Another Radial - this time 18 Cylinders

[DICKEYBIRD]

Dunno how I missed this masterpiece 'til now, kudos to you sir! I want to express my sincere appreciation for the patience & time you've invested in sharing your build process and the many clear and practical tips you have presented here.

You have a gift for making extremely complicated operations look easy & within reach of the rest us hacks.

 

[kvom]

I used Loctite 620 on my locomotive axle shafts/drivers, and it's a very strong hold. I didn't measure the degree of fit, but it's a close sliding fit, probably about .001. When I had to remake one of the axles, I heated the joint with an oxy torch, then pressed the axle out. It still took a pretty strong press, and came out with a loud bang. I don't worry about them coming loose.

With that degree of fit I found that it would being to set within 10 seconds; after that I couldn't move the joint by hand.

 

[mayhugh1]

Because there seems to be some interest in the technique I've been using to create bearings in a workpiece, I thought I would also show you what can go wrong. In preparing the workpiece for my rear cam retainer I pressed a bearing bronze slug into an aluminum disk. The slug was about .0004" over the hole size, and against my better judgement I decided to press it in anyway just to see what would happen. Both surfaces were mirror-finished and coated with Loctite 609. When I started the pressing operation I could tell it wasn't going well since the force needed was quite a bit more than I'm used to. Looking at the bottom of the workpiece when it was over, I could see that the bronze had galled the soft aluminum and had peeled and pushed it forward of the slug as it was pressed in. I normally would have immediately scrapped the workpiece at this point but I faced it to its final thickness on the lathe, anyway, hoping I might get lucky. The photo shows the poor result. You can see a slight gap between the bronze and the aluminum where the aluminum was peeled away from the boundary as the slug was pushed through. The Loctite was scrapped away at the same time leaving no hope of filling in the gap. The truth is, if the final thickness of the bearing within the aluminum were too be, say, 3/16" or so, it probably would have been OK, since the two materials are likely intimately bonded over such a thickness. However, in my case the bearing will end up in only about 1/16" thick aluminum, and so I scrapped the workpiece at this point. When done correctly, there will be no gap whatsoever between the two metals. Only a color difference will distinguish the boundary between the two. - Terry

 

[Swifty]

Could you not have bored it out slightly oversize and inserted a new piece?

Paul.

 

[mayhugh1]

Paul,
You're right. I just didn't think of it. - Terry

 

[mayhugh1]

When I modeled the rear crankcase section of my engine from the Chaos Industries photos, I noticed the rear cam retainer was integrated with the engine's rear seal. In these Hodgson-style engines the rear seal is an o-ringed radial plate, just forward of the impeller, that isolates the engine's oiled parts from the fuel plenum. Everything forward of this seal is wet with oil and everything aft is wet with gasoline. I was concerned this combination would further complicate the cam timing adjustments for the rear row of cylinders, and so I included sufficient space in my rear crankcase section to keep these two functions separate. When I realized that machining the mounting spacers integral to the front retainer would simplify the front cam timing adjustments, I felt like this would also work well at the rear. So, I decided to also integrate these two functions since I'll end up with a much more interesting part to make compared with just duplicating the front retainer.
My (2nd) rear retainer starts out as a 1-1/4" slice of 3-3/4" diameter aluminum with a bearing bronze slug pressed and Loctited into position at the required distance from the center to eventually become the jackshaft top bearing. All features that need to be concentric can be machined from the same side of the workpiece, and so the first step is to turn a spigot to hold the workpiece while machining them. This spigot will eventually be shortened and become a hub with a lip seal for the crankshaft. After creating the spigot, the workpiece was turned around and chucked in a 5C collet chuck. Here, the final o.d. was turned and the o-ring groove was cut. I was then able to verify the fit of the retainer to the crankcase with the o-ring installed and the workpiece still on the lathe without disturbing the set-up. The workpiece was then faced to its final maximum thickness which includes the height of the integral mounting spacers. The through-hole for the crankshaft was bored as well as the pocket for the lip seal. While on the lathe, additional excess stock was removed from the workpiece in order to minimize cutting time later on the mill.
The chucked part was then moved to a horizontal rotary on my mill and the center of the the part was located. The part was rotated about this center to bring the bearing slug to its 6 o'clock position. The three mounting holes were then spotted and drilled. The remainder of the retainer could also have been machined at this time; but with so much of the soon-to-become thin part overhanging the chuck, chatter would have been a problem.
So, the semi-finished part was moved to a rectangular fixture plate with a matching set of tapped mounting holes. All the remaining features were machined except for the jackshaft bore. The part was then flipped over on the fixture, and the counterbores for the SHCS mounting screws were plunge-cut. Two additional non-penetrating holes were drilled and tapped for a puller that will be used to install/rotate/remove the o-ringed retainer from the engine.
All that remained was to match drill and ream the jackshaft bore through the retainer and rear main bearing. This was done with the retainer and main bearing mounted together within the rear crankcase section and with the o-ring in place. The three mounting fasteners, themselves, can't be relied upon to accurately and repeatably locate the retainer on the rear main bearing. In addition, it isn't reasonable to expect the jackshaft to act as a locating dowel to help them find their proper positions as was done for the front retainer. This is because the radial forces of the compressed o-ring will cause the retainer to locate itself in the center of the crankcase bore. If the jackshaft bore in the retainer doesn't perfectly align with the bore in the rear main bearing, this force will have to be overcome when jockeying the retainer over the bearing and onto the jackshaft. This will make assembly much more difficult and perhaps even cause a bind. If the boring operation for the jackshaft is done with the parts inside the crankcase while the o-ring is compressed, this radial force can be used to help align the jackshaft with its bore in the retainer during assembly.
The jackshaft was turned from 303 stainless and is similar to the one in the front cam assembly. Because the top jackshaft bearing is somewhat shrouded from the oil windage, I'll later open up a port through the rear main bearing to encourage oil mist generated by the rear row of cylinders to enter this area and help lubricate the bearing through an oil groove cut into the top of the shaft.
The jackshaft bore is the only unsealed penetration through the retainer and into the fuel area. The positive crankcase pressure vs. the expected slightly negative pressure in the fuel plenum means that a bit of oil may be pushed into the air/fuel stream rather than the other way around.
Delrin bumpers were pressed into holes in the retainer plate in a manner similar to the front. The lip seal will not be installed until final assembly to avoid damaging it during the remaining construction and fitting. After the rear cam subassembly parts were completed the rear main bearing was installed in the rear crankcase section along with the crankshaft. The jackshaft was installed on the rear main bearing and the o-ring was installed on the retainer plate. A pair of long SHCS were threaded into the puller holes and used to help install the snug fitting o-ringed retainer. It easily went down over the jackshaft whose running clearances are .001". The crankshaft still turned freely with with no drag or bind, and the cam disk turned smoothly as hoped. After verifying the fit and operation of the cam with the o-ring in place, it was removed it until final assembly.
The next step will be to tackle the impeller. I really want to make a complex one with compound curved fins like the Chaos Industries version, but I'm not sure I can design one that can be cut in any reasonable amount of time on my Tormach. My 4th axis CAM capabilities are somewhat limited, and so I'll probably be in some deep water trying to make my CAD/CAM work. I plan to break up this 'skull' time with the machining of some of the more mundane parts like the tappets and tappet bushings. I need to make 40 each of these, and if the truth be known, I really don't like making more than one of anything. For some wondering why, then, I would ever get involved with this particular engine I have to confess that I never really expected to get past the crankcase and crankshaft. - Terry

 

[kvom]

Nice solution to clamping the crankcase to the table in one of the above photos. I need to remember that one.

 

[Tinkerer58]

WOW what an amazing build and the skill and workmanship is mind blowing. I don't think I will live long enough to be able to obtain the same skills presented here. I don't know what else to say, I'm speechless for once in my life.th_wav

 

[mayhugh1]

An impeller is located on the rear of the crankshaft in the fuel plenum of this model engine. In some full-scale radials there is actually a supercharger in this area delivering a pressurized air/fuel mixture to the cylinders for increased power. In these model engines the impeller is usually not sealed to the walls of the plenum and is not spinning fast enough to generate any significant boost. It does, however, perform an important function by helping to keep the fuel mixture diffused in the plenum and limiting the amount of fuel that falls out of suspension.
Making an impeller is an neat CNC project and is often used as a demo part to show off multi-axis mills. I made a rather simple straight blade impeller for my 9 cylinder model that was similar to the one in the original plan set. I designed mine with a fully filleted 3D profile to give it a nice finished look, but for this engine I wanted to do something special with more complexity. The complexity I added is primarily cosmetic and probably won't add any performance. In fact, all the extra effort will end up buried inside the engine, hopefully never to be seen again after final assembly.
The impeller I visualized was one with compound curved blades. With blades curved both radially and axially, and the impeller would be reminiscent of some of the demo parts I've seen. As a test I managed to design a single compound blade in SolidWorks; but after spending many hours trying to generate the tool cutting paths with my CAM software, it eventually became obvious that a 5-axis mill was required. Since I'm only interested in parts I can make on my own equipment, I finally gave up on the compound blades. This decision reduced the mill requirement to only 3 axes and greatly simplified the machining. I settled on simpler curved blades that mimic those of a centrifugal pump. My design is similar to the Chaos impeller, but I axially profiled the blades to closely follow the profile of my plenum air guide just to add some challenge back into the project. There is only some .015" clearance between the air guide wall and the blades over their entire length. It's just possible with both of these features working together I might achieve a slight increase in air/fuel velocity through the plenum. With this impeller design, the total volume of the plenum greatly reduced, and the fuel path is primarily through the spaces between adjacent impeller blade pairs and the air guide wall. I also decided to reduce the number of blades from nine to seven. The real reason for this is that seven blades results in a spacing that allows me to use my Tormach .312" spherical profiling cutter to finish profile the entire impeller. I used this cutter (modified) earlier to finish profile the rear cover of this engine and it performed beautifully. I don't believe there is any real functional reason to match the number of blades to the number of cylinders. The intake duration is 208 degrees. With a nine blade impeller five blades will rotate across and partially block the intake port while the cylinder is filling with fuel. With a seven blade design only four blades pass by the intake port in the same amount
of time. There's no issue with sharing of the air/fuel charge between two cylinders whose intake durations overlap since they are too far apart to be affected by either number of blades. My humble conclusion is that, if anything, the seven blade design may be marginally better than the nine blade design.
The impeller started out as a slice of 3-1/2" aluminum round. I began the machining on the lathe where I removed a good bit of excess stock which conveniently left me with a spigot. I turned the workpiece around and used the spigot to hold the part while turning all the concentric features to their final finished values. This included the final o.d. of the impeller as well as the bores for the crankshaft, a height locating step, and a clearance pocket for the cam retainer hub. The milling cutter needs access to the entire top portion of the impeller as it mills the finished profiles in the upper blade areas. Therefore, I milled a rectangular boss on the bottom of the workpiece, where I had left 1/4" excess stock, so the workpiece could be held in my vise during the milling operations. This boss will be removed later from the finished impeller
using a mandrel to hold it in the lathe. In use, the impeller will be secured to the crankshaft with two grub screws 104 degrees apart, and so their holes were drilled and tapped on the still-simple workpiece before any milling was done. This required me to generate the blade profiling tool paths with my CAM with the workpiece is properly rotated to insure the holes actually end up between two blades. Before moving the workpiece to the mill, I began having second thoughts about the small clearances I was planning between the blades and the air guide. The interior profile of the air guide which is integral to the rear cover, was machined on my 9x20 CNC lathe using the same CAM software that I use on my Tormach. In the past when I have tried to match complex lathe and mill cut profiles on pairs of parts I was sometimes disappointed with results. I suspect that either I or my software is not always handling the lathe tool geometry properly, but I haven't done enough testing to fully understand the problem. Therefore, I decided to measure the existing internal profile of the air guide at some dozen points to verify its dimensions with those in my current CAD model from which the impeller was designed. The results were pretty close, and so I felt safe in starting the milling of the impeller.
The first milling operation was a coarse waterline roughing operation which quickly removed the bulk of the excess stock. The second operation was a finer roughing operation with a smaller diameter cutter and smaller steps that left some .015" stock for a final finishing operation. This finishing operation was done with the .312" spherical profiling tool using cutting parameters I had already derived while machining the exterior of the rear cover. The profiling tool paths were generated for a .0003" maximum scallop, and this left a bright and smooth surface finish that needed no manual cleanup. The total run time was about 2-1/2 hours. The gripping boss was then removed by turning a mandrel and facing it off in the lathe. I verified the actual blade to air guide wall clearance with modeling clay since there was no other direct way of measuring it.
I haven't yet gotten to the tappets or their bushings since the impeller ended up absorbing all my interest during the past week. I hope to make progress on them while working up the design of the distributors which is my next step. - Terry

 

[petertha]

Fascinating stuff!
Can you elaborate on your clearance. Is it the same (0.015") amount at the outer diameter ring edge (blue line) as along the curve edge blade profile (red line)?

I just naively assumed gap clearance would have to be very small on something like this, so there is no pressure leak-off betwen individual impeller chambers. Kind of analagous to valve sealing. But I dont have a good grasp of impeller dynamics. Maybe pressure differences are very low to begin with & nothing really happens flow wise until any given inlet valve becomes open, then any adjacent chamber is going to 'feed' it? Is that the general principle?


>There is only some .015" clearance between the air guide wall and the blades over their entire length. It's just possible with both of these features working together I might achieve a slight increase in air/fuel velocity through the plenum.

 

[mayhugh1]

Fascinating stuff!
Can you elaborate on your clearance. Is it the same (0.015") amount at the outer diameter ring edge (blue line) as along the curve edge blade profile (red line)?

I just naively assumed gap clearance would have to be very small on something like this, so there is no pressure leak-off betwen individual impeller chambers. Kind of analagous to valve sealing. But I dont have a good grasp of impeller dynamics. Maybe pressure differences are very low to begin with & nothing really happens flow wise until any given inlet valve becomes open, then any adjacent chamber is going to 'feed' it? Is that the general principle?

Peter,
The clearance I'm talking about is between the edge of the impeller blade (the red line in your drawing) and the air guide integral to the rear cover that fits down over this section of the crankcase. If you look back in my earlier post where I was machining the rear cover you'll see a photo showing the interior profile of the rear cover. This profile forms a smooth funnel-shaped volume which guides the air/fuel to the intake ports. I designed the profile of the impeller blades (your red lined ends) to fit up into this funnel profile and to clear it by .015" along the length of most of your red line.
When a valve opens, then the pressure will drop and the 'chamber' over the intake port will feed the valve instead of the entire plenum. This may (hopefully) result in a bit higher velocity in the plenum or it may not. The numbers show that, even with only seven blades, 3-4 chambers will pass by the intake port during a 208 degree intake duration. - Terry

 

[mu38&Bg#]

Radial flow impellers require a high tip velocity in order to generate significant pressure. In this case running at crank speed (max 5000 RPM?) the pressure will be negligible. I was under the impression that most model engines benefit from the impeller by improving fuel distribution. If the carb flowed into a large plenum, fuel flowing on the walls would tend to fall to the bottom and make the lower cylinders run rich and the upper cylinders lean. I don't really follow model radial designs, is it typical to omit the diffuser for the impeller? I see a big chance for separation of flow on the leading edge of the intake port.

The machining is great on the impeller. I'm surprised to see a radius on the top of the fins, usually this is kept sharp. The photo doesn't really do it justice. If you 3D mill parts you'll know that the camera exaggerates the tool marks. The finish is much smoother than it looks. A swipe with 1000 grit paper should give a mirror finish.

I'd love to see a screenshot of the engine in SW. I just started using 2013.

Greg

 

[mayhugh1]

Greg,
Thanks for the comments. I know pretty much nothing about impeller air flow and so I hope I haven't messed up. I believe, though, from my research that the earliest versions of the Hodgson 9 cylinder had no impeller or diffuser and some of the early builders complained about hard starting and poor fuel distribution. The design was later revised with a half-size diffuser and then shortly after that it was changed again to a full-size (essentially an impeller) diffuser. This is the one I built for my 9 cylinder, and I had no starting problems and only a minor mis-match in upper/lower cylinder fuel distribution. The consensus among builders seems to be that the impeller performs better. I don't know what is being used in the 18 cylinder twin plan set. The Chaos Industries approach, though, seems to be a full-blown impeller/air guide approach which is what I'm trying to mimic. I don't know what their blade-to-air guide clearance is but I'd bet it's pretty close from studying their photos. I don't know how easily their engine starts; but they built two identical models and, according to the YouTube video, they run really well. - Terry

 

[mayhugh1]

Greg,
Here are some shots of my (ancient) 2007 SolidWorks model. The first is a complete model of where I'm at so far. The second is a sectioned model in the area of the rear crankcase where I've been working recently. I tried to rotate the impeller for this shot to show the blade profiles with respect to the air guide wall, but since they are curved there is no single section that displays it in any really meaningful way. - Terry

 

[mu38&Bg#]

That looks great. The diffuser in terms of radial flow compressors is a set of vanes around the impeller, or on a turbo a scroll. I didn't mean to imply the arrangement wouldn't run well. I was going to write something about that, but must have posted it before doing so. In terms of high speed power, the sharp inlet might not be the best, but these models rarely bench race.

 

[ozzie46]

Masterful work that is!!! th_confused0052th_confused0052th_confused0052

Ron

 

[mayhugh1]

Before starting on the distributors, I thought I'd better tie up some loose ends and also tackle the tappets and tappet bushings. The first photo shows the completed rear crankshaft section which now includes the distributor drive gear. The gear was a purchased part, and it has a cross slot for a drive pin. Even though the distributors won't present any appreciable load to the crankshaft, I decided mate the slot with a dowel pin inserted into a cross-drilled hole in the crankshaft. I turned a pair of stainless steel spacers which slide onto the crankshaft on either side of the drive gear. I contoured the one between the gear and the impeller, and it is purely cosmetic. The one at the rear is pinched between the gear and a button head screw threaded into the end of the crankshaft and holds the gear tight on the crankshaft. Using a dummy distributor body I was able to determine the length of the distributor shaft that will later be required to properly mesh the drive gear with the pair of driven gears.
The tappets, as expected, turned out to be miserable parts to make, and I had to make 40 of them. They're 3/4" long and 3/16" in diameter. One end is spherical and the other end has a beveled spherical recess. I carefully modeled the valve train components during my 9 cylinder construction, and I'm duplicating that geometry in this 18 cylinder model. I wrote a program for my 9x20 lathe which turned the spherical end and then allowed me a brief manual polishing interval with a Scotchbrite pad before it parted off the tappet to its final length. When I got underway, I was hand feeding 3/16" O1 drill rod through my 5C collet chuck 0.8" at a time and making a new part every 4 minutes. Due to the long stick-out, I used a very sharp cutting insert and took only .003" doc. However, this still left a tedious secondary operation on the other end. This end requires a spherical recess with a beveled edge. I'm sinking the pushrods a little deeper into the tappet than was specified by the H9 plan set. Another H9 builder wrote that he had experienced issues with his rods popping out of place, and so I sank mine a little deeper into the tappet and, to date, have had no issues. This deeper cavity, however, creates the need for the addition of a beveled top inner edge.The machining of this end of the tappet was done manually using the lathe tailstock. The process was 1) remove the nub left over from the previous parting operation, 2) spot drill the end, 3) rough drill the cavity to partial depth with a undersize drill bit, 4) finish recess by plunging a ball mill to final depth, 4) cut the inner edge bevel with v-cutter, 5) deburr the outer edge, 6) polish the recess with a Craytex bullet, 7) check o.d. fit with go-no-go gage, and finally 7) repeat 39 more times. Since the tappets are unsealed oil leaks, I want as close of a sliding fit to the tappet bushing as I can get. The valve springs are fairly weak, and so any bind or drag is a problem. To maximize consistency I made all my tappets from the same piece of Enco drill rod. The drill rod o.d. measured on the high side of 0.1875" but after the brief polishing, the tappets were consistent sliding fits to the gage I made with my +.001 reamer. This is the same reamer that will be used later to finish the i.d.'s of the bushings. Several test parts had to be made to get the spindle speed optimized for the best possible finish from the ball mill at the bottom of the recess. I took some microscope photos because I was curious about the surface finish at the bottom of the recess. Even though, to the naked eye, the finishes on both ends of the tappet are brilliantly mirror-like, the high power photos show a somewhat different story when viewed close-up. I hardened the tappets to about Rockwell 55. My original plan was to make them a bit softer than the hardened cam; but an article I recently read recommended that the lifters be 5 to 10 points harder than the cam because, being smaller, they absorb more heat abuse and wear. (The author was referring to automotive applications which may or may not apply to model engines.) After heat treating, to my dismay, the o.d. of every tappet at the recess end had swelled slightly and would no longer slide easily through my gage. I had to spend a few more hours carefully polishing the (now hardened) o.d.'s of each one back down a few tenths to its original dimension. While I was at it, I made a tiny spherical wooden lap and polished the recess of each tappet with 600 grit valve grinding compound. In total, I spent some 15 hours of mind-numbing work on that little pile of parts. Even though the mirror finish in the recess was now even more brilliant than before, there was little change in the view through the microscope.
The phosphor bronze tappet bushings were not so bad. I broke the machining of these parts up into two steps. I turned the o.d., added a cosmetic bevel on the outer edge, and then parted off the bushing with one scratch written lathe program. The o.d.'s were carefully controlled as I wanted to insure a Loctited slip fit into the crankcase to avoid any chance of distorting the i.d.'s. After creating all the semifinished parts, I then re-chucked each part and used the tailstock to manually spot/drill/ream the i.d. of each part. The i.d. of each bushing was verified using a worst-case completed tappet as a gage. I didn't combine these two operations because I was afraid the hole center would wonder around since the previous drilling operation would end up creating the drill spot for the next operation. With my equipment, history has shown that the reamed hole would then likely end up inconsistently oversized after running the first few parts. The bushings were finally Loctited into the two crankcase sections. I'm going to have to start interspersing these tedious high volume parts among the more interesting one-off parts to keep from getting burned out during this project. Unfortunately, most of them won't be defined until I get to the heads and then suddenly it's 40 of nearly everything. - Terry

 

[mayhugh1]

Another smaller batch (10 pieces) of parts that I can make at this point in the construction are the compression nuts that are used to support the intake pipes in the fuel plenum in the rear crankcase. Initially, it wasn't clear to me from the Chaos photos how these were used. They didn't seem to be pipe threads nor did they use ferrules or flares. I eventually decided they must compress a rubber o-ring through which the intake pipes pass. So, I threaded the nine bosses in my rear crankcase section with simple 1/2" fine threads. The nuts in the Chaos photos look to be brass to match the tubing they used for their intake pipes. I'm still planning to use stainless steel tubing, and so I made mine from some (questionable) 303 stainless I had laying around. I'm using drops that I think were left over from the material I used to machine the valves for my 9 cylinder, but this particular material doesn't seem to turn like the free machining stainless I've used in the past.
This part was particularly interesting because my plan was to write a single lathe program that would turn, groove, thread, and then part off the nut after I had manually reamed the through-hole for the tube. A simple milling program would then cut the hex head, and the part would be completed.
The program was easy to construct and called three tools: a turning tool, a parting tool, and a threading tool. This was actually to be the first time I've used multiple tools in a single lathe program because it seems I'm always trying to keep a thousandth accuracy in my lathe parts, and my inexpensive tool post is just not that consistent. Even though the tool compensation software within MachTurn - my lathe control program - should easily be capable of handling the task, I had no end of problems with re-referencing each new workpiece. The software seemed to be 'remembering' the tool's work-offsets from the previous run. As a result, the whole batch of parts took longer to make that if I had just machined them manually. And, I never did figure out what I or the software was doing wrong. Also, midway through the batch of parts, I broke my next-to-the-last parting insert and had to start sawing off the semi-finished parts in the lathe chuck. This run of bad luck really broke up the nice workflow I was hoping to achieve. I eventually ended up with a dozen parts which includes three spares. I used my toy USB microscope to get a close-up view of my suspicious material in the root of the threads and was happy to see no tearing. If you look closely at the crankcase photos with the nuts installed you'll notice the axes of the fuel pipes will not intersect the diametrical center of the crankcase as they did with the 9 cylinder engine. From the Chaos photos It can be seen that this is necessary to avoid interference between the rear cylinder row pushrods and the two-into-one
intake pipes. Incidentally, it might also provide an improved transition for the air/fuel entry from the plenum to the intake pipes.
While I was in a batch building mode I decided to do what work I could on the pushrods. Of course, forty of these are needed. Although I'm expecting them to be the same length as those in my 9 cylinder, I can't be 100% sure since I haven't yet designed the heads. For safety, I machined the spherical head on only one end and left them all about .200" longer than my 9 cylinder rods.
This wraps up the all the high volume parts that are sufficiently defined at this point in the project, and so now it's back to the more fun and interesting one (or two) offs.
The last photo is an entrance I made to our backyard during my wife's and my seven summers landscaping project behind my shop. (I sometimes like to work on big
metal projects also.) I welded it up totally from scrapyard steel, and I cut the art panels on my plasma cutter. - Terry

 

[John Rus]

I'll probably be in a coma by the time this is finnished. :big:

So excited to see this run!
John

 

[mayhugh1]

I just realized that I forgot to mention something in my earlier post about the tappets that I had made. Although the end that mates with the pushrod does have a truly spherical recess to match the spherical end of the pushrod, the other end that rides on the cam ring is not truly spherical. A comment by Petertha on another thread about the potential issues at the contact point of such an interface got me to thinking about the problem. I actually turned a slight partial elliptical surface on the very end of the tappet to add a bit more contact area to the interface with the cam. The purpose of the microscope photo of the convex end was to show this. It, in fact, comes across looking like a "hole" in the very end of the tappet in that photo. - Terry