Merlin engine limits & failure behaviour

[Viper2000]

I've been experimenting with the Merlin as part of zero g cut investigation.

I find its behaviour quite strange.

For example, take a Rotol Hurricane and try climbing at less than full boost; say +4 psi and 2800 rpm, being careful to keep temperatures within limits.

You'll find that constant boost is maintained until roughly the rated FTH.
You can then open the throttle and maintain boost to some higher altitude.
Eventually you'll run out of throttle movement.
If you then operate the boost control cutout, you'll be able to go a bit higher again.
Engine failure will follow on swift wings, despite the fact that boost, rpm & temperatures haven't gone into the red.

This is questionable for several reasons.


The boost control cutout just disables the ABC; it can't magically increase the FTH of the engine.
The engine itself doesn't care where the cockpit throttle is; it cares about boost and rpm; there is therefore no reason for the engine to be more stressed at say +4 psi and 2800 rpm with the boost control cutout operated and the cockpit throttle firewalled than at sea level with the cockpit throttle partly closed; indeed at altitude the engine's life is somewhat easier because the intake manifold temperature will be considerably lower.
With a double-acting ABC, the throttle position for +4 should remain constant irrespective of altitude; the ABC should progressively open the throttle valve during the climb to maintain constant boost until I reach the FTH for +4 psi and whatever rpm (rpm matters because supercharger pressure ratio varies as roughly the square of rpm). When the boost starts to fall off, opening the cockpit throttle should make no difference, because the ABC should already have fully opened the engine throttle valve.
With a single-acting ABC as fitted to really early engines, there would be very little throttle movement at sea level before the +6.25 normal operating limit was reached; we'd then need to progressively open up to maintain boost throughout the climb, because the automation would only close the throttle to enforce the boost limit, not open it to maintain boost (hence "single acting"). We can therefore rule out the possibility that the intention was to model a single acting ABC.


Also, I suppose that this has already been reported, but the mixture control in the Hurricane is very strange.

Rich mixture should be forward, not aft, and the mixture handle should be ahead of the throttle handle. This means that moving the throttle handle forward would automatically move the mixture handle towards the rich position, protecting the engine from detonation/preignition.

Throttling back would leave the mixture rich, so the pilot would have to manually lean the engine. However, the degree to which the mixture could be leaned would remain a function of cockpit throttle position.

Incidentally, this arrangement is strongly suggestive of a double acting ABC, because it would only protect the engine if cockpit throttle position directly correlates to a given maximum boost, such that a fixed mixture handle position would always provide a sufficiently rich mixture to guard against detonation/preignition.

[IvanK]

Good observations Viper.

Mixture issues with both Hurricanes and Spitfires has been reported directly to the Devs.

However in both the Hurricane and Spitfire the Mixtures were non incremental but two position either BACK for RICH and FORWARD for LEAN.
In effect these were AUTO RICH and AUTO LEAN. So in either position they will lean as the aircraft's climbs. In both the Spitfire and Hurricane there is an interconnect (or physical contact) between the throttle and mixture lever such that pulling the throttle back will also drag the Mixture lever BACK to the RICH position. This is all clearly documented in the respective pilots notes and maintenance manuals.* I am happy to provide scans of these if required. It is the reverse of what we have become used to.

Boost limits were placed (+4 on 100 Octane and +2 1/2 for 87 Octane) for LEAN operation .. to prevent Detonation/pre-ignition.

A further complication in COD is that in the case of the Hurricanes it would seem the the effects of mixture control have been reversed in relation to the lever positions.


*Note on some restored aircraft the Mixture controls have been set in a conventional sense (Forward to rich) this as a Flight Safety issue in that the pilots fly multiple types and keeping things standard is a sensible concession to 100% historical accuracy. Though in Peter Vacher's Hurricane I R4118 it has been restored with 100% accuracy in this regard.

---------------

Here are a couple of excepts from Hurricane I pilots notes and Hurricane II maintenance manuals:

http://img851.imageshack.us/img851/6535/hurrimix1.jpg

http://img715.imageshack.us/img715/2927/hurricontrols.jpg

[Viper2000]

How impressively, deeply, illogical.

I never would have guessed. Thanks for the references.

However, this does raise some interesting questions about the lean cruise case, because obviously the interference between mixture and throttle control will limit the options.

I can see this making sense for infinitely variable mixture and fixed pitch prop (since you want a richer mixture at a lower rpm), but it raises interesting questions about the sanity of Camm et al in case of a 2 position auto mixture, not least because you could just use a toggle switch to set mixture if you wanted to take that path. Since Camm was a very clever man, I think that we need to dig deeper here.

[Viper2000]

Careful reading of your 2nd document is quite interesting.

"The knob on the mixture control lever projects into the way of the throttle control lever to ensure that the mixture control is pulled back to RICH on closing the throttle, thus preventing the mixture from being excessively weakened at small throttle openings."
[Emphasis added]

Clearly this is in conflict with the concept of a 2 position auto mixture; the control might have had 2 positions, but if so then this strongly implies that there were major problems...

I think that in the context of the sim, in order to avoid scaring off new players, we should try to ensure that the engine controls displayed in that little alt-window are standardised (ie forward = more throttle, more revs, richer mixture). Personally I'd be inclined to say that forward = closed radiator, because forward = fast = caution, but perhaps I'm in the minority there.

In any case, we need the non-cockpit controls to be standardised in order to bring on new players with the minimum pain.

[41Sqn_Banks]

For example, take a Rotol Hurricane and try climbing at less than full boost; say +4 psi and 2800 rpm, being careful to keep temperatures within limits.

You'll find that constant boost is maintained until roughly the rated FTH.
You can then open the throttle and maintain boost to some higher altitude.



According to AP 2095 Pilot's Notes General:


Note 2. - Merlin Engines.-The throttle valve cannot open fully unless the throttle lever is advanced to a certain point on the quadrant. It is therefore necessary, when climbing, to advance the throttle lever when the boost begins to fall.


So sounds like this behaviour is correct.



Eventually you'll run out of throttle movement.
If you then operate the boost control cutout, you'll be able to go a bit higher again.
Engine failure will follow on swift wings, despite the fact that boost, rpm & temperatures haven't gone into the red.

This is questionable for several reasons.


The boost control cutout just disables the ABC; it can't magically increase the FTH of the engine.



AP 1590B Merlin II and III Aero-Engines:

In its unmodified condition, the boost control cut-out valve permitted the boost pressure to rise very considerably. Mod. Merlin 154, however, limits the pressure to +12 lb. per sq. in. for use under emergency condition (see para. 359A et seq.).


So a Merlin that wasn't modified for +12 boost the boost control cut-out will disable the boost control. A modified engine however will not disable the boost control but increase the allowed maximum boost to +12.

[Viper2000]

According to AP 2095 Pilot's Notes General:
[...]
So sounds like this behaviour is correct.
Presumably this is a function of the mod state of the engine/ABC combination. What date is your copy of AP2095? Also, what throttle opening are we talking about? +4 is quite a lot, and the passage you quote clearly implies that the ABC becomes double acting once the throttle is opened beyond a certain point; you could equally perform the test I suggest at say +6 and you'd get pretty much the same behaviour; I just picked +4 to make it obvious that the engine wasn't being stressed beyond its limits.

AP 1590B Merlin II and III Aero-Engines:
[...]
So a Merlin that wasn't modified for +12 boost the boost control cut-out will disable the boost control. A modified engine however will not disable the boost control but increase the allowed maximum boost to +12.

The mod was to drill holes in the valve such that the maximum boost comes out at +12; but the ABC isn't regulating in the normal way any more once the cutout is operated. AFAIK the eventual long term solution was to set up a gated throttle and keep the ABC in all the time, because drilling holes is a one shot deal which makes it very difficult to subsequently change the combat rating of the engine.

In any case, the fact remains that operation of the boost control cutout cannot increase the FTH of the engine. Once the throttle valve is open all the way, it cannot open further; there is no such thing as 110% throttle opening in a physical sense (though you can denote some arbitrary rating as 100% and then have some higher rating that you call 110% if you want, the fact remains that you cannot physically open a valve more than 100%).

Supercharger pressure ratio varies as rpm the square of its tipspeed. The FTH for +12 is given as 10500 feet or so. Ambient pressure at 10500 feet is 68.354 kPa. +12 psi is about 184.039 kPa. Supercharger pressure ratio is therefore about 2.69.

Meanwhile, at 17000 feet, ambient pressure is 52.752 kPa. Multiply by 2.69 and we get 147.282. Knock off the 101.325 kPa datum and we have +6.67 psi boost.

So I probably should have picked 10000' for the +12 FTH. You get the point though? The supercharger just sits there producing a pressure ratio of about 2.7 at 3000 engine rpm, and the ABC operates the throttle valve to control boost. The actual FTH you get is a function of the boost you ask for, but the FTH for any given boost level is set by the supercharger and ambient conditions, not the ABC.

[Sternjaeger]

How impressively, deeply, illogical.

..that's British engineering for ya matey ;-)

[klem]

Does any of this explain why I get the following results when I push/pull the Boost Cut-out in the Spits and Hurris? (NB I have to jiggle the throttle or hit the WEP key to make it actually go into higher boost)

Hurricane DH 5-20 (AKY) @ ~8,000 feet in level flight
Full Throttle boost: 6.25lbs
Pull Boost Cut-out: 6.25lbs

Hurricane Rotol (FTN) @ ~8,000 feet in level flight
Full Throttle boost: 5.2lbs
Pull Boost Cut-out: 5.5lbs

Spitfire I (ZPE) @ ~8,000 feet in level flight
Full Throttle boost: 6.2lbs
Push Boost Cut-out: 6.2lbs

Spitfire Ia (YTR) @ ~8,000 feet in level flight
Full Throttle boost: 6lbs
Push Boost Cut-out: 6lbs

Spitfire IIa (OVC) @ ~8,000 feet in level flight
Full Throttle boost: 5.5lbs
Push Boost Cut-out: 8lbs

In the Hurricane DH 5-20 and the MkI/Ia Spitfires I don't see an increase in boost and the Hurricane wit the Rotol prop only gives an extra 0.3lbs.

Is that correct?

[Viper2000]

According to AP 2095 Pilot's Notes General:
[...]
So sounds like this behaviour is correct.

If you read further in AP 2095, it says:

"NOTE - In Merlin engines the throttle valve cannot open fully unless the throttle lever is moved forward of the setting which gives maximum weak boost at the lowest altitudes. Therefore, at higher altitudes, the throttle must be set forward - up to the climbing position if necessary - in order to get the maximum permissible, or obtainable, boost for economical cruising."
[page 37 (http://www.scribd.com/doc/40498672/A-P-2095-Pilot-s-Notes-General-2nd-Edition-UK-1943)]

The Spitfire I notes state that maximum weak cruising is 2600 rpm, +2.5 psi.

Therefore, at +4 psi, the throttle valve should be able to open fully and the boost should not fall until the FTH for +4 psi at 2600 rpm is reached.

Forward movement of the throttle lever at this point should make no difference because the throttle valve should be fully open, and therefore the only way to increase power output would be to increase engine rpm (which naturally increases supercharger rpm and thus supercharger pressure ratio).

///

klem, I think the consensus is that the Spitfire IIa gives +12 psi boost when the cutout is operated, but since the gauge only reads to +8, you only see +8. However, I note with interest that the "no cockpit" gauge does read to +12, and yet this also only displayed +8 last I checked, so it's not obvious whether or not this is really a bug or a feature.

The Spitfire I notes say that operating the cutout with 100 octane fuel should give +12 psi as well.

I don't know exactly what is supposed to happen with 87 octane for several reasons:

#1 the maximum permissible boost should be less, but
#2 prior to the boost control cutout mod (http://www.spitfireperformance.com/ap1590b.jpg), operating the cutout at low altitude would give more than +12; I'm not sure how much, but if the boost control was literally disabled then WOT at sea level would give about +25 psi, which would consume the overhaul life of an early Merlin in about 15 minutes flat (which they actually did with N17 using special fuel when attempting to chase the world speed record; output was about 2500 bhp, FTH obviously 0).

It is implied in AP 2095 (http://www.scribd.com/doc/40498672/A-P-2095-Pilot-s-Notes-General-2nd-Edition-UK-1943) page 41 that the early cutout did indeed literally disable the ABC. I bet that must have been fun...

Really we need to try to find an old set of Notes for the Spitfire I from the 87 octane days to work out what the pre-mod boost control cutout was actually for, given that climbing boost and maximum boost are the same in the Notes I have which came with the CE.

The Hurricane basically had the same engine, as the Spitfire, and therefore I wouldn't expect to see different boost figures.

FTH for +12 was about 10000' at 3000 rpm for early Merlins, and therefore you should be able to get as much boost as you're allowed at 8000.

[klem]

.....................

klem, I think the consensus is that the Spitfire IIa gives +12 psi boost when the cutout is operated, but since the gauge only reads to +8, you only see +8. However, I note with interest that the "no cockpit" gauge does read to +12, and yet this also only displayed +8 last I checked, so it's not obvious whether or not this is really a bug or a feature.

The Spitfire I notes say that operating the cutout with 100 octane fuel should give +12 psi as well.

I don't know exactly what is supposed to happen with 87 octane for several reasons:
...................
Really we need to try to find an old set of Notes for the Spitfire I from the 87 octane days to work out what the pre-mod boost control cutout was actually for, given that climbing boost and maximum boost are the same in the Notes I have which came with the CE.

The Hurricane basically had the same engine, as the Spitfire, and therefore I wouldn't expect to see different boost figures.

FTH for +12 was about 10000' at 3000 rpm for early Merlins, and therefore you should be able to get as much boost as you're allowed at 8000.

Thanks for that.

My concern was that in three cases the boost didn't change at all (Spit1s and Hurricane DH 5-20) and the Rotol Hurricane boost barely moved.

Did I miss somewhere that these are 87 Octane and boost isn't effective?

[fruitbat]

Thanks for that.


Did I miss somewhere that these are 87 Octane and boost isn't effective?

only plane that appears to be 100 octane is the spitII, all the rest 87 octane:-|

[Viper2000]

Merlin marks II, III, IV and V were covered by the R.M.1.S. rating, which in its final state was:

Combat power = 1310 bhp/3000 rpm/9000' at +12 psi

or

1440 bhp/3000 rpm/5500' at +16 psi (Sea Hurricane only)

Takeoff power = 880 bhp/3000 rpm/SL at +6¼ psi.

There was also an R.M.2.S. rating which allowed increased takeoff power for 100 octane fuel before the +12 psi combat rating for R.M.1.S was introduced:

Combat power = 1000 bhp/3000 rpm/15500' +6¼ psi. (provisional rating, not confirmed by type test)

Takeoff power = 1000 bhp/3000 rpm/SL at +8¼ psi.

[This rating is a bit strange; my suspicion is that the +8¼ psi may actually be the 87 octane combat rating with boost cutout engaged, because otherwise if everything was rated at +6¼ psi boost then there wouldn't have been any purpose in fitting the cutout before 100 octane fuel was used...]

Cruise and climb were the same for both ratings:

Cruise power = 2600 rpm +4½ psi
Climb power = 2600 rpm +6½ psi

Power isn't specified for either of these ratings; you'd get what you got!

The Merlin XII had a higher supercharger gear ratio (9.089 vs 8.588), and it therefore achieved a higher FTH (but less power due to the increased supercharger drive requirement). It was covered by the R.M.3.S. rating:

Combat power = 1280 bhp/3000 rpm/10500' at +12 psi

Takeoff power = 1175 bhp/3000 rpm/SL at +12½ psi.

Cruise power = 2650 rpm +7 psi
Climb power = 2850 rpm +9 psi

Data from The Merlin in Perspective - the Combat Years, by Alec Harvey-Bailey, 4th edition, Rolls-Royce Heritage Trust.