Another Radial - this time 18 Cylinders

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.
Congratulations Terry

I can't stream video here at work so I'll have to wait until I get home before I can view it

Thanks for the explanation regarding the cam ring concentricity. That's one hell of a big bearing surface to keep them concentric with the crank!!!!

Just one more question

I see you used a free cutting mild steel to manufacture your crankshaft assembly, 12L14. This is low carbon steel with added lead for good machinability but it can't have its mechanical properties improved by hardening. I 'm surprised by this choice of material for a crankshaft, even though it machines very nicely. Does Hodgson specify this steel for the crank?
 
Thanks for sharing all the process and the final music I was checking the weather reports in central Texas waiting for this.:)
 
Fantastic. I've been following and I liked all of the build and now such a wonderful run. Nothing like the sound of a big radial: As gbritnell says, Goosebumps.

--ShopShoe
 
Congratulations Terry


Just one more question

I see you used a free cutting mild steel to manufacture your crankshaft assembly, 12L14. This is low carbon steel with added lead for good machinability but it can't have its mechanical properties improved by hardening. I 'm surprised by this choice of material for a crankshaft, even though it machines very nicely. Does Hodgson specify this steel for the crank?

Mattsta,
I don't know what material Hodgson specifies for his 18 cylinder engine, but the planset I have for the H-9 calls out stainless for the crankshaft. I didn't have a piece of stainless that was big enough to make my crank the way I wanted to make it, but I did have chunks of 12L14 as well as lots of common mild steel. I used the 12L14 because it's tensile strength was a bit higher than mild steel (surprised me too). Since I don't have a way of grinding a finished crankshaft I wouldn't have attempted hardening even if the metal was capable. Getting the crankshaft layed into the crankcase with absolutely no binding was the original feasibility requirement I set for myself to actually do the project. Frankly, I never expected to be able to do it and was sure that today I would be looking at a cool looking paperweight instead of a running engine. - Terry
 
Exceptional engine. The attention to detail is beyond believe.
The BIG question is what is next???.
 
Mattsta,
I don't know what material Hodgson specifies for his 18 cylinder engine, but the planset I have for the H-9 calls out stainless for the crankshaft. I didn't have a piece of stainless that was big enough to make my crank the way I wanted to make it, but I did have chunks of 12L14 as well as lots of common mild steel. I used the 12L14 because it's tensile strength was a bit higher than mild steel (surprised me too). Since I don't have a way of grinding a finished crankshaft I wouldn't have attempted hardening even if the metal was capable. Getting the crankshaft layed into the crankcase with absolutely no binding was the original feasibility requirement I set for myself to actually do the project. Frankly, I never expected to be able to do it and was sure that today I would be looking at a cool looking paperweight instead of a running engine. - Terry

Totally understand your methodology bearing in mind the complexity of this project

I guess hardening would create potential distortion problems in addition to having to grind the journals and crankpins to the correct diameters

How about an R-4360 for your next project!!!!!!

I just watched the video. It's an awsome build Terry. Many congratulations
 
Another question Tony

Which cylinders numbers do the master rods fit inside. From memory, on the full size R-2800, they aren't positioned where you'd expect them to be!

Interesting article about R-2800 crankshaft development here.

"As it turned out, the main factor contributing to the 1X
torsional vibration was master rod spacing. In the
original experimental test engines as well as the “A”
and “B” series production engines, master rods were
positioned 100 degrees apart (in cylinders 8 and 13)
to reduce the effects of second-order inertia torque."

http://www.enginehistory.org/NoShortDays/Development%20of%20the%20R-2800%20Crankshaft.pdf

I don't suppose this is relevent in a 1/5 scale engine but it's interesting nonetheless
 
Mattsta ,
Thanks a lot for the article. It was very interesting. I used #1 and #10 which seemed logical to me, but according to the article it was evidently the worst-case choice for minimizing vibration.

Terry
 
Mattsta ,
Thanks a lot for the article. It was very interesting. I used #1 and #10 which seemed logical to me, but according to the article it was evidently the worst-case choice for minimizing vibration.

Terry

I'm sure it's far less important on a scale motor with much smaller inertial forces but I'd like to purchase a set of Hodgson's drawings and to see what he specifies, (and much else besides!!).

The R-4360 is ever more bizarre, having master rods in A7, B4, C4 and D1 and to complicate things further, the master rod journals are staggered, not in a straight line like a conventional in line engine. The crankshaft was a single piece forging and the master rods were split like a modern car connecting rod. So are the main bearings.

How they designed these engine with the design aids and computational technology of the era beggers belief :)

Someone had a go at building one though!!!!!

http://www.nyemachine.com/pratt_whitney_r4360.php
 
Terry, I've been following all the way, and would like to say, Absolutely Fantastic, amazing job!
Cheers, Keith.
 
Hi Terry,

You are my hero. I admire the superb skill.patience and diligence building this fantastic engine.

Now taking my time studying the Howell V-2 Prints before jumping in.
 
A few days after my last post I pulled the spark plugs for a baseline reading and was a little surprised by what I saw. The carb had only been partially tuned, and so sooty plugs weren't totally unexpected. What was unexpected, though, were the locations of the richest running plugs.
Model radials, similar in design to the Hodgson with its centrally-fed rear carburetor, typically suffer from an uneven mixture distribution between the upper and lower cylinders. In order to improve the H-9's distribution, an impeller was added to the original design and then revised a few years later. The result was a performance improvement, but builders still experience rich-running lower cylinders when the carb is tuned for the upper cylinders. This mirrors my own experience with my H-9. The consensus is that fuel tends to puddle in the lower portion of the plenum and enriches the mixture drawn by the nearby cylinders. The H-9 design even includes drain holes in the lower plenum to reduce this problem.
The plugs on my T-18, however, showed a completely opposite result. Its rear row cylinders have intake tubes similar to those on the H-9. The rear row plugs, though, clearly showed the lower cylinders had been running leaner than the upper cylinders. The plugs in the lower cylinders had a slightly rich to almost ideal color while the plugs in the top cylinders showed those cylinders had been running rather rich with obvious soot on the plugs. The front row cylinders with their much longer intakes showed a similar distribution but, overall, they looked as though had been running very rich. The top plugs in the front row were all excessively sooty while the lower plugs were comparable to those in the top cylinders in the rear row.
These first baseline results were obtained using the 1401 Perry carb (.312" Venturi) with its high speed needle set at 8 o'clock (just over a half turn open from fully closed) and the idle disk still set at stock neutral. For fuel I'm using a product called Four Cycle Tru Fuel which is available from our local Lowe's store. This is an ethanol-free 92 octane gasoline designed for lawn equipment, and it contains an added stabilizer for long shelf life. What caught my eye about this fuel for my application was a claim that it also contains friction modifiers which I thought might be beneficial to the distributors' lower bearings and gears. Of course, this fuel could also be mostly advertising hype in order to justify its premium price.
A majority of the 5 or 6 minutes runtime that produced these plug results was between 2000 and 2500 rpm, but it also included periods of idling as well as a pretty ragged restart. The restart was attempted during a hot soak period with the throttle insufficiently open. The engine eventually started, but it required the drill starter to spin the engine for several seconds. The sooty plugs indicated that either the carb needed a lot more leaning, or the unfortunate re-start had badly fouled the plugs; and the engine hadn't been run long or hard enough afterward to burn off the excess accumulated carbon.
To prepare for the next set of runs the plugs were cleaned, and the high speed needle was leaned another 1/8 turn by turning it CW from 8 o'clock to 10 o'clock. (Since the needle on the Perry carb actually rotates with the throttle, my needle positions are referred at w.o.t.) This setting turned out to be too lean, and the engine consistently died several seconds after being choked and started. The engine needed at least another 1/8 CCW turn in order to sustain smooth running. As a result, I returned the high speed setting to its previous 8 o'clock position.
I then broke the set on the idle disk and turned it about 1/8 turn in each direction while the engine was running but could see little or no effect on the idle. It's possible that I was fooled by the grip of the o-rings as this adjustment, semi-unique to the Perry, isn't all that user-friendly. I then decided to rev the engine up to 3500 rpm and hold it there while burning the remainder of the fuel in the tank. The reason I did this was to color the plugs for the next reading with a substantial high speed run with no restart nor idling time. I was still curious about the previous upper/lower plug differences and now really bothered by the earlier overly rich plug readings. My plan was to revisit the idle adjustment after the second plug readings. I then would be in a position to connect any further changes in plug conditions to changes in the idle setting.
At this point the engine has plenty of throttle left to rev past my self-imposed maximum of 3500 rpm, and so I'm confident the carb size is adequate. When starting the engine with the drill starter there is a neat sounding intake hiss that's louder than the fuel pump. Also at this point the engine likes timing advance up to about 20 degrees.
After the engine cooled, I once again pulled each spark plug and recorded the results. Fortunately the heavy soot was gone indicating what I previously saw was more of an operator problem than an engine or carb problem. The upside-down mixture distribution differences remained though. The four photos show typical plugs from the tops and bottoms of the front and rear row cylinders. The important part of the plug is the white insulator. The photos continue to show the lower plugs in both rows are running leaner than the upper plugs. This may be a result of my impeller/plenum design, but it's not at all obvious to me how I was able to invert the result compared with the H-9.
A model builder used to running alcohol may find these 'filthy' plugs appalling; but, for a gasoline model, they really aren't that bad. A proper heat range plug in a modern full-size fuel-injected engine typically won't show any insulator coloring when running on unleaded gas. In a model engine, though, carburetor tuning is always a compromise; and plugs with heat range options aren't available. Carb tuning and, to some extent, the reach of the plug into the combustion chamber are the only tools available to control carbon deposits on the insulators. If the carb is leaned as much as practicable, but the insulators still accumulate deposits that eventually affect the spark then a shift to an alternate fuel may be the only option. Personally, I've come to prefer gas over alcohol because of gasoline's easier starting due to its higher volatility as well as its lower production of water in the crankcase due to combustion.
The cold rainy weather has settled back into our area. When it clears out I plan to work on the idle mixture.
While tying up some loose ends I also added a crude oil filter to the engine's oil line just after the drip feeder. It was made by drilling an array of .020" diameter holes through an inline bronze plug. It's purpose is to prevent chips that might have been inadvertently left behind from being recirculated. The most difficult part of its construction was clearing the chips left behind during its own construction. - Terry

893.jpg


894.jpg


895.jpg


896.jpg


897.jpg


898.jpg


899.jpg
 
At this point the engine has plenty of throttle left to rev past my self-imposed maximum of 3500 rpm, and so I'm confident the carb size is adequate. - Terry

I've been following this thread since the beginning and its simply fantastic work.

Have you considered that perhaps the "load" (prop) on the engine is insufficient for the power the engine develops? That would give max rpm with a lot of throttle opening left.

I am sure the engine is a beast in its power developed, the prop load is just a thought. It would affect tuning.

George
 
Terry,
Just out of curiosity could the Y in the intake tube be having an effect on the fuel mixture between the front and rear bank of cylinders? That wouldn't affect the top and bottom mixture but might affect front to rear. But then again with a little fine tuning you may have the balance as well or better than can be expected.
Art
 
George,
Thanks for the comment. You're very correct about the engine not being significantly loaded during any of my testing. The actual output horsepower from a static test like the ones I'm doing is dependent upon the diameter and pitch of the propeller, the engine rpm, and some air density parameters. When I was in the process of obtaining my prop I used the static thrust calculator located here: http://personal.osi.hu/fuzesisz/strc_eng/ to input its parameters (3-blade, 28x12) and came up with an expected 29 pounds of thrust and 3 output hp while running at 3500 rpm. The prop I'm using wasn't selected for its performance but for its availability. I realized at the time it was much too small to pull a significant fraction of the expected power available from this engine at the maximum rpm I was willing to run. But, that was OK because I wasn't building a flyer but a running display model. With my own imposed maximum 3500 rpm and the current prop, 3 hp is all the engine will ever be asked to put out. My comment about the Perry carb being a good size for this engine was made entirely in this context. If I were building an actual flyer with a performance-selected prop I'm sure it is too small.
...
...
Art,
Indeed, I was concerned about the quality of the Y-connection between the front and rear row intake tubes. I knew that if I didn't have gradual and smooth splits in these tubes there would be fuel distribution differences between the cylinders in the front and rear rows. If you'll remember, I put a lot of effort into getting the interior and exterior seams matched as well as I possibly could. I even pulled a few test joints apart to make sure the solder wasn't leaving a ridge on the interiors of the joints that could disturb the flow.
I was really concerned when, after all that, I saw a significant difference in richness between the front and rear row plugs during my first good runs. It now seems the plug readings from those runs were flawed by the operator, and the last more methodically-done test showed no difference between the front and rear rows. This will still need to be verified, though, at lower speeds and during idling. The persistent differences between the upper and lower cylinders is still a mystery, however.
Terry
 
Terry

What is the spur gear spec used in this engine?

Are they metric spur gears or imperial. I guessing they are about metric module = 1 but the imperial sizes 24 and 32 DP both seem to give the wrong diameters
 
Mattsta,
They are all imperial gears 14-1/2 deg pressure angle and assorted DP depending upon the particular gear. Did you have a question about a particular gear? - Terry
 

Latest posts

Back
Top