Mike's P.M. Research No. 6 Steam Engine

Nice work as usual; space shuttle attention to details. I'm going to have to read my DRO manual for the sub-datums; great idea on the tool positions. Pretty sure mine is one of the Chinese-generics with 200 sub-datums.

Bruce

Yup, mine is the same. Basically, if you've used a CNC before, it is the same as entering G54, G55, etc. You can zero in each SDM and it doesn't affect the other SDMs. You can use them in a lot of different ways, but on my lathe, I use one for each tool on the job. Zero out the tool as usual, then if you come back to that tool, enter the SDM number and you're right where you left off, no remeasuring needed.
 
OK, bearing blocks are done. Boy were these more work than expected, about 18 hours in total for the pair. Let me show you the journey.

I started with a 1/2" carbide roughing endmill (Kyocera) to clean up some of the casting remnants from the gate.

image133.jpg

The pair of blocks (and attached caps) were laid relatively flat using a digital level from Shars.

image134.jpg

In hind sight I should have done them one at a time, but it worked out OK. My poor mill flexes all over the place when doing this heavier roughing.

image136.jpg

The bearings were then clamped to my tooling plate and faced to height. I used a height gage to measure them in-situ.

image139.jpg

I also kissed the oil cup bung in that same setup. From there the parts were hacksawed apart and we get two rough bearing blocks.

image142.jpg

I opted to lightly finish each side so I had some reference surfaces to clamp in the vise. I also cleaned up the sides of the caps in the same manner. It was extremely non-trivial to decide how much material to remove from each side so the cast features ended up in the center of the caps. I got it pretty good, but I would have approached it a bit differently if I were to do it again.

image145.jpg

The oil cup bung was drilled and tapped to the center of the feature, while the bolt holes were centered on the block. These features were 0.005" and 0.007" out of alignment on the two caps respectively. There is no hard requirement for their alignment and it is not visually apparent.

image150.jpg

Here is following finishing both caps. A 3/8" diameter spotface is added to each bolt hole so the mounting screws sit against a flat surface.

image152.jpg

From here on out, the bearings were machined with their caps firmly bolted on. Each cap was stamped across the parting line to help match them up after being separated. An extra long 3/8" HSS endmill was used to finish both faces in one setup to the specified 1.31" dimension. Since the caps weren't centered perfectly to the casting, these faces ended up not centered to the lower casting either by about 0.005". The mounting holes were also shifted so it shouldn't affect assembly.

image155.jpg

The mounting holes are drilled and spotfaced as well. There is also a 1/8" hole through the middle of the lower block, not part of the plans. This will allow a retaining pin to be added which will prevent rotation of the bronze bearing relative to the housing.

image159.jpg

This post is getting long so I will follow up with more details.
 
In the next part, I machine the bore of the bearing blocks.

Out of some scrap aluminum, I drilled and tapped the bolt hole pattern for the lower bearing block.

image160.jpg

It was rotated 90 degrees and indicated true in both yaw and pitch within 0.0002" over the surface. The block measurably bowed under the vise clamping force (caused problems later).

image161.jpg

The bearing block was bolted on with a mill fixture used to back up the part and resist cutting forces. Roll about the X axis was indicated within 0.0002" over the length of the machined block.

image166.jpg

The block was drilled to 0.500" and then bored to 0.7522" in 0.020" increments. This was an incredibly tedious process.

image172.jpg

Thankfully, I used the MDI on the CNC to automate the fine feed and retract.

image168.jpg

The fit was supposed to be a 0.0005" interference fit to the bronze bearing, but that ended up being a bit heavy. The bearing really isn't very round or even. It had variations of almost a thou which messed with the fit I was going for. I ended up hand reaming it to an awesome fit to the shaft. Super free with almost zero wiggle in the shaft.

image176.jpg

Here is where the problems hit. Since the shaft is such a close fit to the bore, any misalignment would cause binding. I could get the shaft through both bores but when they were bolted down, the shaft was locked tight between them. An inspection later would reveal that one bearing block had the bore out of parallel by 0.0002" over the 1.31" width of the block, and the other was 0.0016". The blocks were mismatched in height by 0.0004". The main issue was the out of parallelism of the one bore to the base by a thou and a half. Multiply this by the 2-3 bearing block widths between the bearings and you got a shaft misalignment of nearly 5 thou.

image180.jpg

How will I fix this? Stay tuned!
 
Bearing blocks part 3...

I needed a way to correct for the out of parallelism of the bore to the base of the block and to match the height of each block. I figured I would try to scrape them in, because why not?

DISCLAIMER: I am not an expert scraper. I am not trying to teach anyone how to scape. I do not claim to do it the right way. I am doing this for a hobby. I learned what I know on YouTube. Apparently this needs to be said on this site :confused:

Here is my measuring setup. The granite plate is a grade B, flat within 2um (0.00008"). The flex arm (not ideal, but it is all I have) is holding an Interapid 312B-3 0.0001" DTI with the tip at approximately 12 degrees to the surface giving a true reading. The granite was cleaned with alcohol and allowed to come to room temperature. My spotting dye is High Spot Blue.

image196.jpg

My technique is to manually lift the indicator tip, slide the part under and ease the tip onto the part. The part is kept in good contact with the plate and swept to find the lowest point (center of the bore). Both sides of both parts were measured and the lowest one zeroed out the indicator.

image195.jpg

I first scraped for good bearing contact as the milled surface was poor. In all these pictures, my bluing is very heavy. This was often done for the camera as the thin bluing was almost impossible to see.

image185.jpg

I was using a file with the tip ground to a negative rake with a wide radius. After 2-3 passes, I am starting to break up the mill ridges. I am working at 45 degree angles relative to the part and rotating my scraping 90 degrees each pass.

image187.jpg

Once the surfaces were in decent bearing, I could evaluate the geometry. One block was 0.0004" high of zero on one side and 0.0006" high on the other. The other block was 0.0000" on one side (arbitrary) and 0.0013" high on the other. I focused on this crooked one first.

image189.jpg

This was done by selectively removing high points from the higher side. I mentally divided the part into 4 bars across the short width. The whole surface got one pass, then 3/4 of the surface of 2 passes, 1/2 got 3 passes and the final quarter got 4 passes. I would correct about 0.0001" of angular error per pass. I would continually check the part for hinging to determine if I was developing a twist or bow as I worked. On the first part, it stayed pretty flat the whole time. After 15 passes or so, I was able to remove all the angular error from the part, with both sides of the bore measuring within 1/4 graduation of the indicator of eachother.

image192.jpg

The process was repeated for the second part, First to correct for angular error and then to start bringing the height down. During this process, I noticed the part would hinge directly on its center leading me to believe I had developed a bow to the part. Some scraping on the center only corrected the hinging so it occured at the 2/3 point (centered on the width) opposite where I was pushing it.

image199.jpg

Once the blocks were within a tenth of eachother, I went back to the first and ground a smaller file. Using extremely light bluing, I tried to increase the bearing contact. It was pretty easy to identify the burnished locations and mark those with a Sharpie to scrape. The smaller file gave me great control and allowed minimal material removal. After 10 passes of this I was bearing at 30 points per square inch or so.

image197.jpg

I used the first block as a reference and scraped the second within 1/2 graduation (0.00005") on the indicator relative to the first. I got the bores on each block even side to side within 1/4 graduation (0.000025)". It was extremely frustrating to take measurements at this level as any dust on the plate would throw the measurements all over the place. Also simply holding the part for a minute or two would cause it to grow by nearly a tenth. My final measurements were taken with gloves on and after an hour of letting the parts come to room temperature. My measurements are for reference only since my indicator was last calibrated 11 years ago and my surface plate has a flatness limit of 0.00008", but man was I excited to be able to control the removal of material to this level.

image200.jpg

Remounting the blocks yielded a shaft that spun freely in the bearings. I thought I needed to scrape the base as well, but this was good enough.

View attachment 63382109301__80CC3441-46D8-4D91-939F-AE8C3EC93BB1.MOV

EDIT: No oil in the bearings yet in this video, just showing how easily it spins.

So there went 18 hours of my free time (4 of which were scraping)! Getting closer.

Mike
 
Last edited:
Had a bit more progress since the last update that I'd like to share.

I focused on the crankshaft next. Remember that this is a 5 piece press-fit assembly (two shafts, two discs, and the offset pin). My buddy got a start on one of the discs which I lightly cleaned up

image202.jpg

The second disc was cut to match the first. I had to make the OD slightly undersized from the print to get them to match. The CCMT insert worked great for this heavily interrupted cut. The bore was reamed to 0.624"

image214.jpg

I had to take some time to think of how I wanted to cut the offset pin bore such that each disc had the exact same throw and the webs in the castings were aligned. Any error in the throw of the cranks (relative to eachother) would create a misalignment of the end shafts and they would bind in the bearings.

To accomplish this, I hand punched and drilled a tiny hole in the middle of each of the webs visually.

image220.jpg

I then turned and sanded in a 0.6238" arbor that was a tight sliding fit to the bore of the discs. Using a tiny drill bit shank to align the holes, I super glued the discs back to back.

image221.jpg

image222.jpg

The super glue firmly held the parts together for the remainder of the machining operations. I set this up in the mill. Unfortunately I did not have any Vee blocks that were big enough to hold the part, yet small enough to fit in the vise (this was a problem later on). I added a hold down clamp to make a hard stop I could push the part against. I cut off a section of the arbor to use as a pin to indicate the bore.

image225.jpg

I first indicated at the base of the pin to find center of the bore, then used a sharp tipped chamfer tool to visually clock the part such that the webbing would be centered on the to-be-drilled hole. I did not yet take any care to accurately align the part to the machine since any effort there would be lost when I loosened the vise to clock the part.

image227.jpg

Next, using a tenths indicator, I swept the pin up and down to measure the angular alignment of the part in the vise and tap it into true with a lead tapper. This was time consuming and took about an hour. At the end of the day, I got it within 0.0001" over the roughly 1.2" length of the pin. Once it was aligned, I found center again.

image228.jpg

Out of concern that the part might move when machining it, I gently placed a strap clamp on top, tightened to finger tight. This was a huge mistake I would learn later.

image229.jpg

The hole was rough drilled

image233.jpg

Then I threw in a under sized reamer to cut the precision .3115" bore. When I first mounted the reamer, it had 0.009" runout. Using a tenths indicator, I was able to tap the tool into runout <0.0005". I figure that the flexibility of the tool shank will allow the tool to center up once it hits the drilled hole.

image230.jpg

Unfortunately, once I removed the part from the vise, I could see something was off. The hole wasn't right. I setup a measurement cradle on the surface plate, using the center bore as a datum and measured the hole was crooked by 0.002" in the radial direction and 0.008" in the tangent direction. I suspect between the addition of the hold down clamp, the drilling, and the reaming, the part must have shifted in the vise.

image234.jpg

This amount of misalignment was not tolerable and would cause issues with the center shaft alignment once the assembly was pressed together.

It needs to be corrected... To be continued.
 

Attachments

  • image230.jpg
    image230.jpg
    900.4 KB · Views: 4
OK, now to correct the parts.

I opted for a new work holding method of a 3.5" 4 jaw scroll chuck clamping the bottom boss. Again, I setup the part with the tight fitting dowel and clocked the part using the tenths indicator. Once it was darn close, I indicated the vertical of the pin and tapped the setup to have about 0.0001" over 1.2" of angular misaignment.

image236.jpg

Once the part was sitting true, I indicated the existing location of the small bore and verified the throw of the crank was accurate and within specification using a bit of trig.

image238.jpg

The hole was colored with steel blue and fine bored only to full cleanup.

image241.jpg

Once done, I setup the measurement cradle on my surface plate a verified the hole was parallel to the bore within 0.0002" in both directions :p

The new bore (now slightly oversize from the print was measured with telescoping and small hole bore gages. I really hate small hole bore gages and don't trust them under 0.001. I trust my telescoping gage skills within 0.0002".

A new crank pin was turned identically to before except with larger ends to match the oversized holes.

image244.jpg

The parts were cleaned of blue and oil then heated over open flame for about 10 minutes to break the super glue bond. It was super strong. Yes I know our stove deserves some TLC. I cook a lot!

image245.jpg

Using the arbor pin, I slowly pressed the parts together on my mill vise. I removed it several times to inspect on the surface plate to verify they were going together straight. All good!

image247.jpg

image248.jpg

The shaft ends were pressed together on the lathe. The part was held in the 4 jaw, while the shaft was in a big drill chuck in the tailstock. An aluminum plug was placed between the discs with the exact size of the gap to prevent the assembly from collapsing under the pressure.

image250.jpg

The reamed hole was a bit light of a press, so Loctite 609 was added to strengthen the bond.

image251.jpg

And Done!

image252.jpg

The shaft is a great fit between the bearings and it spins very smoothly. A bit of run-in will likely make it even better.

image255.jpg

Here is a short clip of the flywheel and crank shaft spinning in the bearings with a light coat of Vactra #2 way oil for lubrication.

View attachment 63545291795__FF3B9F80-9106-42EB-A14A-FB3DEB7E4B18.MOV
 
The crankshaft is going to be my next major part to work on since it will let me cycle the cylinder by rotating the crankshaft.

The casting was extremely rough (the worst of all of the ones in this kit) and they two mold halves were not lined up well at all. I spent quite a few exhausting hours draw filing and sanding to get this clean enough to work on.

I plan to leave most of this casting rough after machining the features, so I'd like it to be relatively attractive.

image257.jpg

image259.jpg

image261.jpg

I jumped back to the cylinder casting to get the bolt holes mounted. The casting was held in the vise against a hard stop and clocked so the bores are in line along the X axis. I got it within a thou. The hole patterns are different on each side and I was very careful to drill the holes correctly lest I repeat my mistakes on the frame.

image264.jpg

The holes around the big bore are tapped 10-24 while the holes around the valve bore are tapped 5-40. I centered on each bore before drilling each bolt hole pattern so the pattern was perfectly centered. I did clock the part extremely close, but it is more important to center the pattern on the bore rather than getting the center to center distance perfect.

image266.jpg

All drilled and tapped. I lightly stoned the surface with a hard Arkansas stone to remove the burr around the tapped holes and keep the surface flat.

image267.jpg

The cool thing now is the bolt holes allow me to assemble a big chunk of the hot end of the engine. Here is the frame, head, cylinder, outboard head, cross head, wrist pin, piston rod, piston rod packing nut, piston, valve, inboard valve head, and valve packing nut. I had zero assembly issues.

image268.jpg

image269.jpg

Here is the engine as it stands!

image270.jpg

image271.jpg

I made up some gaskets with computer paper, and the piston holds air pretty darn well. It needs better gaskets than oiled paper but I'm very happy.

Major components remaining:
  • Connecting Rod
  • Eccentric Ring
  • Outboard Valve Head
  • Cylinder (drill/tap steam inlet, drain bungs)
  • Bearing Sleeve (turn to length, add oil holes)
  • Frame (drill and tap for oil cup)
 
Here is a list of machined parts and my status:

Crank Disc (Qty. 2): Done (4.5 hours)
Cross Head: Done (8.5 hours)
Crank Pin: Done (4 hours, 1 Scrap)
Lower Valve Head: Done (2 hours)
Piston Rod Packing Nut: Done (1.5 Hours)
Lower Linkage: Done (2.25 Hours)
Upper Linkage: Done (2.25 Hours)
Eccentric Hub: Done (8 Hours, 1 Scrap)
Wrist Pin Bolt and Nut: Done (1 Hour)
Outboard Valve Head: Incomplete (1.25 Hours, 1 Scrapped)
Valve Rod Packing Nut: Done (1 Hour, 1 Scrap)
Oil Cup (Qty. 2): Done (1.75 Hours, 1 Scrap)
Piston: Done (5 Hours)
Lower Valve Rod: Done pending final fit (0.5 Hours)
Pillow Block & Cap (Qty. 2): Done (17.5 Hours)
Valve Eccentric Ring and Cap: Incomplete
Upper Valve Rod: Done pending final fit (0.5 Hours)
Valve: Done (3 Hours)
Crankshaft - Short: Done (0.75 Hours)
Crankshaft - Long: Done (0.75 Hours)
Piston Rod: Done (4.5 Hours, 1 Scrap)
Connecting Rod and Rod Cap: Incomplete (2.5 hours)
Base: Done (5.5 Hours)
Flywheel: Buddy did this one, I just skimmed it true. Done (1 Hour)
Frame: Done pending drip oiler selection (10 Hours including repair)
Cylinder: In Process, steam port and drain port needed (8.25 Hours)
Inboard Head: Done (3.5 Hours)
Head: Done (1.5 Hours)


Recorded hours to date: 102.75 Hours
(probably close to 115 including cleaning, bench work, and planning)
Estimated Completion: 80%
 
Last edited:
A steam engine is on my bucket list, thanks to your detailed build notes including the time and effort log, I have a great idea of what I will be getting into when I decide to move forward on this.
 
A steam engine is on my bucket list, thanks to your detailed build notes including the time and effort log, I have a great idea of what I will be getting into when I decide to move forward on this.
I'm having a blast with it. Honestly I don't care about the hour count, it is all for fun.

The No. 6 is tied for the biggest steam engine they make. This is fun for me for sure. Only limitation is that is uses a LOT of steam/compressed air, so you need to have a good compressor to run it. Mine may not be able to keep up continuously.

The smaller engines look like less work, but have a ton of tiny parts. Gas hit and miss engines are the next level up and they seems to have a massive jump in complexity and difficulty. I have been follow an Edwards #5 radial engine on here that is super impressive.

I've learned more about machining doing this engine than all of my projects before it combined. Seriously
 
Back
Top