270 Offy

Home Model Engine Machinist Forum

Help Support Home Model Engine Machinist Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
Terry, once the craft vinyl (Oracal 651) is cut, do you make any attempt to dissolve the adhesive prior to installing on the metal face? I can see on one hand where some tackiness might be good in that it stays put, but then maybe the opposite when you remove. Are they reusable?

The cumulative gear backlash during throttle variation is a really good point. I'm still not clear on your mitigating procedure though. Was the test fixture somehow variable that you could tweak the center to center gear distances right? And then once satisfied, you somehow measured resultant shaft centers & replicated these using DRO to the real part? In the test fixture do you measure play gear by gear, like hold one gear, wiggle the the other, adjust... then proceed down the line?

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.
 
Petertha,
I don't try to reuse the gaskets. It's easier to replace a damaged one with new material. Once it's in place, I treat it as a new permanent surface on the part.mSo far, I've used this stuff only in unpressurized areas to prevent oil leaks where sealer might have typically be used.

With respect to the gears, the particular test set I used was only to verify the gears would turn smoothly with a hint of backlash at their theoretical spacings. If they hadn't passed that test they would have been re-made.

I have made test sets in the past where I could vary the distance between the gears to see what my margins were around the theoretical spacing. There were too many cascaded gears in the Offy tower to start playing with spacings to conform to poorly made gears though. - Terry

p.s. Another effect of gear slop can be large valve timing changes during sudden speed changes. Their effects are probably overwhelmed by the carburetor vacuum changes though.
 
mayhugh1 !
Can you show me how to make a surface like that?
That surface is great !
22.jpg
 
mayhugh1 !
Can you show me how to make a surface like that?
That surface is great !
View attachment 111737
Minh-thanh,
Thanks for the complement. After machining, the surface was glass bead blasted. There's more information about my particular setup, which is nothing special, several posts above this one where someone else asked a similar question. - Terry
 
The backside of the front half gear tower contains a mirror image of the rear half's six bearing and gear pockets. An additional complication is the magneto drive block which is a forward extension of the tower and contains the right angle drive components needed for the magneto driveshaft. In the full-size engine, this block was an integral part of the tower's front-half casting. To simulate the original engine's appearance, Ron suggests bolting the magneto block to the tower and blending the two together with aluminum-filled epoxy fillets.

I came up with two more approaches that were considerably more complex but avoided the need to color match epoxy to the tower. The first approach was to split the drive block housing into two pieces. The rear piece, including its blending fillets, would be machined integrally with the gear tower. The front half, a separately machined part would be bonded to its rear half with screws and a nearly invisible Loctite seam.

The second approach, and the one I ultimately used, was to machine the entire block as part of the tower. This required considerably more machining on a hefty workpiece that was turned mostly into chips. Since the block stands over a portion of the tower, the workpiece had to be repositioned three times to get the needed spindle accesses. The third setup was required to machine some cosmetic features on the side of the block. Inevitable errors accompanied the re-referencing and tool changes that were needed, and a couple seams that couldn't be hidden with bead blasting required some manual cleanup. The long reach finishing cutters that were needed inside the deep workpiece created their own issues with surface finish and machining time.

The front-half tower's first machining operations were the bearing and gear pockets on its backside. After verifying their fit with the already completed rear half with all the gears, shafts, and bearings installed, the workpiece was flipped over for its lengthy front side machining. More than half of the nine hour machining time was spent roughing in the shape of the tower. After the finishing operations, a quick-dry epoxy was poured over the periphery of the part to keep it connected to the workpiece during its final machining.

A portion of the workpiece had to be cut away to gain access to the port side of the drive block. This side contains some contouring added to the design to smooth the transition of the block into the tower. Once this area was machined, the part was returned to its original position with its backside facing up so the periphery could finally be machined. The front half of the gear tower has a convenient bead around its periphery that allowed the part's front and back surfaces to be machined and the cut free of its workpiece without leaving an annoying seam. This bead, a part of the original engine's casting, was probably used to help remove the front tower's half during maintenance/disassembly. This final operation separated the part from its workpiece with a slot milled around the outside of the bead. The epoxy kept it safely attached to the workpiece until it could be removed with heat. After a half hour in a 300F oven, and while still hot, the (Devcon 5 minute) epoxy could literally be rubbed off the part. - Terry


127.JPG
128.JPG
129.JPG
130.JPG
131.JPG
132.JPG
133.JPG
134.JPG
135.JPG

136.JPG
 
You just made my morning!
I always enjoy following your builds. Very inspirational.
I guess I should make something this morning before they turn my power off.
California.
 
I have the drawings and the book for building the "OFFY" and is going to be my next build. After seeing your build though I don't know if I can do it justice but will give it a damn goo try.
 
The cam gear covers bolt onto the upper ends of the gear tower and will eventually enclose it around the camshafts' driven gears. These covers are small in size and rich in detail and will be troublesome parts to machine. They'll wind up in prominent locations at the top of the engine, and so the effort they'll soak up won't go unnoticed. I scrapped four parts and the better part of a week coming up with a pair of covers that I was satisfied with. Again, my efforts were focused on CNC machining them from single blocks of metal. Ron describes a somewhat simpler method for creating them from several manually machined pieces.

I started with a pair of 2" x 2" x 3/4" aluminum blocks and was able to mill most of the parts' intricate features while working on their topsides. Once I completed this step on my first pair of parts, I noticed the covers' thicknesses were .005" less than that of the assembled gear tower. Both the covers and the tower seemed to match my understanding of the drawings, but I felt the thicknesses should probably match. So, I added another .005" to the cover design and started over with a second pair of parts.

Work holding issues showed up during the second step which involved opening up the sides of the parts to accept the cam box and end plate covers. This step included cutting the parts free of their workpieces on their exact centerlines. I built up temporary epoxy bridges to hold the parts in place during the milling operation that would free them from their workpieces. But, with little excess stock to adhere to, the epoxy failed on both parts and the cutter scarred them as they were freed. Yet another pair of parts had to be started.

The final machining steps opened up the interiors of the covers for the clearances needed around the cam gears. This was done in two operations using 1/8" diameter end mills and working down through the bottoms of the parts. The first roughing operation was performed with a cylindrical cutter, and the finishing operation done with a ball cutter. The finished diameter of the opening had to be increased slightly beyond the dimension given in the cover's drawing in order to accommodate the fillet left behind by the finishing cutter.

After machining, the covers' mounting screw locations were transferred to the temporarily assembled gear tower, and its eight 2-56 holes were drilled and tapped. The bores for the end plates were also finished using a boring head, and their 0-80 mounting screw holes were drilled and tapped. The pair of openings on the backside of the assembled tower shown in the last photo may look mis-machined, but they're designed to fit around the ends of the sheet metal covers that will later enclose the cam boxes. - Terry

137.JPG
138.JPG
139.JPG
140.JPG
141.JPG
142.JPG
143.JPG
144.JPG

145.JPG


146.JPG
 
I do enjoy reading your build threads, all very informative, excellent photography, lots of great ideas, thank you for sharing your knowledge and techniques!
 
I was hoping to tie up the ribbons on the gear tower and have at least one fully machined, gasket'd, and working assembly with moving parts that I could show off. However, the plans call for a trough to be machined into its backside for the purpose of carrying pressurized oil to the engine's top end. Since I'm not comfortable with beginning work on the engine's oiling system by starting in its middle and working outward, the tower will have to be torn down later for some additional machining.

The tower/drive block combo has a number of cover plates that are easy to machine and, after polishing, add nice contrasting touches to the tower's simulated cast aluminum surface finish. The more challenging parts, though, are the internals of the magneto drive block. Ron's parts list calls out what are now obsolete Small Parts numbers for the miter gears that he originally used. However, I wasn't able to cross-reference them to currently available parts. The BOM shows them as 48 DP, but the G479-Y and G461-Y eighteen tooth 48 DP miters currently available from Boston gear appear to be too small to handle the modifications implied in the drive block's assembly drawing.

I happened to have a pair of Boston Gear G462Y 32 DP 16 tooth gears left over from an earlier project that I decided to use. I increased their original .188" bores to .2495" and turned down their o.d.'s from .536" to .480" in order to fit them inside the drive block's half inch bore. My modifications were very similar to the ones shown in the drive block's assembly drawing, but I was initially puzzled about why the drawing also seems to show most of the output shaft's gear hub machined away.

It was then that I realized that my drive block's housing is wider than it's supposed to be. There are several dimensions associated with the drive block that are left for the user to fill in, and evidently the value that I'd estimated for its width from a Xeroxed photo in the manual was incorrect. The reason why this width is important is that there must be room between the drive block and the magneto for a faux coupler between the two. Embedded in this coupler will be a pair of magnets needed to trigger a Hall device located on the side of the magneto. (Although the manual recommends making the magneto's input shaft from stainless steel in order to minimize field distortions around the magnets, I made the drive block's output shaft from stainless as well.)

Rather than re-machine the drive block, I designed a new inboard cover that would allow me to move the output shaft's inboard bearing entirely within the housing. I'm fairly sure I picked up enough space for the coupler, although it too may require a minor redesign. These changes turned out to require shortening the hub of my miter gear after all. It's probably important that these gears be bored for tight press-fits onto their shafts since, after shortening, there may not be enough surface area for Loctite to adequately hold them, if loosely fitted, in place. The inertial loads of the coupler and magneto won't be insignificant during abrupt speed changes at the rpm's this engine will be capable of.

Fitting the gears inside the drive block would have had to be done blindly, so I cross-bored a scrap piece of metal to try to duplicate it. After opening an inspection window on its top, I was able to iteratively modify the gears and begin the shimming process for an optimum engagement. The fixture got me into the ballpark, but I had to complete the process inside the actual housing. In previous miter gear (distributor) applications, I've had gravity to help out with setting the engagement depth for silky smooth low-backlash operation. Gravity seemed to work against the right angle drive's engagement, however, and setting up its gears was a lengthy and tedious process.

The drive block's components complicate the assembly and disassembly of the gear tower. Although the miter gear and its 20 tooth drive gear must be hard pressed onto the block's input shaft, the bronze shop-made bearing which is captured between the two must turn freely on the shaft and slip into the drive block during assembly. Simultaneously, the rear end of the input shaft must slip into the ball bearing located inside the rear half of the gear tower. I've included photos of my two shaft assemblies showing the shims I used. I didn't include any dimensions since they would apply only to my particular (mis-machined) drive block. The output shaft's outboard bearing is a slip fit inside my drive block, and so I added a spacer between it and the nose of the miter gear to prevent the bearing from walking inward during normal operation.

The next step will likely be to machine the cylinder liners so I can finish up the block. - Terry

147.JPG
148.JPG
149.JPG
150.JPG
151.JPG
152.JPG
 
Hi Terry
As always, beautiful work !
But I really wanted to thank you for sharing your mistakes as well :) and the really awesome and clever engineering used to right the wrong. You always come up with an elegant solution.

Many thanks for sharing !

Scott
 
The Offy's cylinder sleeves will be sealed inside the coolant-filled engine block using Loctite. Each sleeve has a machined ring around its top end that will fit inside a recess in the head. The bottom ends of these rings will be sealed inside similar recesses that have already been machined around each bore in the top of the block. Unlike the Merlin that used these rings for metal-to-metal seals, the Offy will use a head gasket. When machined and assembled according to the drawings, up to .125" of each ring will wind up inside the head depending upon the thickness of the head gasket. This means that matching the heights of the sleeves above the Offy's block won't be critical as it was in the Merlin, but each sleeve will have three diameters and an inside corner that will require careful machining.

The drawings specify Stressproof for these sleeves which, in comparison with 12L14, offers improved corrosion resistance in a liquid cooled engine. During efforts with the Merlin's Stressproof liners, I struggled to obtain bores with surface finishes that wouldn't require inordinate amounts of honing.

Stressproof turns easily on both my 9x20 Wabeco and 12x36 Enco lathes, and fine surface finishes are easily obtained using tooling with carbide inserts designed for steel finishing operations. When boring, however, the rigidity of the cross slides on both lathes is lacking, and the resulting chatter at the ends of typical boring bars that fit my lathes limits the quality of a finished surface in Stressproof. This isn't an issue with similar operations on 12L14 using the same machines, but Stressproof's 2x hardness and .5x machinability creates problems for boring operations in my shop.

The solution I eventually worked out began with drilling out the cylinders to within .050" of their final diameter using a series of drills. My Enco lathe's back gear was then used to reduce the spindle speed to 100 rpm for use with its slowest possible (.0047"/rev) power feed rate. A much slower feed is available from the Wabeco and, even though I replaced its lightweight cross slide long ago with a large block of metal, its VFD doesn't produce usable torque below 300 rpm.

I used a 3/4" diameter boring bar that had to be notched to fit my lathes' tool holders. The insert was a Korloy TCGT32.51-AK with a .032" radius nose and three polished razor edges designed to turn aluminum. This combination, along with dark thread-cutting oil, produced an acceptable bore inside Stressproof, although the insert's life was extremely short. I was able to complete two sleeves per edge, with the second one having somewhat poorer surface finish.

Depth of cut is important. If it's too shallow, even the Korloy insert will skip across the surface and spoil its finish. If it's too deep, the insert will wear out even faster and produce a poor result. I found .020", or so, (diameter) to be a sweet spot. With .050" excess stock remaining after the drilling operations, I had a couple chances during machining to calibrate the lathe's DRO for consistency between parts. The goal was to bore all the sleeves to within a thousandth of one another in order to minimize the amount of (messy) honing that would be needed to bring them all to the same diameter. Since the piston rings will be machined to fit, the final honed diameters aren't important so long as they're identical.

I had enough 1144 of the correct diameter to make four parts plus a spare. The spare was finished along with the rest of the sleeves for later use as a light-test fixture that will be used to select the rings that will be used in the engine.

Before beginning machining, a Delrin plug gage was turned to exactly match the two diameters already machined into the block. These were mic'd and their measurements recorded as targets for the sleeves' o.d. machining. This same gage will be used later to machine the cylinder recesses in the head.

Using a convention steel finishing insert, the sleeves' o.d.'s were turned with fine but unpolished surface finishes before starting the drilling and boring operations. A parting tool was used to turn the sharp inside corner needed to seat each sleeve inside the block. The top outside edges were slightly beveled to later aid insertion of the head down over the block. A short taper was also machined in the bottom of each cylinder's bore to ease insertion of the ringed pistons during final assembly.

When machining was completed, each sleeve was marked with a unique number so its progress could be tracked during honing. Record keeping isn't all that important with a small batch of cylinders, but when honing a sizable lot, wear on the lap itself can drive the process around in frustrating circles. Initial bore gage measurements showed four of the sleeves were starting out within a few tenths of one another, but one outlier was a full thousandth over.

A barrel lap and 280g Cloverleaf paste was used to bring the four cylinders up to the outlier's i.d. Once this was accomplished, faint machining marks left over from boring could still be seen on three of the five cylinders - probably the second bores from the insert's edges. These machining marks with their familiar period of some 200 spindle revolutions, often show up on my Enco during power-fed finishing passes even when turning. I've inspected the lathe's easy-to-get-to gears for defects but have never been able to pinpoint their exact cause. The marks left during turning polish out quite easily, and so I've just been living with the problem. They're a real PITA inside a cylinder bore, however.

I spent a few hours trying to clean them up with 600g paste, but ultimately had to return to using 280g to make any real progress. Final polishing was done with 600g, however. Some .003" eventually had to be honed away from the cylinders' i.d.'s to obtain five clean bores within a tenth of one another. The actual machining time for each bore was about an hour, and honing wound up adding another. - Terry


153.JPG
154.JPG
155.JPG
156.JPG
157.JPG
158.JPG
159.JPG
160.JPG
161.JPG
162.JPG
 
Hi Terry. What Loctite flavor will you use for bonding the sleeves into the block & what target annular gap? Reason I ask is I'll be bonding my bronze valve cages into aluminum heads in the near future & I just happened to notice some new product lines. Or at least new to me since I don't really check in all that frequently. Not sure they will be much difference over what I have, but I've also gotten in the habit of dating my adhesives & mine might be getting on a bit with age. I'm not even sure what shelf life is on these products, but I had a run of 'let-go's' in another hobby that I traced back to ancient Loctite, so not an area of expense I want to gamble on.

I think you have done all your liner builds without a tool post grinder preceding lapping stage. Do you feel it doesn't offer a worthwhile advantage over your current methodology, or yet another nice-to-have that Santa is not taking the hint? LOL

Do you have any plans to lap the liner OD's or they are coming off the lathe well enough?
 

Attachments

  • SNAG-11-13-2019 0000.jpg
    SNAG-11-13-2019 0000.jpg
    180.3 KB · Views: 390
Peter,
The plan is to 680 Loctite. The annular gap is between a half and one thousandth. I'm using the o.d.'s as they came off the lathe. The finishes were good enough and they fit the block without polishing.

The toolpost grinder is an interesting idea. Have you seen any commercial units suitable for a cylinder bore or are they typically shop-made? - Terry
 
I had good results with the loctite 648 as sealing the liners into the block AT my bugatti bild. No leackage so far.
Michael
 
The toolpost grinder is an interesting idea. Have you seen any commercial units suitable for a cylinder bore or are they typically shop-made?

Terry, I have a Themac TPG. What little testing I did on my prior liner prototypes suggest it will work for me to bridge the gap between lathe single point turning & lapping, hopefully with more control & consistency. I think we have chatted on this before but I think my plan might be a bit more bass-akwards to the 'better' approach, which is to make the liner bores the same ID/deviation/finish by whatever means & size the rings from that. I'm trying to simultaneously hit a target bore dimensional specs with the final finish, which is probably added complexity. Part of the reason for this PITA process is to use commercial (OS-FS-56) rings, at least for my first build. Ultimately I will make my own rings because that is the way forward for any subsequent engine & then I will have a comparative reference.

Once my heads are done, liners will be the next thing to tackle so I will be in better position to show results & comment. One of the challenges of TPG's is general lack of appropriate stones of proper OD, thickness, grit & type (aggravated by the country I live in). So I bought a 1/4" thick surface " grinding wheel & had some discs water jetted out. That seemed to work but I may have chosen too fine a grit out of naivety. I learned later on that amazing finishes can be had with what I would have considered coarse grit. They cut better & run cooler etc. although I was taking squeak passes off like less than 0.0005". I hav some more things I want to ry in terms of dressing angles though. Progressive DOC measurement is another another vital ingredient to success I discovered. I could not count on my cross slide DRO & for sure not the dial on grinding ops. I mounted a tenths DTI on the cross slide to monitor progress & its rather disturbing how much things move around just tightening the cross slide & under power. So more to come on all this later I hope...

I have looked at lots of homebrew TPG designs. I'm no expert but I think what drives the cost of commercial units are attention to detail on the bearing assembly. Usually premium grade angular contact bearings, geometry, lubrication & methods to compensate under running/temperature conditions. Also ID boring means the rpms are getting up there, which means vibration & balance & (typically flat) vs pulley belts come into play. So add all this up & it starts looking a lot like a Dumore or Themac. There is one set of plans that looks decent though & the fellow seemed to know his stuff. i can provide details if you like. Brand name TPG's come up in Ebay & the like but how does one judge condition. I've heard, at least for Themac, they will do a teardown/rebuild to their specs for X $.

There are probably lots of less stringent grinding applications where brand name TPG's are probably overkill. They are rather big & clunky, but that suites their purpose. I'm mulling buying a decent quality Asian motor/spindle with integrated ER collet for 'make it shiny' applications. But I don't think it would be up to task for a series of liners, even factoring in new bearings every so often. Although I would love to hear findings from others who went this route (and measured the results).

Thanks for Loctite info.
 

Latest posts

Back
Top