DCS Spitfire Mk IX performance

[DSR_T-800]

I'm sure the Fw-190D9 would handle a Griffin no problem. :)

[Hummingbird]

I'm sure the Fw-190D9 would handle a Griffin no problem. :)

 

With the way that thing handles ingame atm? Not a chance mate. Unless ofcourse the Spit gets equally porked, then the Dora might stand a chance, but I do see the Spit XIV as superior regardless :)

[IronJockel]

With the way that thing handles ingame atm? Not a chance mate. Unless ofcourse the Spit gets equally porked, then the Dora might stand a chance, but I do see the Spit XIV as superior regardless :)

 

What about it? it handels great.

However i can't wait to see cambat between the Dora and the mk9. Even if it is a 1943 version, it will be quite the opponent for the german fighters.

[Ala13_ManOWar]

Whatever the argue is (I don't get really why the argue), if those are the charts for the DCS Spit IX module we'll be getting another piece of art milestone in simulation and that's great news :beer: .

 

I can't wait to experience myself the feeling of a really accurate Spitfire flight dynamics and behaviour for the first time in PC history. That's the closest we'll get without expending 4000£ in a 15 minute ride.

 

S!

[flare2000x]

I normally don't post here, but I must say I think we don't have to argue.

 

So what if the FTH is a thousand feet higher? So what if the speed is 5% better at certain altitudes? These differences are minute and frankly and slight advantages will just help level the playing field a bit more between a 1943 and a 1945 plane. Will the simulation still be the most accurate Spitfire simulation to date? Absolutely!

 

Besides, didn't we have the same arguments about over and under performance when the 190 and 109 came out?

 

However it is an enlightening discussion - thanks for the straight and to-the-point information, YoYo! Can't wait to fly the Spitfire.

 

Flare

[Hummingbird]

What about it? it handels great.

 

Sadly not, we tried to bring the issues to attention but were ultimately ignored. To be fair though I'm not so sure its just the Dora - there's definite issue with the Mustangs CL curves as well, but its hard to debate with people who dont want to share the data they claim to have.

 

As for the Spitfire IX, I think it will eat all the other fighters we have ingame atm alive.

[NineLine]

Sadly not, we tried to bring the issues to attention but were ultimately ignored. To be fair though I'm not so sure its just the Dora - there's definite issue with the Mustangs CL curves as well, but its hard to debate with people who dont want to share the data they claim to have.

 

As for the Spitfire IX, I think it will eat all the other fighters we have ingame atm alive.

 

One, the Dora is off topic here.

 

Two, you were not ignored, Yo-Yo answered all your concerns, you just didn't get the answers you wanted. So time to drop it. That is the last warning I am giving on that.

 

Please... discuss the topic at hand, and dont make these discussions an excuse to complain about something or someone else.

[rel4y]

http://www.spitfireperformance.com/spitfire-lfix-ads.jpg

 

If performance (410 mph at 21500 ft) of the DCS model relates to the quoted spec sheet, there is a discrepancy of ~6 mph (1.5% higher) as it states that max. speed of 404 mph is reached at 21000 ft with an aircraft weighing 7106 lbs. The relative error in FTH is 2.4%. I wouldnt really call these percentages a significant relative error.

 

Now looking at higher altitudes, namely 36k ft and taking an approximate mean value (~375 mph) of the best high alt performing tests MA 648 and BS 543 only, the relative error will be ~5.3%, which in term is significant. The spec sheet doesnt give any numbers at these alts, so theres no comparison possible. The spec sheet data seemed to be more in agreement with the BS 543 test though, for which the resulting relative error would be in the range of 6.5-7.5%.

 

I am quite sure in practice not many people will ever fly at these altitudes in DCS, but accuracy to real world data has been the premiss for a DCS flight model so far and it should stay this way.

[Yo-Yo]

If performance (410 mph at 21500 ft) of the DCS model relates to the quoted spec sheet, there is a discrepancy of ~6 mph (1.5% higher) as it states that max. speed of 404 mph is reached at 21000 ft with an aircraft weighing 7106 lbs. The relative error in FTH is 2.4%. I wouldnt really call these percentages a significant relative error.

 

Now looking at higher altitudes, namely 36k ft and taking an approximate mean value (~375 mph) of the best high alt performing tests MA 648 and BS 543 only, the relative error will be ~5.3%, which in term is significant. The spec sheet doesnt give any numbers at these alts, so theres no comparison possible. The spec sheet data seemed to be more in agreement with the BS 543 test though, for which the resulting relative error would be in the range of 6.5-7.5%.

 

I am quite sure in practice not many people will ever fly at these altitudes in DCS, but accuracy to real world data has been the premiss for a DCS flight model so far and it should stay this way.

 

The post FTH region is very tricky for measurements both in real life and in simulation - any small deviation in altitude or speed lead to changes in static and ram pressure and require long time to stabilise the speed.

For example, this is the third way of plotting this area.

 

THe chart is from the very detailed report regarding the methods of neutralising ram air leak to the engine cowling through the air filter.

The solid and dash-dotted lines are for blind and louvred plates that cover air filter elements preventing the ram air from leaking. So, these tests were conducted for Mk IX with Merlin 66 having universal air intake.

 

As you can see, this plane shows more than average results...

Edited March 28, 2016 by Yo-Yo

[Friedrich-4B]

The post FTH region is very tricky for measurements both in real life and in simulation - any small deviation in altitude or speed lead to changes in static and ram pressure and require long time to stabilise the speed.

For example, this is the third way of plotting this area.

 

THe chart is from the very detailed report regarding the methods of neutralising ram air leak to the engine cowling through the air filter.

The solid and dash-dotted lines are for blind and louvred plates that cover air filter elements preventing the ram air from leaking. So, these tests were conducted for Mk IX with Merlin 66 having universal air intake.

 

As you can see, this plane shows more than average results...

 

In addition to this, the universal air intake could be fitted with a denser "Tropical" air filter, that reduced ram pressure cf the temperate version; as a result, the top speeds suffered; BS310 was tested with both the temperate and tropical variants in late 1942:

 

(NB: the Vokes Universal air intake system was standardized on the Mk VIII and introduced to the IX in late-1943-early 44.)

Edited March 29, 2016 by Friedrich-4/B
Reinstate image

[Kurfürst]

2 Kurfurst

 

I will answer step-by-step:

 

First of all, all speculations about weight effect to level speed are useless. As you must know, the drag polar is a quad parabola. For high pwered planes like fighters at high speed and low CL numbers (for moderate and low altitudes) the 5% of weight difference DOES NOTHING. The report I have contains the weight corrections, they are not more than 0.5-0.7 mph that is much less than instrument error. It is not more than an engineer's "flees catching" that is completely useless in this context.

 

Part 1.

 

If the results originally re-calculated/achieved with only 95% weight (as they are - see your own test papers and yellow specs paper) they cannot be used as a basis of reference for a model for 100% weight, obviously. It will introduce increasing margin of errors with weight in level speeds - and indirectly will also skew other related specs - see the difference on airspeed for another aircraft shown in the manual.

 

I find it interesting because the DCS model also shows the very same symptoms that only the blind cannot see - it quite close the real world reference but as altitude increase, the DCS model also increasingly diverges from the reference flight trial results (which I consider a dubious choice for already explained and proven reasons, but may just aggravate the problem, but is not the cause of it).

 

"Instrument/measurement error" is not really an explanation since its impossible that ALL (good or bad) results were effected by such a series of "instrument errors". Moreover, a single or two error it would not shows a clear trendline.

 

On the real world data I just posted, this is quite clearly caused by difference in weight and resulting required increased/decreased Angle of attack and resulting drag change. The question is, what causes the divergence on the DCS dataset...?

[Pilum]

I think Yo-Yo is quite correct about the significance of a 5% weight change: My C++ simulation estimate of the difference in speed for a 5 % weight change at FTH for the Spitfire Mk9 is 1.41 Km/h or an astounding 0.2%. Getting late now and busy on Wednesday but will post "flea" figure on Thursday if needed. :smilewink:

[otto]

I think Yo-Yo is quite correct about the significance of a 5% weight change: My C++ simulation estimate of the difference in speed for a 5 % weight change at FTH for the Spitfire Mk9 is 1.41 Km/h or an astounding 0.2%. Getting late now and busy on Wednesday but will post "flea" figure on Thursday if needed. :smilewink:

 

Not that it's a topic that i know much about. But there's no whey to know if your C++ simulation is correct .

[otto]

Chances are that Pilum, as a qualified aircraft engineer, with several years of aircraft industry experience, might know more about calculating aircraft performance than either you, or Kurfurst, for that matter. Chances are also excellent that Yo-Yo knows more about modelling accurate flight FMs than those who think they know better. :joystick:

 

I'm disappointed that the spitfire is not a 25lbs version so if it's one of the "best" no 25lbs versions if fine by me. But i'm a blunt guy and: my c++ simulation shows.. is just an empty argument .

[Friedrich-4B]

Part 1.

 

If the results originally re-calculated/achieved with only 95% weight (as they are - see your own test papers and yellow specs paper) they cannot be used as a basis of reference for a model for 100% weight, obviously. It will introduce increasing margin of errors with weight in level speeds - and indirectly will also skew other related specs - see the difference on airspeed for another aircraft shown in the manual...

 

"Instrument/measurement error" is not really an explanation since its impossible that ALL (good or bad) results were effected by such a series of "instrument errors". Moreover, a single or two error it would not shows a clear trendline.

 

On the real world data I just posted, this is quite clearly caused by difference in weight and resulting required increased/decreased Angle of attack and resulting drag change. The question is, what causes the divergence on the DCS dataset...?

 

Rather than providing a chart showing the comparative differences between a (presumed) 109G at take-off weights of 113% & 95%, it would be more useful if Kurfurst could calculate and provide a dataset and performance curves for a Spitfire L.F. Mk. IX's performance at 100% T/O weight, taking in all the information Yo-Yo has tried to explain to him.

 

That way we can all see whether Kurfurst's argument holds any water or, whether, as Yo-Yo has explained, the real world difference was minimal

 

Note also that the A&AEE calculated that the average speed difference due to the different FTH's of the 4 Mk IXs tested was 3 mph (5 km/h) - ie; average human walking speed. ( see:http://www.spitfireperformance.com/ma648.html )

 

Owing to large variations in the performance of the four aircraft tested it is not possible to obtain an accurate value for the improvement in performance but the above quoted figures for increase in full throttle height give an average increase of about 3 mph. around the full throttle height.

 

Part 1.

I find it interesting because the DCS model...also increasingly diverges from the reference flight trial results (which I consider a dubious choice for already explained and proven reasons, but may just aggravate the problem, but is not the cause of it)

 

Reviewing Kurfurst's "proven and explained reasons" for considering the flight test results as "dubious": From the first page of this thread:

 

the engine of the prototpye BS 543 was operating on too rich mixture, therefore it had an abnormally high full throttle height and improved low/medium performance at low medium altitude. Very higher altitude performance suffered somewhat.

 

BS 543 also had an experimental propeller that was never serialized and achieved performance far in excess of ALL other Spitfire IXs/VIIIs tested.

 

See: http://www.spitfireperformance.com/bs543.html

* BS543's Merlin was not operating on a too rich mixture because while the A&AEE report mentions that BS551's Merlin 70 was operating on a too rich mixture, there's no mention of BS543 having similar problems. Kurfurst is merely reading this into the report, based on his coloured interpretation of the performance curves.

 

* BS543's Rotol propeller was a "serialized" propeller that was proven to be slightly less efficient than others tested on other Spitfire IXs - it did not confer a magical performance boost. See:

2.1 General. Except for the propellers the condition of the three aircraft was identical. The propeller details are summarised below:

and http://www.spitfireperformance.com/proptypes.jpg.

 

*BS534's level speed and climb rate NOT Full Throttle Height was used as a reference by Yo-Yo.

 

MA 648 is an even poorer choice since its representative only of an experimental plane with a type of injection carburetor that never saw service in the end (all Merlin 66s were produced with the Bendix Stromberg injection carburator, not the SU type carb), hence the much higher altitude performance than normal. That plane never existed in service.

 

http://www.spitfireperformance.com/ma648.html

 

* The only "experimental" feature of MA648 was the S.U. Injection carburettor that conferred a maximum FTH advantage of

200 ft. in MS gear and 800 ft. in FS gear.

 

For the rest, MA648 had the Vokes Universal carburettor air intake fitted to production standard L.F. Mk. IXs operating over Europe.

 

What Kurfurst fails to mention is that Yo-Yo also used JF275 (FTH 20,200 ft, 405 mph), an L.F. Mk VIII, as a reference.

[Buzzles]

Guys, I understand you're really all passionate about the Spit, I'm really looking forward to it as well, but this thread was simply Yo-Yo offering a quick comparision of the current DCS model to some publically available existing data.

 

Would it not be more sensible to actually wait until we get the Spit in DCS so we can build up our own data set for the DCS Spit and compare against a much larger and wider source of data than one(!) report?

[rel4y]

Well, in all honesty I have some doubts about the MA 648 test specifically. Maybe someone could help me out about info on the S.U. pump injection carburetor tested. The engine spec sheet posted by Friedrich as well as this scan on spitfireperformance (below) list the normal Bendix-Stromberg carburetor as the standard injection mechanism. This is also the way I remember things to be.

 

wwiiaircraftperformance.org/Aircraft_Engines_of_the_World_Rolls-Royce_Merlin.pdf

 

So my question is: when was the injection carburetor design employed in Mk IX production, because I am not sure it was. :huh: Furthermore as stated in the report this S.U. pump version is a non production model.

 

This particular pump incorporates a rubber diaphragm which will probably be replaced by an improved metal one in later pumps.

Which in turn makes this statement in the test, very relevant.

 

These results compare favorably with those of other Spitfire LF Mk. IX aircraft, which fact is attributed cheifly to the higher full throttle height obtained with the S.U. pump.

THe chart is from the very detailed report regarding the methods of neutralising ram air leak to the engine cowling through the air filter. The solid and dash-dotted lines are for blind and louvred plates that cover air filter elements preventing the ram air from leaking. So, these tests were conducted for Mk IX with Merlin 66 having universal air intake.

 

As you can see, this plane shows more than average results...

 

Thanks for the response. This is an impressively perfoming Spit indeed. My main concerns respective high alt performance is, that the DCS LF variant performs better at high alt than the HF variants tested by the A. & A.E.E.

 

(http://www.spitfireperformance.com/bs310.html)

 

The graph of the current DCS model hits ~401 mph at 33.000 ft and the test of Mk. IX HF BS.310 give exactly a number of 391 mph at same altitude. That makes the LF 10 mph faster than the HF! Take a look at the table at the very bottom. And from interpolating the curves at altitudes above 33k ft the DCS LF will perform close to a 2 digit percentage better than the HF variant tested by A. & A.E.E.

 

This makes me wonder why the change in supercharger gear ratios from 7.06 to 8.03 for the second stage was jusified in the first place when the LF one performs better than the HF one at high alts. Yo-Yo I am with you that a 2.4% relative error in FTH is perfectly fine, but please take a look at high alt performance again!

Edited March 30, 2016 by rel4y

[Yo-Yo]

 

The graph of the current DCS model hits ~401 mph at 33.000 ft and the test of Mk. IX HF BS.310 give exactly a number of 391 mph at same altitude. That makes the LF 10 mph faster than the HF! Take a look at the table at the very bottom. And from interpolating the curves at altitudes above 33k ft the DCS LF will perform close to a 2 digit percentage better than the HF variant tested by A. & A.E.E.

 

This makes me wonder why the change in supercharger gear ratios from 7.06 to 8.03 for the second stage was jusified in the first place when the LF one performs better than the HF one at high alts. Yo-Yo I am with you that a 2.4% relative error in FTH is perfectly fine, but please take a look at high alt performance again!

 

As you can see, RL LF Spitfires were faster than real life HF model... that's real life measurements, things can happen.

[Kurfürst]

So my question is: when was the injection carburetor design employed in Mk IX production, because I am not sure it was. :huh: Furthermore as stated in the report this S.U. pump version is a non production model.

 

The short answer is that the Bendix-Stromberg injection carb was introduced with the Merlin 66 in 1943, and was standard (this was also the first carburattor that effectively solved the negative G problem, the previous "solution" was more like bandaging)

 

However the S.U. type of pump injection carburetor never was introduced on the Mark IX, or Merlin 66, although several types (or Marks) of it were proposed and tested for use.

 

As I understand the SU type as opposed to the Bendix type is a very different and simplier design, and as an advantage the SU carb improved the full throttle height of the engine under rammed conditions, by at least 1300 feet compared to the Bendix - probably because the carburetor caused losses in the SU design rammed air were less. That's one part of it of course, and this experimental design was anything like the SU carbs used later on cars, it had several disadvantages such as inaccurate fuel metering during cruise conditions and more sluggish throttle response.

 

The trials with MA 648 was just a test bed for this new type of proposed, better, but never introduced SU carburetor. This fact is very clear from the report. It also clearly states that BS 543 results were odd (BS 543 report itself states the same). Also there is nothing wrong with that, most trials during the war were about finding out how a proposed new equipment changes things, if its worth the trouble of making it standard, and not about how "in production" type performs.

 

In fact you do not have to look further than the results of BS 354 to get just how odd the BS 543 test run yielded when it came to FTH and altitude results, quite similar to the DCS Spitfire IXPerimental FM model which also yields a 2000 feet increase in the FTH, even in a 170 mph climb...

Edited March 30, 2016 by Kurfürst

[rel4y]

As you can see, RL LF Spitfires were faster than real life HF model... that's real life measurements, things can happen.

 

Well thats what I figured as well. As said the first problem I have is that a LF variant should not outperform an HF at high altitude (by such massive amounts btw) and it should not have higher FTH (which isnt the case to be clear). This would make the very existence of an HF supercharger obsolete, but the RAF used these HF models specifically as high altitude escorts and interceptors. So the RAF at least must have seen a benefit in the HF over the LF in terms of high alt performance.

 

To get more into the physics context: It is obvious why an increase in supercharger gear ratio will positively effect air compression and thus oxigen supply to combustion. It is also clear why this will have adverse effects at lower altitude. Therefore physics tells me that a supercharger with lower compression output can not deliver the same or even higher combustion efficiency with - mind - the same engine parameters.

 

Secondly, I do not have access to this air filter test so I cant really comment much on details of the subject. However I can see that this is an obvious outlier compared to the other results and thus my scientific sense tells me to rather question the outlier, than all the other results being in good agreement to each other. Now I dont know what, but I am pretty sure something had to be different in this test compared to the others.

 

The red line you plotted is for the cleaner elements blanked off. More appropriate would be the Louvred plate fitted one which performs worse, but resembles actual configuration of the universal filter design. Above 34k ft the DCS one will even outperform this outlier by ~9 mph max. :noexpression:

 

Dont get me wrong, I love DCS and all the flight model work you have done and feel it is as close as it gets to real life. However the community is not uneducated and has contributed several times to improvements of a DCS module and this time I feel its no different. Though I dont know what may be the cause, I still say something is not exactly right here.

Edited March 30, 2016 by rel4y

[Yo-Yo]

Part 1.

 

If the results originally re-calculated/achieved with only 95% weight (as they are - see your own test papers and yellow specs paper) they cannot be used as a basis of reference for a model for 100% weight, obviously. It will introduce increasing margin of errors with weight in level speeds - and indirectly will also skew other related specs - see the difference on airspeed for another aircraft shown in the manual.

 

I find it interesting because the DCS model also shows the very same symptoms that only the blind cannot see - it quite close the real world reference but as altitude increase, the DCS model also increasingly diverges from the reference flight trial results (which I consider a dubious choice for already explained and proven reasons, but may just aggravate the problem, but is not the cause of it).

 

"Instrument/measurement error" is not really an explanation since its impossible that ALL (good or bad) results were effected by such a series of "instrument errors". Moreover, a single or two error it would not shows a clear trendline.

 

On the real world data I just posted, this is quite clearly caused by difference in weight and resulting required increased/decreased Angle of attack and resulting drag change. The question is, what causes the divergence on the DCS dataset...?

 

The graph you posted does not meet two "if"s I hedged my statement about weight effect... even if we close our eyes and do not see 18% difference instead of 5% that is not a fair play, yeah?

I mentioned that this feature is valid only for high-powered plane capable to fly at high IAS, so, low CL, and for moderate altitudes where maximal IAS is not close to the IAS of best climb (maximal L/D ratio).

The plane you posted the chart for has 480 kph at SL that is not even close to the Spitfire IX speed, so I can presume that it is not so well-powered as Mk IX.

And, as you can see, the speed difference really increases with altitude reaching eternity near the ceiling, because the weight affects both climb rate and ceiling.

[Yo-Yo]

Well thats what I figured as well. As said the first problem I have is that a LF variant should not outperform an HF at high altitude (by such massive amounts btw) and it should not have higher FTH (which isnt the case to be clear). This would make the very existence of an HF supercharger obsolete, but the RAF used these HF models specifically as high altitude escorts and interceptors. So the RAF at least must have seen a benefit in the HF over the LF in terms of high alt performance.

 

To get more into the physics context: It is obvious why an increase in supercharger gear ratio will positively effect air compression and thus oxigen supply to combustion. It is also clear why this will have adverse effects at lower altitude. Therefore physics tells me that a supercharger with lower compression output can not deliver the same or even higher combustion efficiency with - mind - the same engine parameters.

 

Secondly, I do not have access to this air filter test so I cant really comment much on details of the subject. However I can see that this is an obvious outlier compared to the other results and thus my scientific sense tells me to rather question the outlier, than all the other results being in good agreement to each other. Now I dont know what, but I am pretty sure something had to be different in this test compared to the others.

 

The red line you plotted is for the cleaner elements blanked off. More appropriate would be the Louvred plate fitted one which performs worse, but resembles actual configuration of the universal filter design. Above 34k ft the DCS one will even outperform this outlier by ~9 mph max. :noexpression:

 

Dont get me wrong, I love DCS and all the flight model work you have done and feel it is as close as it gets to real life. However the community is not uneducated and has contributed several times to improvements of a DCS module and this time I feel its no different. Though I dont know what may be the cause, I still say something is not exactly right here.

 

You make a very common mistake - the COMBUSTION itself is a function of a carburettor or injector metering automatics. They measure the actual air mass flow and meter the required amount of the petrol keeping air/fuel ratio at the desired level depending on power rate - rich mixture for high power, weaker - for cruise.

Engine power depends primarily on the air mass flow, so, more Pascals and less Kelvins :)

If you increase blower ratio, the result will be in reduced power before FTH (due to overheated and then throttled air), at FTH (due to overheated air and increased power to blower) and increased power beyound the FTH.

If you take a look at the chart you will see that even with louvred plate the LF is faster than HF by 7 mph.

I agree it is not according the theory... but you must understand the thing I am always trying to bring to the minds - there are a lot of reasons making the measurements dispersion. If we are trying to compare, for example, samples in two cases as the values itself have dispersions greater than the difference between the values - sometimes the samples of the values can be twisted.

As the max TAS chart getting more data, you can see that the different tests give a lot of different points.

 

And, by the way, I really do not know yet if the air cleaner was used for Normandy Spifires, or we must have it blanked. And if our cockpit has a control for the air cleaner shutter. I think that it does not have... so the air cleaner will be blanked. :)

[Friedrich-4B]

And, by the way, I really do not know yet if the air cleaner was used for Normandy Spifires, or we must have it blanked. And if our cockpit has a control for the air cleaner shutter. I think that it does not have... so the air cleaner will be blanked. :)

 

The soils in the region of Normandy where the Allied airfields were constructed, are full of hard silicates that severely reduced the life of an aero-engine when ingested, so it's more likely that the air filter was used in Normandy; eg: in this photo, the intake shutter near the mouth of the intake is down.

 

 

The bypass was used was during ground running & takeoff or landing, or when flying in a dust or sand laden atmosphere

 

Spitfire VIII Pilot's Notes, also relevant to 1944-45 L.F. Mk. IXs (ignore the details for early aircraft):

 

 

As it is, unless any future Normandy airfields on DCS maps can reproduce the dusty conditions, whether or not the filter was used is moot.

[rel4y]

I am enjoying the discussion Yo-Yo. :)

 

If you increase blower ratio, the result will be in reduced power before FTH (due to overheated and then throttled air), at FTH (due to overheated air and increased power to blower) and increased power beyound the FTH.

 

You are absolutely correct. But to be fair I did say "It is also clear why this will have adverse effects at lower altitude." While not going into detail, this is exactly what I was referring to. Now we both know the Merlin 66/70 etc family also used an intercooler to adress the problem of compression heating.

 

If you take a look at the chart you will see that even with louvred plate the LF is faster than HF by 7 mph.

 

Yes and I am not sure how and why that is possible. You say it is measurement dispersion, that may certainly be one explanation. Looking at the data distribution of the individual test flights I clearly see a trend when looking at your graph (http://forums.eagle.ru/attachment.php?attachmentid=137589&d=1459344523). The red air filter test above 14k ft doesnt really fit into the distribution. I dont have the charts here in excel to calculate exact numbers, but just by eyeballing it can be seen.

The blue DCS line fits perfectly up until FTH to MA 648, to the spec sheet data, to BS 543 and all other tests. Above it develops an unusual bump at 36k ft and also surpasses every real life test above 34k. Now we have already elaborated why MA 648 can not really be considered a production line aircraft. Sadly I still know nothing about the air filter test, I cant find it anywhere. But something has to be different about this test!

 

And, by the way, I really do not know yet if the air cleaner was used for Normandy Spifires, or we must have it blanked. And if our cockpit has a control for the air cleaner shutter. I think that it does not have... so the air cleaner will be blanked. :)

 

The supercharger intake sat below the engine prone to dust and dirt at the ground. The mechanism was as you said pilot operated and there were three different switch types IIRC. It definitely had to be used on unpaved airfields, otherwise one would risk major engine or charger damage. Since there were different filters for tropical usage as Friedrich pointed out, the normal filter certainly was employed as standard in the field and not blotted off. The small gain in performance would not justify risking engine failure or unnecessary deterioration when it can easily be avoided.

Edited March 30, 2016 by rel4y

[Friedrich-4B]

Since there were different filters for tropical usage as Friedrich pointed out, the normal filter certainly was employed as standard in the field and not blotted off. The small gain in performance would not justify risking engine failure or unnecessary deterioration when it can easily be avoided.

 

A correction on my part: Morgan and Shacklady, page 282, shows there were three types of Vokes filter systems: Type A, without a louvred plate under the filter element; Type B w/louvred plate (mod. 1068 ) and Type C with an extra plate with square holes on top of the filter element, plus modifications to the seals. There was no change in the filter element itself. What did change was the louvered plate reduced the performance loss from having the filter element open directly into the intake, as shown in Yo-Yo's attachment.