Discussion in 'A Work In Progress' started by mayhugh1, Aug 12, 2019.
No grass growing under your feet.
For the machining of the crankcase's lower half, I used the same tooling and tool paths used in my experiment with the upper half. With the probe error corrected, the centerline of the final result ended up within a thousandth of the workpiece's centerline. The holes for the cap screws that will secure the bearings in the webs will be transfer-drilled later when the bearings are in place.
The two halves were then assembled so the access ports could be machined into both sides. The sizes of those ports didn't look at all unreasonable on my computer screen, but I was struck by just how tiny they really are once I had actual parts in my hands. My hat's off to the builders who actually assembled their engines through them. They really aren't necessary with a split crankcase, but I didn't want to change the engine's appearance by leaving them out.
Although there's plenty of machining left to do on the crankcase, most of it will alter the shapes of its workpieces and make them more difficult to hold in a vise for later operations. The remaining crankcase machining won't be done until after the bearings are installed and the block is machined.
I iteratively machined some disks to use as gages for measuring the actual diameter(s) of the bearing bores. Remarkably, all three came out to be 1.5005" which was just a half thousandth over their target value. Measurements using the same gage disk in each web showed the bores to be in line to within a few tenths, and their center axes to be flush with the top surfaces of the webs. This was good news since it means the crankcase won't have to be line bored before installing the bearings. If I'm careful, and with some luck, I may also get away with not having to line bore the installed bearings.
The next step will be to machine the bearings. - Terry
Wow, impressive start. What alloy did you select for the crankcase?
Sorry I missed your question. I don't know for sure what the alloy is since I used a piece of unmarked scrap, but it appeared to machine like 6061. - Terry
SAE660 bronze was used for the split crankcase's main bearings. Machining began by skimming the 1-3/4" o.d.'s of a pair of three inch long rounds. After facing away half their diameters, they were bolted together and machined as an assembly to create a set of three clamshell bearings. Reserving one inch for a work-holding spigot, the material I used was long enough for a fourth (spare) bearing.
The top halves of the bearings came from the first workpiece set up on the mill. About half of its diameter was machined away except for its spigot. There were no critical requirements on this operation except for its surface finish, but the same setup was to be used for the second workpiece whose machining was a little more demanding. While fine-tuning the setup's accuracy and rigidity in preparation for the second workpiece, I managed to screw up the first workpiece. An already well-worked drop was scrounged up for a replacement that had barely enough excess stock to be usable.
In order to end up with a uniform height for the centers of all three bearings, the second workpiece had to be carefully fixture'd under the spindle so precisely half its diameter could be removed over the whole area of interest. Before starting, the y-axis DRO was zeroed over the axis of the workpiece so it could be used as a reference to locate the bearings' mounting holes.
Without disturbing the finished second workpiece in its setup, the first workpiece was temporarily clamped to it using a pair of hose clamps, and the bearings' mounting holes were drilled and counterbored. All eight holes were tapped through both workpieces so screws could be used to hold the assembly together for the remainder of its machining
The assembly was moved to the lathe and its o.d. finish-turned to match the bore measurement made earlier in the lower crankcase. The center was also through-drilled and reamed for the crankshaft's .500" journals. Unfortunately, the drill wondered off course in the deep bronze material leaving behind an unacceptable .006" runout. This was corrected using a sharp insert in a long skinny lathe boring bar. This operation was run under power feed with a spindle speed of only 60 rpm to eliminate chatter. When completed, the measured TIR was essentially zero, the bore was smooth and concentric with the bearings' o.d.'s, and I had a new spec for the crankshaft journals.
The boring bar was used to open the bore up to .514" since I had a length of .513" drill rod that I could use as a test bar to check the alignment of the installed bearings. It will also be used later to pickup the axis of the installed bearings when it comes time to bore the front and rear of the crankcase for its outer ball bearings.
Before parting off each bearing, its accessible face was finished using a gage pin in one of its bolt mounting holes as a center reference for the bearing's width. An internal oil groove was also turned inside the bore. Because of a small kerf allowance, each parting operation was started with a thin parting tool and then finished with a hacksaw (stationary spindle, of course). An expanding mandrel was then used to grip each bearing so it could be faced to its finished width.
Before separating them, the bearings' halves were numbered to insure consistent reassembly. The bearings passed a quick sanity check while seated, but not bolted, inside the crankcase with the test bar rotating freely.
The lower crankcase was returned to the mill so shallow counterbores could be machined at the locations for the bearings' mounting bolts and provide flat starting surfaces for their spotting, drilling, and tapping operations. The temporarily tapped holes in the bearing halves were reamed for clearances needed around their mounting bolts and then carefully checked under a magnifying lens for burrs that might interfere with the bearings' fits.
I found it best to install all three bearings along with the test bar as an assembly and to tighten the bolts as though I were installing a head. Just before finally tightening them, I used my upper body weight to bear down on all three caps which seemed to snap the bearings into place.
The test bar is only a thousandth under the bearing bores, and even though it was snug, it could be spun using a two-finger grip. I installed/uninstalled the combination several times to make sure its assembly was consistent. I did discover that all six bolts have to be fully tightened in order to spin the test bar indicating that one or more of the bearings is springing back out of its seat when its cap is loosened. After the assembly had been allowed to sit overnight, seating evidently improved a bit since the test bar spun much more freely. Measurements on the test bar showed it's in the center of the crankcase and parallel to its sides and bottom to within tenths.
I doubt that I'd get any better result with a final line boring operation, and so this wraps up work on the main bearings until the crankshaft is available which will likely require some minor fitting. Since its the safest place to store them, the bearings will be left installed in the crankcase for the remainder of the build. - Terry
I received an email asking how I'd go about making and installing the main bearings in the split crankcase without line boring and without the use of CNC. Here is how I was originally planning to install a set of square bearings before deciding to make things more difficult by using the circular bearings:
Set the crankcase up in the mill, and using a cylindrical end mill, machine a rectangular slot through the three webs on the centerline of the crankcase. Machine a set of bronze blocks for press-fits in the slots. With the blocks installed, drill/thread a pair of bearing mounting holes through the blocks and into the crankcase for temporary holding screws for the next operation.
Using a .500" diameter ball end mill, mill a semicircular slot down the crankcase centerline that's .250" deep.
Machine a set of three bearing caps and secure them in the mill vise. Mill the same hemispherical slot through the tops of the three caps as a group, and then drill/ream the clearance holes for the bearing mounting screws.
Install the caps with either a test bar or the actual crankshaft in place and fit if required. - Terry
Thank you for posting this incredible work..It got me moving again..We all benefit from your efforts...……...
There are two remaining operations related to the support of the crankshaft. These include boring the front of the crankcase for a front bearing and boring its rear for a circular housing that will contain a rear bearing and oil seal. Both of these bearings, though not identical, are ball bearings. The test rod running up through the central bronze bearings will be used to pick up the centers of both of these operations.
As long as it's properly located, the exact diameter of the rear bore isn't critical since the rear housing can be turned to fit. In fact, it'll be turned a few thousandths oversize to allow the housing to locate itself on the crankshaft before being tightened down. The rear bearing will be positively retained in a counterbore in the rear housing, and will be used to limit the crankshaft's aft thrust.
The diameter of the front bore is more important. The front bearing in the original design is retained in the crankcase by a one thousandth interference fit. I'll likely try to reduce this fit to a couple tenths to accommodate the split crankcase and then take advantage of the oil pump housing located inside the gear case to retain this bearing. This bearing can then be used to control the crankshaft's forward thrust.
In order to continue progress while still studying the gear case design, I tackled the simpler rear operation first. After indicating its center, the previously roughed-in bore was finished on the mill using a boring head. The mounting holes for the rear housing were also drilled/tapped. The mounting hole locations in Ron's design were selected to overlay the eight very deep mounting holes for the removable bearing webs. In my case, I used just four holes that were equally spaced about the crankcase halves.
The rear housing could then be machined. A starting workpiece was created having a healthy work-holding spigot that wasted more material than was actually used. Machining began in the lathe on the forward side of the housing so all its critical features could be completed in the same setup. This ensured the bores for the bearing and oil seal were concentric with the housing's mounting lip and that all three were normal to the flat mounting surface that will end up against the crankcase.
While still set up, I added a face groove for an o-ring immediately adjacent to the mounting lip in order to seal a potential oil leak between the two. This required a special tool ground from a .040" drill bit that was soft soldered to a piece of steel. This tool was originally made for a similar operation in the Knucklehead build.
After the rear housing's frontside operations were completed, the part was flipped around and gripped by an expanding mandrel inside the bearing bore so the rear machining could be started on the lathe. The rear side contains a circular groove with vertical walls that would have required a pair of right and left hand boring tools to complete on the lathe. Since the part had to be moved to the mill for a number of hole drilling operations, I decided to mill the channel instead of turning it.
The bearing was pressed into the rear housing, but the oil seal won't be added until much later at final assembly. With a bushing pressed onto the rear of the test rod in order to bring its diameter up to match the i.d. of the rear bearing, the rear housing was trial-fitted to the crankcase. After tightening the housing down to the crankcase, the test rod continued to freely rotate as it did before. - Terry
Terry - The Design/modifications, tool making/modifications, and actual build are consistently amazing to me. This one too is looking good!
The front-end machining on the crankcase includes a bore for the front crankshaft bearing and a pocket for a gear case containing bearing recesses for a pair of driven gears. One of these gears will eventually connect the crankshaft to the gear tower, and the other will drive the water and oil pumps. The Offy uses a dry sump oiling system, and its pressure and scavenger pumps will eventually be located inside the gear case.
The crankcase was set up in the mill vise and indicated for access to its front end. The location of the bore for the crankshaft bearing was picked up from the test rod running through the three bronze bearings. The rear bearing in the rear housing and the starter bearing in the not-yet-machined front cover each have a bit of 'wiggle' room since they're bolt-ons to the crankcase. The front bearing, though, will be hard fixed to the crankcase and must be on the same axis as the main bearings.
I was concerned about attempting an interference fit for the front bearing since I was unsure of just how much interference could be tolerated while drawing the crankcase halves closed. If the flat surfaces of the crankcase halves don't completely close against each other, not only will there be an oil leak, but the axis of the front bearing may end up displaced from the axis of the main bearings and create a bind in the crankshaft. On the other hand, too loose of a fit will allow the bearing to spin in its bore and damage it.
Boring the perfect fit would require more luck than I was willing to risk and couldn't be tested without removing the crankcase from its setup. So, I opted for a one thousandth oversize bore that would provide a clearance that I could later shim out - something that's possible with a split bore. Aluminum foil as thin as .0005" is used by some candy manufacturers on their chocolate treats and can be a sweet source shim stock.
The remainder of the gear case machining was completed including bores for the two driven gear bearings which were also bored a thousandth over. Although I'd have preferred zero interference fits for these bearings, the front cover will eventually contain a matching pair of bearings, and there's no way to align bore them in pairs. The front cover's wiggle room will likely be taken up by the starter bearing, and unavoidable errors in locating the positions of the two bearing bores will make the cover difficult to assemble unless the bearings have some wiggle room of their own. When assembled, there will most likely be enough friction created by positioning errors to prevent the gear bearings from spinning in their bores.
A trial assembly of the crankcase halves around the front bearing was successful. I found that adding a thousandth shim around only the upper half of the bore allowed the crankcase to close tightly around the bearing and allow the thousandth-under test rod to freely rotate inside all five bearings.
One of the things that attracted me to the quarter scale Offy is its faithful adherence to the original engine's great looking appearance. The painstaking detail in many of the model's individual parts will provide a number of interesting mini-projects with their own short term satisfactions that'll help keep me interested in such a long term project.
The first of these parts is the front cover. It encloses the gear case and contains the starter bearing which will be the sixth crankshaft bearing. Other than rearranging its mounting bolt pattern to accommodate the crankcase split, I duplicated Ron's design. It's finished periphery will provide a template for the later machining the crankcase's lower sloping sides. The magneto mounting bracket was an integral part of the casting for this part in the original engine. Ron attached a separate bracket to the cover with hidden screws and blended the seams with fillets of metal-filled epoxy. I used my Tormach to machine the cover and bracket as a single part. There isn't room for a front shaft seal, but Ron included a groove for a 12 mm x 1mm CS o-ring around the bearing's i.d. that should be effective against oil leaks.
My first serious mishap in this project occurred while removing the overhanging excess stock from the rear of the cover in preparation for its rear face machining. The large multi-insert facing cutter that I was power feeding in my manual mill grabbed the overhanging lip and pulled the end of the part partially out of the vise. This stalled the cutter until I was able kill power to the spindle. There was no damage to the already machined top surface, but there was a deep gouge on the part's back surface. With all the effort invested so far, I felt there was nothing to lose by trying to salvage the part with a tig-welded repair. The welding created some of its own damage to the topside surface, but the final result was much better than expected with no visible trace of the repair.
As it turned out, my biggest concern was with the mill itself since I had to hammer the R8 cutter out of the spindle using a long drift in place of the draw bar. The cutter had spun inside the spindle bore and was jammed against what remained of the collet key pin. Fortunately, TIR checks on the spindle bore showed there was no apparent damage, and re-tramming the mill seemed to return things to where they were before the accident.
With the cover installed on the crankcase, there's still no sign of binding of the test rod, but the friction of the cover's o-ring has added significant drag. - Terry
Can't you visualize Terry in a chocolate shop - "I'll take 20 feet of that foil please .... aaand I guess 2 of those Eau Claire's to go... in a box please!"
Glad your scary incident didn't damage the machine. Nice TIG repair.
Again precision and results beyond my comprehension capabilities - AMAZING! What is the surface finish inside/out on the front cover after the TIG repair? Looks perfect - almost like bead blasted?
It is bead blasted. The texture left behind by the grit I have (whatever it is) seems to do a good job of simulating a casting at this scale. Plus, it hides machining marks with minimal effort. - Terry
If you wouldn't mind could you briefly outline your equipment/materials/method for bead blasting? Please and thank you. Your builds, for me, are like an educational PhD!!
I have a 24" x 24" x 36" bead blaster cabinet bought 25 years ago from Eastwood Products. It's no longer sold, but equivalents are available from Harbor Freight. I use glass beads purchased from the Local Harbor Freight which are, I believe, 80 grit. I've been using the same media for over 20 years and so its probably a bit finer than when originally purchased. The air supply is 90 psi from a large compressor.
So far, I've been focused on parts associated with the crankcase in order to look for problems that may have been created by the split. The cylinder block should be the last of these and, when completed, the remainder of the build will pretty much follow Ron's original design.
The reason for replacing the long studs running through the block with separate upper and lower mounting screws was to simplify the engine's assembly and make head gasket servicing less traumatic. This creates the need for flat surfaces under the heads of the head bolts that will eventually run up through the roof of the block. Since most of the block's interior will be gutted, the real issues are only with the screws in the block's extreme outside corners. The heads of these four cap screws will wind up partially inside the front or rear walls of the block, and so one of the block's first machining operations was to create surfaces for them.
After preparing a starting workpiece with finished outside dimensions (the block's sloping sides will be machined later), four holes were partially drilled up through its bottom for the heads of the cap screws that will be in the corner head bolt positions. Their flat surfaces were created by running a same-size drill bit with a ground-flat nose back into the holes. The holes were then back-filled with aluminum plugs that were turned, coated with Loctite, and pressed in place leaving about an inch gap inside the workpiece for later insertion of the screws.
After the Loctite cured, the plugs were blended into the block's bottom surface with a light facing pass. The block was then set up in the mill with access to one of its sides and its interior hogged out.
Sixteen 3-48 blind tapped holes were then drilled into the bottom of the block to attach it to the crankcase. Since the block will eventually contain coolant, these holes don't penetrate its interior. Sixteen mounting screws sounds like a lot, but they may allow me to do away with the o-ring I had been planning to use to seal the block to the crankcase. The cylinder sleeves will be sealed to the block, but not to the crankcase, and so there's a potential for an oil leak through the seam between the two. The top of the block was then drilled/reamed for the head bolt clearances, the oil return tubes, and the water passages. The 5-40 cap screws that will be used for the head bolts were trial-fitted in the four corner positions.
After initially roughing them in, the bores and counterbores for the cylinder sleeves were finished using a fixed-setting boring head. The Offy's head gasket history is a bit tarnished, and some of Ron's own experiences are detailed in a section of his manual. Some of these problems may be caused by the minimal clearances between the head bolts and the tops of the cylinder sleeves. Although I was able to massage the locations of the head bolts and increase these clearances from .016" to nearly .050", after holding an actual block in my hands, I regretted not reducing the diameter of the sleeves (and pistons) to pick up a little more margin. Finally, I machined a couple .020" Teflon head gaskets to make sure I'll be able to hold onto the clearances I do have. - Terry
Would silicone rubber gaskets be an option for model IC engines like the Offy?
Silicone rubber would probably work well around the coolant passages and oil return holes, but I'm not sure how well it would stand up to combustion pressures. Silicone also dissolve in gasoline and so that would need to be considered as well. - Terry
I didn't know that silicone was affected by gasoline so that's a new learned thing for the day.
I was thinking that silicone is easy to laser cut with even one of those hobby lasers and doesn't suffer from cold flow the way that Teflon can, at least to my limited knowledge. They both have similar upper temperature limits as well. That would make it easy to cut complex shapes with less effort. As you say, it could still have application where fuel is not involved.
Beautiful job on the Offy, by the way, and it looks like you still getting good use from your Tormach mill.
Separate names with a comma.