Vederstein's Dancer Engine Build

[vederstein]

In the Plans section, you'll find some musings on my version of a water pumping steam engine located at the London Museum of Water and Steam (a.k.a. the Kew Bridge Steam Museum).

Being that this engine has a some components where the techniques are new me, I felt I'd share my learning experiences. I don't intend on documenting every part, just those components that have difficult or weird setups.

First off is the walking beam.

It started as a piece of 1/4 x 1-3/4" barstock. I drilled in the pivot points to be a close fit for the 1/4-20 bolts.



Then I transferred the blank to my Mini-Mill that I put a CNC conversion on years ago. As many know, a Mini-Mill is well, mini. Therefore I could only profile and pocket the walking beam in halves. I did the left side, then rotated the part and machined the right. Then I flipped the part and pocketed the opposite side. The close fitting holes allowed my to break down and reassemble the setup without losing too much positional accuracy. It's not that important though. Other than the location/size of the holes, the actual profile and the pockets are more for looks than anything else.





After the CNC milling, I transferred the (now looking like a) walking beam, to manual mill. Using it like a drill press, I drilled out and reamed the part. Then, after a bit of deburring, the part is complete. Then make a second...





Being that my Mini-Mill Y-axis motor had a coupling failure, these two parts took all day to make. That's enough for one day.

...Ved.

 

[vederstein]

I try to do something new on every engine I design/build. On the Dancer's Engine, I have the steam admission valves run off a shaft that's run off the crankshaft through 90 degree bevel gears.

I designed the engine without giving much thought as how to make bevel gears. This is especially true because I have a Chinese Mill-Drill and not a real knee mill like a Bridgeport.

After giving it some thought, I realized without making some tooling, It would be impossible to cut a 45 degree gear without rotating my dividing head 45 degrees.

Years ago, I had a cheapo milling vise on a swivel that gave up the ghost:



I never threw it away because I thought it might be useful in the future. Today was that day.

Also, an inexperienced engineer had a bunch of part pallets made that never worked. When the decision was to scrap them, I took a few home:



So the idea was to remove the bars and and modify it to mount to the vise swivel to support my dividing head/tailstock. The bar would be barely long enough, so I wasn't sure this would work. But I'd rather screw up on a piece of scrap versus some virgin aluminum that actually cost me money.

So I cleared off the mill and mounted the bar and located center. The bar had the 5mm dowel pins pressed in so I used those pressing up against the milling table slots to maintain alignment. I then drilled/bored a hole for the center pivot.




Then I drilled clearance holes for the swivel clamps and milled some slots for the dividing head.




After the mounting holes/slots, I milled slots for the alignment "nubs" on the dividing head.



All machining complete, I cleared the table and mounted the bar to the swivel Then the dividing head/tailstock to the bar.






Checking for fit, I performed the same functions on the other end for the tailstock.

At this juncture, I see that I cannot use the tailstock because the gear cutter is going to cut right into the rod. But the dividing head idea seems to have merit.

I'll need to make another jig with a similar idea. It''ll be more rigid that this, but as I said, I'd rather screw up material I didn't pay for.

That is all.

...Ved.

 

[johnmcc69]

You really have nothing to lose, & on the "Cheap". Nice thinking! Looks like you have a winner..

john

 

[vederstein]

Well it was time to give making those bevel gears a try. Note that I've never done this before and I have no idea what the hell I'm doing. I expected (and was successful) to screw up at least my first try.



Some people here on their builds only show their successes. But on can learn just as much from how people have failed as well.

Anyways, here it goes. (This is probably be a couple of successive posts):

Being that these are right angle bevel gears, I needed to set the lathe cross slide to 45 degrees. I used my combination square for that purpose:



Next was to make a gear blank.

Mistake #1: When making the blank, I couldn't get even my small lathe tools to cut the two 90 degree angles. So I had to remove the part and press it onto an arbor to continue to blank. Hopefully the following pictures are enough explanation:







Off to the mill with the set up from the previous post. I again used my combination square to get the rotary table to 45 degrees:



Next Post.....

 

[vederstein]

...Continued from previous post:

I pressed the gear blank onto the arbor and proceeded to cut the teeth.



Unfortunately, the gear blank spun on the arbor and I lost position. Upon investigation, the teeth weren't deep enough and I think because my mill setup was too shakey. Part 1 = Scrap.



On to Part 2:

Ok, Arbors bad, so I took my 1" diameter brass stock, cut the angled tooth face, drilled/reamed the bore:





Back to the mill, I tore down the setup and bolted the rotary table directly to the table and proceeded to cut the teeth. I messed up on my rotary table indexing and my teeth were off. CRAP!!!! (Part 2 = Scrap.)



I completed the part because it will be valuable for fitting later on the engine, but I know I'll need to do this again.

On to Part 3:

Same deal as Part 2, just this time I didn't mess up the indexing. So after the teeth were cut, I faced off the gear to clean up the rough tooth edges:



I used an 1/8" wide parting tool to whittle away the back side of the gear then it was parted off.



Tore down the milling setup again and mounted the rotary table square to the machine to drill/tap the set screw threads.



I did the same to Part #2, and I have hopefully have 1-3/4 useful gears. At least they'll be good enough to check for fit when the rest of the engine is assembled.



And that was enough for one way considering how fussy these things were.

...Ved.

 

[vederstein]

On to the valves.

With some mystery3/8" brass roundstock, I first used a parting off tool to cut the admission and exhaust ports and then drilled/tapped the end.





I flipped the part, faced it off to length and drilled out the end for the steam exhaust.



Using the rotary table, I cross drilled the six exhaust holes in each exhaust port. Repeat for the second part.



The valve tapped consisted of turning the thread to the major diameter then using a threading die, cut the threads. To keep the die straight, I back it up with a block and the tailstock. Then I used a parting tool to cut off. I'm leaving off the flats for now and will only mess with those if I need them when the engine is assembled.



And I have two spool valves that if I'm lucky, won't require too much fussing to work.

That's all for today.

...Ved.

 

[vederstein]

The stand components are decorative to put a bit of flair into the engine. Each stand is a piece of 1/2" brass tubing with turned ends inserted. The assembly is held together with 10-32 allthread. There are four stands.

The tubing is easy enough. The ends would be ease if not for the decorative flutes.

First was to make a blank:



For functionality, this is enough, but the flutes make it look much better. Do do so, I needed a fixture in my rotary table. So I reamed a 7/16" pocket and threaded the bottom of a scrap piece of aluminum round:

.



From there I can put the fixture into the rotary table and clamp the component into the fixture. Machining the flutes (one flute every 30 degrees) could then commence:



Do this eight times and we have a set of stands:





That's enough for the day.

...Ved.

 

[vederstein]

The Main Bearing Housings are, like the stand ends, mostly decorative. They could've been made much simpler, but I wanted some frivolous details.

There are four Main Bearing Housings so each operation was times four plus 1 for my prototype piece for a total of five.

First I machined the outer angle with the lathe's compound set at 20 degrees and the engagement boss.



Next was to drill/ream the 0.500 bore and then remove the part with a parting tool.





With the blanks done, I need to make a jig to mount the parts into the milling machine/rotary table.





Off to the mill. I offset the rotary table by 20 degrees then indicated in relative to the angle of the blank.



Then I milled in the flutes. The depth of the flute isn't specified on the drawing and it's to whatever the builder (me) wanted. After creeping up on the dimension, 0.030 flute cut was my preference.



There's a 0.093" hole on the back of the part so I then squared up the rotary table again. Ideally the hole is to be in line with one of the flutes. I couldn't figure out how to fixture this accurately, so I eyeballed it relative to the drill bit.

.

After making some brass 0.093" pins (not shown), I test fitted the parts to the walking beam. It works for this guy. This was about three hours of work.

...Ved.

 

[johnmcc69]

Looking good Ved!
The flutes & beam look very good!

John

 

[vederstein]

Sometimes a part is difficult not because of its design, but because of what material is available. The rod end seal caps I could've made from barstock and been done with it. But No!!!!! My ******* thought "I have some flatstock that I haven't used in years, I'll make that round!"

(Bad idea. I wound up using more material than if I has just parted off some roundstock. But they're done. Here's go the documentation:

I first drilled/reamed the gland and bolt pattern into a piece of 1/8" thick brass flatstock. Eyeballing the center would be good enough.



Then I made a mandrel with a boss and the bolt pattern to mount the plate to. Note that I marked the chuck and the mandrel with a Sharpie so I could put it in the same position.





The mandrel was chucked back into the lathe and the plate mounted. I then turned down the part to the 0.75" diameter. Rinse - Repeat.

(Actually I scrapped my first plate and had to make two mandrels because the first one broke with a single bolt instead of all four.)

...Ved.

 

[vederstein]

Today was making the crankshaft end connecting rods. The ends of the rods are two different lengths of the same design so I was able to make both types of end with a single setup.

I changed out the lathe with my four jaw and centered up some 1/2" square stock. The stock was then faced.



Drill the center and I used a 1/4" endmill to create the pocket to receive the center section of the rod.





Then, with the compound set at 15 degrees, I cut the taper. These were surprisingly accurate.



Then I tapped the center after drilling out some more.



Being that I'm building my own design and this is a prototype I made some changes on the fly. The taper was better than I expected and the bore worse than I expected. (The bore was to be .250" and the endmill apparently wobbled and the bore was .265" on three of the four parts.) The center bar section of the connecting rod was to be 1/4" CRS rod, but I changed it to 3/8" Brass because of my bore fit issues.

I don't know if to leave the ends square, chamfer, or to attempt to radius them.

Whatever the outcome, when I publish the design, the changes will be documented.

Being that my mill is still set up with the rotary table, this was a good place to stop.

...Ved.

 

[vederstein]

New day, continue on with the connecting rods.

I had new gumption, so I decided to go ahead and properly radius the rod ends. After drilling/reaming the pivot points, I set up a pivot rod. From there I can set the end mill in position and rotate the part to get the radius.

First I plunged cut about 0.030 outside the radius for roughing cuts, then I set to the proper radius and cut to the final radius. For those that attempt this DO NOT EVER CLIMB MILL WHILE DOING THIS. YOU WILL LOSE YOUR FINGERS!!!! (BTW: The brass gear was a piece of scrap that I used as a spacer.)





The next couple of pictures should be sufficient to show how I milled out the clevis web. The major item of note here is to cut out the opening portion of the web last. Otherwise the part is so flimsy, that it will not be accurate.

....to be continued what I feel like it.....

Ved.

 

[James Barker]

I am still watching and following along with you on your journry. I just went and bought more popcorn for the trip. Nice work!

BC1
Jim

 

[vederstein]

Now to work on the walking beam structure.

I clamped a length of 5/16 brass round into my V-blocks. After indicating into position, I drilled/spot faced per the print. The only precaution was to just start the drill and not drill through. I didn't want to drill into my Vee Blocks. Easy...





The offset stands for the bearing shaft was originally to be 3/8" square stock, but I didn't have any. So I changed the design to round stock. Either way is acceptable. It's of the utmost importance that the distance from the bottom of the stand to bottom of the radius be consistent. The parts can be a bit off, as long as they're all off the same amount.

So I fixtured the part in a Vee Block with a back stop so that my cut was always in the same relative position.





The build is far enough along now that I can start to assemble some of the components other than having a bowl of metal components. Thus far, I'm pleased with the look of the walking beam now that I've assembled it.

To be continued.....

Ved.

 

[vederstein]

The material I have for the upper and lower plate are sheets of 1/8" thick aluminum. The only method I have to cut them into manageable pieces is to cut the sheet with a Sawzall. A Sawzall certainly isn't the most accurate cutting method in the world.

So a lot of the upper plate was just getting the stock to size and squareish.






After that indicated to center and drilled out the bolt pattern and the cutout corners.



Then I plunged my endmill to rough out the cuts and then ran a final cut to size.




I assembled the upper portion of the engine for ***** and giggles and I'm really liking how this thing is looking:

Bye for now.

...Ved.

 

[vederstein]

I made the crank bearing housings without any pictures thinking that the parts would be easy. In fact they were quite fiddly. The valve driveshaft bearing housing are smaller, but of a very similar design. The design requires three of these housings. Knowing how fiddly the crank bearing housings were, I thought I'd try something new (to me):

I would machine all the functionally important features into a single piece of barstock. Then I'd cut the pieces apart and finish the sides. In doing this, I left a 1/4 inch space between what would be each part for the saw cut.





I scribed the saw cut line and cut out the three pieces.



Using the tap drill bit as an indicator, I re-index the part in the mill to qualify one side to print.





I could then clamp all three pieces in the mill and finish the part to width.



After tapping the holes, I then used the same technique as in a previous post to round the end. It was a little off (particularly the first piece), but the important details are the location of the holes, not the radius on top.

...Ved.

 

[vederstein]

Making the Cylinders:

The two cylinders are very similar to each other and it would be redundant to show the manufacture of both.

After squaring the block to size, I drilled out the holes for the cylinder head threads and the steam port:



The cylinder has a .875" bore. I have a .875 reamer, but what I don't have is a drill bit about 1/32" less than the reamer. So as strange as it may seem, after drilling out as big as I could, I used a boring head to get the diameter close. Then I finished the bore with a reamer:







After machining that side, I removed the part and made marks with a Sharpie to where the following details were to be. It's very easy to get confused and spend and afternoon making scrap because the part was drilled on the wrong side.



Then is was a matter of drilling/tapping/reaming the remaining sides that need machining:



I actually did make scrap from my first part, but the second one came out ok. As I write this post, I still need to make the second cylinder but it's too late in the afternoon to continue...

 

[vederstein]

The Base Plate should be simple. After all, it's mostly a rectangular plate with a bunch of holes drilled in it. It would've been simple if not for that my mill isn't big enough to machine the part in a single set up.

As before with the Upper Plate. I have some 1/8" thick aluminum that's too big for my 4x6 bandsaw. So I used a Sawzall to cut it.



Next: Off to the mill and while playing musical clamps, try to get as many qualified edges as possible. In this case it one and three quarters sides. My mill just doesn't have enough Y-axis travel.



From there I performed the necessary drilling operations on as many holes as I could reach. Even though the finish hole are larger than 1/8", I drilled 1/8" for reasons that'll be explained in a moment. After that was milling a clearance pocket.



Going as far as I could, I unclamped the material and inserted some 1/8" pieces of roundstock and used those pins to relocate the part square to the milling table. After indexing my position I could finish drilling the holes.



After all the holes were at least located with a center drill, I then finished the part to size. I used the existing, 2nd, set up and then I had to position the part a third time to get all all the outer edges.



There's some "bling" details I left off the part for now. I'll decide later if to put those in or to leave off if the engine is installed on a wooden base. I couldn't resist assembling what I have thus far accomplished.

It's starting to look like an Engine!!

Six hours today.

...Ved.

 

[vederstein]

The Piston Rod Clevis was much more difficult that I anticipated. It's not that it's requiring any machining operations that I haven't done before, it's just that the part is small.

Thinking about it, I made a "Faceplate" for my rotary table with some scrap roundstock. I made a boss to index to the pivot center and I used an already existing 10-32 thread @ 1.000" from center to clamp the part.



I purchased a new toy: An axial center indicator. This was the first time I got to use it. It's a Shars item. I could've purchased one cheaper from Ebay, but there are in metric, and my tooling is Imperial. I know it's from China, but I've generally had good luck with the quality of Shars stuff.



Now that I had center located, I loosened the part and moved the table to 1.000 from center. I clamped onto a bolt from with the drill chuck then re-tightened the rotary table jaws to get the existing holes square. I don't know if there's a better way, but it worked for me.



From there I clamped in the material for the part. I had previously drilled/reamed the pivot center and a clearance hole for the 10-32 clamping bolt. I drilled undersize the transition from radius to straight.



Then I milled the pivot radius oversize 0.010" and milled the flat sides 0.010" oversized as well.



I then milled to the finish dimensions.
After that milled the bottom flat which released the part from the 10-32 clamping bolt.



Rinse & Repeat and I have two Pivot Clevises. (I actually had to do this three times because my first part was scrap due to a stupid setup error on my part.)

I still have to put in the .188 pocket and the thread, but that's was enough frustration for today.


...Ved.

 

[vederstein]

It's been some time since I've last posted about this engine. Quite simply, the remaining parts just weren't that interesting to document their machining.

I've had to remake the following parts:

• (1) Bevel Gear, but that was anticipated.
• Both pistons/piston rods
• The cylinder heads are warped and leak quite badly. They need faced off flat (warped barstock)
• The valves were very finicky to get adjusted in the correct position.

At this stage in the game, it does run (see video), but there's a lot more tweaking to do. Also, I need to paint the thing and put it on a base to make it pretty.

But yesterday (when it ran for the first time) was a big milestone in this project.

Video for your enjoyment or ridicule: