Joy's Valve Gear Horizontal Engine

[Dave Otto]

Thanks for the link to the video.

I'm not really a steam guy but that was really cool!

Dave

[kvom]

After preparing the stock for the conrods, I confirmed my suspicion that even the 12" vise jaws are too short to hold it the same was as for the valve link arms.  I need to buy some steel for a fixture, so in the meantime I started on the main bearings this afternoon.

Since 660 bearing bronze seems to be available only as round rods, I bought a foot of 1.5" diameter rod in  order to be able to mill the .75x1.25" cross section.  Only in hindsight did I realize that each half of a split bearing is .75x.675, so a smaller rod would have worked.  Duh.  Anyway, cut of 3" of the stock and faced each end on the lathe.



Then rough squared it using y-axis of the mill;  with soft metal this is the fastest as I can cut each side in one pass.



Scribed to saw in 4 pieces:



This is a technique I came up with to cut small pieces on the vertical bandsaw:



Further milling yielded these 4 pieces:



Then glued together using 3M high-strength adhesive.



I know it's common to soft solder bearing halves for boring the center hole, esp. using a 4-jaw lathe chuck, but I will drill and ream the .75" center hole on the mill.  With the halves clamped in the vise, the glue is mainly a means of keeping each set separate.  After the holes are done, I'll wait for the conrods to fit the bearings.

[kvom]

Quick bit of CNC milling today.  Used some of the 1.5" bronze rod for the first ops on the valve packing glands, as shown here:



Then chucked in the lathe by the round 1/2" boss and turned to final thickness.  Results shown with the split bushings; here I interpolated the holes on the CNC mill and reamed .75".



I was planning to drill the holes for the valve rods in situ after attaching to the steam chest, but that's probably overkill.

[kvom]

Instead of actually cutting metal, the past two days have been spent in learning basic Solidworks, since I was able to get a free student edition version as a US veteran.

Here's my first cut at modeling parts and an assembly of the conrod, bearing, and the two valve links:

[kvom]

Finally got to cutting metal on the first conrod, using the same method for holding the round stock as for the valve link rods.  "Fixture" starts as a length of 1/4" HRS to which I bolted a pair of parallels to one edge.  The collet block and its partner are held against the parallels with "Mighty-Bite" clamps, and both ends then clamped vertically.  Here's the workpiece on the fixture after the first ops:



Rotate the setup 90 degrees and reclamp, then further ops:



Finally flip 180 to do the details on the inverse side.



The surface finish isn't as nice as I'd hoped coming off the mill, so some hand work will be needed after the rest of the milling.

[kvom]

The conrods have been a "struggle" given the size of the parts, their odd size, and the number of ops needed.  Here's a week's worth of effort.

After machine the four profiles, I machines some soft jaws to hold the shaft:



Then I could do the machining ops on each end.





Only a few ops to go:



On the first one I made a mistake, programming the bottom profile as a .375 endmill but using a .5" endmill.  The result is still functional but is not symmetric.  Given the amount of work to get this far I'm going to finish it, reserving the option to remake in the future.  I've learned enough making these two that the third would take considerably less time, though still not trivial.

[maury]

Kvom, really nice work on a complex part. I was wondering if you could elaborate a bit on the programming part of making this part, and did you rough it out manually first. I can see me building one of these engines in the not too distant future.

I wonder if a casting would be a better rout to take with this part, ductile iron should be a good material for it.

maury

[kvom]

I'd decided to use stress-proof (1144) steel to avoid warping, and that was available  only in round bars.  I ordered a 3' piece for about $65 from McMaster and sawed it into 3 1' pieces.  Made these two and crapped out another.

The first ops were just profiles 2 sides in one orientation, then rotate 90 degrees and profile the other 2 sides.  I found that the best way to hold the stock was not using the fixture with the collet block, but rather to clamp in the Y direction between the ends of the 12" vise jaws extending out from the 6" Kurt vise.  My mill doesn't have quite enough travel to machine the entire length, so I finished the crank end with later ops.

On the last one I did the profiles cutting full depth with a small .025" stepover.  This gives a better finish, and also eliminates the chance of the workpiece rotating in the jaws.  I used a 1/2" 4-flute carbide cutter with 2" flute-length (stock diameter is 1.75).

After the side profiles are cut, then you need the following ops using the soft jaws to hold:

1) drill and ream the center hole
2) drill and ream the hole(s) in the crosshead end.
3) Machine the round bosses in the crosshead end.
4) Remove the extra material from each end and open the crank end.  I did rouging with a 1/2" EM and finish pass with 5/16.
5) Round the rest of the crosshead end
6) Mill the V shape in the crank end, each side

The last ops, which remain for me to do, is to open up the crosshead end.  For this I'll just clamp the shaft between two pieces of steel in the vise (the shaft is slightly less than 6", so can't use the vise jaws alone.  Then use a large square to make sure it's parallel, and obtain the proper X zero using the crank end.  I also need to drill the holes in the crank end for the keeper screw.  Eventually i'll need oil holes for the split bearings once they're fitted.

I'm not sure a casting would be much easier in the long run, but would present it's own and different challenges.  I looked at having one cast in aluminum at Shapeways using the STL I drew up of the conrod, and each one would cost over $300.

I could also have gotten some square bar HRS that would have likely worked OK for a bit cheaper price.

I should note that my mill's configuration prevents me from clamping the stock to the table directly, hence the need to use the 6" vise.  Using the fixture with the square collet block has some problems, mainly clamping the block securely enough.  I had it pull loose a couple of times resulting in the 2nd try being ruined.

Using the round stock with a 4th axis/chuck and tailstock would be an even better solution as almost the entire piece could be machined without removing the work once attached.  Only finishing the end openings would need vise work.

[smfr]

Kvom, really nice work on a complex part. I was wondering if you could elaborate a bit on the programming part of making this part, and did you rough it out manually first. I can see me building one of these engines in the not too distant future.

I wonder if a casting would be a better rout to take with this part, ductile iron should be a good material for it.

Maury, in my Muncaster build I made the conrods on a manual machine from 1144SP without too much trouble. Details in my build log here:
http://www.modelenginemaker.com/index.php/topic,2470.msg53710.html#msg53710

Simon

[maury]

Kvom, thanks for the detail. I haven't made a large part from steel on my CNC, so I expect this will be one of my first. I take it you only used 2D  tool paths.

Simon, thanks for the link. I poked through quickly and will get back to study it when I have more time. I see you have quite a bit of detail there. It's going to be a big help. I feel more comfortable hogging big sections of metal off using my Bridgeport than with my CNC, not sure which way I will go when the time comes.

maury

[kvom]

Everything was 2D toolpaths.

I considered doing some roughing on the Bridgeport, but in reality you don't gain much if you're cutting a full length profile.  The metal removed in roughing means the CNC mill is cutting air anyway, so machining time is not saved.  Using full depth and a .025 stepover on a .5" endmill at a conservative 18ipm feed, each profile takes less than 20 minutes.  If I had a piece of stock currently I'd likely make a new one using the experience I've gained.

These were the biggest parts I'd done on the CNC mill.  I faced each end of the stock on my lathe first, and then used the  crosshead end face to establish the logitudinal zero for each op until the end, when the faced material was removed.  Afterwards I'd use the big end. My mill's limit switches have been disabled so that I can't reference the axes between power ups, so I either leave the spindle in a known position after use, or else re-zero after startup.

[kvom]

As exercises in learning Solidworks I've modeled a good number of the parts.  It occurred to me that this engine could be 3D printed in rigid plastic with some sliding surfaces added in metal or Teflon, and it would likely run quite well on air.

[kvom]

Some slow progress the past two days:  yesterday finished the machining ops on both conrods, and today fitted one of the split bearings to its rod.



The bearing machining was manual and was lots of "whittling" away at the bronze until it fit.  I measured the thickness of the big end and calculated the offsets from center to cut the top and bottom slots.  Using a 3/8" endmill with an estimate of +/- .110", I found that I got a nice sliding fit at +/- .114", so all the rest of the slots were cut at that width.  First task was to cut the top and bottom slots to depth equally so that the bearing could enter the conrod opening.  Then I split the bearing halves and did the front and back slots individually.  The far end slot is cut at a 5-degree angle from vertical to allow the tapered keeper to widge the bearing against the crank.

As assembled initially, it's very stiff to turn.  I need to relieve the main bearing.  It appears that the conrod is slightly off from being straight front to back, as the big end is off center in the crank opening.  This could either be warpage in machining, or else having the holes for the crosshead pin being slightly off from perpendicular.  However, there's enough flex in the rod to allow it to be assembled to the bearing and crank as shown in the picture.

I still need to drill 1/16" holes for oiling.

[Roger B]

That's some complicated machining on the con rods    Looking good 

[kvom]

Got back to the engine a bit this afternoon by machining the end journals of the crankshaft (slots for the gib keys) and then loctiting them to the rest of the crank.  I also turned the conrod upside down, which made it line up much better.  Using the flywheels to turn the crank, the entire drive of crankshaft, conrod, crosshead, and piston operate quite smoothly for a first cut without lubrication.