3D model, either removing features or adding material in keep out areas, to force the software to give me tool paths where I want them.

Keep up the good work, I am looking forward to you making chips. :)
A couple of questions regarding the issue of the tool paths cutting your hold down hardware: does the software allow you to set a "Clearance Height" or a "Retract Height"? Alternately, does the software allow you to select a start Z height and a finish Z height? If so you could make your model taller to include the hold down bolts, then start the actual cutting lower so as not to cut air before you get to the part.

It has a clearance height, which I thought behaved as the retract height, but I hadn't thought of distorting the model by making it taller. Setting the Z-height where the cutting finishes? I think there's a way to do that. I think that the program by default cuts the top with the paths at Z=0.000 because it has no way of knowing that you don't want to do that.

I just downloaded the demo version of DeskProto 7. I'll play with that at some point.

I have been planning to cut it in one operation, but I've done rough and finish passes before and I'm thinking about it for this. The milling is going to use a 1/2" diameter cutter and is largely taking thin cuts off the vertical sides of the piece. There's one place in the blank that turns into cutting a slot, that area on the right front, in front of the surface that angles up to the mid-line. I use a FogBuster for cooling, not flood coolant, and I'm a little hesitant about cutting a slot that's an inch deep. The fog buster has a hard time getting into deep slots.

I could cut that corner off in a couple of ways; probably the easiest would be cutting the corner off in a separate operation with another end mill. It's just going back and forth between two points, dropping the cutter for each path. I could even do it with a saw. That would leave the main operation as all cutting along vertical edges.

I'm looking forward to making chips, too.
Just a small addendum.

I've cut many things just by using the command line in Mach 3. A few simple Go To statements (G01) back and forth, dropping the cutter every change. So I set up to do the same with cutting off the stock between the angled edge and the rest of the blank.

Because of the angle and needing more than just X and Y, I drew a few cylinders in the CAD drawing, found their centers and made a tool path. This is side plate 2 and the big rectangle is a estimate at the size of the stock I'm cutting it from. The reason 2 and 3 are farther off the edge than 1 and 2 is that I tested the points I had and the cutter didn't get off the blank. I drew the rectangle wrong. So I moved 2 and 3 further along the angle.


The idea is go in numerical order, and since the endpoints are all off the work piece, when I got from 4 to 1, I'll lower the cutter, too.

With this approach I never cut a deep slot.
It wouldn't be a shop project without little things to overcome, right?

I did that little cut shown in the previous post. It didn't take long to notice that the cuts didn't seem to be as deep as the Mach 3 display (DRO) was saying. When the file was done, instead of being through the 1" thick piece, it was cut about halfway. The cutter had driven the rest of the way up into the collet. I thought I had tightened that as tight as always. Oh, well...

I repositioned the cutter (a 1/2" carbide, 4-flute end mill), tightened very tightly and started the cut again. This time, it cut farther down, but still left a sizable ledge on the work - maybe .050 thick. I trimmed that back but every time I cut another pass, it seemed like the cutter pushed farther up into the collet.

Since the cut had pretty much achieved its purpose of trimming away most of that corner, I said it was close enough to done and then decided to do the actual cut. I re-zeroed the cutter in the collet yet again and thought I'd try the real outline cut. The real outline file also left a thick ledge again.

At this point, I took the carbide cutter out and replaced it with one of my older HSS cutters that I had used several times before. I suspected that the cutter was too small in diameter and that's why it couldn't grab it securely enough. It was a cutter I got from some dood's table at Cabin Fever in '15, so maybe it was surplus because it was imperfect? My mistake in retrospect was that I never stopped to analyze the problem and didn't really ask, "why is this a problem now and never has been before?"

I did the same exact file with the HSS cutter, spent about 10 minutes cutting air, and it also left part of plate uncut. By this time, it was late enough to shut down in the shop and I said I'd look closer at the cutter and collet today.

The cutter diameter was fine. The carbide cutter measured 0.4995 to my manual micrometer while the HSS cutter measured 0.4985. The diameter clearly wasn't the issue, so I cleaned out the collet with a paper "shop towel" (the blue paper kind) and some mineral spirits. I don't know how it got oil in there, and I didn't see anything on the paper towel, but I put the carbide cutter back in the collet and did the last couple of passes around the side plate (I edited the file to remove everything above cutting to -0.950). This time, it left a thinner margin on the very left side of the blank. Re-checking the zero showed it hadn't moved, so that meant the work probably did. I lowered the bottom cut from the CAM set (-1.001) down to -1.004 and ran it again, this time resulting in a proper looking side plate. I think the reason that previous try left some material on the left side is I inadvertently didn't position a backing plate under the left edge.


I think that I'm going to mirror the files to re-create that path to cutoff the corner and try to get the other side plate done tomorrow.

These still have a lot of machining to go, but the first chips are always good.
Last edited:
The other side plate is cut to size, and it had a peculiar aspect to it. The CAM program neglected to cut one little area to the proper depth.

In the red box here. You can see the ledge better at the bottom.

That uncut area on the top is supposed to be 0.124 above the floor. That's 4 cuts at .031; the rest of that surface is the right depth; where you see the ledge and across the right side, it's .093 - three passes. I ran the file again, just to watch through the first four layers and the cutter just never went there. I was concerned it might have been from my editing the file, but that was just to keep the cutter out of the uncut area. It went into this area to cut it down to -.093, it just didn't go back to cut the last pass. I've been using Deskproto for about 10 years and I can't think of another time it did something like this.

I have a way to use a Logitech Rumblepad - a game controller - as a pendant-like interface for Mach3 and cut that area down to 0.124 "by hand" driving the mill with the buttons on the Rumblepad. It needs a little file work, but should be OK.

Next step is take both sides off this plate and move into my plans to bolt them together and drill the holes for the guide pins Brian designed in.
Picking up where I left off last Monday, now it was time to take both pieces off the mill, de-burr them, then bolt them together, smooth them and remount to get ready to drill the holes for the two tooling pins that go into both sides to hold them in the right positions. That makes it easier to take them apart and put them back together precisely, which I'm going to have to do. Other than breaking the one fine sanding belt I own before I was fully done with sanding, there were no issues.

This is where I made a silly screw-up and spent a day trying to find it. The issue centers on the dimensions in red in this drawing.


The problem showed up when I went to position the center drill over the hole for the tooling pin circled in red. It looked too close to the top edge, and instead of measuring what's in the red oval, I got more like .150 to .160 instead of .220. That led to studying it for hours, looking at other measurements and just being confused. The overall length and width of the plates were right. The dimensions going left to right looked fine. The dimensions from the edges of the raised area to the part's edges were also offset. IIRC, the top one, 0.696 was too small and the 0.625 on the bottom was too big.

What this would mean was that the CAM program had shifted the centerline X-axis upward in this view. I couldn't figure out how it could do that and not mess up the whole part.

Late one night last week, after puzzling over this for hours, I took yet another look at the work and realized my mistake. I had put the two pieces on mill upside down and backwards. Notice the ~45 degree angle in the drawing is at lower left. Here you can see it's on upper left. The top of this stack is side 2. You can tell it from side 1 by the machining marks


Since the next operation was just positioning the mill and drilling a total of 11 holes of different sizes, it didn't take as much time to finish this phase of the machining as it did to figure out what I did wrong setting up to do it.


There are five large holes with a counterbore; those had to be done in three steps; the two tooling pin holes had to be prepared in three steps also and tooling pins cut to length. So it wasn't simply drilling 11 holes, but it wasn't so much hard as just a lot of motions.

These are complex pieces. The vertical side facing you, the right and the far vertical sides all get drilled and tapped holes. Two per side on both side plates. The far side gets a 1.1" diameter hole bored into it, centered on the plane where the two side plates touch. And lots more. I'll be working on these for a while already (it only seems longer than Brian spent on the whole engine), but I think the way most of the next operations should get done is with the machining vise on the mill's table. For some of the large features that get bored out, perhaps the four-jaw chuck on the lathe.
Since I've been doing Monday updates and it's Monday, here's an update.

As the week began, the first step was to take the two sides apart. I tapped the five #10-24 holes in plate 2 that will get screws holding the sides together, and then it was time to take my aluminum tooling plate off the mill and put the milling vise back on to start working on some of the internal complexity of these parts. It was time to hog out the three different cylinders inside side plate 1.

The largest of these cylinders is 2.376" diameter and 0.557" deep. Like most of you, I have a rotary table, a boring head for the mill, and a boring bit for my lathe. How best to do that? Since I think a strong feature of my shop is the CNC control, that's where the approach went.

The basic thing this is doing is cutting a circle, and the ability to cut smooth nice circles is built into G-code. Plus, while the commercial CAM program I have can cut circles, it ends up being a fairly coarse approximation - I can see and feel the steps the CAM generates. I've hand programmed circle cutting while making the parts for the big mill CNC conversion and while making the connecting rod for my Webster. This post has some details on it while I found an incompatibility between the software I use to verify G-code and Mach 3. I decided to hand write G-code using the circular interpolation (G03 - cutting counterclockwise) instead of using the CAM program.

The first thing I did was use the biggest end mill I have, a 3/4" diameter cutter, in the centers of the cylinders and not cutting to final depth. Just removing bulk metal and creating a starting point. That's two holes.

The basic approach is to use that 3/4" hole in the middle as an entry into the cut. That hole is where I lower the cutter (so I'm not drilling with an end mill) and then move back out to the rim of the circle to cut around the perimeter. I have to keep straight in my head that the tool path marks the center of the cutter, not the diameter of the hole it's cutting. When using the CAM program, it will work from the diameter it needs to cut keep track of the tool radius offset.

After testing a couple of ideas, I figured the way to cut the big cylinder is by cutting two passes around the circle, one with cutter center at the edge of the 3/4" hole, which doubles its size to 1.5", followed by a second cut with the center of the cutter 0.375 in from the final size. Since the radius is 1.188, the cutter goes at 0.813 from the center. This is after the first operation on cavity 1. I have a video here showing the last few passes.


The changes in the way the bottom of the cylinder reflect the light can't be felt, and you can see marks from both passes around the circles.

With that nice result, I wrote toolpaths to enlarge and deepen the center of that cavity and then enlarge and deepen the small cavity on the right, bringing both to final size. Except I misread the drawing and made cavity 3 too small. It's always easier to correct making something too small as opposed to making it too big, and with the file that enlarges and deepens that cylinder taking about 15 seconds, it took me more time to fix the G-code than fix the part by running it (video of the process - 28 seconds long at actual speed).


I think I can just revise the X and Y coordinates for the other side and be done fairly quickly.
I'm late on my usual Monday updates because I wanted to be able to say I'm finished with the two side plates.

Hollowing out the second plate using the same basic approach as shown in the last post worked fine and really didn't take much longer than running the files a week ago. Once I found a typo that made them just a bit wrong.


Then it was a matter of drilling and tapping holes everywhere. On side 2.


and the rest. The top has a 1.100" diameter hole that I cut using the exact same approach as I did for the internal hollowed out areas, as well as four tapped holes.


The right side has a tricky feature, a hole drilled at 40 degrees to the bottom, 1/2" reamed. This really caused me about a day's delay of "think about it 30 times, measure it 20 times, machine it once"


You can see what I think is my big mistake on this one - I drilled and tapped two holes on the back side plate (2) that are extra. I think I can fill them with a set screw and use some red LocTite to keep them sealed.

And since you've seen every other view of it, here's the bottom side, which has a 3/8-16 tapped hole in plate 1.


They clearly need some cosmetic work to pretty them up a bit, and some deburring, so they're not Done done, but I think all the features are there. I honestly think this might be the most complex little assembly I've built. There's nothing equivalent in the Webster or my Duclos Flame Eater.

I think I'm moving on to the crank shaft next, but it's also a good time to make sure I have all the raw materials. I need to order a few pieces of metal I didn't order last time I shopped.
Picking up where I left off last Monday, now it was time to take both pieces off the mill, de-burr them, then bolt them together, smooth them and remount to get ready to drill the holes for the two tooling pins that go into both sides to hold them in the right positions. That makes it easier to take them apart and put them back together precisely, which I'm going to have to do. Other than breaking the one fine sanding belt I own before I was fully done with sanding, there were no issues.

This is where I made a silly screw-up and spent a day trying to find it. The issue centers on the dimensions in red in this drawing.
How do you check your part program? There are some pretty sophisticated programs for checking that will backplot solid models but a quick, easy check can be done with CNCdiscriminator Discriminator CNC Editor The download is free and as long as you're not expecting it to interpret macros and such it's free. It only plots cutter paths, not solid models but it's pretty good for even some 3D profiles, depending on what you expect from it, I use it all the time and if you want to edit the program it's better than the solid modeller in the CAM package.
How do you check your part program? There are some pretty sophisticated programs for checking that will backplot solid models but a quick, easy check can be done with CNCdiscriminator Discriminator CNC Editor The download is free and as long as you're not expecting it to interpret macros and such it's free. It only plots cutter paths, not solid models but it's pretty good for even some 3D profiles, depending on what you expect from it, I use it all the time and if you want to edit the program it's better than the solid modeller in the CAM package.

I use the GWizard Editor. It gives the cutter paths on a grid and not on the model. Since the grid is 1/4" squares, I can get the dimensions of what it cuts out to a good accuracy with my hand calculator, but it still wouldn't find the problem. No software would have.

I mounted the part upside down and backwards. That sort of "short in the headset" just isn't going to be found by software.
The last ten days or so have been focusing on making my crankshaft. Since I've never made a real crankshaft (supported at both ends), like this one, I was interested in getting experience at it.

The final size of the rectangular stock the shaft is made from is 1.000 x 0.500, and while Brian recommends making it 1/32 on each side oversized to start with, I left it .025 on each side oversized. Mainly because I do everything CNC from the command line and counting by .025 at a time is more familiar than counting by 0.0313


By the way, I actually threw away more steel in those chips, 7.77 cubic inches, than I ended up with. The volume of the 1.25" diameter, foot long bar to start with was 14.73 cubic inches. The volume that's left was 6.96 cu. in.

After this, I did the usual trim and de-burr of all the edges, and then put the two center drill spots in both ends that will hold the piece so it gets turned to the proper sizes and shapes. This is how that was done - verifying zero on the left and after the work on the right. It was just about all I could do with my mill to put the center marks on the ends of an 8" long bar. The head was just about at the top of its column.


At this point, I had a real bump in the road. I have two lathe dog sets for my Sherline lathes, but nothing for the big lathe. It was the weekend, a holiday weekend here in the states, so now what?

First, I tried to turn between centers without a dog to see if it could work but I could barely scrape anything off the piece. After looking for options, I thought I could make one out of a 2" diameter piece of steel or aluminum, about an inch long. Bore a 1.5" diameter hole in the middle. Cut away everything that doesn't look like a lathe dog. My immediate problem was that while I had a cutoff piece of 2" aluminum, it wasn't long enough and I had no other junk stock big enough to work on. I eventually found a piece of pipe in my junk stock that was smaller than 2" diameter but big enough to work, so today I made it into the lathe dog. It's some sort of soft, gummy aluminum alloy, maybe the stuff they make shower curtains out of.


It isn't strong enough, probably because the walls are too thin. When I tighten the setscrew more, the ring stretches. I made some test cuts with it and it seemed to worked well enough, though. If I had some steel or something that would resist stresses more than this one, I'd make a replacement.

I'll go ahead and try it, though. Within a few days, I'll know if it holds up to the stresses of making this crankshaft.

Latest posts