Radial Engine Master Rod Design Query

[Vixen]

Back in the day. If the pilot detected detonation in his engine, he had to act quickly. He had two options, one was to pull back the throttle and reduce boost. The second was to increase the mixture strength, to rich, in order to supress the detonation. Imagine the situation: your climbing and the engine goes into detonation. If you close the throttle, you will not have sufficient power to clear that mountain range ahead; if you richen up the mixture, you will not have sufficient fuel to reach your destination. Mmmm. Safer by train, back in those days. 

Mike

[AVTUR]

I have a feeling that this compensation business is about detonation. It probable does not happen in a model engine (because of size) but causes problems with larger normally asperated engines. With forced induction it is instant death of the engine. In its final development the Centaurus used about 15 psi(g) of boost.

I note that the Wright Cyclone had its master rods positioned away from the top of each cylinder bank.

AVTUR

Just noted Mike's comments about detonation. I would love to consider it more but this topic has already got horribly complex before one considers combustion.

[Vixen]

I have a feeling that this compensation business is about detonation. It probable does not happen in a model engine (because of size) but causes problems with larger normally asperated engines. With forced induction it is instant death of the engine. In its final development the Centaurus used about 15 psi(g) of boost.

The final iteration of the Centaurus was the Mk 373. It delivered a massive 3200BHP and needed Direct Petrol injection of 100/130 octane Avgas (which contained enough TEL lead to sink a battleship) together with the addition of water/ methanol injection to keep the detonation under control under high power (take off and climb) conditions.

Mike

[Laurentic]

Mike - thank you for the confirmation re lubrication and No.6 cyl for the master rod on Bristol engines, there is much logic in that!

On some engines the use of water/methanol injection was also used to cool the air between blower and cylinder, cooler air = greater mass of air = more power.

Chris

[AVTUR]

After some head scratching I have results for the model Bentley BR2.

The two graphs are for the cylinder with the master rod and the next four cylinders on one side. The analysis assumes the engine is a radial, not a rotary. It does not make any difference. The remaining four cylinders are a mirror of cylinders 2 to 5.

On the graphs Series 1 = Cylinder with master rod, Series 2 = Cylinder 2 and so on.

The first graph appears to suggest that compensation makes little difference. The more detailed second shows
1. There is a small difference in stroke.
2. There is a large variation in angle as TDC is approached. The would lead to problems - if the ignition was timed to, let's say, 24 degrees before TDC on the master cylinder the actual ignition on cylinders 3 & 4 was happening at 15 degrees before TBC (retarded ignition = very hot running). Conversely setting the 24 degrees on cylinders 3 or 4 would give 33 degrees of advance on the master cylinder (over advanced ignition = detonation). There would be a large, but less important, actual variation in valve timings.

There are a lot of questions here like how do you time a big radial engine. I sure you don't fix a timing disc to the front and illuminate it with a strobe.

The compensated case will be looked at after I have measured a Hercules slave rod next Tuesday.

AVTUR

[Vixen]

Hello AVTUR,

Thanks for taking the trouble to prepare that spreadsheet and share the results for the uncompensated Bentley BR2 with us.

The second graph zooms in and clearly illustrates the problems associated with an uncompensated nine cylinder 'round' engine in terms of CR and TDC/timing errors.

It would be very instructive if you were to publish a third graph, showing only the master cylinder, cylinder #4 and it's mirror; cylinder #7. This will probably be close the worst case condition as far as uncompensated radial engine geometry errors are concerned.

Mike

[petertha]

Very nice AVTUR. At some point, some of us nerds would like to learn more about your tool. I inherited an Excel spreadsheet a long time ago, possibly off the old Yahoo R&R forum but even that is getting foggy. I wish I could remember the author's name so I could provide him proper recognition. He basically used geometric formulas to solve resultant parameters in 1-deg steps, allowing different permutations of MR/LR layouts depending on the goal. I modified it along the way with charts much like yours & even managed to loose most of my own work. Looking back on it now, one could take this further & pick of min/max's, points of interest, or 'solve-for' certain target conditions.

On my glow plug 5-cyl radial, I was just interested to keep CR's equal because ignition happens when it happens. Once I validated the layout to achieve this, I didn't have a good reason to return to the spreadsheet or take it further. But the next engine will go 'spark-bang' so no avoiding the issue ha-ha. This subject is very interesting to me, so look forward to more discussion. I did have an appreciation for just how much timing could vary depending on the MR/LR layout as your examples demonstrate. And I was always curious as to how this was addressed or compensated for in FS 'round' engines. Maybe the distributer itself?

I found my old cheat sheet of how to graphically lay out a compensated MR/LR assembly, PDF attached. Apologies, it was never intended for public distribution so may not be clear (but hopefully correct!). More of a personal note-to-self screen grabs if I had to do another. If anyone spots any boo-boos be sure to let me know so I can retract & no metal shall be harmed.