Quarter Scale Merlin V-12

[bigrigbri]

May I say the cogging action is the sticksion of the valve stems when tying to open and being acted upon at an angle overcoming the initial valve spring load. Try a small smear of motor oil on the stems before assembly.
Running in carefully will loosen things up with adequate lube.
Heavenly work though.

 

[mayhugh1]

I was planning to machine the stud tubes next, but after thinking about it for a while I didn't feel like screwing around with the lathe motor just yet. So, I decided to tackle the long flimsy intake gaskets that I hadn't been looking forward to dealing with either.
If I were were building a billet engine I'd already have a CAD model for the intake ports. Both heads would be identical, and I'd be able to create a CAM program for the drag knife on my Tormach to make multiple copies of the gasket from a common design.
The intake flanges on the Merlin's port and starboard castings are essentially identical although not quite parallel to the heads' axes. However, I match-drilled their mounting holes to the bosses on the intake manifold halves which have some variations of their own. As a result, most of the holes did not end up perfectly centered in the mounting bosses on the heads, and the port and starboard hole arrays are not quite identical. They're different enough that two slightly different gaskets had to be created.
Each 10" long gasket requires 46 holes clearance'd for 2-56 mounting screws. I was concerned that cutting so many tiny holes with a drag knife might develop into a hassle, and so I decided to drill them using a metal template.
I started by clamping each head in the mill vise, and with an optical microscope in the spindle I used the DRO's to manually record the center of each mounting hole. I also recorded the locations of some key dimensional features of the ports so I could build simple shape models of them in SolidWorks. Software is available for Mach-based cameras to automate a process like this; and, in fact, Tormach used to sell such a package. My personal experience with the raster-to-vector conversion software that I've tried to use on my plasma table has been disappointing, though.
I then created a pair of gasket models and a CAM program to machine a matching pair of drilling templates from 1/8" aluminum plate. Using the drag knife, I cut out three copies of each gasket from the same automotive gasket material that I used months ago when I match-drilled the intake manifolds to the heads. It was important to use the same gasket thickness that was used during the match-drilling in order to maintain the original fit of the manifold between the 60 degree heads.
The two sets of three gaskets were then stacked under the metal templates which were screwed tightly down to a piece of MDF. The holes were drilled through the gasket stacks using a drill press, and they came out looking as though they had been punched. At this point, these are the only gaskets I currently have planned for this engine, but that may change as the build progresses. - Terry

 

[kvom]

Some good ideas here for the future. I need to remake some gaskets for my loco and will use this technique.

 

[mayhugh1]

Pressurized oil will eventually be supplied to the valve train components through flanged fittings located at the outside rear of each head. Internal 90 degree flanged elbows are needed to connect these external fittings to the rear-most cam block in each head. From here the oil will be injected into the collinear arrays of hollow rocker arm shafts for distribution to the rocker arms and cam bearings. The waste oil will be collected in a trough running the length of each head and eventually be returned to the crankcase through the stud tubes.
The elbow needed in each head is rather small, and the space around it is very limited. In addition, each will cross over the top of one of the studs; and so the fittings must be easily removable so the stud nuts can be accessed.
The flanges were machined from brass plate, and soft 3/32" o.d. copper tubing was used to form the elbows. I've found that most small diameter tubing can be easily formed around simple wood mandrels. I usually chuck a wood dowel in the lathe and use a grooving tool to turn a tight fitting slot for the diameter of the bend that I need. The slot helps to prevent the walls from deforming. I've used this technique for copper and stainless tubing less than 3/16" diameter, but I've not had much success in applying it to full tight radius bends in thin wall brass tubing. The 3/32" tubing turns out to be too small even for my miniature tubing cutter, and so I trimmed the elbows to length with an abrasive cut-off disk in a hand-held Dremel-type tool. A few hundred rpm is sufficient to cleanly cut the soft tubing and helps limit the shrapnel field if (when) the disk fractures. The disk will try grab the tubing, and so it needs to be contained in a slotted cutting block. The tubing also gets really hot during cutting and should be sandwiched between the cutting block and a top cover block.
These two fittings are the first of many that will have to eventually be made since most of the Merlin's oil distribution system is plumbed externally. Some time was spent now experimenting with construction techniques for use later.
The real difficulty in making these fittings is coming up with a soldering fixture to hold the components in exact alignment so the completed assemblies will slide into place in the heads with no gaps and with all the mounting holes aligned. With no CAD model or dimensioned drawings to work from, careful measurements were taken of the actual hole locations while each head was resting on a surface plate. Fortunately, both heads' measurements agreed to within a couple thousandths, and so only one fixture had to be made.
The assembly was soldered using low temperature 60/40 solder plus an activated rosin flux. 'Activated' means the flux also contains an organic acid which makes it a better, although more corrosive, cleaner. Since it was only available in gallon containers when I purchased it several years ago, most of it will be around long after I'm gone. The flux was sparingly applied with a toothpick to only the ends of the copper tubes in order to keep the solder from spreading across the flange surfaces. I formed small solder ringlets around the tube ends and pressed them flat against the flanges so I wouldn't have to feed the solder by hand and make a mess of things. The assemblies were allowed to air cool and then were pickled in dilute sulphuric acid (drain cleaner) for a few minutes to remove the flux. Pickling was followed by a neutralization bath of baking soda dissolved in water and then followed by a water rinse. A Scotch Brite pad finally brightened up the surfaces.
The cam block mounting screws in the heads were removed one at a time and re-installed with low-strength threadlocker to seal the coolant passages they penetrate. After an elevated temperature cure of the threadlocker, a temporary flanged injector was cobbled up and screwed to the rear of each head so a syringe could be used to inject oil. The plug in the front cam block was temporarily removed to verify that oil had traveled through all the rocker shafts to the front of each head. Ample tell-tale seepage between the rocker arms and the cam block ends indicated the rocker arms and cam bearings were likely being well lubricated. - Terry

 

[ddmckee54]

When are we voting to elect the Grand Poobah of engine makers? When we do, you've got my vote.

On final assembly are you going to put a little high temperature silicone on the mating surfaces to ensure the oil only leaks out where you want it to leak? It doesn't look they don't have gaskets so I'd think that you'd want to goop up the connections on the outside of the engine at the very least.

Don

 

[mayhugh1]

Don,
Thanks for your kind words. I'll likely make some tiny paper gaskets for the flanges on the fittings external to the engine and just leave metal against metal for the internal fittings. The flanges on these engine fittings are usually small enough that it isn't difficult to get nice flat machined surfaces that leak minimally with oil which isn't a big problem inside the engine. In the oil test I did above there wasn't any significant leakage, fortunately. Coolant is another issue, though. Metal-to-metal flange seals usually don't work with coolant, and so all those will probably be gasketed. - Terry

 

[10K Pete]

I've been watching this build with great interest and am fascinated by the
work you are doing. I have had nothing to contribute until now.

Have you ever used Permatex #2 Aviation Form-A-Gasket? It is made for
metal to metal seals in fuel and oil systems. Isn't washed away by them,
doesn't harden very much, easy to disassemble and does clean up with
alcohol. I've used it for 50 years and my Dad for a long time before I came
along.

It was used on the Packard V-12s in PT boats (Dad was a MotorMac)
and Merlins and Allisons.

I suspect it would work perfectly for your application. If the metal to
metal fit is even close you won't need a separate gasket!

Pete

 

[mayhugh1]

Pete,
Thanks for the tip. I just looked it up, and decided to order a tube. I'm going to need something like that to seal the stud tubes if I can't figure out a way to do it with o-rings. I've used regular Permatex #2 on automotive gaskets, but it wasn't the aviation version you recommend. I'll run some tests on it when it comes in. - Terry

 

[10K Pete]

Pete,
Thanks for the tip. I just looked it up, and decided to order a tube. I'm going to need something like that to seal the stud tubes if I can't figure out a way to do it with o-rings. I've used regular Permatex #2 on automotive gaskets, but it wasn't the aviation version you recommend. I'll run some tests on it when it comes in. - Terry

You should be able to find it at the local auto parts store. Most all of them
carry the stuff.

Pete

 

[Davewild]

https://www.facebook.com/BattlefieldsPast/posts/827987097324228

This may be of interest to you Terry.

Dave

 

[golfpin]

got to admire you sir stay with it hope I will see u finish it...all the best to all on the sight for Christmas and new year golfpin

 

[mayhugh1]

I received and bench tested the replacement motor for my lathe, and it seems to run as it should right out of the box. The cooling fan on my older motor comes on immediately when the unit is energized, but on this newer version it evidently switches on only when the motor reaches an elevated operating temperature. This might be one of the 'improvements' made by Wabeco to reduce the amount of swarf blown through the fan to accumulate over the motor. With my new cabinet ventilation system, I'd rather have the improved electronic component reliability expected with a continuously running fan. A rule of thumb I used during my real working days was that electronic component lifetime generally doubles for every 10C that its operating temperature is reduced. Of course, fans have their own reliability issues; but they're typically cheaper and easier to replace. My current plan is to limp along with the old motor for a little while longer - at least until after the holidays. I guess I'm still hoping to stumble upon a fix before completely giving up on it.
The next step in the build was machining the 56 stud and coolant transfer tubes that run between the cylinder blocks and heads. The coolant tubes were trivial - just 28 short lengths of metal tubing that connect the coolant passages around the cylinder liners to the main passage in the head. These .160" long 5/32" o.d. tubes were parted off in the lathe from a length of thin wall aluminum tubing.
The 28 stud tubes, on the other hand, are a bit more complicated. The long studs that will tie the heads and cylinder blocks to the crankcase pass through the head coolant passages, and so the head stud holes are sealed by flanged metal sleeves. These stud tubes double as conduits for top-end waste oil to return to the crankcase. Openings in the tops of the tubes above the coolant seals allow waste oil to enter and trickle down and around the studs to the crankcase. These openings are really only required on the outside seven stud tubes in each head since the lower sides of the heads are where oil will tend to accumulate.
The first photo shows two possible designs for these tubes. The Quarter Scale documentation provides the design on the left which is made from a length of 7/32" o.d. thin wall aluminum tubing with one end spun to form the flange. The notes call for the bottom of this flange to be coated with a 'suitable' sealant. A steel slotted washer provides a durable surface for tightening the stud nut as well as an entry slot for the oil.
I tried my hand at lathe-spinning this flange on some test parts and was surprised at how easy it was to do using a hand-held sharpened wooden dowel. An issue I ran into, though, was an inevitable radius left in the corner underneath the flange that prevented it from sitting down flat over the sharp corners that I'd left on the reamed stud holes in the heads. I experimented with deforming the radius using a press and a scrap block, but even with re-annealing I could see tiny cracks in the stretched metal. Since I didn't want to radius the corners on the already completed heads I decided to re-design the stud tubes.
I machined my single-piece design shown on the right side of the photo from stainless steel. The tube's o.d. was turned for a close slip fit in the head, but a slight taper on its bottom end provides a couple thousandths clearance to aid assembly with the cylinder block. A .020" wide groove, machined in the bottom surface of the flange, will be filled with the Permatex sealant recommended earlier by Pete10K to form the upper seal. The tube's wall thickness ended up at just over .020".
The Merlin heads were not designed to fit down against the top surfaces of the cylinder blocks. Instead, shoulders inside the combustion chambers will be sealed to pressed-in liners which protrude slightly above the decks of the cylinder blocks. These liners create a .050" gap between the top surface of the cylinder block and the bottom surface of the head. All 28 tubes must also be sealed inside this space between the block and the head. Although the full-size engine used custom pocketed seals, the Quarter Scale uses simple o-rings placed around the metal tubes bridging this gap. There are no machined pockets for these o-rings in either of the two surfaces, and so the .070" thick o-rings will be compressed by about .020" within the gap.
A challenging part of the assembly is to engage all 28 o-ringed tubes in both the head and block while simultaneously engaging the liners with the sealing shoulders in the combustion chambers. Once this is done the head and block pairs are bolted together with an additional 24 head bolts (whose holes must also align) in order to form a standalone subassembly. This subassembly will be much easier to deal with than the individual heads and blocks when it's time to slip them down over the studs and the ringed-pistons.
Back in April when I drilled all these tube holes I realized it wasn't possible to match drill any of them. They all, including the combustion chambers, had to be referenced to a common datum on each pair of castings. Some of the holes even had to be 'moved' because they had been pre-cast in the wrong locations for my 'short' crankcase. To make things even more challenging, the holes for the coolant transfer tubes were reamed for light press fits in the heads and close slip fits in the blocks.
I was thrilled and totally surprised when the long awaited trial assemblies of each head/block pair with its 28 tubes and 6 temporary Delrin separators went together per plan without having to enlarge any of the holes. I wasn't even sure that over the ten inches I had to work the DRO's on my mill were up for the task. For me, this milestone was even more significant than getting the crankshaft laid in. But, of course, I don't yet know the subassemblies are leak-free.
Inconsistencies in the head counterbore depths, though, came around to bite me while trial fitting the stud tubes just as the inconsistent counterbores for the valve guide flanges affected the valve fitting. I had to re-machine most of the stud tube lengths and create four different groups of parts based on length to accommodate the range of head thicknesses I inadvertently created for myself. It seems that just as with Jerry Howell's IC engines there are very few quick-and-dirty machining steps allowable in the Merlin. - Terry

 

[10K Pete]

All the superlatives have been used......

So Wow!! Tiny, tiny little parts.

Pete

 

[petertha]

I'm sorry to be the one to inform you Terry, your pic & others like it show FS valves with some funky recessed dish shaped recess. Are you sure you are OK with flat-bottoms? At the rate you crank out complicated parts, a new set could be delivered by Miller time Kidding of course, its coming together fantastic. BTW why were the FS valves shaped this way?

 

[ICEpeter]

Hello Terry,
Good to hear you got the new lathe motor and it works out of the box. As far as the aux. fan / cooling fan is concerned, it appears that there are two versions of the cooling fan operation. It depends on the version of your motor. I copied two pages out of the manual I received for my Hanning motor and it shows two versions. In the first version, the cooling fan comes on as soon as the motor is powered up. In the second version, the cooling fan is controlled by the motor's CPU and appears to be controlled by a temperature sensor that monitors the CPU operating temperature and presumably controls the cooling fan operation as well.

The drive motor is also equipped with what appears to be a thermal switch embedded in the winding and connected to the CPU. I presume the thermal switch will shut down the whole system, indicating a drive motor overheating situation.

I could not find a parameter in the parameter set up in my manual that would permit setting a operating temperature range for the cooling fan. It is apparently an internally set parameter that can not be accessed.

Question: Did you receive an operating manual together with your motor that offers more detaisl about the internal set up of your motor? If not, you may be able to find the applicable manual at the Hanning website.

Peter J.

View attachment Varicon.pdf

 

[mayhugh1]

Petertha,
I'm really not sure why the RR engineers contoured the valves as they did. Maybe they thought the smooth contours would help with pre-ignition. My understanding is that the Merlin's wartime fuel was eventually 100 octane and even though the c.r. was only 6:1, in an emergency the supercharger boost could run as high as 13 psi. - Terry

 

[mayhugh1]

ICPeter,
Thanks for the diagrams. I didn't receive any documentation with the replacement motor except for a sketch with a comparison of the old vs new control wiring changes. Once I get the new motor installed I plan to take the original one apart to see if there is anything obvious that I can repair. If not, I'll likely try to adapt another VFD to it since there may not be an issue with the motor, itself. To me it smells like the pre-check that the processor runs through to measure the motor's parameters is failing for some reason as the motor heats up. If so, it may be a portion of the measurement circuitry failing. - Terry

 

[mu38&Bg#]

..... FS valves with some funky recessed dish shaped recess.....

These are called tulip valves. They were common at the time in all engines. I've not found specific details about how the shape was developed. It seems to have started with the idea that a larger radius between the head and stem improved flow, and the head was relieved to reduce weight.

Greg

 

[mayhugh1]

There were a few loose ends on this build that I wanted to tie up before taking a holiday break. The rear-end adapter on the engine stand that I've been using up to this point will soon get in my way as construction moves toward the rear of the engine. I considered building a scaled-down version of the rotisserie used to assemble the full-size engines, but the more I thought about it the more it seemed too awkward for bench-top use. So, instead, I made a new adapter for my existing stand. With the engine's front cover work completed, I should be able to finish the remainder of the engine with it supported in this new front-attached adapter.
After completing the cam chain enclosure several weeks ago, I wanted to do a little work on the wheel case so I could verify the fit of the chain cover. This would have included machining and installing the rear drive adapter on the crankshaft. While planning it out, though, I noticed that I had somehow neglected to finish machining the rear hub on the crankshaft. This hub was supposed to have been bored to accept a beveled adapter that will drive all the engine's accessories including the water and oil pumps, magnetos, camshafts, and supercharger; and it will be driven, in turn, by the starter drive. Back in July, I had to make up a special steady rest for my 9x20 lathe in order to support the crankshaft while its ends were being finish-machined. Evidently I became distracted and didn't finish boring out the rear hub, and so I had to pull the crankshaft back out of the crankcase to complete the machining. I knew this was coming up, and it was something I wanted to complete before tearing down the lathe to replace its motor. While the crankshaft was back out on the bench I also sealed the 24 oil end caps, as well as the threads of the 12 bolts securing them, with a removable thread-locker.
Finally, I needed to do another production run of some tiny parts so I could finally test the fits of the cylinder blocks to the crankcase. Just as the Merlin heads are offset from the cylinder blocks by short gaps, the cylinder blocks are similarly offset from the crankcase decks; and so, the studs must also be sealed across these gaps. Oil return holes previously angle-drilled through the threaded stud holes in the crankcase allow the oil trickling down and around the studs to eventually return to the sump.
These seals are lathe-turned Delrin sleeves that fit snugly in matching counterbores previously machined in the top two surfaces of the crankcase and the bottom surface of each cylinder block. The alignment of these counterbores is important so the cylinder liners in each block can slip into place in the crankcase openings with all 14 oil seals in place. I was able to verify the fit of both blocks using some snug fitting Delrin spacers in the bores for the liners to simulate the gap.
That's likely all for a while. To everyone on the forum, have a great holiday and my best wishes for the new year. - Terry

 

[stevehuckss396]

Absolutely beautiful work. Been watching from day one. Keep up the great work and I cant wait to see what next year will bring.