Another Radial - this time 18 Cylinders

[gbritnell]

Terry,
I've been following along since day one and like your son am anticipating the start-up. I know full well what you mean about start-up gremlins. It's the same as going to a show. I check and test all of my engines prior to leaving home and wouldn't you know it when there's a crowd around something happens. The worst was always the fried Hall transistor. Thankfully that hasn't happened for a long time.
I have to say again that the detail on your build is outstanding.
gbritnell

 

[mayhugh1]

I wired up the firewall after installing all its components, hopefully, for the last time. I then moved the engine from its assembly stand to its running stand and bolted it in place in front of the firewall. Due to the engine's excessive weight and awkward envelope this was a tricky two-step transfer that required an intermediate stand that I made to support the engine vertically on its front cover while the running stand was bolted to the rear of the engine. I re-tested the fuel loop after installing the carb on the engine and connecting the fuel lines. I then connected the throttle and spark advance linkages between the firewall levers and, for the first time, the actual engine. Space is pretty limited between the rear of the engine and the firewall, and this makes dealing with the the tiny rod end fasteners and my fat fingers difficult at best. I was really glad I was able to get most of the linkage work completed before involving the engine.
I then completed the connections between the ignition modules and the engine. Each module has its own dedicated braided ground connection to the engine crankcase. My ignition design provides several inches of separation between the Hall trigger cables and the high voltage distributor wiring in order to minimize emf issues. The engine's oil intake line was connected to the oil tank, but the engine's output line was temporarily run to a plastic catch bottle. My plan is to flush one or two quarts of oil through the engine for disposal before finally closing the oil loop and recirculating it. I'm using a straight 40 weight Shell aviation mineral oil purchased from a small local airport. This additive-free oil is designed for aircraft engine break-in.
After installing the three blade 28x12 prop the engine suddenly seemed too big for my shop, and so I took it outdoors for the final assembly photos. The whole thing weighs just over 70 pounds, and it is pretty difficult to move around even with my firewall handhold.
Today, in central Texas, we're having an exceptional one day winter heat wave with the outside temperature in the low 60's F; but the rest of the week is expected to only reach the mid 40's F. Since this is a radial with extremely long fuel intake tubes I suspect it will be very difficult to run it outdoors anytime soon after today. Unfortunately, the engine is much too big to be safely run inside my little shop. And so I will have to wait until the weather warms up, hopefully next week, before getting serious about running and tuning.
However, with only an hour of daylight left on the only day I'll have this week to play with it, I just had to see if I could at least get a first pop. I was surprised when I got much more. After choking the carb with my thumb for a few manual revolutions and then switching the ignition ON, I spun the prop with my hand; and the engine started right up and ran for about ten seconds. I didn't attempt any carb tuning since I was unprepared and currently unfamiliar with the Perry carb. I didn't receive any tuning information when I bought it, and so I have some research to do. But, I was able to restart the engine several times and eventually ran a dozen ounces of gasoline through it in several short runs lasting as long as 20 seconds. Currently, the engine doesn't have much throttle response, and the tach shows it tends to run at 1000 rpm.
I also re-tried my drill starter, and with a newly-charged battery I can get about 20-30 seconds of spin time if needed. As the engine finally consumed its second tank of fuel, the rpm surged to a very smooth 3000 rpm as the mixture leaned out out telling me the carb is probably currently too rich. I found on my H-9's Super Tiger carb that both the high and low speed fuel adjustments had to be adjusted nearly OFF for optimum performance, and so this shouldn't be too big of a surprise. The elapsed time counter showed an accumulated run time of only 4.2 minutes when the sun began to set and the temperature drop, when I muscled the engine back into the shop.
While running, the engine spit/slung a lot of oil initially, but the amount decreased significantly after the first run. I believe I got carried away with the oiling while I was testing the oil pumps and running the compression tests a few months ago. The engine seemed to continually drip oil after that even though I had drained the sump. Some of the oil was obviously from the lower lifter bushings, as expected, but some was also coming an indeterminate source that left me wishing I had used a gasket between the crankcase sections. Several days later I also noticed oil dripping from the exhaust pipes on some of the bottom cylinders. This was also typical of the H-9 and was due to oil seeping into the combustion through the ring gaps on the lower cylinders. On this engine I have several more of these PITA cylinders to deal with.
The only real issue I ran into during these brief initial tests was with the LH threaded nut that keeps the prop on the crankshaft. This nut continually loosened during running creating a bone-chilling prop rattling noise. I was probably fortunate that these first runs were short or the prop might have come completely off. This wasn't an issue on my H-9 since this nut was integral to the spinner and was continually re-tightened during starting. I changed the design of the spinner on this engine to remove the T-18's excessive starting torque from the weak portion of the crank nose and transfer it to the keyed prop on the beefier section of the crankshaft. At the time I didn't think there would be any significant torque requirements on this nut, especially since the prop is keyed to the shaft, but I was evidently wrong. I'll re-visit this later in the week after today's excitement dies down. For the present I solved the problem with a hefty lock washer.
The next task is to familiarize myself with the operation of the Perry carb and, when the weather permits, go through the tuning process and verify it will actually perform on this engine. I'll then video the final result. If the weather doesn't warm up I wouldn't be surprised if I fire it up inside the shop against my wife's warning and my own better judgement. - Terry

 

[Swifty]

What an exciting time it must have been when it came to life. Looking forward to the video.

Paul.

 

[petertha]

Awesome Terry, that must be a great feeling. Its a big, beautiful Beast! Take your time on the home stretch & look forward to the hearing the roar when its ready.

 

[Ethan D]

Beautiful work! Can't wait to listen to the roar!

 

[gbritnell]

Hi Terry,
Not much more to say. Superlative job! The details on the controls and the engraving on the firewall are just icing on the cake.
I too can't wait to hear it sing.
gbritnell

 

[kvom]

The most impressive build thread I've ever seen here. Wow!

 

[Scott_M]

WOW
Just beautiful, right down to the last screw. Very impressive Terry !

Hoping for warm weather in Central Texas http://www.kwtx.com/weather

And again...thank you for taking the time to document your build for the rest of us to see. I would bet that you also have an impressive number of hours into the Build Thread.

Scott

 

[vascon2196]

Very impressive...amazing job.

 

[mayhugh1]

Thanks, everyone, for your kind comments.

Terry

 

[kvom]

Time to change your avatar.

 

[brendanf]

Terry telling us you got it running and not posting a video, and worse yet, telling us we have to wait until Texas gets better weather...

Your killing me here! Practically drooling on my keyboard, desperate to see a video of this thing run so I can show my dad..

 

[kuhncw]

Beautiful piece of work Terry. Thanks again for the very educational build thread.

Chuck

 

[mikemott]

Your engine is one of the most inspiring builds, the firewall presentation really is as has been noted the icing on the cake. Thank you for all the work it takes to share such a complex build.

Mike

 

[johnny1320]

That's just beautiful Terry! Work of art!

 

[gus]

Hi Terry,

You are my " Sifu" aka '' GrandMaster'' ..

 

[mayhugh1]

When I built my first two IC engines, Jerry Howell's V-2 and V-4, I also built the carburetors for them. Jerry's carb is fairly complex, having high and low speed needles as well as a functional Bernoulli butterfly valve. The three copies I made (the V-2 required two) worked fine, but it seemed like I was continually adjusting the needles to not only start the engines but also to keep them running. With my limited experience I wasn't sure if this was due to the carb design, the engine designs, my machining, or whether constant fiddling is just the nature of the beast.
When I built my third engine, the Hodgson nine cylinder, I noticed the Howell carb design with its .25" bore had been licensed and included in the plan set I had purchased. After talking with Lee about the performance of such a tiny carb with the H-9, it wasn't clear that he actually had much personal experience with it. His carb of choice is a larger and very difficult-to-find unit from a vintage Cushman scooter. In fact, he mentioned that he maintains membership in a Cushman club in order to get access to them.
My research into what other builders were using for their H-9s was also disappointing. Those who were satisfied with their carbs were using models no longer available. Being intimidated by the complexity of the H9's induction system as well as the large number of machined parts that need to play nicely together, I decided to go with a commercial carb in order to reduce the number of unknowns that would be dependent upon my machining abilities.
After a few false starts I ended up with a Super Tigre #12163145 - a dual needle barrel carb with a .35" bore. Although my final gasoline settings seemed extremely lean, the adjustments were not overly sensitive; and the carb seemed to work well on my H-9. Once dialed in, the transition from idle to full speed was smooth, and I never had to mess with the needles again.
The induction system on my T-18 is even more intimidating. With no real expertise I designed my own diffuser and plenum based more upon good looks than any flow theory. The front row intake tubes are nearly twice as long as the rear row intakes which, themselves, are pretty long. In addition, the thermal characteristics of my T-18 heads and cylinders which affect the intake temperatures just have to be much different from those on the H-9. I felt Jerry's carb was too small for this engine; and so, again, I decided to go with a commercial unit. My tried and proven Super Tigre, though, hasn't been available in this country since the Japanese tsunami of 2011.
Choosing an RC carb for a multi-cylinder model engine from a distributor's website is a bewildering experience. The selection is typically large but the availability is in constant flux since the carbs are designed for particular RC engines that seem to come and go with the frequency of women's fashions. And, RC engine size isn't a sufficient spec to use when shopping for a carb for a multi-cylinder model engine. The requirements of the two engines can be very different. For example, I'm more interested in a carb that can provide my 18 c.i. display engine with a quality idle and smooth transition to mid-range rpm compared with an RC enthusiast who wants his .60 c.i. acrobatic single cylinder engine to make reliable high rpm power.
During my search, I came across Perry Carbs now owned by Gary Conley who is a member of the model engine building community. I was attracted to his website because it includes an inventory listing by Venturi size which, to me, is a more useful spec than a recommendation for a particular single cylinder engine. I spoke to Gary about my requirements which included a maximum 3500 rpm. He assured me one of his carbs was capable of performing even better than the Super Tigre.
I purchased his model 1401 which has a .312" Venturi. Perry carbs are also barrel carbs with high and low speed adjustments. A needle valve is used to control the fuel flow at high speed. But, an adjustable disk with a cat's eye opening uncovers a slot in the spray bar to regulate the amount of fuel at idle as well as in the mid-range region. Because I plan to use gasoline instead of typical RC methanol-based fuel, Gary supplied a special disk with an opening better calibrated for use with gasoline. For some reason, no documentation arrived with the carb, but two of the photos show the information I was able to locate online.
The design has evidently evolved a bit since the version shown in the exploded diagram. The throttle arm is now plastic, and the high speed needle no longer requires a tool to make adjustments. After installing the carb on my engine and hooking up the throttle linkage, the sloppy feel resulting from flex in the plastic throttle control arm was a bit of an annoyance. So, while looking for something to do while progress on the radial is currently weather stalled, I machined a brass replacement that feels much better. While the carb was still off the engine I decided to disassemble it to better understand its internals, and it was probably a good thing that I did. The area between the o-rings on the aluminum disk was covered with a sticky gunk - maybe dust encrusted o-ring grease - and there was a piece of aluminum swarf stuck in the narrow fuel slot. I don't believe any of this could have from my end earlier in the week since I'm running a fuel filter just in front of the carb.
The instructions mention that a only few thousandths movement of the idle control disk can make significant changes in the idle performance. I noticed the two o-rings tended to stick and slip when I attempted to rotate the disk a small amount (the documentation warns about this). So, it feels like this method for making small low-speed adjustments may be more difficult compared with a more conventional low speed needle. I'll reserve final judgement, though, until I actually go through the process. My T-18 carb adapter is designed around the standard carb dimensions of the Perry as well as a couple carbs salvaged many years ago from my younger son's RC junk box. If the Perry doesn't work out, I may have some other plug-n-play possibilities.
Perry's recommended starting point for the high speed adjustment is one turn open from fully OFF. After checking, I found that mine had been set 2-1/4 turns from fully OFF while I was first-starting my engine a few days ago. As I suspected at the time, the carb was probably set ridiculously rich; and I probably should have waited until I was familiar with it before attempting to start the engine. Anyway, it seems no harm was done; and now, with a bad case of cabin fever, I'm ready for the weather to warm back up. -Terry

 

[cfellows]

Congratulations on a beautiful build and successful first run, Terry. You're truly an inspiration.

Chuck

 

[mu38&Bg#]

The engine looks incredible. I'm looking forward to hearing 18 cylinders firing.

I didn't know Conley offers gasoline compatible metering for the Perry carbs. That's good to know.

Greg

 

[petertha]

I was trying to find that gasoline/methanol distinction myself off the Conley website, but couldn't. I see 'standard' and 'pumped' version for the same engine application, but not fuel type? Maybe Terry can elaborate - did you swap out any Supertigre squishy bits to prepare for gasoline? Maybe Perry uses body/seal/o-ring materials are compatible with both fuels?
http://www.perrypumps.com/prod01.htm#mvvs

I can understand the frustration trying to size carbs from (lack of) RC engine venturi / orfice specs. They just don’t catalog that dimension unless a helpful reviewer bothered to make note. Plus, there are just so many variations that may affect size even assuming a target displacement; sport 2S vs piped/boosted 2S vs 4-S vs boosted 4S vs pressurized fuel inlet vs muffler/pipe assist...

The Conley/Perry venturi size gradations are certainly helpful, but I assume the lookup table must be primarily derived by 1) matching OEM carb hole size 2) maybe orfice:displacement correlation? I'd be interested to know myself. Anyway, you could use this info to your advantage if the Perry didn't pan out - just back-track the Perry model# to the matching engine OEM part. (Personally I found Perry's somewhat fussy, particularly in the mid-range, but that was a different era & completely different application).
If I had to suggest an alternative, it would be OS-4-S. Just going down this path myself, but luckily an easier matchmaking exercise than this 18C.