Discussion in 'A Work In Progress' started by mayhugh1, Aug 12, 2019.
Thanks for the reply on bead blasting.
Much of last week was spent modeling the complex gear tower which includes a pair of front and rear halves, end caps, and a tightly integrated take-off block for the magneto. After a lot of frustration, I still don't have an assembly that I trust to begin machining. My wife would say it's because I can't follow instructions, and to an extent she'd be right. I've found a few online screen shots from those that have gone before me to be invaluable.
I took a break from the tower and returned to removing chips from the crankcase since its modeling had been completed. With the foundational machining done, I felt it was safe to finish up the external profiling that will finally give the crankcase its distinctive shape. Other than several o-ring grooves that are still planned, its bottom and both sides were finish machined. The bottom was milled using a tiny ball cutter in order to create an array of cooling fins with rounded tips and filleted roots.
The exhaust fan on my bead blasting cabinet is currently out of service, and so the photos show the machined surfaces straight off the mill. Some cleanup will be required, but that will become more obvious after the surfaces are bead blasted for the first time. - Terry
The crankcase side covers are nearly identical to those on the full-size engine and add realism to the model but a couple tricky parts to machine. The covers contain an array of tall cooling fins like those on the bottom of the crankcase but, instead of being flat, they're smoothly contoured over a pair of bulging pockets that hold the crankcase breathers. Ron provides instructions and a design for the shop-made cutter that he used to make his covers, but I wasn't certain of my ability to duplicate his manual effort. Examples are out there from others who have, however.
My initial modeling was done expecting to follow Ron's method, but the radius'd fins on the crankcase turned out so well that I decided to try something similar on the side covers. The design I eventually arrived at created so many surfaces that my ancient XP computer had barely enough resources to handle the resulting CAM load. The total machining time worked out to be nearly eight hours per cover with much of it required by a 15k rpm 3/32" ball cutter running at 17 ipm.
A piece of MDF temporarily glued to the bottoms of the starting workpieces prevented the finished parts from being damaged when they were finally cut free. It also dampened some potential chatter-causing vibrations during the large-tool roughing and semi-finishing steps. Due to record setting 100F+ days we've been having down here in Texas, the machining was spread out over several evenings to reduce the probability of me or the Tormach suffering a heat stroke. The photos show a few of the machining steps.
The covers not only seal up the engine's access ports, but they also contain the crankcase breathers. Ron scaled the engine's original crankcase ventilation scheme right down to the oil baffles inside the breathers. Unfortunately, its performance didn't scale as expected, and the atmospherically vented breathers created a frustrating oil control issue. Ron's eventual solution was to vent all four breathers into the engine's oil tank.
The breathers were machined from a single piece of stock without their internal baffles. In order to minimize the number of plastic hoses later on, I connected the two breathers on each cover together with a short length of thin wall stainless tubing. Hoses will eventually connect the rear breathers to the oil tank. The o-rings that I had originally planned to use to seal the covers to the crankcase were replaced with simple .010" thick teflon gaskets. - Terry
I've often wondered, do you happen to know if Ron had access to a real Offy to touch or take closeup reference pics? Or was this design entirely based on books & such? Certainly captures the striking features that's for sure.
Ron told me he didn't have access to an actual engine but used published photos and info about the engine to come up with his design. I have to agree - it had to have been an incredible amount of work.
After putting up my last photos, something about my breathers doesn't look correct. I just discovered that I transposed two dimensions and machined them too short. So, I'll be doing them over. It's a good thing it wasn't me coming up with the original design. - Terry
Terry: I thought I would let you know that I only vented the front cover of the engine to the oil tank. The rear breathers were blocked off to the atmosphere. The direction of the engines rotation tends to throw oil against the front cover. That is where oil wants to escape while running. Venting the rear cover to the oil tank wasn’t necessary. Can’t hurt though and keeps things symmetrical I guess.
As far as what I had to work with designing the engine; I had only photos and the article in Hot Rod magazine showing a 270 being assembled by an Asian fellow. I believe his name was “ Chickie Hiroshima” or something close to that. I am bad with names. He apparently worked for Offenhauser at the time. This would have been back in the 50’s. The article was most informative. It gave many of the dimensions of journal sizes, cam timing diagrams, listings of various things that could be ordered special to make an engine unique to the person ordering one. I should be able to come up with the name of the guy who put the article together, but as I said, I am bad with names. He wrote for Hot Rod magazine and wrote a book on souping up the small block Chevy V-8 engine. My older brother used to drag race and he relied heavily on this book when building his engine. He even aquired a Lathem blower that was next in line to be installed but he was drafted before he ever got around to that. This was in 1963 or 64. Good times with many memories.
I checked with my brother. The guy’s name was Don Francisco.
Thanks for the reply. I've misunderstood your email to me last month:
'The next thing I changed was the design of the crankcase breathers. Again trying to follow the prototype, I tried controlling the oil entrapment in the blow by with internal baffles. Due to the reduced distances involved with the model, oil would come out the breathers especially when throttling down from high speeds. It takes a while for the scavenge pump to catch up with all the oil in suspension and the oil level in the crankcase would rise temporarily, and would push out the breathers. I solved the problem by eliminating the baffles and piping the breathers on each side together and running a fuel tubing line from one breather back to the oil tank under the base. Air holes to the atmosphere were eliminated. The engine breathes into the oil tank."
But just to be clear, since I'm getting ready to re-machine the breathers, there is a left and a right cover and a front and a rear breather on each one. Are you saying that you vented only the front breather on the left cover to the oil tank and then blocked off the other three breathers? Thanks. - Terry
Sorry for the mix up. The two breathers on the one cover are tied together as you have done and vented to the oil tank. I call the cover on the intake side of the engine the front cover and the one on the exhaust side the rear cover. That isn’t correct terminology, but just how I mentally think of the engine. I am always looking at the engine from the intake side and think of that as the “ front” of the engine although it is not. I should be more careful as to how I explain things.
Thanks Ron. That clears it up. - Terry
I re-machined the four breathers so they're now at the proper height above the side covers. Both breathers on the port side are functional and plumbed together so they can be vented into the oil tank per Ron's recommendations. The breathers on the starboard side are now only cosmetic.
The block covers, like the side covers, are finned and were machined similarly; but the fin spacing is only 1/16" compared with 3/32" on the side covers. This spacing was designed for a slitting saw, and so the 1/16" ball end mill that I used to radius the fins extended the machining time even further. Since it's flat, the port side cover is fairly simple, but the starboard cover is complicated by its internal coolant manifold and the integral flange required for the water inlet pipe. The manifold is a lengthwise milled cavity in the backside of the cover that's permanently enclosed with a J-B Welded ported plate. I left a .020" thick glue line all around the cover for the epoxy - something learned during my previous build. My cover plate has a dog ear covering up a small machining error in the cavity.
The block will be full of coolant, and so the cover seals require thought. Each cover will be secured with forty 0-80 cap screws, but those by themselves may not be enough to provide a water tight seal. Since I prefer using sealers only as a last resort on models, I originally planned to o-ring the covers to the block, and my initial modeling included grooves in the block for .040" o-ring cord. With so much machining already in the block, I decided later to move the grooves to the covers. However, after completing its machining, the thin port side cover wound up with some warp that would have made its grooving difficult. In the end, I settled on a pair of .010" thick Teflon cover gaskets.
The sides of the block aren't parallel but have 1.7 degree sloping sides that may have their origins in the draft angles required by the original engine's sand casting process. I delayed machining these sides until after all the other block operations were completed since the resulting trapezoidal shape would have been difficult to fixture. I machined the block's sides on my manual mill and then used the Tormach to drill the eighty .047" cover mounting screw holes. In an attempt to keep the fixturing consistent, I first machined a pair of 1.7 degree angle blocks that I used to support the part in both mills' vises.
When it came time to drill the holes, I found fixturing the block to be more difficult than expected. With the part resting in the vise on the stacked pair of 1.7 degree angle blocks, its small-area ends weren't sufficient by themselves to maintain alignment of the block along the mill's y-axis as the jaws of the vise were tightened. Even after clamping a machinist square to the fixed jaw of the vise for use as a y-axis reference edge, some trial and error with paper shims was required to get the part indicated properly.
I picked up a lot of small hole tapping experience during the Merlin project, and so the block's eighty 0-80 tapped holes weren't a concern. One of the photos shows the simple tap holder and support block that I used along with plenty of WD-40 to tap all eighty holes with the same tap and without incident.
Finally, the water inlet pipe was bent from 1/4" 303 tubing using a three wheel Rigid bender before silver-soldering it to a stainless steel flange. A Teflon flange gasket completed the assembly. - Terry
I just LOVE that bead-blasted finish!
Harry Miller knew how to make an engine pretty as well as functional. A design that was hard to beat and has stood the test of time.
Looking Great! Thanks for the detailed description of how it's done. Then looking at the pictures one can really appreciate all the hidden precision detail and effort required to produce such an outstanding piece.
The gear tower is a fairly complex assembly of gears designed to connect the overhead cams to the crankshaft. The tower also includes a 'magneto block' containing one of two sets of bevel gears that will eventually drive the magneto. The water and oil pumps will be driven from a separate 60 tooth take-off gear located below the crankshaft. The Offy's gears span four separate subassemblies not including the split crankcase, and their exact locations will be affected by the use of any gaskets or sealers between these assemblies. Because of the large number of high rpm gears and their loads, accurate tooth profiles and spacings will be important for their longevity.
Before finishing the tower's modeling, I thought it best to have the actual gears in my hands so their running spacings could be verified and adjustments made to the model if necessary. I've twice before run into issues with new gear cutters producing poor tooth profiles although these were probably a result of mislabeling. Except for the scavenger oil pump, the Offy gears are all 48 DP. I have several known good 48 DP cutters, but unfortunately a new one had to be purchased for this build.
Several of the gears are identical, and so it made sense to slice multiple copies of the same gear from long pre-machined blanks. Ron recommends making the gears from casehardened mild steel as was done in the full-size engine. However, I have no experience with casehardening and was concerned about using the gears as learning tools. With no way to control the process or to measure the final result, I was worried about over-hardening and embrittling the tiny gear teeth especially those on the 60 tooth gears. Instead, I machined most of the gears from 1144 which has a Rockwell C hardness of around 25. The hardness of mild steel gears (even those purchased from Boston Gear) are equivalent to only 1 or 2 if they were to be compared on the same scale. Stressproof's tensile and yield strengths (important specs even under a hard skin) are twice those of mild steel.
The stresses that the crankshaft sleeve gear will see are a bigger concern to me especially since physical limitations require it to be attached to the crankshaft using only Loctite. I machined this gear from O-1 drill rod which, after a 1475F quench, was tempered at 375F. Its final hardness and tensile strength should be roughly twice that of Stressptoof, leaving its Loctite bond to the crankshaft as its weakest link
As often seen on commercial gears, I typically chamfer the corners of o.d.'s of my shop-made gears. I left the Offy's gears, though, with full width tooth contact for a bit more durability. The widths of the gears were finally finished on a surface grinder.
The oil pump gear stock was machined from 360 brass. The pressure pump gears are also 48 DP, but the scavenger gears are 32 DP. The individual gears will be parted off from their blanks later when I'm more familiar with the lubrication system. A set of shafts and spacers finished up the nuisance parts associated with the gears.
The nose of the test rod that I've been using as a dummy crankshaft was machined to temporarily accept the hardened sleeve gear so the fits of the two 60 tooth driven gears could be verified. The meshes of these two gears with the crankshaft gear span the split between the two crankcase halves, and from previous measurements I knew the distances between the gear centers were on the order of a thousandth within theoretical. The three gears turning freely with minimum backlash was something of a minor milestone for the project.
A fixture for testing the 40 tooth gears was also machined to verify the meshes and spacing between them. Its design was small subset of the gear tower and was also used to tweak the end mill parameters for the bearing fits. These three gears also turned freely with minimum backlash at their theoretical spacings.
Finally, a bearing removal tool was ground from a long hardened hex wrench. Counterbores will be added behind the bearings in the gear tower so the bearings can be removed as necessary using this tool. These counterbores and the tool were also part of the testing done in the above fixture. - Terry
Terry: after about a year or so of running with the case hardened gears, I had a tooth break off one of the gears. That tooth went through most of the gear train and wiped out most of the rest of them. I machined another set from 1144 . Ten years later, the second set has held up very well. Like you said, one overhardened tooth was all it took to make a mess. Live and learn.
Thanks for the reply. I may re-think the crankshaft gear now. - Terry
Hi Terry. What is your procedure for verifying tooth depth of cut on these little gears? For example do you zero contact the cutter to the stock, in-feed X amount according to the book value & that's it? Or maybe cut some teeth on either side of the stock at some slightly shallower depth, then (somehow?) measure & adjust in=feed vs. the target from there? I've never done any, hence the question.
So the conclusion on material is to use 1144 SP and leave it at that unhardened - no heat treating of any kind?
I cut gear teeth according to the book. I make sure the diameter that I assume for the cutter includes its runout in its holder. The blank is fixture'd so its runout is less than a thousandth. The total depth of cut for the Offy gears was .045" which I did in three passes: two at .020" and a final at .005". I always verify the result afterwards by making sure there is no more than a degree or two of backlash with the gears meshed at their theoretical running spacing.
I'm still deciding whether I'll re-temper the crankshaft gear to make it a bit softer, but the others will definitely be left in 1144 especially after Ron's comment. - Terry
Separate names with a comma.