The One Valve Engine
Page Two, Fixing The Valve Train
Rush back HOME
22 May 2011:
Over the last few months, I've been thinking (I know, it's a dangerous thing for me to do!) about my next project and have decided to do something sorta original.
When studying antique engines (most of which I can't afford to own), I've been struck by the pains some inventors/designers went to in order to have something different or novel or to get around someone else's patent. Now "different" doesn't necessarily mean "better" or "simpler".
Take for example the McVicker Patent Engine with it's exhaust pressure operated valve. Then there are the various cam-stopper sideshaft engines. Add to that, engines like the Edwards with it's retro (even for it's time) oddball valving and mixer. Then there are the gearless and camless engines, some of which are more complicated than geared engines. There are some really unusual carburetion and ignition arrangements and some of them even work!
With these engines in mind, I decided to design one that is four cycle but has only one valve. This idea uses no ports but has a single poppet valve that does both intake and exhaust functions. Then, I make the valve operation gearless with an unusual method of operation.
Here are some basic specifications:
Bore: 1.5" (38.1 mm)
Stroke: 2.5 Inches (63.5 mm)
Displacement: 4.42 Cubic Inches (72.43 cc)
Compression Ratio: 4:1
Governing: Hit and miss with flyball governor.
Power Output: Who knows?
RPM: Won't know 'til it's done.
Ignition: High Tension Jump Spark.
Fuel: Gasoline By Direct Injection.
Rod And Main Bearings: Sealed Needle.
Everything but the valve, valve spring, bearings, wrist, crankpin and bolts will be made from scratch.
I've still got a couple of days of design work to do and I have as yet not decided on the actual layout of the injection pump and pintle but the drawing is coming along nicely.
24 May 2011:
I'm still at the CAD when time permits. I left out a feature in the last posting. I'm going to try using PTFE (Teflon) "0" rings for the piston. It's designed so I can re-machine the grooves for cast iron rings if the Teflon doesn't work out.
Today, I've been figuring out the fuel injection pump and about have it.
A really confusing rendition of the engine timing side.
If you work at it, you may be able to figure out what I'm up to here. Note that all of the parts are X-ray views so you have to use colors to separate the parts.
The injection pump is the yellow thingy above the cam/ratchet. So far, I haven't got the pump valves drawn, nor do I have the bleed-off port drawn.
Since I needed a detent for the cam when the governor holds the valve follower up (this is so it can go through the cycles without the valve working), I simply put another follower at 90 degrees from the valve follower. The second (vertical) follower is used to disable the fuel pump during latched cycles and when the exhaust valve is open. The upper follower pushes a pin up to swing the fuel pump pushrod up and out of the way of the pump plunger on the strokes when it is not needed. It is shown in the valve-closed and operating position with the pump pushrod riding on the pin as shown.
I'm using the funky curved leaf spring to apply constant pressure on the detent/fuel follower so it will hold the cam still when the governor is holding the valve follower away from the cam.
I'm not sure when I'll start whittling parts for it but, before I do that, I've got to get a parts list ready and go "junking" for steel then pay a visit to my friendly machine shop with the water-jet outfit to blank the frame parts, crankshaft cheeks, etc.
29 May 2011:
One has to start a job somewhere so, yesterday and today, I made a fixture to machine spheres. The spheres will be the governor weights.
Parts of the fixture. Fixture in place. Machining first hemisphere.
The theory of the fixture is that you can make a sphere by pivoting a tool on the equator of the sphere. The distance from the center of the pivot to the tool tip is the radius of the finished sphere.
Machining pin in place. Starting second hemisphere. Finishing second hemisphere.
The way I made these balls was to first drill and tap a shallow 5/16-18 hole in one end of the 1" diameter bar stock. Then, after forming the first hemisphere, the bar is removed from the chuck and a threaded 5/16" rod is screwed into the hole. The work-piece is then held in the chuck by the rod and the other hemisphere is formed. It's not necessary to have the bar stock cut to exactly the diameter of the sphere. A little more length allows for any slight miscalculations. You simply keep moving the tool toward the chuck in increments until the ball is round.
After finishing the aluminum "test" or "high speed" ball, I cut a couple of lengths of 1" diameter brass bar stock and made the governor weights. The drilled and tapped holes will be used for the arms of the governor.
The aluminum ball. The two brass balls and the aluminum ball.
A little help from some emery paper and polishing compound and we have a nice set of low-speed governor balls. If I need to run the engine faster, I can always make another aluminum ball and use that pair for faster operation.
I've been working on the fuel injection arrangement. One of the complications of the fuel injection pump is tha, although the piston moves at each revolution of the engine, it must only operate on the intake stroke. I think I've got it figured out.
Drawing showing new iteration of fuel pump.
The 24 May drawing shows a previous design that I didn't like. The newer design shown above uses the second (vertical) cam follower (used as a detent) to also release a spring loaded bypass check valve in the fuel pump on the intake stroke (engine valve open).
The bypass valve arm (shown in yellow) has a flat spring at it's end which contacts the detent lifter. When the engine valve is open, the vertical lifter will be in the up position, forcing the bypass valve arm to rotate counterclockwise, releasing the holding pressure on the bypass valve. In this position, the bypass valve acts like a low pressure spring loaded check valve. When the engine valve is closed (as shown), the return spring for the detent lifter presses the bypass valve arm spring down, holding the bypass valve securely closed.
Since I'm thinking of using a poppet valve pintle for the injector, I can set the cracking pressure of the pintle at a higher pressure than that of the bypass check valve when it is in the bypass position. When the engine valve is closed, the bypass valve in the pump is held hard closed by the stiff flat spring attached to the bypass valve arm and thus allowing the pintle to do it's thing. At least, that's the theory.
31 May 2011:
Something just seemed wrong with the drawing as seen in the 29 May posting. Last night, I figured it out. The fuel injection pump was pumping every other stroke whether or not the governor was latched. I fixed this little problem by adding to the latch rod.
Newest Governor Latch Details.
You can click on the image above to see more detail. What I've added is a vertical bar that connects to the lifter end of the Governor Latch. When the engine is up to speed, the end of the Governor Latch that is at the Engine Valve Lifter end of the latch moves away from you. The end at the Engine Valve Lifter latches against the lifter body to hold the engine valve open. At the same time, the vertical arm of the Governor Latch moves underneath the screw on the Pump Bypass Latch Arm, keeping it from rotating clockwise in order to press the Bypass Valve Ball against it's seat so it can move fuel to the injector during the intake stroke. Now, I think I can safely say that, barring any other goof-ups, the engine is almost ready for dimensioning. Yeah, yeah - I didn't get all of the hidden lines removed before I posted the drawing but I think you get the idea.
The ignition control will probably be something simple like a magnet on the crankshaft that operates in conjunction with a grounding contact on one of the lifters. If the grounding contact is closed (valve open) ignition is disabled. This will both keep from having a wasted spark at the end of the exhaust stroke and will be a "spark saver" for when the governor is latched.
1 June 2011:
The design is getting there and today, the valve was whittled out of a bigger valve.
I went to the local landfill that I was told would let me root through the pile of junk lawnmower engines and pick some pieces. I was after valves and keepers but, when I got there, I was told that they discontinued that practice so I was outta luck.
On the way back, I dropped by the local rental, Snapper dealer and lawnmower service shop and the mechanic gave me a couple of used valves, one of which I "whittled down" to the size I need for the engine. That little chore took about all afternoon!
Tomorrow, if I can find the steel (or cast iron), I'll start on the valve cage.
2 June 2011:
The cage is done and the valve is seated.
Valve cage assembly.
I made the valve cage out of steel. Although cast iron would have been ideal, I didn't have any to whittle on. It should work okay as-is.
Tomorrow, I'll start on what I call the port. The port bolts to the side of the head and the valve cage mounts in it.
4 June 2011:
Well, I've gotta admit that yesterday, I goofed. I had a nice piece of steel and was into making the port assembly about three hours when I discovered that I'd shaved about 0.100" too much off of one dimension. It's now residing in the junk box again.
Today, since I'm fresh out of any steel pieces that are big enough to make the port, I decided to see how it will work made in aluminum. On my last scrapyard visit, I found a 5" piece of 2-1/2" diameter aluminum round bar stock. It had been run over by trucks and generally was pretty warty looking but, I figured I'd need it sometime. Today was "sometime". Since it machined nicely, it must be something like 6061-T1 alloy.
Starting with the raw aluminum. The "port" finished.
I had some difficulty visualizing the port because two of the dimensions were very close to one another. It helped when I drew the features on the block with a felt pen so I could keep track of where everything goes.
Here's the port with the valve cage mounted.
It'll be interesting to see how the aluminum holds up to the heat of the engine. I think it could be okay because, after the exhaust finishes flowing through it, cool air will be immediately sucked back in through the same path. The oval passageway is the connecting passage to the combustion chamber and I may mount the injector so it sprays in this passage. Gas flow will be relatively swift in this location and it may help atomize the fuel.
11 June 2011:
Wow! I didn't know that a week has passed. I must be having too much fun. I have been doing some work on this project, up until today, I've been dimensioning the drawing and making up material lists among other things.
Since I had the material handy, I went ahead and got a good start on the eccentric assembly.
Drawing of the eccentric, eccentric ring, eccentric hub and pawl.
I had a piece of 1/4" thick brass from the scrap yard. I cut it to shape and plan to anneal it and bend it around the eccentric ring. It'll be interesting to see just how it fits after bending. If the brass is really soft after annealing, I may just re-harden it.
I just noticed that the program I use to convert the ACAD .DWG files to JPEG files doesn't know how to convert the "diameter" symbol (an 'o' with a line diagonally through it). It substitutes a question mark.
Raw material for eccentric ring. Ring cut-out. Eccentric blank.
The eccentric itself is made from a water-jet cut out that was left over from The Non-McVickerish Engine. The hardest part was machining the groove for the eccentric ring to run in. The 1/2-13 threaded hole in the center of the eccentric was so I could machine the O.D. and the groove. I suppose I could jam a piece of 1/2-13 all-thread into the hole but it would still show, so why bother.
Here's the product of today's work.
12 June 2011:
It started out okay but soon degenerated into an "Aw-Shoot". I put a Mapp torch to the brass piece and got it as hot as I could then quenched it in water. I thought it would be annealed enough to bend easily. Not so.
Bending the brass eccentric ring. AW-SHOOT!!!!
I had the sneaky suspicion that the brass was going to be hard to bend around the eccentric. I -almost- got it around the eccentric before it broke at a stress point.
I started over doing what I probably should have done originally. I got a piece of pipe that I could whittle into the proper shape to make the eccentric ring and worked it over.
Starting over with a scrap of steel pipe. Here's the assembled eccentric.
I turned the O.D. and the I.D., then I cut it off longer than needed. I made a couple of ears out of 1/4" steel. One of the ears, I milled to half thickness for joining to the pawl. The other ear was drilled through with a #4-40 tap drill then re-drilled for 4-40 clearance half way through. After tapping the hole, the ears were welded to the ring then the works was put back into the lathe, cleaned-up and faced to it's finished thickness.
The ear with the screw hole in it was then saw cut and a notch was filed in the end of the notch for a spreader. The ring was then assembled on the eccentric and lapping compound was applied and the screw tightened incrementally until I had a good bearing surface on both pieces. I'm not sure how well the eccentric and ring will wear due to the similar metals but, since there's a pretty good bearing area and there will be no load during most of the rotation, it should work out.
If it wears poorly, one fix would be to machine the I.D. of the ring about 0.031" larger and make a 0.015" thick brass strip 1/4" wide to fit between the ring and the eccentric. I don't think it would hurt anything to just let it float in there.
13 June 2011:
Today was the junking and water jet cutting day. I decided to try two scrap yards. The first one, which had been promising about a year ago, had moved all of the good stuff out so I moved to scrap yard number two. The second yard came through.
My list included 1/8", 1/4", 3/4" and 1" hot rolled plate. After climbing a couple of piles I had it all. I suppose I could have picked a cooler day but seem to have forgotten how hot a junk yard can be on a nice warm spring day in Florida.
Anyway, I took my CAD disk and hauled my materials to the shop were I get the water jet cutting done. This time, I remembered to take my camera.
Cutting the 1/4" steel base plate. Cutting the 3/4" steel main bearings. Cutting the 1" steel crankshaft cheeks.
The water jet cutting in progress.
I'm still amazed at how steel and even ceramics can be cut with a simple stream of water with a little garnet powder in it. Of course, the water is under about 50,000 pounds per square inch pressure! The input to the machine is a computer CAD file of the parts.
The man who ran the machine, Jason Scott ([email protected]) made programming the part look simple after I made files of parts drawings for each thickness of steel. Jason can make single plane parts of iron or steel up to six feet by eight feet by four inches thick! The accuracy is very precise. In fact, he says he can make large spur gears that have teeth accurate enough to require no machining if the gear is to run slowly. That means the things like pinion (and bull) gears can be made with this machine without the requirement of patterns to be made! Any finish work (shaft bores, tooth polishing, etc.) can be done conventionally.
It seems to me that water jet cutting is a natural for parts that ordinarily have to be cast, like gears for tractors and traction engines. This is a blatant plug but, if you are in need of something like this, Jason can make it for you.
What happens when you get too close. The finished parts.
When I was photographing the process, I gave the crew a good laugh when the jet got to the end of a cut and sprayed the area. Luckily, the camera survived the bath! Anyway, Jason wants to prove how versatile the machine is and, when I send him a CAD drawing of a one-piece connecting rod, he'll make it out of left-over steel.
Today was a good, although HOT, day!
15 June 2011:
Yesterday and today, I worked on the head. I had a piece of malleable iron rod that was 5-3/4" in diameter so I cut off a piece, faced it to length and turned the diameter to 3.8". A lot of chip making but I had the material around.
Combustion chamber side of head showing port. Port side of head.
Here's the head with the port and valve cage mounted.
On Monday, my rotary head arrived so I had my first chance to use it to layout the head bolt holes.
Since I now have a critical mass of parts (my McMaster order arrived today), I'll probably start on the crankshaft tomorrow.
16 June 2011:
The crankshaft is coming along.
Turning Crank cheek O.D.
Working big stuff with little machinery can be done, it just takes time. Note the watch-making tool. It just fits the holder. That's the only way I could turn that diameter. The work-piece will not fit over the carriage so the big tool was needed in order to reach over far enough. It did a good job when taking 0.005" cuts at a slow feed rate.
17 June 2011:
The crank is finished, awaiting the rod so it can be assembled.
Rod bearing and thrusts on pin. Pulley end main bearing and thrust on pin.
The rod pin in the left photo is pressed in place. The shaft sticking out of the main pin bore is a slip fit and will be used during the pressing of the other end of the rod pin into the pulley side cheek to assure that the cheeks are properly aligned, The pulley end main shaft in the right photo is just tapped partway into place and will be removed prior to pressing the other end of the rod pin into the pulley end cheek..
The way I'll assemble it will be to assemble the rod bearing into the rod and stack it and the thrusts on the crankpin. Then, I'll lay the governor end cheek on the guide pin and start the rod pin. It'll then go into the press to finish pressing the cheek onto the rod pin.
I'll then knock out the alignment pin and measure to make a steel spacer exactly the distance the cheeks are apart. Placing the crankshaft in the press again, I'll insert the spacer then press one and then the other main pins into place.
19 June 2011:
Yesterday, I did some dimensioning and got ready to start on sizing the frame pieces and main bearing blocks.
Today, I got started on the mains and frame.
Sizing main bearing height. Side plates laid-\u\p for dimension checking.
I squared-up the main bearing blocks then drilled them full depth with a 1/4-28 tap drill. Then, I re-drilled the holes from the top to a little past what would be the parting line with a 17/64" clearance drill. The caps were then sawed along the parting line and the bases and caps were milled clean at the parting line. A 1/4-28 tap was run down the holes in the bases. One of the last operations on the mains will be finalizing the location of the bores and boring them for the roller bearings after welding them to the sides.
I then put the two side plates in the mill and machined the main bearing line and the bottom line of both sides (clamped together to assure that they were the same dimensions) to make the crankshaft height to spec. Next time, I will finish up the dimensions of the side plates then will do the front plate where the cylinder will bolt on and the small back plate.
Once everything is properly sized and the main bearings welded on, I'll drill and tap the side plates and bolt them together and to the bottom plate with 10/32 socket head screws. There is one vertical joint that will have to be welded. I'm still not sure if I'm going to weld the rest of it together. On the Non-McVickerish engine, there was significant warpage from the welding that I had to machine away.
21 June 2011:
I picked up the rod this morning and the only machining I had to do was to bore both ends and deburr.
Rod, straight from cutting. Bearings in place. Pinned and ready to press crankpin.
Since the middle part of the rod is very straight, all I had to do to prepare to finish the ends was to clamp the rod in the mill vise and find the center of the wrist pin end. Once the wrist pin bore (5/8") was done, I simply had to step the mill table six inches and then bore the big end. Oh, yes - before you make any snide remarks, The note "BIG" on the left-hand photo notates the large diameter ends of the bores, NOT the big end of the rod.
The rod was assembled to the crankpin with the thrust washers and the pulley side cheek was slipped over the temporary main pin. Just to be sure it was truly lined-up, I knocked a 1/2" steel rod between the cheeks in one of the 1/2" balance holes. Snugly pinned on both sides of the crankpin, it was ready for the press.
Pressing the crankpin. Pressing the mains.
The crankpin was TIGHT, just about all the press could do. The main pins were easier. As you can see in the right-hand photo, above, I made a 3/4" spacer out of an aluminum end to keep from springing the crankshaft as I pressed the mains into place.
After the greased main thrust washers were put in place, the shaft and the hardened sleeves were cleaned with lacquer thinner and Loktite was applied to the shaft then the sleeves were slipped into place. Since I may have to someday remove the sleeves, I used the thread-locker type of Loktite, which isn't quite so tenacious as the bearing set type.
Now that the crankshaft is finished, I can work on getting the frame assembled and sized.
24 June 2011:
Today, I worked on the cylinder and headers.
Headers, bored and ready for spot drilling. Using rotary table to locate drill spots.
The frame header is bored for a snug slip fit of the projecting 3/8" of the cylinder to assure it is aligned with the crankshaft. The headers are turned for a press fit to the cylinder.
Test bolting the parts after drilling and tapping.
Gee willikers, Mr. Wizard, using the rotary table to lay out the circular bolt patterns is really cool! Everything lines-up really well.
Using homemade boring bar for cylinder bore. Finished cylinder.
Because the cylinder is 5-1/2" long, I had to make another boring bar to clean up the piece of seamless pipe used for the cylinder. I thought it was going to chatter but, by feeding slowly and taking 0.002" cuts, I got a nice finish. The only problem is that, in order to clean it up, the bore ended up being 0.025" over the design. That means that the PTFE "O" rings I got to try as piston rings are a loose fit. I may try turning the ring grooves in the piston so the ring will contact the cylinder wall. Before pressing the whole works together (headers and cylinder), I will get the bore as smooth as I can by using 400 grit carborundum paper with the hone.
You will note that I turned the O.D. at the ends of the cylinder to assure a good fit in the headers.
25 June 2011:
It's beginning to look like it might end-up being an engine.
First test assembly of main components.
The cylinder is finished and the oiler tube is done. After I've wrapped the 0.06ho0" steel around the headers, that part will be finished. I'm still considering how to attach the sheet metal to the headers. I could weld it but the difference in thickness of the metals to be joined may make getting a good bond to the thicker metal troublesome. I could also drill and tap the headers then drill and screw the hopper tin to the headers. Either way, leaks could be a problem. I'm a-thinkin' on it.
After final sizing of the frame and cylinder parts, I assembled what I have done so far to see how it lays together. I've drilled and tapped all the pieces so I can bolt them together. If the frame loosens-up when the engine's run a bit, I may just go ahead and tack weld the parts together. As it is, there is one welded joint (beside the main bearings) that I will do when I'm sure all the dimensions work out.
26 June 2011:
The frame is assembled, the mains are done and the crankshaft is in for a test fit.
Boring the mains.
After bolting the frame together and taking dimensions for the main bearings, I disassembled one side and trimmed it so the crankshaft would be perpendicular to the cylinder. Then the mains were aligned and welded into place.
The engine was put into the mill and, after carefully checking location, one main bearing bore was sized. Since my boring head won't reach far enough to get all the way to the other bearing, I had to unclamp the engine, turn it over and go through the locating process again before boring the other bearing.
It's starting to look like an engine.
Next, I'll finish the cylinder hopper and make the piston. It looks like there will be a little interference between the rod and the end of the cylinder at two points in the stroke so I may have to use the die grinder to relieve the rod and the end of the cylinder.
27 June 2011:
Spent all day making the hopper. Mashed my finger, too. Real productive!
Before starting on the hopper, I put in the main bearing grease fittings and, using a shaft collar and a brass washer, made a shield for the pulley side main bearing.
Layout, drill, tap, bend, mash finger.
The hopper cover took pretty much the whole day. It involved drilling a bunch of 8-32 tap holes and tapping them. Then, I cut a piece of the 0.060" steel and screwed it to the cylinder headers with eight screws. Using some elbow grease, the trusty vise and a mallet, I formed the steel around the bottom then pounded, squeezed and cussed 'til I had the cover all the way around. I slobbered a thick coat of silicone gasket sealer over the surfaces then tightened it down.
The hopper cover was bent over the top of the hopper. A hole was punched for filling and another for the oiler tube. There's a nice decorative 8-32 screw in the "aw-shoot" hole I made for the oiler on the wrong side of the fill hole.
You will note in the third photo from the left, a small discolored spot in the cylinder. I decided to tack weld the cylinder to the rear header in a couple of places just in case it attempted to work loose.
The hopper is done. Serviceable but not very pretty.
I've got to modify an oiler I have in my "stockpile" to fit the tube on this engine then I'll be through with the cylinder. I guess the piston is the next thing on the agenda after that.
Since the cylinder bore cleaned-up 0.025 over, too large for the Teflon "O" rings, I'll call Dave Reed at Otto Gas Engine Works tomorrow and order a couple of iron rings. Dave is good about advising on the best width, etc. so I'll use his recommendation as to ring groove geometry.
28 June 2011:
Got the oiler fitted and have a good start on the piston. The oiler had a 1/8" NPT male end on it and the oiler pipe is threaded 10-32. Rooting around in the junk found one of those thingies that are used on ceiling fans and stuff to screw on the end of the threaded tube that holds it together. The thing looks like brass but is really diecast aluminum. It was easy, though to clean the threads with a 1/8" NPT tap for the oiler end. Running a 10-32 tap through the other end adapted it to the oiler tube. Now that it's mounted, it leans a bit but I'm not gonna start hauling on that diecast doohickey 'cause it'll surely break.
Turning the O.D. of the piston. Oiler and partially finished piston.
Doing a careful measurement of the bore gave me 1.542", a little over what I thought it was. Called Dave Reed and discussed how to manage the oddball size. He thinks that -maybe- a 1-9/16" (1.5625") would work but may not break in. He's looking for something closer, maybe a 39.5 mm (1.545") ring will work fine (for a 3mm, 0.118" groove) but would have to special order it. Since there's no great hurry and I won't finish the piston until I have the rings in hand, there's no problem.
A problem did crop up when I fitted the piston to the bore. I've got 0.005" clearance machined into the piston skirt with another 0.010" clearance in the ring area. It went in about half way then tightened-up! I've got 0.005" clearance machined into the piston, so there's more than 0.005" necking at the middle of the bore. I have no way of measuring a bore except at the ends, so I guess I'll use the piston and feeler gauges and just spend a few hours with the hone opening up the middle of the bore.
Oh, well - tomorrow's another day.
29 June 2011:
I had a revelation last night. When I went to the shop this morning, I chucked up a piece of bar stock in the lathe and, indicating the end, made sure it was straight. Then I ran the carriage back and forth with the indicator on the bar. DANG! In 18 inches (the length of the bar), there was 0.007" non-parallelism! No wonder the piston wouldn't fit in one end of the cylinder! The dang thing bored about 0.004" tapered.
The manual for my Chinese lathe didn't have anything showing how to cock the headstock, I removed the threading data plate and found the bolts that hold the headstock frame to the ways. After a bit of jiggling with a brass mallet, I had it indicating within 0.0005" parallesism. Methink that's something I need to check after doing heavy interrupted cuts.
Since I already had the cylinder assembly finished and there was no way to get it back into the lathe, I spent some quality time with my brake cylinder hone. I quit when I had the taper down to about 0.001". Then, I put the piston back in the lathe and turned the skirt until I had about 0.010" of clearance. I then turned the ring area 0.010" smaller than the skirt.
Piston, finished except for ring grooves.
After getting the piston fit right, I indicated the perpendicularity of the quill to the mill table and took out about 0.001" per foot of error.
I then put the piston back into the mill and carefully indicated the wrist pin location. Drilled, bored and reamed, it's ready for assembly. The wrist pin, a 1/2" dowel pin was ground to 1.3" length and a couple of PTFE disks were cut to use as cylinder wall protectors in case the pin drifts. Since the pin is a little better than 0.0005" bigger than the 0.500" reamed hole, it should stay put. I also drilled the oil hole, located to be right over the wrist pin.
30 June 2011:
I ordered the rings yesterday, so today I finished the piston by cutting the ring grooves and hogging-out material inside the piston to give clearance for the rod.
Cutting the ring grooves.
Clearance relief on cylinder and rod.
After the piston was hung on the rod, the assembly was put in the frame. The rod and cylinder were marked where they interfered. A die grinder was used to relieve enough metal so there was no collision.
2 July 2011:
Awaiting the piston rings, I've made the cam/ratchet post and the cam/ratchet retainer. The retainer also is the pivot for the latch rod and the mount for the governor spring.
Cam/ratchet post and retainer/governor latch pivot/spring holder.
Well, everything was going along just hunky-dory until, in the last operation on the cam/ratchet retainer, I broke off a 6-32 tap! This was supposed to be one of the spring mounting screws. Since the tap was bottomed-out when it broke, I didn't even try to get it out.
In theory, the engine speed is controlled by a hit and miss latch operated off of the governor. The latch pivots on the cam ratchet retainer and, between the retainer and the big "washer" will be a loosely held brass leaf spring that holds the latch disengaged until the force from the governor overcomes the spring tension and allows the latch to move. The "washer" screws are not going to be tight but will allow an adjustment (note the small nut) of the spring to control speed.
Anyhoo, the "washer" ended-up with only three holes. The tap resides in the retainer where the fourth screw should be located. If it doesn't work because of the missing screw, I'll just have to carve a new one.
6 July 2011:
Now that the 4th of July weekend is over, it's back to the engine. The piston rings came in and I filed the gaps for 0.004" clearance and mounted them on the piston. Then, while wrestling the piston into the cylinder, one of the rings hung-up and broke! Dang! I went ahead and put it together with only one ring, figuring if it doesn't hold compression, I can always get another one.
I made the ratchet and cam blanks, preparatory to cutting the profiles. It took all day to do this because I didn't have any 2.5" diameter round stock so I had to cut a piece from a hunk of 3/4" plate. The reason it came from 3/4" is that the cam is that long. It took a lot of whittling to make it but it's now ready for the rotary table and the mill. The ratchet wheel was easier to blank out because it only needs to be 1/2" thick and I had some rounds left over from when the parts were water-jet cut that cut down very nicely.
So far, showing the cam and ratchet blanks mounted on the cam post. The broken ring is in the foreground.
The ratchet pawl is to be made from 1/8" steel and I've printer plotted a few so I can cut one out and glue it to the steel for sawing to rough shape. As you can see, the engine is coming right along. I anticipate the fuel injection system is going to take a while because some of the parts are going to be very small due to the tiny amount of fuel injected on each intake stroke.
7 July 2011:
I worked on the ratchet and cam today. They're finished.
Setup for machining cam.
To machine the cam and ratchet, I needed something to hold it to the rotary table. First, I turned the head of a 1/2-20 bolt so it would seat into the center hole. Then, I turned a long 1/2-20 nut so one end just fit into the Morse taper hole on the face of the rotary table and then turned another diameter on it that allowed a snug slip fit to the cam blank. Tightening the bolt and adapter in the table forced it to be centered. With the cam blank bolted-up, there was only 0.002" runout, well within my requirements.
Cutting the cam flats. Cutting the detents. Cutting the ratchet.
The table was then mounted on the mill and, using a 3/8" diameter cutter, the flats were cut 90 degrees apart. Then, the cam was turned 45 degrees and the detents were cut 90 degrees apart.
The ratchet was mounted to the rotary table using another made-up adapter that centered it. The adapter was threaded 1/2-20 so it could be tightened to the table. This time, the blank was pressed onto the adapter because I didn't have a long enough bolt to be able to hold it with a nut.
The ratchets were milled at 45 degree intervals using a 3/4" end mill.
The finished cam and ratchet. Cam and ratchet mounted on engine.
Tomorrow, I'll work on the ratchet pawl, the pawl arm and the fuel pump arm.
8 July 2011:
I got the arms done. Nothing specially complicated about it, just time consuming.
With pawl and fuel arms.
I started on the pawl until I found out that I hadn't plotted it actual size. Back to CAD for a new plot. Tomorrow, I'll get that made and probably start on the cam followers.
9 July 2011:
The pawl is done and the cam works!
Valve beginning to open at near bottom of stroke. Valve fully open at bottom of stroke.
Beginning of intake stroke and fuel injection. Valve beginning to close at near bottom of stroke.
I may have to adjust the eccentric and ratchet timing a bit to optomize the valve timing but it's close enough now to see that it will work. You will notice that I had to do some profiling on the pawl arm in order for it to miss the ratchet at certain parts of the cycle.
Also, I'm going to have to study how it works in reverse rotation. I haven't tried rotating the engine backward to see if there are any binding points. If so, I may have to modify the mechanism so it doesn't tear itself up in the case of an engine kickback.
Starting on the lifter guides and lifters.
Once I have the lifters in place, I can really tell how it's going to work. I may have to swap valve springs to a lighter one if it seems to take too much force to open.
10 July 2011:
I did some CAD on the governor parts this morning and have got a good start on the cam followers.
Cam followers and guides semi-done.
The photo shows the little ball bearings (1/2" O.D.) that I will use for the rollers. As the more astute of you will notice, there are a couple of oddities on the right-hand guide. One is that, I miscalculated with the piece of scrap and have a water jet start hole showing. The second thing is that I had to relieve part of the top surface of it due to interference with the pawl arm.
12 July 2011:
Finished the lifters and guides today and mounted them. The valve train works.
Start of compression. Exhaust stroke..
The valve opens and closes at about the right times but, when the engine runs, it's going to be very noisy. Oh, well. I never said the design would be perfect.
I may start on the fuel pump tomorrow. That is, if the "honey-do" projects don't take up my day.
13 July 2011:
The valve train noise (and the force required to open the valve) have had me thinking. This morning, I removed the cam and milled ramps right at the top of the lift on each "lobe". This decreases the force required to finish opening the valve. Then, I had a major brainjolt. What is a strong exhaust valve spring needed for if there will never be any suction on the valve (as in a two valve engine). Rooting in my spring box produced a much weaker spring that has enough force to close the valve quickly.
Progress to date.
In the photo above you can see the previous spring on the engine baseplate. If the new spring works all right, I'll modify the keeper so it is held centered. I also modified the upper follower return spring by drilling a pocket into the top of the lifter for the spring. That way, I can use a longer spring and have more even force over the stroke of the lifter. It works much more smoothly now and is quieter. Another advantage of the new set-up is that, if the engine backfires, the ratchet can turn backwards against the weaker springs without stressing anything.
After all that, I started on the governor. You can see the trunion mounted on the crankshaft. That little part took almost four hours to make because the deep slot required very light cuts. You can see the raw governor latch fork on the base.
14 July 2011:
The governor's done. I suppose I could say, "It's a boy!", but I won't.
If you look closely, you'll see that I'm using some little (0.312" O.D.X-.125" I.D.) ball bearings to ride on the governor pins. These came out of one of the dead hard drives in my "Deceased or Redundant Hard Drive" box. You'd be surprised how much good stuff is inside those little buggers. My biggest worry is that I'll use-up all of my old MFM and RLL half-height drives. Although the newer drives have about the same stuff in 'em, it's all smaller.
Tomorrow, I may do the latch rod. That may be fast and it may be slow. We'll see. After that's done, I may belt it up and motor it to seat the ring. It has very little compression due to the unseated ring. After a couple of hours of motoring, I should begin to see some compression. If not, I'll have to purchase another ring.
16 July 2011:
The latch rod is done.
The semi-finished latch rod. Latch rod on engine.
The latch rod is mostly finished. The only thing that needs to be added is the fuel pump lockout bar which will be mounted on the latch rod. I won't make that part until I've got the pump in position and working. The flat spring that controls speed is a piece of steel strapping. It is more than springy enough for this purpose. Speed is adjusted by turning the brass nut on the cam hub where the latch rod pivots. Turning in the nut pushes the spring against the cam hub, making it press harder against the latch rod which makes the governor weights have to spin faster to overcome the spring.
Motoring the engine.
The rest of the day was spent converting the motoring pulley I used on The Homebrew Hvid. I had to turn the diameter down to about 8-1/2 inches so it would clear the bench with the belt on and make a bushing to adapt it to the crankshaft of the engine. I bored a small diameter flat belt pulley O.D. to 1/2" so it would fit the shaft on the variable speed gearmotor. The engine drives directly off the of the motor.
I was pleasantly surprised when, after only a few minutes of motoring, the engine began to get some compression. One comment I have to make is that it's a noisy little rascal!
Next is the fuel pump and injector. Last is the ignition system which will be trivial.
17 July 2011:
I got a pretty good start on the fuel pump today.
Aluminum bar stock end. Pump body roughed-out. Pump body machined minus internals.
I made the command decision to make the pump body out of 6061-T1 aluminum because it is easy to machine. I needed the free machining properties because of a couple of deep small diameter passages that had to be drilled.
In the right-hand photo, you can see the pump plunger. This is a piece of #65 drill rod (0.035" Diameter). The return spring is from something long since junked.
As I worked on this part and others, I made running changes. This is usually pretty tedious because, when building a prototype of a production piece, each change has to be carefully documented so the final drawings will accurately depict the piece when they go out to production. Here, since there will never be but one example, I can just pencil changes-into the shop drawings. Unless there is some dimension conflict, I don't have to go back to CAD and make changes.
One of the changes I made in the pump was to leave the inlet check valve out of the pump body. I will make a fitting that screws into the inlet port that will have the check valve built-in. I've found that I can thread 1/16" pipe with a 10-32 die and use a 10-32 plug tap for the female thread and the joints work very well. The inlet, outlet and bypass ports are threaded and ready for the fittings to be made.
If I'm lucky, tomorrow I should be finished with the pump and can see if it will actually pump.
18 July 2011:
Well, this is one of those good news - bad news days. The good news is that the fuel pump is finished. The bad news is that it doesn't work.
Pump plunger, spring, etc. Relief valve parts.
Everything, including the very small #65 (0.035") 1/2" deep hole went well. No broken bits or cussing. Not shown above left is the small PTFE (Teflon) pump plunger seal that goes in first in the installation.
The check valves are made have brass housings which screw into the pump body. A short length of 1/16" pipe is threaded 10-32 and screwed into the other ends of the housings. Inside the housings are a small spring and ball. The inlet has the ball seating on the 1/16" pipe end which was lapped to the ball. The outlet valve has the ball seating on the brass housing.
When I hooked up the plastic tubing and stuck the inlet tube into a cup of mineral spirits, it didn't pump at all. Didn't even try! At about quittin' time, I discovered that the threaded joints in both of the valves were leaking. Tomorrow, I'll revisit the whole valve arrangement.
The finished pump, not quite ready for prime time.
I sorta figured that the pump would be problematic but was wrong in thinking that the pump itself would be hard to make. The problem is in the valves.
19 July 2011:
Well - I spent the entire day fiddling with the pump. Most of the time was spent trying to get the valves to quit leaking. I made a lapping fixture by soft soldering a ball to the end of a piece of 1/8" copper tubing. Then, with some fine lapping compound, I worked over the seats until I could see a good seat.
Once the valves were lapped, it pumped a very small volume. I think I'll have to drill a bleed port to get all of the air out of the plunger bore. I think there's a bubble trapped between the Teflon seal and the outlet/bypass port. This will require drilling a hole horizontally until it meets with the plunger bore right below the seal. Then, I'll have to drill another hole from the top of the pump housing to intersect the horizontal passage. The horizontal passage will be plugged at the outside end. A screw will be used to seal the top of the bleed passage. Doing this will get rid of virtually all of the air in the system.
If it still doesn't pump, I may have to enlarge the pumping plunger, increasing the pumping volume to the point it can overcome any tiny internal leaks. While I was fiddling with the check valves, I took a loupe and inspected the steel balls. Up close, they don't have a perfect surface but they should seal against the lapped seats.
20 July 2011:
I got the bleed port made and tried the pump again. First, though, I'll show the check valve seat lapping tool. It was made from a bearing ball the same size as the check ball. The ball was tinned and soft soldered to the end of a piece of 1/8" copper tubing. A little fine lapping compound was put on the lapping ball and the ball was rotated in the seat until a good seat pattern appeared. The rough seat was simply the end of the drilled hole.
The check valve seat lapping tool.
Pump with bleed screw mounted on engine.
You can see the bleed screw in the photos above. It is plugged at the top with a 4-40 machine screw and fiber washer. The access hole to the top of the pumping bore can be seen just above the outlet port. It is plugged with an 0-80 screw.
After bleeding every bit of air out of the pump, I motored the engine to work the pump. Although I could see fuel moving very slowly in the plastic tubing, it took it a couple of minutes to pump a single drop to the end of the outlet tube. Methinks I have to increase the pumping volume to get a practical fuel pump. To do this, I'll use larger diameter drill rods for the pump plunger.
I'm going to order a couple of #50 (0.070") and a couple of #39 (0.0995") drill rods as well as some check valve balls. I probably won't do much on the engine until I get the new parts.
21 July 2011:
Did some more on the fuel pump today.
Outlet fitting. Fuel pump with non-compression bypass in place.
I fiddled with the pump all day today. The outlet fitting adapts the 10-32 thread in the pump body with 1/8" copper tubing. I also moved the outlet check valve to inside the pump body to minimize the volume of the pump for better operation.
After trying several outlet valve spring strengths, I finally got the pump to work. By my rough calculations, the pump, as it is presently configured (0.035" plunger moving about 0.125" at maximum stroke) should pump about 0.0001202 cubic inches per stroke which was, by reckoning just a bit more than required to run the engine. Now, I think my reckoning was wrong. Way lean!
I've just done a calculation that may make a fool of me for designing the pump for the volume I did. Without going to the books, I recall the stoichometric ratio for gasoline/air to be around 14:1. Now Assuming the ratio is in volumes (the engine has a swept volume of approx. 4.42 cubic inches. Dividing 4.42 in3 by 0.0001202 in3 gives a ratio of 36772:1, a little on the lean side!
Using this calculation, the pumped volume of fuel should be 4.42 divided by 14 or 0.315 cubic inches of fuel per stroke ....... That sounds like a LOT!! Hmmm.......If the engine were running at, say, 600 RPM and firing every sixth revolution at no load, that means it would be consuming about 100 times 0.315 cubic inches of fuel or 31.5 cubic inches of fuel per minute and that's A LOT OF FUEL!
Using my original 0.035" diameter plunger would yield a fuel consumption of about 0.01202 cubic inches of fuel per minute. Now THAT number seems reasonable so I may have my math screwed-up. I never said that my strong suite was math! :-)
Anyway, I ordered some 0.070" and some 0.100" diameter drill rod in case I have to increase the pumping volume to overcome valve seating losses. The 0.070" rod will give about 0.00048 cubic inches per 1/8" stroke and the 0.100 rod will give about 0.00098 cubic inches per stroke.
23 July 2011:
The fuel pump is almost ready for prime time.
Testing the fuel pump. Note the fuel stream.
It took some doing but the fuel pump now works and the off-cycle bypass valve works reasonably well. I tried the new bearing balls in the valves and they helped a bit but the volume pumped was still very small.
What I ended-up doing was enlarging the pump plunger diameter from 0.035" to 0.070". Now, it will pick up from the tank without purging the air when the stroke is at the maximum setting. I think that, when it is adjusted for the proper amount of fuel for running and has air in the system, I will have to hand pump it full stroke to get the air bled out.
In the photo above, the left-hand red cup catches the fuel that will eventually go to the injector. The right hand cup catches the fuel that is bypassed during the off-cycle (when the IN/EX valve is closed).
The injector is another thing. I'm still thinking about it. Since I have a lot of trouble drilling holes as small as #70, I will not be able to depend on a small diameter jet for making a spray pattern. I will try some ideas in the next few days.
27 July 2011:
I'm still thinkin' on the fuel system. I've mounted the fuel tank on the side of the engine so it's about level with the pump. This should make for easier priming. Then, after I got it plumbed-up, "The Law of Unintended Consequences" kicked-in. The bypass valve stem leaks due to the fuel level being higher than it is.
I tried to make a PTFE seal for the plunger but it still leaks a little. I'm thinking of making the entire bushing out of PTFE. That way, I can adjust the tightness of the seal by tightening it in the fuel pump body, thus squeezing the neck down.
This morning, I went out and found that the outlet line drips occasionally. The outlet valve is probably not seating very well. I'll have to do something about that, too.
30 July 2011:
Lately not a lot has been done. Today, I decided to counter drill the outlet fitting so I could fit a small spring to see if I could make the outlet valve work better. Success!
I also made the PTFE plunger guide for the bypass guide. It didn't seal as well as I liked so, since the pump works pretty well now, I lowered the gas tank so the pump is above the fuel level. No more leaks. At least none that I can't fix.
Engine with fuel tank mounted low.
The fuel injector gets my attention tomorrow. It'll probably prove to be problematic too, but I do have a few ideas to try. The injector will mount in the aluminum port piece that is attached to the head.
31 July 2011:
I think I've got the injector working well enough for the engine to run. I said, "I think", because it seems that it was too easy.
The injector assembly. The injector looking from the valve cage bore.
The injector is made from 3/8" bar stock. It is threaded 3/8-24 to screw into the port assembly. The little screw and washer at the top is the bleeder. At the side is a home-made fitting to adapt the 1/8" copper line to 10-32.. The adapter screws into the injector. Not shown in this photo is a locknut with a chamfer on the port side to seat an "O" ring to seal combustion gases. At the tip is an 1/8" ball and a flat spring/keeper. The ball is seated in the injector body and forms a check valve to keep combustion gases from traveling back up to the pump.
When screwed into the port assembly, it is at the edge of gas flow and out of the way of the valve.
It looks like it's getting fuel to the combustion chamber.
When I got it all assembled and fixed a few leaks, I purged the mineral spirits I was using as a test fuel and filled the tank with naphtha. After priming the system, I motored the engine and within a minute, smelled naphtha vapor. As you can see from the above photo, I used a torch to ignite the mixture when it exited at the port. At this time, it makes nice puffs of flame on the exhaust stroke but I have no idea whether it's close enough to 14:1 for the engine to even try to run.
Tomorrow, I'll cobble up an ignition system and see if it will try to run. Depending on my mood, I may take the camcorder out to the shop to make a video of the exercise.
1 August 2011:
Today was frustrating. I got the ignition done and on the engine but spent the rest of the day fiddling with fuel pump and injector valves.
Ready to test. Sensor and magnet. "Battery Saver" contact.
The ignition consists of one of my solid-state ignition modules along with a Hall-Effect magnetic sensor. The sensor is mounted on one of the crank cheeks. I use the "Battery Saver" of my ignition module to sense when the valve is open and disable the ignition on that cycle. As you can see in the right-hand photo above, I've made a spring contact that touches the valve lifter when the valve is open. This also inhibits ignition when the engine is in the "miss" cycle.
After getting it all hooked-up, I turned on the video camera and spun the engine. When I turned on the ignition it made a couple of half-hearted tries then settled in to only fire weakly once in a while. I noticed that there was some blowback into the fuel system when the engine fired so I spent quite a while trying to get the ball valve on the injector to work right but wasn't really successful.
Then, since the outlet valve in the pump wasn't seating well enough to keep combustion pressure from backing-up the system, I fiddled with the outlet valve for a couple of hours to no effect. In an effort to get the engine to run at all, I tried plugging the injector port and simply running the line from the pump to the port. If I pressurized the system to really flood it, it hit a few licks but was really rich. I'm going to think the situation over, probably going to a different type of valves.
2 August 2011:
Today I worked on the fuel pump and injector. The bypass valve in the pump was changed to a poppet type. The outlet valve was improved by making a little spring loaded pushrod to hold the ball valve closed. Also the seats were reworked.
The injector ball check valve was re-seated deeper in the body and a simple music wire retainer was made. The valve is supposed to blow shut upon compression. It's still not working like it should but here's part of yesterday's video and today's video edited together.
YouTube Video of First Try
4 August 2011:
Still stuck in the "fiddly bits". Yesterday, I worked on the injector. Abandoned is the ball valve, replaced by a poppet-type that is held lightly shut with a coil spring against the head. Then, I found that the spring mount on the injector interfered with the valve. I should have placed the injector further away from the valve and may have to plug the present injector bore and move the injector closer to the head in the passage. A temporary fix was to grind away as much of the spring mount, keeping the valve and injector from banging together.
After motoring the engine, I found that the valves in the fuel pump were leaking a little and it took all day today to modify from a ball-type inlet valve to a weighted needle valve. I still don't know why I couldn't get a ball valve to seat under pressure but, no matter what I tried, from lapping with another ball and fine compound to using a punch and forcing a seat in the brass valve body. Right now, I have an inlet valve that works pretty well and should, with use, seat better. Against what I would have thought, I've got a brass valve and seat.
Then, this afternoon, after only getting it to hit once in a while, I finally just maxed out the pump stroke and it hit more often, probably about every other revolution. I think the problem is that I've got the pump designed by theory and didn't take into account things like valve bounce. Tomorrow, I'll make a new, larger diameter pump plunger. The present one is 0.070" and I've got some 0.100" drill rod I can use. I'm thinking that will do the trick then I can go about working out other problems that will crop-up once the fueling is reliable.
6 August 2011:
Yesterday, I made the new pump plunger and the engine tried to hit every time. Now, it only occasionally will fire again after the governor unlatches. I now think it needs more flywheel mass so it can carry over. What I think happens is that, when it hits, it speeds up to governed speed in one power cycle and the governor latches. Then, because it slows so fast, it can't get through a full cycle before slowing too far to recover.
I dug out some scrapyard barbell weights to make into another flywheel. I don't know what's in the cast iron but it sure machines wierd. With my little lathe, it takes a while to get it done, too.
Flywheel raw material. Cleaning up the wheel.
The holes in the weighs are too large for the 3/4" crankshaft so I cleaned-up the bores of the weights and made a bushing so I could press them together then bore it to fit the crankshaft.
When the flywheel is finished, I will broach a keyway in it and mill a slot in the crankshaft so I can use a gib key to attach it. I'm going to wedge the crankshaft in position then clamp the whole engine in the mill to make the slot. It would probably be harder to get everything lined-up if I took the crank out.
7 August 2011:
The flywheel's finally done.
Milling the keyway. Flywheel in place.
I spent the whole day finishing the flywheel, broaching the keyway in it and milling the slot for the key in the crankshaft. My idea of mounting the entire engine in the mill worked like a champ. I used the dial indicator to make sure the engine was square in the mill.
One of these days, though, I'm gonna figure out how to fit a gib key. I now think that I should broach the keyway in the flywheel first then measure it to see how deep to make the keyway in the shaft. This is because you're sort-of at the mercy of the broach, the bushing and it's shims and, even though I could make some shims, I don't think it would come out exactly as planned. As you can see in the right-hand photo, the key is in almost as far as it will go. A couple of extractions and I'll probably have to shim it.
8 August 2011:
I modified the fuel volume adjustment so I can now adjust it "on the run" instead of having to stop the engine to do it.
Fuel volume adjustment. Allllmost running on it's own.
The volume adjustment shown above is a spring-loaded gizmo that is located between the pump plunger and the pump actuating arm. It limits how far up the stroke the plunger can return, thus limiting the stroke. The nut atop the moving piece adjusts the up limit.
Today's run netted a nearly independent (of the motor) engine. It will, on occasion, fire every time it comes up on compression but the fuel mixture varies widely between rich (black smoke) and lean (weak firing). I think the injector is now the culprit. Having tried a free-moving and spring loaded ball valve as well as a free-moving and spring loaded poppet at the tip, it seems to try to run best with the spring-less poppet.
I may try a reed valve or try leaving out the valve entirely and putting a tiny jet in it's place to see how that works. I could also try snaking a line to underneath the intake/exhaust valve and stay away from the combustion chamber.
Working on these projects is like re-inventing the wheel and gives me an appreciation of the engineers and blacksmiths of over a century ago when they developed the various types internal combustion engines.
9 August 2011:
More work on the injector today. This time, I removed the poppet valve which only caused the fuel to dribble out and replaced it with the jet out of a propane torch. The diameter of the hole is around 0.005". When hooked up to the pump, it made a very thin stream of fuel. In the inlet air stream it should break-up into a mist. Not optimum but should have been better than the dribble.
I put it all back together and motored it for about fifteen minutes without even a weak hit. I quit when I could see fuel backing up into the inlet fuel line to the pump. Leaking valves! That must be what's keeping the pump from making pressure. As designed, it should do a couple of hundred PSI easily but it backflows under compression pressure.
I've tried three types of home-made valves. Ball, poppet and needle, all with and without return springs. I've used the valve needles and poppets with fine lapping compound to seat them. The ball valves were seated using a ball soft-soldered to a post and lapping compound. All of them eventually worked into what I thought were good seals but all have leaked under pressure.
I quit early and, while reading, decided that if I could get the inlet valve to seat without leaking, the pump should be able to make pressure even if the outlet valve leaks a little. I'm going to root through my small parts and see if I can find one or more viton-tipped float valves to try.
11 August 2011:
Maybe not. After a few more iterations of the injector and pump valves, I got the engine to almost stay running on it's own this afternoon. At one point, it ran for about fifteen seconds.
The fuel system is maybe working pretty well but other problems have shown up. One is that I think the valve is not seating fully because I hear a bit of a hiss when the engine fires. Now, the sound could be blow-by due to having only one ring.
The serious thing I think is, after the engine gets up to even a slow governing speed, serious limitations in my gearless idea are being felt. I think that, before the engine gets to a speed that it can continue to run, the cam starts skipping. I made a little improvement by making a friction washer for the cam to keep it from over-running.
I'll continue to fiddle. I'd really like to get the engine to run for a while so I can document it on video before I consider changing to a more normal 2:1 reduction cam. I'll keep the one valve idea but it will probably run better without all the jitter in the present ratchet system.
12 August 2011:
After pinning the eccentric on the governor shaft to keep it from slipping and increasing the size of the setscrews in the ratchet for the same reason, I tackled the injector again. The check valve was leaking.
I lapped the valve some more and increased the tension of the valve spring to make the cracking pressure higher and, after re-assembling everything, I motored it. With a few adjustments, it was firing every time it came off the governor so I removed the belt and hand cranked it.
The engine seemed to be really trying to run and would sustain itself for as long as 30 seconds before quitting. Interestingly, it would usually re-start with little effort and run for about the same length of time. I was considering getting out the camcorder to document progress but the engine "sorried-out" and quit, not wanting to re-start.
A little detective work located a leaking injector check valve which was causing compression to fight the fuel pump. The present check valve is made of brass, seating in steel. Depending on what I find when I have the injector apart, I may make a new valve out of steel. The heat could have affected the brass valve but I think the valve spring could have been de-tempered by the heat, even though it wasn't that hot when the engine quit. I'll see tomorrow.
14 August 2011:
As I said a few days ago, I wanted to re-work the inlet valve to the fuel pump. I made a new inlet valve housing and tried a viton tipped float valve I had in my "small junk" drawer. It sealed nicely but tended to fail to lift on the inlet stroke. I believe it was due to the valve being rubbery and jamming itself into the seat.
Viton tipped brass float needle inlet valve.
What I ended up doing was making a brass needle. After I got the weight of it right, it sealed nicely.
Today, I made a steel injector check valve and spent a couple of hours lapping it so it wouldn't leak. The lapping process started with fine carborundum valve grinding compound, proceeded to fine hard-metal lapping compound then to fine soft metal lapping compound then the valve was finished off by rotating it in it's seat for a few minutes using motor oil. What a pain!
New injector valve and orifice.
The injector valve was made from a steel 6-32 flat head machine screw. The brass 10-32 screw is both the spring holder and the jet. The smallest drill I have is a #70 (0.028"). It needs to be about 0.005" but since I dropped the torch jet and can't find it, it'll have to do.
I think it's about as far along as it's going to get with the present ratchet drive cam arrangement. The engine runs for a while but eventually, jitter in the valve train causes it to stop. I don't think it'll ever run much better than it does now without changing the cam and cam drive to a regular 2:1 reduction and a 180+ degree cam.
Also, I think the fuel injection is too early in the intake stroke, just after TDC when air velocity in the connecting port is low, causing poor atomization. Since the fuel pump is driven by the eccentric and the eccentric controls valve timing, there's no way I can appreciably change it. With a regular cam drive, I can make a separate cam for the fuel injection and can adjust it to any point on the stroke. I think that, if I inject the fuel about 45 degrees into the intake stroke when air velocity is relatively high, the fuel will break up into smaller droplets, causing better combustion.
Anyway, after fiddling with the injector, I made another YouTube video of the engine as it is now.
YouTube Video of Second Try
Page Two, Fixing The Valve Train
Rush back HOME
BOY! This is fun!
(Hits Since 22 May 2011)