Elmer's #33 - a novice makes chips

[Thayer]

This build thread began on another machining site and with our local admin's blessing I want to share my efforts here as well. The first several posts will look very familiar if you have been following it already. If not, well, here it is. I have since learned about embedding images, so this thread will be a little tighter than the first round.

Thayer

As a bit of reintroduction, I've had a manual Sherline 4400 lathe for a while and a few years ago complimented it with the A2Z CNC Monster Mill to maintain Sherline compatibility. Knowing little of operating a mill or CNC at the time, after a few basic bits and pieces I christened it with a couple of Elmer's simple engines, a #25 wobbler followed by the #2 twin vertical wobbler, seen here, "My first runners."

While the main use of the tools has been making bits and pieces for my radio control airplanes, I have long meant to revisit Elmer's catalog and tackle something a bit more complex. A couple weeks ago I did just that. I've now begun his #33 horizontal mill engine and began with the cylinder block to see right away if I was up to the challenge. I'm not fast with my projects, but try to build cleanly and accurately. I will update this thread as I go but am rather sure that life will keep it from running this year.

My main goal, beyond finishing the mill engine of course, will be to pass along a bit of inspiration to others like me who are new to machining. The work you see in online forums can be at once both inspiring and rather intimidating. I've found twice in the past, and expect to again, that you don't need to be a master machinist to enjoy turning out a working steam engine.



This first photo shows the basic start. My scrap stock was a little oversize, so I made a several passes with the Sherline flycutter to bring it down to its nominal cross section of 1x1 inches. Note the slight relief in the top of my vise jaws. Inspired by Tryally's slim vise, I inverted the stock jaws and milled the relief as a "permanent" set of parallels. I will occasionally skim them when I can't indicate them accurately. Note the layout ink and scribing. I decided to lay out this project as accurately as possible, despite the availability to just use the DROs and run some G-code.

Of note to those still sitting on the fence is the piece of wood used to contain some of the fly cutting swarf. I have a stash of used breaking boards from our local karate studio, gathered after one of their semi-annual black belt tests. If an 8 or 9 year old kid can earn an internationally recognized black belt after four years of training, surely we can muster enough confidence to carve a little aluminum, brass and steel.

[Thayer]

The trued-up 1x1x1.375 cylinder block back in my lathe and ready for boring. I am using a cheap import drop indicator to set the 4-jaw for the appropriate 1/32 offset for the cylinder. I did an initial center by eye using the concentric rings, then indicated opposite faces to bring it in tight before offsetting for the bore.

I used this same setup to clean up the ends of the rough sawn blocks and face them to length. Careful adjustment of the rough block in the chuck allowed me to maintain it parallel to the bed within half a thou along its length when I originally faced it, and again now. Surely that will be good enough for this project.

FWIW, the shop-made indicator holder is that shape because I had an offcut about the right size.


Step three was opening up the bore. I started with a center drill in the tail stock, then stepped up .040 - .050 at a time until just shy of .400. I then used the largest Sherline boring bar to bring it out to .500. I actually left it a bit shy to give me some room for honing the cylinder after finishing the rest of the operations on the block.

[Thayer]

The outside of the cylinder gets turned round for .25 on each end and I didn't think I could safely hold it to turn the second end without ripping it out of the jaws or mucking up the first.

I ended up making a slip-fit mandrel from 12L14 steel, first using a steady rest to center drill the ends. I then turned it between centers to size.







To hold the cylinders, I left the center in the headstock and replaced the lathe dog and faceplate with the 4-jaw. With the mandrel back on centers, I lightly griped the block to hold it while I turned the second end.



I know better and should have padded the jaws before turning the second side. I didn't tighten them much, but the hammering from the interrupted cut was enough to give me something to polish out later.

[Thayer]

Here are two cylinder blanks bored and ready for the valve porting.



I did a pretty careful job of laying out the ports and it was a good thing too. Somehow I missed on my edgefinding operation by a bit. Before spot drilling the 13 holes for the porting and steam chest mounting, I dimpled the face about .005 to make sure I was good to go with the drilling sequence. My first pass showed the error so I rechecked the edge and got within a couple for the next round. You can see the second round of smaller dimples are a lot closer in Y and nearly dead nuts on for X. They are intentionally smaller so I would know which was second set.



I figured that was close enough and pressed the go button. Spot drilling with a #1 center drill, followed by peck drilling sequences for the intake and exhaust ports and steam chest.



The final holes all fit within my layout lines so I think it is pretty close. The step in the exhaust port bores comes from the center drill being slightly larger than the final bit and going deeper than I needed to. The intakes have an .005 skim as well to check the G-code.

[Thayer]

I've been admiring everyone's dedicated tapping stands on HMEM and can certainly see the appeal of such a project. I haven't gotten to it yet, but did upgrade my tapping capabilities on this project.

Previously I had placed my lathe center in the mill headstock then jogged it down onto the top of my tap handle. Tapping meant one hand turning the handles with the other on the keyboard chasing with the center.  As you can imagine, that was pretty slow and didn't do a great job holding the handle steady.

To improve my tapping going forward, I took an 18x.312 stainless shaft I salvaged from an old printer, turned a 60-degree center on the end and bored out an unused tool holder to just clear the shaft.



Now I set up much the same way as before, but thread the 5+-ounce stainless shaft through the spindle and let gravity feed it as I advance the tapping handle. This works beautifully well and is a lot faster than the old method. In essence it is poor-man's version of the spring-loaded center often used.



If you get the bore in the holder a little looser than you wanted, you can turn a drop-in collar for the upper end of your Z-axis spindle as well. I haven't done that yet and am taking solace in that this new setup has a lot less slop than I had with the previous method.

[Thayer]

Next up is drilling the long diagonal ports. I used a digital TiltBox to set the cylinder at the proper angle. I doubt it is all that accurate, but I don't have a sine bar yet and the tolerance for this op was pretty loose.



I then used my height gauge to determine how deep I wanted to drill. I measured off the notch in the inclined lip of the cylinder bore and the target port.



Last but not least, the current status of the cylinder. Yes, I have a second block but that one has a way to go to catch up. Hopefully I won't need it for this project.



For anyone working on one of Elmer's designs, carefully check the geometry of the notch in the end of the cylinder. He calls for a 1/32 offset* but I could not get the drill to catch in the bottom with that dimension. At 1/16 I had no trouble. The image shown is the end with the larger offset.



That's all for now. I am quite pleased with my work so far and think it looks pretty good, despite the harsh lighting.

*I realized that I had read the print wrong. The indicated 1/32 dimension is for the centerline of the port, not how far into the cylinder wall it is offset. That all makes a lot more sense to me now and I now believe the photos above show the porting as Elmer intended.

[Thayer]

After mucking about with the TiltBox, I came up with the idea for milling this little piece of aluminum. The hole means nothing. It was just in the scrap stock I used.



As you can see from my CAD drawing, it is an alignment aid for drilling the ports. I figured I would make it up as I can see making a few more of these cylinders for other projects.



Here it is roughly aligned with a cylinder.

[Thayer]

Before I show it off in use, here are two little layout / test tips I find useful as a timid beginner. My mill is CNC and I will usually "cut" some air before committing tool edge to metal.  If you have a piece of stock that is a little marginal in size, or just want to see your robot run a bit to give you confidence in your coding, you might appreciate that a standard Fine Point Sharpie can be securely seated in the #1 MT of a Sherline spindle. This gives you a pen plotter of sorts. Seeing ink along the toolpath on your actual stock, secured to your table, buys a lot of confidence. At least it does for me.

Step one is to manually edit your g-code to make one lap around the path at Z0 minus a few thou. I will usually bump the speed up a bit as well. Jam the Sharpie into place, then start the spindle at a slow speed and apply light pressure with your fingernail or some sort of stick to true up the wobble.  You don't have to get it perfect, just slow down the big swirls. Set your Z0 with the tip just touching down and then press the go button.



The path you see above was with a bit of wobble going on, but it left a path about an eighth wide which happened to correspond to the end mill I was using. This made it an easy task to confirm that the mill would not crash into the vise before I started cutting. Yes, this is thinner stock than the part you saw previously. I made two, the first  from .125 stock which I decided was not thick enough. It worked, but I knew it could be better.

If you really want to get fancy, you can draw the part in one color and the tool offset in another. Don't forget to reset your g-code, Z0 and feed rate before you start cutting.

When it comes time to set your cutting tool to Z0, getting it to "just" touch down on your stock can be a little intimidating for a beginner. You read about auto probes, cigarette paper feeler gauges, 1-2-3 blocks and even candy wrappers being used.

Truth be told, when you are going to cut through a piece of stock the Z isn't really all that critical.  I will usually bring it down close, then jog down .001 at a time using my keyboard. While bringing it down I rock the spindle back and forth using the Sherline #2049 handwheel and just watching for the first light mark or curl at the cutting edge. I'll call that first touch Z0 and proceed.

Jogging at .001 as you feel for the surface can be slow going though, so I usually make a slow Z move to get the tool as close as I dare, hopefully just above the surface. A strong backlight provides a nice visual aid letting you easily see the gap between the tool and the stock. If the surface is reflective, don't forget that the tool and its reflection are 2x the actual height apart.  Here is a photo showing the tool hovering just a few thou before touchdown.

[Thayer]

While it's just a simple little cradle, this lets me set the cylinder block at the proper angle for drilling the steam passages between the cylinder and porting. Just press the cylinder into the seat against the base of the vise and carefully it snug up.

Elmer used that cylinder design for several of his engines so the odds are pretty high that I will be drilling these passages again in the future. I wasn't completely happy with the first one and I've already got a second cylinder blank that may go on his #29 Vertical Mine engine next. This seems like a better way to go and likely more accurate than the TiltBox which calls out a +- 0.2-degree tolerance. Even if the angle isn't perfect, it will be more consistent.

Finding Y0 comes from indicating the edge of the block, making a few measurements of the actual part and doing a little math. Centering this passage is not critical so if I miss center by a little it really doesn't hurt anything beyond my pride.



The part I thought was brilliant for a beginner, if I do humbly say so myself, was coming up with the little nub above the cylinder to set my X0.



Touching off that with my edge finder puts my spindle directly above where I want to come in with the 1/16 end mill to start the passage.



Actually cutting this part took about 8 minutes after I had it designed (Ashlar-Vellum Graphite) and generated g-code (d2nc in Mach3). I was feeding at 16 ipm with a .020 doc. The stock was a scrap of 3/16 aluminum tooling plate. The finish of the cut could be better, obviously from letting the chips sit in the slot while I chased around it. I used TapMagic applied with a small plastic pipette and tried to clear the chips as best I could with an acid brush. I could clean it up with a finishing pass but doubt I will.

The cylinder with the dychem on it is the undrilled blank, the one posing under the mill in the A2Z holder is the first one I drilled. I indicated that one just to see how close I could get the mill and it looks near enough with my Optivisor.

[Thayer]

My wife and close friends accuse me of being a perfectionist at times and I certainly trend that way. While I haven't done a lot of machining, I have seen quite a bit of nice work and am in no hurry to get done as the process is every bit as important as the result.

Take for instance that little alignment tool for drilling the steam passages. The first cylinder is perfectly usable and I have no immediate plans for starting the #29 mine engine. Once I came up with the idea for the tool though, I figured I would explore that path a bit instead of blindly pushing fwd on the #33.

Jumping ahead a bit, here is an example of where working slowly and methodically seems to have its benefits.



And yes, the SHCS are merely place-holders to confirm the fit. Eventually the cylinder will have studs and hex nuts. Or at least threaded rods to simulate them.

Anyway, time to catch up a little.  Here's a quick sequence of spot 'em, drill 'em and mill 'em.





I'm using a #1 center drill above, followed below by the appropriately sized carbide drill and end mill from one of those assortments of 1/8-inch shank resharpened pieces you find in a plastic box for $20 or so at the local tool sale.





I am not locating a hole and going through all operations before moving on to the next. Instead I programmed my g-code to return to X0 Y0 before starting each new line of holes to compensate for the backlash. The spotting sequence then goes like this.

start at X0 Y0
spot NW then NE mounting holes
return to X0 Y0
spot left, center, right holes for upper row of valve ports
return to X0 Y0
spot left, center, right holes for middle row of valve ports
return to X0 Y0
spot left, center, right holes for lower row of valve ports
return to X0 Y0
spot SW then SE mounting holes
finish at X0 Y0

Because of the different sizes, opening up the mounting holes and valve ports was then split into two ops before I finished with milling the valve slots. This shot also shows the exhaust porting out the side.



Does that make sense?

Let me know if you have any questions.

[Thayer]

Here is my steam passage alignment widget, not just posing as before, but actually being used for the first time, and successfully so.



In the second image below I am applying cutting fluid using one of those cheap plastic syringes you usually see in the hobby shop near the CA glue. I can get the fluid right where I want without getting nervous about them getting chewed up.

[Thayer]

Now on to the cylinder heads. I started with the outer one for practice as concentricity here is not critical. At least not until you get to the bolt circle.

First I chucked up a piece of 1-inch stock a few inches long. I used the 4-jaw not only to center it reasonably accurately but also to get a good grip on it due to the overhang.





After facing and a cleanup pass, I brought it down to a few thou over the cylinder OD and then turned a registration spigot that is just over the cylinder bore. This gives me a way to hold the part to clean the outer face.



I then parted it off using the parting tool for most of the cut, followed by a hacksaw to finish.  The diameter reduction at the tip is a slip fit in the cylinder.

[Thayer]

Reversing the head, I gripped the spigot with the 4 jaw and swept it with my test indicator to ensure it was directly under the spindle.  To get this done accurately you actually have to sweep it twice. First you hold the spindle stationary and center the part in chuck while spinning the RT for reference. Once you have that dialed in you then rotate the spindle to sweep the part with the indicator. This time you are ensuring alignment with the spindle.



But you say you already had the part true on the lathe before you moved the chuck? Yeah, it may have been, but nothing ensures that screwing the chuck onto the chuck adapter in the RT will keep it true. I find my combination of tooling results in a "lathe-centered" part being about .0015 TIR off when I move the chuck to my rotary table. It doesn't take long to bring it true, so I pretend .0015 is a big miss. In actuality, a total indicated runout is pretty good for that combo, with a +- deviation of only .0007-8. This would be fine for most things, especially if I drilled generous clearance holes in the head. On the other hand, if you can get it better reasonably easily, why not do so?



Notice the use of a calibrated Blue Raspberry wrapper for locating Z0.



Once I had that, I lightly spotted the holes to visually confirm alignment.



Finally I spotted them deeper, then drilled through.

[Thayer]

Here's an abomination. Truly, a wrecked part waiting to happen.



Even though the head OD was a little oversize I tried to protect it with pads while mounting it. I had already trued up the hidden outer face at this point, so it was just a matter of turning that .300+ spigot down to .030 high and a little smaller in diameter. The lathe tools were supposed to hold everything in alignment while I snugged it up.



When I mounted the chuck and tried to indicate it I was able to get the tip or the base true, but not both at the same time. I don't know what I was thinking with this lash-up, but at least I gave up on it before I ruined the part.



A few minutes later I had a nice split bushing turned from a piece of thick-wall 6061 T-6 tubing.



As you can see, this worked a treat after I slotted it with a hacksaw and cleaned it up.

[Thayer]

Repeating the operation with the cylinder and a little hand tapping gave me the results you see here.



It was late at night at this point, and I almost neglected to index the cylinder for proper orientation. I ended up hand cranking my RT until the test indicator showed zero deflection when tracing the valve port face of the cylinder block. A few minutes after taking these I snapped the photo with the SHCS that you saw above and went to bed.