Ohrndorf 5 Cylinder Radial

[petertha]

Some trial & error futzing, eventually I arrived at combination where the liners could be re-inserted & the bores were undersized in the 0.0005 – 0.0010” range.

[petertha]

So once again, the liner bores were re-lapped to target ID, this time as a mated assembly. I will now call these good. I had the bright idea to submerge the assemblies in my ultrasonic cleaner again to remove trace lapping compound. But very shortly thereafter I noticed some corrosion stains starting to form on the liner surface, so I immediately pulled them out. Swabbing & rinsing with mineral spirits is a better way to go. Then a light coat of oil for storage.

Next engine I would leave the liner bores unfinished (undersized) at the boring bar stage. Some of the interference surface conditioning could probably be minimized, particularly if there was no need to pull & replace the liner. The cylinder assembly would now have to be jigged so that the grinding / lapping / honing operation could proceeded on the mated surfaces.

[Roger B]

That was quite a journey      I do like Acrolaps   

[Admiral_dk]

Woa - that part of the journey ended up being a lot more winding than anticipated ....

So I hope that you are really satisfied with the result !!!
I think they all look good and should work very nicely on the finished engine.

Per

[petertha]

A question came up on the other forum, so pasting reply as further elaboration if my saga verbiage was not quite clear.

I don't think there is an advantage to excessive interference, in fact several potential disadvantages. For example, in my case where the cylinder shape is tapered & thicker wall near the top can contribute to different shrink force on upper vs. lower liner. Pretty much every RC engine I have disassembled requires oven heat to install & remove the liner, even when brand new. My initial heat testing was based on a separate spare test cylinder which was the original design, not my subsequent modified design. But I think either reamed ID surfaces were not as consistent, or possibly they stress relieved a bit. In this kind of application, a half thou one way or another seems to make a big difference. My longwinded story was just to say in hindsight I would not bother grinding & lapping the liner bore until mated to cylinder & completely stabilized. Unless you have good shop methods to very tightly control both OD & ID dimensions & finishes. Jung provides these instructions in his 5-cylinder engine which is probably not far off how mine ended up after lapping the cylinder ID (0.02mm = 0.0008”).

To ensure optimum heat transfer, the cylinder liners must be shrunk into the cylinders. The inner diameter of the cylinder is about 0.02 mm smaller than the respective outer diameter of the liners to unscrew. After uniform heating of the aluminum cylinder by means of gas burner or hot plate (to about 200 ° C), the cold liners are inserted into the cylinder.

[petertha]

Carrying on with what surely must be the most drawn out & convoluted engine build post in the history of model engineering, now onto the head assembly. I hope I can accurately recall what I did at the time. Here goes.

I suspect heads are one of the more challenging build aspects of most 4-stroke engines. Maybe because they are the intersection point where so many important & inter-related components must come together. Internally are the valves, valve cages, port passages. Externally the valve spring assembly, rocker assembly, ignition plug, cooling fins, INT/EXH piping connection, mounting holes & fasteners. There is a lot to get right. This is also where many important tolerances, clearances & motions occur.

The reason I mention this is because I tried to stick with the original design intent for the most part, but there were several mods along the way. I pretty much had to make prototypes of most parts beforehand. With heads, a change in one component inevitably has a domino effect to other things. To name a few, the O9 had wider valve seats than what I preferred. So that changed the valve cage, its placement in the head & valve itself. The threaded portion of INT/EXH ports were quite short. Very few threads for tubing couplers which was another motivation to increase the head diameter slightly. Plans are metric & which included tubing sizes I could not get. I found a few (minor) dimensional errors only by transcribing the drawings into 3D CAD. So, what you will see here is more or less the end results for each the various components. Some will also appear out of order construction wise.

[petertha]

Here are some CAD assembly pics of where this is going. The heads are made from 2024 aluminum (or was it 6061, sheesh!). First operation was in the lathe. Turning the blank OD, turning a lip boss which fits snugly within the upper liner ID, the hemispherical combustion chamber profile and the radial cooling fins all in one setup. I acquired a used Radii internal/external spherical turning tool which has come in handy, but it’s a bit fiddly to set up. The combustion chamber bowl profile has a specific radius & depth combination which needs to be dimensionally correct to achieve target CR. After turning, the head blank was parted off & finished to length. After several prototypes leading up this point, I made 7 heads ‘production’ heads. 5 for the engine, one as new spare and one sacrificial spare to commence each operation.

[petertha]

Next operation was in the mill. I first drilled the 5x clearance hole pattern for M3 hold down screws. One hole occurs at different counterbore depth as a function of head orientation. I made an angle fixture/jig which was used for several subsequent operations as well. Alignment of head to the fixture depends on the combination of the cylinder lip mating the fixture’s center hole and constrained radially by the 5 hold down bolts. Because the heads are in & out of this fixture, this normally may not be an optimal method to align things because of slight bolt/hole clearance. But with all 5 bolts in place & the collective machining deviations, I found they the head didn’t have much if any free movement at all, so called it good.

The glow plug hole, counterbore & thread was done in one operation. Once I figured out how much to offset the hole from the head edge, all the heads were done in one setting. Then remove the head, replace with another, preserve the setting & repeat.

Next was milling the facet faces which become the area to which the rockers boxes are mounted. The valves will exit perpendicular to this face. Once the depth was established, it was just a matter of lopping off the material with a face mill. The fixture allowed me to rotate the head blank 180-deg & repeat left & right side.

[petertha]

At this point, I had already made the bronze valve cages and valves in advance so that when it came to drilling the valve cage hole, the target diameter & bore depth plan was fully defined.

I was aware of a few different paths to take when installing valve cages/seats. Some press or shrink the cage into the head with interference fit & valve seat is subsequently profiled. Some strive to have valve/cage sets fully meeting seal test outside of the head (usually by vacuum test), then slip fit the cage into the head Loctite to try & maintain the seal condition as much as possible avoiding any distortion. That was actually my initial plan too, but for unanticipated reasons explained later I ended up doing the seat cutting once in the head. There is only about 0.010” of seat contact chamfer to be removed from the cage lip so it’s not like a lot of material. The risk of course is that if you screw up the seat & cannot get seal or lap your way out of the problem, because the cage would have to be torched out in order to be replaced, hopefully leaving the head in decent shape.
 
I bored the valve cage hole from the combustion chamber side, out through the top side facet using the same angle fixture, but now inverted. I worked out the required offset enter distance in CAD using the jig edge itself as reference. First a smaller hole was drilled through the head, this mates to the valve guide segment of cage. Then the larger counterbore for the valve chamber body. It was a bit tricky to establish this bore depth because of the curved hemi shape and tool changes. It has to be a flat bottom hole so that when the cage is inserted & bottomed out, the seated valve ends up just beneath a flush position, conforming to the hemi shape as much as possible. For that I undersize predrilled most of the material away and then used a flat bottom end mil to make the hole ID and to target depth, pre-referencing the EM off the fixture as a datum for repeatability

Things went mostly to plan but somewhere between the hole ID, cage OD & surface finish/dimensions I had a few with slightly tighter feeling fit on the cage. Still within the Loctite gap allowance, but mostly I didn’t want to risk having them hang up during glue assembly because bronze triggers accelerate the Loctite cure significantly faster than steel alloys. End mills are not reamers so maybe that should have seen anticipated. Possibly using a boring head might have assisted here, but it seemed like more work & still leaves the issue of a flat bottom hole requirement unless one has an automatic (Wohlhaupter style) head. The end mill was a special order (metric flat bottom) style. In the end I made a simple lapping tool to condition the hole & that worked well.

The cages were glued with Loctite 680 high temp retaining compound. Lastly, I should mention that aside from the Loctite, there is no other mechanical retention of the cages in the head. I did consider pinning them or threading them but I just couldn’t see a good method I could accommodate or successfully pull off with the dimensional constraints. I did a torch test on a dummy assembly & the cage stayed mated, so I guess time will tell.

[petertha]

Machining pics

[petertha]

The heads & fixture was again flipped to do operations on the facet side of head. The valve hole is first located & centered, then counterbored to accommodate the valve spring body. Then holes drilled & threaded for mounting rocker cages.

A flat fixture was used to hold the head to mill the center groove using a ball EM. Then flipped on its side for sawing the vertical cooling fins which are at various progressive depths. I used a 0.045" slitting saw ~ 0.200" deep about 0.025" DOC per pass. About 200 rpm felt right on 3" OD cutter. I’m still not super confident with this operation. I kept a steady feed rate, used lots of WD-40, cleaned the teeth & trench of swarf between passes because there is very little clearance & aluminum is a gummy material. I've read horror stories where guys folded up the cutter & destroyed the part on the proverbial 'last pass'. At this stage a lot of work has gone into the part.

Lots of edge deburring & cleanup. I use these cheapo rubber abrasives in the Dremel, they work quite well. This pic shows the same tool held in a needle file handle for hand work.

[petertha]

The heads at this stage were cleaned & the valve cages installed using Loctite 680 HT retainer. Permanent installation is required at this point because the next step is to cross drill the port passages through the aluminum head & into the bronze cage. As you can see, the valve seats are cut. I will discuss this later on.

[petertha]

Fixtures were machined to hold the head in the correct orientation for making the intake & exhaust port passage holes in the lathe. This took some CAD work to figure out the correct orientation & placement. Two 2 separate fixtures were required for intake & exhaust because the offset was different. The fixture was worth it though, my earlier attempts in the mill failed miserably.

[steamer]

Watching another air cooled engine come together!     Nice job Peter!    I'm watching those heads intently!

Dave

[petertha]

Once the head was mounted to the fixture, the gas passage hole was spotted & drilled through the side of valve cage. It broke through with no drama, the Loctite joint held. This hole was then opened up with enlarged counterbore & threaded for the tubing retention nut. At this point I already had a prototype coupler nut & flanged tubing prepared to test the fit.

Tapping the 7/16-28 threads was a bit concerning because it enters the head at an angle where only a portion of the hole is getting threaded, maybe 3-4 threads being introduced on one side until the tap starts cutting the entire hole. The threads also encounter interruptions from cooling fin grooves on one site.  I held the tap in the tailstock & basically pushed it in along the bed by feel, turning the spindle chuck with the other hand, using only partial chip breaking reversals & lots of cutting fluid. Because the threaded hole is so shallow it required a bottoming tap. I sacrificed another plug tap by grinding the leading threads off & carefully repeating the operation while in the same setup. Single point threading was not a viable option to me.

Overall, I hate these threaded holes even though they are predominant on commercial model engines. They are fussy. The problem is one of scale, it’s more challenging to make typical flanged tubing fittings with smaller yet flange bolts into the head. Especially at some of the entry angles somewhat required with radial engines & their manifolds, not intersecting other holes or features etc. Anyways, I mentioned earlier that I increased the diameter of head slightly from original plans & this was one of the reasons why. Catching just a few full threads on a coupler nut without eventually striping or losing seal or crudding up just didn’t seem appealing, at least from my own RC experience.