PDA

View Full Version : Actual operation of the Merlin Boost controller



camber
Dec-17-2013, 02:26
Ages ago I wrote a detailed post on the British Merlin boost control and the boost control cutout in the ubi forums and was ably helped out by 41Sqn_Banks, lane and others. I thought I would summarise the original post here updated with the extra info that was posted. There is lots more great documents scanned into the original thread if you are interested:

http://forum.1cpublishing.eu/attachment.php?attachmentid=8264&d=1326414965

I have a background in control engineering and my interest is: how did the real boost control and override work exactly? What is being measured, what is being manipulated and how?

In modern process control, we have sensors (measuring the process), actuators (intervening in the process), and computer logic driven control loops feeding electrical information between the two. I am fascinated with 1940s and earlier control engineering, where to control a parameter like actual boost at a rated boost setpoint the control loop is a purely mechanical system where the sensors, actuator and control logic are real interconnected parts like valves, pistons and control linkages.

The funny thing is that the handle marked "throttle" in the Spitfire really is not a direct throttle control at all. The handle is the input of a mechanical control loop that maintains a boost (manifold pressure) setpoint, based on where the handle is. The control loop output is the throttle valve position (between carb and supercharger). The pilot does not actually have direct throttle valve control unless he pulls the boost cutout switch.

The problem with giving the pilot a direct link between his "throttle" handle and throttle valve is that a fixed gear supercharger powerful enough to give altitude performance will develop way too much manifold pressure on the ground unthrottled. 87 octane fuel can only tolerate around +6psi boost without engine-damaging predetonation, however the Merlin II/III supercharger can develop around +20psi at full throttle valve opening on the ground. The BF109 DB601 actually uses a fluid drive to alter the supercharger speed instead of fixed gear, an arguably more complicated but elegant strategy.

American planes of the same period had direct throttle valve control, so pilots had to carefully set appropriate manifold pressure (by not opening the throttle valve too much) and then continuously adjust manifold pressure (which would decrease as the plane gained altitude). Analagous to a contant speed prop or automatic temperature control, combat performance improves if you take routine workload away from the pilot. For this reason the British gave their Spitfires and Hurricanes automatic boost control. But how did it work?

The best explanation for a suitable boost control I came across was this document:

Rolls-Royce Merlin Automatic Boost Regulator, Courtesy of Dave Birch and the Rolls-Royce Heritage Trust (http://www.enginehistory.org/Piston/Rolls-Royce/R-RmerlinABC/R-RmerlinABC.shtml)

I recommend a full read of the document if interested, but here is a simplified version of the system with some of the detail stripped away (Figure 1). Basically the boost cylinder valve is plumbed so that a relatively high pressure (from the supercharger output itself) is routed to one side or the other of a piston attached to the throttle via linkage. When actual boost is equal to rated (desired) boost, both sides of the cylinder are sealed and the piston stays where it is. There is more detail about how this is actually done in the original document. The piston has a linkage which alters the mechanical relationship between the pilot's throttle lever and actual throttle plate. This is a rather simplified version, in fact in the real Spitfire the linkage is done through a differential gear but with equivalent mechanical logic.

http://i406.photobucket.com/albums/pp142/wjhawaii/spithrottle_sm.jpg

There was actually two versions of the boost control, fixed datum (Merlin II) and variable datum (Merlin III and retrofitted Merlin II). The essential difference was that the fixed datum could only control at rated boost (+6.25psi for 87 octane). The variable datum was better because it could control boost at any intermediate value below rated boost based on the pilot handle position. If you selected +2psi using the handle and climbed, boost would remain controlled at +2psi with the controller automatically opening the throttle valve as height increased.

Boost cut-out override (BCCO)

The Boost control override did not originate as an emergency power setting, but was adapted to be so by the ingenious Brits. In original form it was just a way of disabling the boost controller in case of malfunction, thus making the system directly link the pilot handle to the throttle valve and giving him the ability to set any boost the supercharger was capable of (but without control, boost would change with altitude).

CloD shows the Spitfire red tab rotating a little cam allowing the throttle handle to go further, which is not the actual case and confuses the red tab with the throttle gate which appears as an additional overboost system on the (reality) Spit II. In fact the red tab in Spit I/II pulled a cable which opened a channel around the valve, which applied suction to the valve piston and forced it to the left in Figure 1 and stay there, thus disabling the controller. The Hurricane is correct in that the red tab is replaced by a knob that pulls the cable (the "tit").

Although it is hard to find references on this, it is easy to see how the BCCO could become an unofficial emergency power switch. A pilot could pull it and try for a bit more boost than the rated 6.25psi, and hopefully get a bit more power without damaging the engine.

Modified BCCO Overboost

When the British went to 100 octane fuel which could tolerate +12psi boost, they adapted the BCCO from it's original purpose (disable boost controller) to be an official overboost switch, and they did it by drilling a single hole of precise size into the system at squadron level. This hole changed the cutout action so instead of totally switching off the controller, it incremented the upper controllable boost level from +6.25psi to +12psi. Of course the supercharger is not spinning any faster, so this is only possible if you are low enough with sufficient rpms for the supercharger to deliver +12psi boost (full throttle height is lower for +12psi than +6.25psi). It was a very impressive system modication, really. The hole was on the suction side so it it did not vent pressurised fuel-air to the atmosphere (which would not a good safety feature even without incendiary bullets around).

The throttle gate overboost system

The real Spit II had (in addition to the BCCO red tab) a take off power gate on the pilot handle that gave +12.5psi. This seems a bit strange, an alternative overboost system that seems redundant! This is the "gate method" alternative overboost system described in the Rolls Royce document above and summarised below:

The gate system does not alter the boost controller action like the (modified) BCCO red tab, it actually overrides it. Effectively it allows a maximum throttle valve position to be held even with the controller piston fully to the right in Figure 2 (i.e the controller piston is sensing "boost too high" and is trying but being disallowed to close the throttle). In the case of the Spit II gate this throttle position is calibrated to give +12.5psi at sea level.

http://i406.photobucket.com/albums/pp142/wjhawaii/spithrottle_b.jpg

The advantage of this system is that it is real take off only overboost. In normal conditions (take off gate, red tab not pulled), the Spit II would take off with around +12.5psi depending on airfield altitude, which would immediately decay by itself as the plane climbed even if the pilot didn't touch the handle. It would decay back to +9psi (Spit II rated boost) at which point the controller regained effectiveness, and boost would be maintained at +9psi in the climb until such point the supercharger could no longer deliver it.

Anyway feel free to comment,

Cheers, FS~camber

9./JG52_Meyer
Dec-17-2013, 03:36
Way over my head my friend , but thank you for posting

Bewolf
Dec-17-2013, 05:15
Hey Camber!

I am no Techie, but I found these descriptions highly interesting. You never stop learning. Thanks for posting! http://1.2.3.11/bmi/theairtacticalassaultgroup.com/forum/images/smilies/thumbup.gif

Cheers
Bewolf

Roblex
Dec-17-2013, 10:38
Fascinating! Thanks for posting.

You mentioned US engines being directly linked. Here is a description of just how hard that made them to control

Not as Easy As It Looks (http://thanlont.blogspot.co.uk/2012/05/not-as-easy-as-it-looks-ii.html)

I will never complain about how hard it is keeping a Merlin from overheating again! :D

ATAG_Snapper
Dec-17-2013, 11:22
Great post, Camber! :thumbsup:

I've read tech manuals on this topic that had me totally baffled, but your description above lays it out beautifully. Thank you. :salute:

camber
Dec-18-2013, 07:18
Thanks for the feedback!


Fascinating! Thanks for posting.

You mentioned US engines being directly linked. Here is a description of just how hard that made them to control

Not as Easy As It Looks (http://thanlont.blogspot.co.uk/2012/05/not-as-easy-as-it-looks-ii.html)

I will never complain about how hard it is keeping a Merlin from overheating again! :D

Fascinating article there, I didn't realise that a late war model like the Hellcat was still so difficult! It sounds like although there is some automatic MAP control, the ins and outs of manually selecting blower speed were still terribly complicated. I remember reading (can't quite remember the reference) that P-38s were notoriously high workload to change configuration. An officer in theater pointed out that many times that when formations were attacked, pilots did not appear to react (and were killed), and it was thought they were going through the many steps required to go from long-range economical cruise to combat configuration. Lockheed pointed out that they had proposed solutions to these problems but the USAAC upper brass had not shown interest. The Germans certainly had the right idea with the automatic engine management Kommandogerate system in the FW190. The Brits certainly were OK with automatic blower shifting once they had two stage in the Spit IX, although pilots reported being half scared to death when the blower suddenly shifted by itself the first time, apparently it was not a smooth transition!

ATAG_NakedSquirrel
Dec-23-2013, 13:55
Fun read, thanks!

69th_Zeb
Nov-25-2017, 00:35
Camber, that article was a great read. I learned a bunch of new stuff! You can see more complex versions of the same applied technology still being used on PT-6/Allison turbine FCUs.

I'd make a minor clarification (knock on wood), 87 octane ships were limited to 7(6 1/4 depending) inches. 100 octane allowed the engine to be operated without knocking at 12.5". The ABCs were modified to limit (A.P.1590b) a governing range up to 9". If the red boost cutout lever at the quadrant was rotated, that would move the piston to max governing range of 9". If, in addition, the throttle was moved past the gate, you would gain an additional 3.5" beyond the controller's governing range.

In order to reach 12.5" inches of boost, both conditions would need to be true as the mechanism travel is cumulative (BCO+Gate). Correct?

Another interesting note, there was another aneroid inside the carb that was connected to the pressure side of the supercharger. As pressure increased, the aneroid would pull the main metering valve open to automatically enrichen the mixture. Pilot's notes showed a 40% increase in fuel consumption @ 12.5".

69th_Zeb
Nov-25-2017, 22:51
Was just going through my copy of the Merlin Series II manual. Here's some additional diagrams of the boost controller.
32790
32791

More images...
32793
32795

And some more discussion...
https://forum.il2sturmovik.com/topic/29968-spitfire-engine-control-management/page-2

Baffin
Nov-30-2017, 10:43
Camber, that article was a great read. I learned a bunch of new stuff! You can see more complex versions of the same applied technology still being used on PT-6/Allison turbine FCUs.

I'd make a minor clarification (knock on wood), 87 octane ships were limited to 7(6 1/4 depending) inches. 100 octane allowed the engine to be operated without knocking at 12.5". The ABCs were modified to limit (A.P.1590b) a governing range up to 9". If the red boost cutout lever at the quadrant was rotated, that would move the piston to max governing range of 9". If, in addition, the throttle was moved past the gate, you would gain an additional 3.5" beyond the controller's governing range.

In order to reach 12.5" inches of boost, both conditions would need to be true as the mechanism travel is cumulative (BCO+Gate). Correct?

Another interesting note, there was another aneroid inside the carb that was connected to the pressure side of the supercharger. As pressure increased, the aneroid would pull the main metering valve open to automatically enrichen the mixture. Pilot's notes showed a 40% increase in fuel consumption @ 12.5".

At the risk of sounding picky, please confirm that you're writing [9"] when Merlin boosts were expressed in [9 PSI]. I remember in the CFI training days, of working hard to get manifold pressure inches of mercury straight in my head... Then CLoD came along with Pounds/Square inch to express manifold boost pressure. :S

Even as I write this, I only THINK that I've got this correct. The OP and Zeb are obviously extra well qualified in piston engine induction systems... This thread has been free tuition for me! :lecture:

Tibsun
Nov-30-2017, 13:36
Baffin... PSI Pounds per Square Inch. Inch = ".
He pretty sure ment PSInch and not inches of mercury

Camber wrote this long time ago, he might be not there answering to you guys.

Baffin
Nov-30-2017, 20:46
Baffin... PSI Pounds per Square Inch. Inch = ".
He pretty sure ment PSInch and not inches of mercury

Camber wrote this long time ago, he might be not there answering to you guys.

Thanks Tibsun. It's important for the new readers that we get the abbreviations right. " is probably not the best way to say PSI when potential confusion is possible. :salute:

69th_Zeb
Nov-30-2017, 20:59
Haha. Naw good catch. Just a bad slip of terminology.
I knew Camber was probably gone for a while, but I figured some of you guys would chime in and find nuggets to fix.

Would be fun to make an animation in Blender someday.

camber
Dec-24-2017, 01:35
Hi Zeb, I only just noticed that this ancient post was getting some new interest. I am still around but usually flying as L.E.O Charlton.

I was definitely using Pounds per Square Inch (PSI), where atmospheric is about 15psi (although 'boost' is PSI above this) instead of inches of mercury manifold pressure.

That is a interesting point about how the gate, BCCO and boost controller work together, say in the actual Spit II. My understanding when I wrote it that going through the gate pulls the boost control piston (see the original diagrams) to the "boost too high" stop, moving the throttle valve onto a fixed setting which corresponds to 12.5psi on the ground. For the sake of discussion let's say this is 60% open (the supercharger can make ~20 psi at sea level). So it will give 12.5psi regardless of whether the BCCO red tab (Spit) or tit (Hurri)is activated.

So what happens if we pull through the gate, take off and just keep climbing? The piston will stay stuck in the 'boost too high' position for a while, but boost will decay from 12.5psi takeoff boost as the supercharger output reduces due to the fixed throttle valve position + decreasing ambient pressure. At some point the supercharger output pressure will decay enough with altitude for the piston to move off the 'boost too high' stop, throttle valve can now increase above 60% open, boost is controlled again and should stay constant up to full throttle height (i.e. when throttle valve is 100% open). The controlled boost value will be lower or higher at this time depending on whether the BCCO is pulled.

But if you look at the original linkage diagrams I posted the controlled value (say 6.25psi or 9psi) should be mechanically incremented by a fixed value due to the fact you have left the throttle handle in the gated position. But I am not sure if that would really happen because the linkage system in the diagram is a simplified version of the real mechanism to explain the concept, which in reality looks more similar to a differential gear. So I'm not really sure. Obviously you were probably not supposed to gate the throttle, pull the BCCO and climb to full throttle height.

Just to summarise I don't think you need to pull the BCCO for the gate to work properly, but it will make a difference to what (probably) lower boost value the system settles on as you climb.

Interesting about the mixture aneroid, I didn't know that but it makes sense.

Cheers, camber



Camber, that article was a great read. I learned a bunch of new stuff! You can see more complex versions of the same applied technology still being used on PT-6/Allison turbine FCUs.

I'd make a minor clarification (knock on wood), 87 octane ships were limited to 7(6 1/4 depending) inches. 100 octane allowed the engine to be operated without knocking at 12.5". The ABCs were modified to limit (A.P.1590b) a governing range up to 9". If the red boost cutout lever at the quadrant was rotated, that would move the piston to max governing range of 9". If, in addition, the throttle was moved past the gate, you would gain an additional 3.5" beyond the controller's governing range.

In order to reach 12.5" inches of boost, both conditions would need to be true as the mechanism travel is cumulative (BCO+Gate). Correct?

Another interesting note, there was another aneroid inside the carb that was connected to the pressure side of the supercharger. As pressure increased, the aneroid would pull the main metering valve open to automatically enrichen the mixture. Pilot's notes showed a 40% increase in fuel consumption @ 12.5".

69th_Zeb
Dec-25-2017, 00:10
Hi Camber!

I love hangar-flying threads like these.

I forgot to add it to my images, but I neglected another image showing linkage installation in the Spits. The red throttle gate that's safetied with brass safetywire is itself a separate control lever connected directly to the pneumatic side of the boost cut out (like the red pull lever on the hurri), making the piston go to its stop. The open differential setup would allow the throttle lever to input to the butterfly beyond governing range...when moved further beyond the gate. Is that correct?

In other words, for aircraft approved for 12+ psi, moving the piston to max would only give you 9 psi, beyond the gate on the throttle would give you 12+.
And yeh, you would only reach these levels at sea level to 1000 ft only on standard day.

Another neat detail that I noticed in the engine modelling (last I checked), is that at full throttle, prop RPM does not affect boost pressure at the gauge, because there isn't a restriction at the throttle while the engine is sucking/boosting. When slight restriction is induced such as part throttle/closed, changes in RPM (prop pitch) are more apparent on the boost gauge.

I'm away from home, but I'll try to dig for that lever diagram, as well as the one for the mixture aneroid.

~S~!