Ridder's Stirling Bobber

[bent]

My next project will be the Stirling Bobber as shown on Jan Ridder's website:

http://ridders.nu/Webpaginas/pagina_stirling_jaknikker/jaknikker_frameset.htm  (this is the Dutch version, you can click the link at the top for English, but the picture in the frame then disappears...)

I don't know why, except that it looked intriguing.  But, to me, the motion of his model seemed a bit...hesitant?  I don't want to use the term jerky, but there is too much variation in flywheel speed for my taste.  Though maybe it's necessary for that hesitation to allow for the expansion and contraction of the air to occur?  So I thought, why not increase the flywheel speed via a gear train, maybe that would smooth out the motion a bit.  Well, one way to find out.  A bit of modeling, and tearing apart an old electric valve actuator to find some suitable gears, and voila (image below).  Gears modeled as simple discs, and the shafts aren't modeled yet either, but enough to get the idea I think.  Of course, some weeks later from starting this, I have come across another video with somewhat the same idea as mine!  Ah well, great minds and all that.

https://www.youtube.com/watch?v=gOClpeq8yVI&t=75s

The mill is trammed, and the new machine vice is mounted and squared to the mill table, test cuts have been made, acrylic chip shields with magnet bases whipped up and ready to go, bearings and test tubes have arrived, and the weather in the garage has gone from glacial to just chilly and damp.  Wish me luck!

[Admiral_dk]

this is the Danish version

Well I'm a Dane and I can tell you that this isn't Danish .... pretty sure the site is in Dutch just as Jan Ridder speaks it.

[bent]

Wups!   
fixed it...

[bent]

Have not progressed much on this over the past few weeks, other than to finish truing up the blanks for the uprights.  Before drilling all the holes in those, and for the base plate, I wanted to get my DRO's installed (it's been 20-some years since I've had a machine with DRO's, and forgot how handy they are, until reading other's posts on this site).  I did decide to purchase a cast flywheel for this project, just because.  Bought a 3in. "heavyweight" MartinModel.com casting, it looks pretty good.  Pic #1 is raw, Pic #2 after some fettling with a round file to remove the flash between the spokes.  After reading Jo's posts, I think my next steps are to spray a coat of primer on it, then chuck it in the 4-jaw to center on the inside of the rim, and start truing it up. 

I do like that the vendor anneals their castings, and provides a decent sized spud on one end to help in machining.

[b.lindsey]

Looking good so far. Nice looking flywheel too.

Bill

[bent]

A little bit of work done Sunday afternoon on the first part (the flywheel support standard), and some learning experiences with the DRO's on the mill (it's been a long time since I've used such devices).  I'd laid out hole centers with some blue sharpie marker and light scribe marks, but they got wiped away from handling, so I ignored all that and used the DRO's directly to spot all the features.  Pic #1 below is the part with side notches milled, and holes bored.  First notch milled really nice, but forgot to set the cutter deep enough and only had half a slot.  Argh!  Back up and recut, and hit the same cut depth with the dro - and hey, no visible step in the finished notch, aren't dro's neat!!  After milling the notches and swapping the collet out for the chuck, and re-zeroing the DRO's I noticed that I'd mislocated the notches...figured it out finally - I'd forgotten to adjust for the thickness of the edge finding in spotting one of the pairs of notches.  So yeah, the notches are about 0.1" too close, so spotted the holes with the same offset so it wouldn't look too obvious.  Yeesh, what a noob!  Oh well, those are decorative bits anyways, and luckily I'd figured out my goof before spotting and drilling the bearing hole.  So, something to remember for next time.

But I was pretty happy using the z-axis DRO to hit the depth of the bearing pocket within a few thou - dropped a bearing in the pocket and it sets back at just the right depth (you can see it doing so in the photo).  Same for spotting and controlling the depth on the hole c'sinks, these came out pretty darn symmetric all around, happy with that.

Flipped the part on edge and spotted, drilled and tapped the two 6-32 holes.  Went slick as snot - even tried "powering" the tap for the first few threads, using the speed control.  A little scary, so I stopped doing that, and instead used the spindle wrench to turn the tap (held in drill chuck) to start it (holding down pressure on the z-axis with t'other hand), and then finished the hole by hand after getting a few turns in.  Pic #2 shows the second hole tapped, with the tap I used - a spiral fluted, TIN coated tap I swiped borrowed from work, this worked beautifully, it ejects chips out the top, leaving a much cleaner hole with no shavings buried down in the threads.  Will need to get a few more of these in other sizes.

Finally, set the part on a 45-degree angle, eyeballed using a trusty old drawing triangle, to mill the chamfers - this too is where the DRO's are handy, the chamfers came out pretty symmetric just by touching off with the side of the end mill and dialing in the cut depth with the DROs.

Pic #3 is the part, completed, sitting on my working drawing with all the markups (note the 3/8 radius modified to 1/4" for the notches, as I have not got a 3/4" mill...).  All in all, am happy with the new toys, though the operator needs more training, and the engineer should work on his drawings before releasing them to the shop.

Lessons learned:
1.  Always double check the offsets when zeroing the dro's.  And lay out the holes, just as a triple check.
2.  Don't drill big holes with the mill set in high range, or it will stall.  Repeatedly, if you are that dumb to keep trying.
3.  Like when using the lathe, have a set place to put the chuck key, and make a habit to put it there every time; don't carry it around the shop with you when looking for drills, or calipers, or checking the drawing, because you'll put it down somewhere stupid and forget about it and spend 10 minutes searching for it.  Again.  I should probably paint all my keys different colors (drill press, mill, lathe) to keep them straight.
4.  Almost forgot.  When clamping parts in the screwless vice, make sure the clamping screw has a good bite in the slot chosen, or the part will slip even though you think the jaws are tight.

 

-Ben

[zeeprogrammer]

Looking good.

I'd never seen a tap like that. I'll need to check it out.

Always nice to see 'lessons learned'. 
I'm sometimes embarrassed to mention a 'lesson' but shouldn't be. Everyone starts out learning 'lessons'.

[MMan]

Hi Ben,

I know I have been guilty of all your lesson learned. what I try to do now is to: Zero the DRO on the edge finder, wind over by the radius and then zero again. Key for me is for it to be a habit. Of course I get distracted and forget, mostly I catch it because the tool looks to be in the wrong place but not always. Sigh.

Chuck keys always find a place to hide, I guess that is why keyless chucks are so popular - just feels wrong to me on a machine tool.

Spiral flute taps are great - lovely to watch the chips climb out of the hole and no need to think about 2-in-1-out. Although thinking back to when I used to drill holes as a sallow kid in production, the tapping machines did not do 2in1out and didn't eject the chips either, no idea how that worked hindsight.

Al the best,

Martin.

[bent]

"Key for me is for it to be a habit. Of course I get distracted and forget, mostly I catch it because the tool looks to be in the wrong place but not always. Sigh."

Argh.  I get you, Martin - I did it again last night (twice, caught it before cutting chips the second time at least).  

Oh well, once again nothing but the cosmetic details was affected, so decided to press on.  Got the uprights all finished over the last couple nights.  I hope to start work on the flywheel next.

Pic #1 below is the cylinder standard, finishing milling the fork cut-out.  Did this all in one pass.  Originally drawn as an odd size U-shaped cut (full radius on the bottom) which would have required a boring head that I don't yet have, so made the rectangular cut instead.  Expediency and all that.  Also left the dustpan in the shot, you can see how thing the swarf is, as the nervous new mill owner was loathe to crank the handles too quickly.

Pic #2 is the same part flipped up for drilling and tapping the mounting holes - you can see the chip ejection from the spiral fluted tap, here in #8 size.  Love those things.

Pic #3 is the crank/gear support, after final planing to size with a 1/2" end mill.  Not super happy with the surface finish.  Add a flycutter to the shopping list.

Pic #4 is a family shot of the completed parts so far (all 3 upright/support thingies), though I will probably do some sanding on the faces yet.

[bent]

Another few days, a little further along. 

Pic #1 below is a family shot as of Saturday, with the base plate (shown upside down with the counterbored screw holes) and 3 uprights finished (less some sanding/polishing).

After reading Brian's posts on his flywheel casting, I decided to get started with mine.  First a coat of paint (thanks Jo!) to better keep the graphite in check from the cast iron surfaces (Pic #2 below).  Gaudy red, because I had a nearly empty can of it, and no primer; not sure if I'll keep it this color or not.  Then into the 4-jaw chuck, and start getting it aligned.  I was checking with a dial indicator on the vertical face of the inner rim, and along the radial direction of the step between the inner and outer rim.  With a bit of fussing, and a piece of 0.060" weld wire for a shim, I managed to get the total play on the indicator down to about 0.010 to .020" FIM.  Not too bad, and the indicator motion was mostly random, with a couple of "wows" (non-circularity) about 90 degrees apart (casting distortion).

I'd seen Brian's and zeeprogrammers posts, and the suggestion of using a mandrel block to support the casting from the inner rim/spokes face.  And that's where the real problem showed up.  This flywheel casting had over .100" of "float" (one side of the mold tilted relative to the other during casting), so there is a pronounced visible wobble when the parts are rotated.  There was also a noticeable shift in the casting (the two mold halves had slid along the parting line) of about 0.06", so another axis of the part was going to look wobbly.  Set down tools and went inside for dinner, as I was getting tired and wanted to sleep on the matter. 

Woke up Sunday, and after a bit more dithering and re-measuring, realized that there wasn't much else to do but start cutting and see if I could make things work.  I could choose to try and balance out the wobble, shimming the part in the chuck, to get a more or less equal amount of wobble on both faces - but then both faces would look like heck when it spun.  Or, dial in on one face, and let the other one wobble.  I took the latter approach, figuring one face was going to be somewhat hidden behind the crank and the rotating gears.  Pic #3 is the flywheel in the 4-jaw chuck, shimmed to get the outside face as centered and as flat as possible.

Machined the hub "spud" to clean it up, and the od and outer face of the rim.  Then reversed the part, exchanging the 4-jaw for the 3-jaw now that I had a cleaned up spud to turn it on.  Wow was that ugly, the wobble on the face and rim were very pronounced relative to the first side.  Oh well, press on, press on!  Cleaned up the o.d. and rim face using very light .005 and .010 cuts due to the heavy interrupted cuts, and then made a series of passes with a 20 degree taper on the cross slide, to get the corner of the rim, where it steps down to the inner rim, to be circular, and carried that across the face of the inner rim until it too was no longer wobbly.  Finally, a pass with a 30 degree taper, to get the corner of the inner rim (where it steps down to the spokes) to be circular.  Pic #4  is the "finished" view of the "bad" side of the wheel.  Finished in quotation marks, I still have to face and bore the hub, and I may cut a 45 or steeper chamfer where the hub machining runs out to the spokes.  I'm not happy that I've lost so much of the cast finish for this part, but better it looks straight/true spinning than not.

Pic #5 is a shot of the "good" side.  I went ahead and trimmed the inner corner of the rim to get that line circular and true, and match somewhat the machining done on the bad face.  Again, need to face the hub and probably chamfer the corner where it meets the spokes.

FWIW, as an engineer that designs castings, this casting comes close to exceeding the limits for float and shift in a production foundry.  Oh well, still beats the amount of work I'd have had to do machining a spoked wheel out of bar stock on a rotary table, which was my original plan.  And as I don't own a foundry or CNC mill, I'm hesitant to invest that kind of time just for one wheel, I'm not retired (yet) .  Martin Models is not too far a drive from home, so maybe I should plan to do an on-site inspection of the wheels I want to buy?  Hmm.  They'd probably slam the door in my face, now.  Seriously, this was a nice casting to machine, and in the end, I think this wheel turned out pretty good, though it took some work and more machining than I wanted to get it there, so I'd probably do it all again.  Some variation is normal in castings, and our eyes will see these defects when the wheel is turning, so having a plan like mine to get as many of the corners/edges to be circular as possible may be the only way?

The last Pic #6 is my attempt to get a shot of the rotating wheel, so you can see the amount of "wobble" on the good face.  The blur lines of the spokes, and the rim edge where it meets the spokes, show where some of the remaining visible wobble is; but the rim edge and inner step run pretty true (no blur lines). 

[zeeprogrammer]

If the rim is running true then I'm thinking you're good to go. (But keep in mind I have had little experience with flywheels.)

I'm curious about the suggestion to paint before machining.
Wouldn't that only help up until the paint is machined off?
Does the same technique apply to metals other than cast iron?

[Tennessee Whiskey]

Looking good so far Bent   Zee, that’s a Mr. Pete trick on the painting: paint it all first and after machining, there is nothing to “tape off and paint”

Cletus

[zeeprogrammer]

Looking good so far Bent   Zee, that’s a Mr. Pete trick on the painting: paint it all first and after machining, there is nothing to “tape off and paint”

Ah. It's about having the spokes and inner rim painted. Thanks.

[bent]

Well, yeah, except for the dings and scratches (sheesh), to fix those I'm going to have to tape it off and hit it again anyway.  Also, the inner rim face on the "bad" side I will repaint just to get it to look like the "good" side.  But in theory, it's a good idea...

It still cuts down on the debris/dust from handling the raw casting, and during cutting I found the contrasting color to help figure out how/where to cut next.

[bent]

A bit more progress this last weekend, but some frustrations too.  Put another coat of paint on the flywheel, to cover the areas where I had machined to get some of the visible wobble to go away.  Swapped the lathe from 4-jaw to 3-jaw to turn one end of the flywheel (holding the spud of the casting), and sand/polish the o.d. and flanges.  Then went to reverse the part, only to realize that the jaws on the smaller chuck wouldn't open wide enough to hold the wheel by the rim.  Grr.  Oh well, I could finish that part on the lathe at work.  But maybe need to think about getting a larger chuck for the lathe...or just get really good at setting up the 4-jaw.

Onwards.  Realizing I'd not printed drawings for some parts I wanted to work on, and missing material for others, I settled on making the con rod, and decided to turn it from rectangular brass bar stock.  So, back to the 4-jaw, and started whittling.  All went well (photos 1 and 2 below) until it came time to thread the end of the part.  Using my cheap Amazon #5 threading die, I very quickly buggered up the end of the rod.  Grr.  Managed to get a partial thread on the end using a #6-32 die, but I may need to remake this part from scratch.  Took pictures, but everything came out blurry (hands shaking from rage? dunno). 

Frustrated, but also realized that not having a die holder for the lathe tailstock was limiting.  Years ago I bought a #2 Morse taper with a 1/2-20 male thread on one end, just to make such devices with interchangeable ends to hold different sized dies, but forgot about it until now.  So, went to cut off a couple pieces of 1.5" steel bar to make die holders from, and promptly broke my one remaining reciprocating saw blade.  Grr.  Oh well, take it to work on Monday and use the bandsaw.  Good thing I waited, after looking about later online, it looks like the #5 die I have is a pretty crappy example; most dies have at least 4 threading points, mine has three, and the lead-in chamfer is visibly off making the rod to be threaded want to walk out of the die.  Also, the outside diameters of my thread dies are odd sized (3 different o.d.'s for 4 dies), and they don't have adjusting screws.  Hmm.  Need to do some shopping for a good set of threading dies.

Finished the flywheel today at work (except for a setscrew, will finish that at home).  Pics #3 and 4 - both sides of the wheel.  You can see the difference between the two sides in the painted areas of the outer rim if you look closely.  I think it won't be noticeable, and the flywheel definitely looks better now (visibly "not wobbly" when turning) than it did when I started.

[zeeprogrammer]

"Grr" - a more polite way of saying 'crap' in my world. 

It would be tough to see both sides of the flywheel at the same time. Looks good to me.

[b.lindsey]

The flywheel came out very well. You have to be happy about that!!

Bill

[bent]

Yeah, happy with the flywheel.  Like Brian, I really liked how the cast iron took a polish with just a few passes with different grit sandpapers.  And found some brass chunks from an old valve stem that will suffice for the cylinder and caps, and got my steel for the die holders cut on the bandsaw.  So, Monday was a good day.  Sunday...well, that was a planning day and we'll leave it at that.  But if anybody know a good source for a quality small die set besides McMaster?

[bent]

Another evening at home, so got a bit more done.  Machined the cylinder, cutting fins with parting tool and then drilled and finish bored to 10mm. [pic #1] 

I had cut the pyrex test tube to length over the weekend with dremel diamond saw blade and drip can, same setup used as for the glass cylinder flame eater.  Getting pretty good at this, though the length of the test tube meant I had to make a couple of cuts, the test tube bumps against the dremel body on any cut over about an inch and a half.  So, a picture of the two parts, looks like it fits! [pic #2].

OD of my stock was on the thin side, so I tweaked the bolt circle inwards a bit.   Hopefully the #3 screws will still fit, if not I may turn some studs and nuts out of brass.  May take that approach just to get the odd lengths needed, and just to try it.  But dialing in the bolt circle was easy - zero on the center bore, then step off the x-y coordinates for the bolt pattern with the DRO's.  That is so slick.  Did also find out one way to make the DRO's read funky - step on the power cord and jostle the AC adapter, and hey, the readings are off.    Oh well, at least I've isolated a probable cause.  May need to get a power strip and fasten to the sides of the mill stand to keep things out of the way of my two left feet.  Pic #3 is finishing tapping the last hole on the cylinder side, then flipped and did the opposite end.  Now what to do next...hmm.



 

[bent]

More progress.  Bored the cylinder heads for #3 clearance holes, 0.950" bolt circle (clamped together in the machine vise, might as well screw up two parts at once, no? - Pic #1.

Then set up and drilled/tapped #3 holes for the cylinder pivots.  Aligned as best as possible by making marks with the mill head centered on the part, then rotating horizontally to give Mk 1 eyeball alignment. Pic #2

Then back to the rod side cylinder head - a 6-hole pattern of 0.040" holes to provide breathers for the backside of the piston.  Not much room for error, about .008" wall thickness on either side.  So, spot drill with smallest center drill (just the tip, after watching Stephen's video) [Pic #3] then carefully peck drilling, one hand on the drill feed and one resting atop the quill to "feel" the advancing drill.  [Pic #4] Shows the last hole just before drilling it. 

Then the piston.  [Pic #5] Made this one from some crappy cold-rolled hardware store bar stock.  It machines horribly, and this time, drilling out and tapping the centerline for #6-32 I swear I could feel all the piping (sulfides and other debris from the cast bar) trying to snap my drill and taps.  But got the skinny side finished, then bored the fat end, and mucked it up (.005" undersized relative to the cylinder bore) somehow.  Sigh.  Will have to redo it.  This time, with some nice clean stainless scavenged from the scrap bin at work.

Set down tools and spent Sunday morning (after Mum's day breakfast) cooking up a new batch of IPA instead of making more scrap metal.

More good news is I found a good local supplier (they will replace if defective) for some HSS dies, #2 thru #6 plus #8, about $5 to $7 each.  Should last awhile and all are one o.d. size, so should be able to make just one die holder for the lathe.  Also put some various fastenings, along with some hex stock, on order from McMaster. 

[Ian S C]

If for some daft reason you have to make piston #3, try and get some cast iron.
Ian S C

[bent]

Yup, will need to make another (3rd) attempt.

Next time with sharp taps (the old ones got really dull trying to tap the crappy steel, and one snapped in the stainless bar).  Have some 12L14 rod on the way.  But cast iron...hmm.  Good idea, thanks Ian!

[bent]

Strange.  Cast grey iron bars running about 2x to 3x higher than 12L14 steel bars at McMaster.  Perhaps due to difficulty shipping (brittle iron breaks more often)?

[bent]

A bit more progress this last week, between the youngest's band concerts and other outside activities.  Got the 12L14 bar stock, and whipped out a new piston, much better machining than the last go with low quality steel.  The piston fits snugly enough to fall slowly (5 or 6 seconds) out of the cylinder, with thumb over the end, so that's sorted.  No pictures, as I'd left the camera at work that day.

Also machined down the gears to fit in the assembly and to get a decent round section to place #4-40 set screws (Pic's 1 and 2).  The white (delrin?) gear machined easily, but other gear (black nylon?) kept fighting back, slipping in the lathe chuck.  I managed to get a 4mm hole bored through, so that I could stuff a bit of 4mm shafting in, and then it clamped a bit more securely, at least enough that with light (.010 in) cuts, I could finish it up. 

Then bolted the tilting vise down on the mill bed and machined the set screw hole (at about a 15 degree angle) for the flywheel (Pic 3).  Also noted a bit of surface rust on the exposed iron faces of the wheel, so spent a little time polishing that off with some 1500 grit wet/dry paper, then wiped some corrosion-inhibiting oil on the surface.  May need to think about waxing it, or clear coating?

Also got some #3 socket head cap screws for the cylinder caps.  The #3 tap I have is so long on the tip, that the threads on the finned end of the cylinder did not fully form before the end of the tap hit the next fin (couldn't get the screws to go all the way through the fin).  So, I re-cut the end of the tap, by chucking it in the lathe and then very slowly hand-grinding the tip with a diamond wheel in the Dremel tool.  Pic #4 shows the result, something much closer to a plug tap.  Also used the Dremel/diamond wheel to shorten one set of four screws to fit the skinny end of the cylinder, where the tapped holes go through the first cooling fin.  Also found during this assembly of the cylinder, that the silicone o-ring Jan uses to hold the glass test tube in place does not provide very secure grip on the tube - the glass tube can slide/rock within the machined sleeve and slowly works loose.  I will probably fix that problem by wrapping a bit of teflon tape around the test tube, to give a bit tighter fit between the cylinder cap and the glass.  Or, if that fails, use some epoxy to fill the gap, as Jan suggests in his build notes.

Waiting on a set of HSS thread dies from local supplier, should be in today, and can then make up a die holder for the lathe tailstock, which will help in making the the cylinder pivots with #3 threaded ends.  Also make up a piston guide bushing (planning to use a bit of the graphite rod I have for that), and then should be mostly done - cut some 4mm axles, and spot the set screws, grind flats...assemble and test.  Then fix the mistakes, re-assemble, re-test, tweak, find the camera, etc.

[Ian S C]

Don't laugh, I use old window weights for making cast iron pistons, you have to be a bit picky, the skinny ones tend to have a higher chance of being clilled. I have only had one failure, that was down to me trying to get the skirt too thin, and it broke off at one of the labrynth grooves(there were 4 on the skirt).
Ian S C

[bent]

Not laughing.  But I haven't seen sash weights since I left my Mom & Dad's home some number of years ago, everything uses multi-pane aluminum-frame sliding windows now it seems.  That said, my sister just finished renovating the old family farmhouse to put it up for sale, can't remember if she said they replaced some windows?  If so, did she happen to keep the weights?  Hmm.

[bent]

Hmm.  Still like your cast iron idea, though Ian, but finding some at reasonable prices.  I think I need to keep my eyes open for returned valves.  When I first started at my current job, all of the valve cases were cast grey iron in big, heavy sections (my first designs were converting them to ductile iron).  If one of those comes back thru, it may need to make a trip past the bandsaw for, um, engineering samples.   

[Tennessee Whiskey]

If you live anywhere near an agricultural area, another good source of cast iron is old tractor weights. The ones in usable condition usually bring a $1/lb and they are usually 50 pounds and up, however, the ones with the attaching structure damaged, bring a good bit less. The grain structure is usually very good and at these weights, chilling isn’t usually an issue. Just a top tip from here in Springbucket.

Cletus

[Ian S C]

In NZ we have what are known here as Demo yards where demolition firms sell the remains of demolished buildings, and thats the place to look.  My last lot had been used as counter weights on sliding doors in my brother in laws' piggery.  The trouble with cast iron is that it is cast to shape using the minimum amount of metal and its some times hard to hang on to to cut a bit out the right shape and size.
Ian S C

[bent]

If I stick with this much longer (there are so many engine designs!) I will just have to suck it up and buy some bar stock.  Or use up all my aluminum swarf for mixing thermite.

[bent]

Whittled out my die holders for the lathe tailstock.  Machined a bit of 1117 steel to 1.25" od, then bored and tapped 1/2-20 (to fit the thread on the end of the MT-2 part I got from McMaster), and cut a 13/16" counterbore to fit the dies. (Pics 1 and 2).  Drilled and tapped a set of #6 threaded holes to hold the dies in the pocket, using a clamping v-block from work to index the piece 180 degrees to match the divots on the dies.

Once that was done, made the two pivot screws (shoulder screws) that hold the cylinder in the support fork and allow it to "wobble".  Pic #3 shows the part turned from a piece of 1/4" 12L14 steel hex stock, going dead slow and shallow cuts on the lathe to keep from bending the part and inadvertently making a taper on the skinny end.  Brian used a cylindrical-headed screw with a standard flat screwdriver slot on his drawings, but a hex screw seemed easier to make and looks a bit better in my opinion.  Then used the new die holder to thread the end to #3-48 (Pic #4), by sliding the tailstock on the ways and turning the lathe chuck by hand.  Finally, used the parting tool to create a thread relief where the die taper runs out (Pic 5), before parting the screw off and finishing the head with a flat file and sandpaper.  Very happy with how these turned out, though I need to go back and re-tap the holes in the cylinder to full depth now that I have re-ground the tip of my #3 tap.

Almost done with this one, just need to make a crank and the piston guide bushing, then assemble and test.

[Brian Rupnow]

Carl--when I built the popcorn engine, I redrew everything Stu Hart had done, basing my work in inches instead of millimeters. Although the crosshead guide looks formidable to machine, it is mostly straight lathe work. If you look at this drawing you will see three places that have a 0.313" radius. That is where I used the 5/8" ball nose endmill. also there is one large radius of 0.787".--For that I used  a boring head in my mill. Hope this helps.---Brian

[bent]

That's a fun looking part, Brian.  Dunno why it's posted here, but no skin off my nose.

Finished up most of the bits yesterday, and have started to test assemble to see what needs tweaking/cutting/hammering/filing etc.

Pic #1 - the crank I whittled out from a piece of 1.5" bar (the only stock large enough in my bin) to get a 5/8" diameter, and then parted off the 1/8" thickness needed (had to finish the cut with a hacksaw, my parting tool blade is now dulled beyond use and needs to go to work with me to visit the bench grinder).  Reversed the part and cleaned up the back side, bored and reamed for 4mm shaft (Pic 2).  Tapped #4-40 for the shoulder screw I bought from McMaster (Pic 3), then tig welded to a bit of 4mm shafting I bought (Pic 4 shows the part clamped ready to weld).

The graphite piston guide bushing also had to come from a larger bit of graphite rod.  This time I took Joco's advice and used SWMBO's spare vacuum that'd not been gathering dust (haha) in the back of the closet to suck up all the graphite dust (or at least a pretty fair proportion of it).  You can see the nozzle held off the end of the toolpost in Pic 5. Pic #6 is the part just before parting off - one last check of the fit to the piston, a nice .002" or so running clearance.  Coated the o.d. of the bush with some cyanoacrylate glue, and pressed it home in the rod-end cylinder cap.

Now off to scrounge the garage and find my .050" allen wrench.  I know it's around here somewhere...just not back in the toolbox drawer.  I can't count the number of times I've berated the boys to do that, you'd think I'd heed my own dictums. 

Hope you all have a good holiday, I'm going to be grilling some kebabs later according to SWMBO, along with some corn on the cob.  Oh, and gotta fill the keg with the IPA that's finished fermenting, and make sure there's some bottles of store-bought in the fridge.  Priorities...I think I'll see to the beer first if you don't mind!

[bent]

Well, finished assembly of the engine last night, and went to look for some alcohol to run it, but I'd forgotten I'd used up the last of it (and no, I'm not gonna waste good scotch for a test) 

But, everything fits and spins pretty good, with a bit of tweaking to get everything lined up ok.  I think if I do another build of Jan's engines I will try to make the bearing supports from monolithic blocks machined out, so the bearing bores can be done in one pass and are fixed/aligned at the point of machining.  Trying to manually tweak the bolted on bearing supports to eliminate any binding is a pain.

Also, I need to remember not to buy (never ever again) shafting material from Amazon.  None of the 4mm shafting I got was straight , and the flywheel (so carefully machined, and such a nice snug fit on the shaft, has a noticeable wobble as a result. 

Hopefully I will get it running tonight and post a YouTube video (fingers crossed).

[bent]

Hmm.  No joy.  Suspect I have too much leakage, either past the piston or at the cylinder head.  Will try sealing the latter tomorrow with some silicone, then possibly make a 4th piston with a tighter fit to the cylinder.  Past that, dunno.  Maybe revert to the direct drive that Jan's original engine used, thinking perhaps there is too much friction and windage with the gear reduction and higher flywheel speed?    

[Ian S C]

The piston is the thing. You have to get it to just slip though the cylinder under it's own weight, but virtually stop if you block off the end of the cylinder. If the piston is a wee bit tight, it slides but not quite loose enough, sometimes it's best to run it in by driving the motor from an external power supply, these days I use another Stirling Engine, but a small electric motor is better. If you use a small motor, you'll notice the speed pick up as the piston, and displacer gland frees up.  At the running in stage, and as long as the piston is NOT graphite, a drop or two of oil is OK.  Graphite, polish it down with paper.
Ian S C

[bent]

Thanks for the encouragement, Ian.

[bent]

Well, whittled out a 4th and 5th piston, one with .0005" clearance, the 5th with zero measured clearance and a very tight fit to the cylinder (subsequently run-in by hand).  And it still doesn't run.  Can only conclude there is too much friction in the gear train.  Will rebuild it without the gears, when I'm ready.  For now I am going to set it aside, and start working on the PMR #5 kit I have.
 

[Ian S C]

The gears you have seem to have a fairly high ratio, I wouldn't  go much above 2:1, from the diagram on P1 it looks nearer to 4:1. Your current flywheel would be quite adequate  for direct running. The advantage of gearing up the flywheel is that you can reduce it's size, I'v forgotten the maths that go with this, did well in Mechanics at school in the 1960s, but it seems to have gone now.
Ian S C

[MJM460]

Hi bent,

Following on from Ian's comment, the kinetic energy stored in a flywheel is proportional to the rotational speed squared.  So that four to one gear ratio is like putting a sixteen times larger moment of inertia.  I agree with Ian, the gear ratio is probably too high, as the original design is almost certainly not that far out.

I am not clear why you felt it was needed at all as your early posts imply that it was a decision made before you started building.  Was it in the basis of a U-tube video perhaps?

The simplest thing is to try the flywheel direct driven if the shaft will fit.  Or another flywheel you may be able to borrow.  Or I will look up the topic in my Talking Thermodynamics thread the post number where I discussed using rectangular bars to get an idea of the required moment of inertia.  Gives you a good basis for determining the flywheel moment of inertia required if the air turbulence does not result in too much resistance.

I have to admit to saying the flywheel could not be too big, providing the friction did not exceed the engine capacity, but you seemed to have produced a counter argument.  I assume it would not keep going if you have it a spin to start, so perhaps friction is the issue.

That little engine in common with most other Stirling engines is only single acting, at least when it first starts.  So the power pulse has to give the flywheel enough energy to take the piston and displacer back the other direction.  This is where the flywheel size would be quite critical.  It has to be big enough to run the engine to the start of the next pulse, but not so big that the pulse does not turn it that critical half turn.  I am suspicious that after a few minutes of running, enough air leaks out the displacer rod gland and past the piston, so that the pressure fluctuates above and below atmospheric, and the engine gradually becomes double acting, and so probably a lot smoother running.  But until then, the engine will not overcome much friction.

MJM460

[bent]

Hi MJM.  Yes, it was a decision to try and get a smoother motion than Jan's original machine.  Probably not a good idea, as the apparent "hesitation" in the motion is likely the reason it runs - it takes time for the gas to cool/heat.  In the video of the gear-train version, the motion is still fairly variable in rpm at the flywheel, but he is also apparently using a much larger piston bore and a longer cylinder to boot...have to scratch my head a bit. 

Friction in the gear train, though, is probably the biggest factor in the failure.  Was gone last week on a trip to NYC for #4 son's award ceremony; will probably get back on this one soon.  Patching up the holes in the base plate is one of my concerns, and has me wishing I'd paid more money for an AC tig welder...oh well, nothing that a shot or two of JB weld won't fix. 

[bent]

Ok, rebuilt the engine over the last week, milling down the larger flywheels support that held the gear train, and shifting the mounting holes.  Managed to get everything squeezed up tight enough to re-use the existing crankshaft.  Assembled it all up and spent some time tweaking the alignment to get rid of binding issues.  Also had to replace the marbles - I used the ones that Jan suggested (slingshot ammo) but after the exposure to heat, one cracked and the other two showed signs of soaking up some of the graphite I had used to lubricate the piston during running in.  So, I cleaned the cylinder and glass tube as well as possible with alcohol and some cotton swabs, getting rid of as much of the graphite as I could see.  A couple of assembled engine pics (from two sides) shown below.

And...she runs! 

<a href="https://www.youtube.com/watch?v=8i_aQ742S_g" target="_blank">http://www.youtube.com/watch?v=8i_aQ742S_g</a>

[Admiral_dk]

Great result and you must be very happy that you turned it into a runner 

[bent]

Thanks.  Pretty pleased with the result, after the previous experiment flopped.