1-48 Spitfire fuselage comparisons

[Edgar]

There is a fundemental flaw in using photographs, and this is something that our French-Canadian cannot/will not see, or acknowledge; the over-worked phrase "The camera never lies" is a lie, as anyone, who's indulged in the hobby of photography, will tell you.

Basically, lenses of a focal length less than the diagonal measurement of the negative make items near the camera seem disproportionately large; photograph a girl close-up, with even a mildly wide-angle lens, and she'll display a nose like Schnozzle Durante/Cyrano de Bergerac. Portraits should always be done with a mildly telephoto lens, to keep items in some sort of proportion; overdo it, though, and perspective gets flattened completely.

When Peter Cooke was verifying the existence of the late-war "bulge," introduced on the upper cowling of the IX/XVI, he photographed VIIIs & IXs from around 100 yards, with an 800mm telephoto, also taking great care to set up at exactly 90 degrees to his subject; this "flattened" the image enough to enable him to concentrate on the outline shape of the nose.

Edgar

[ben_m]

There is a fundemental flaw in using photographs, and this is something that our French-Canadian cannot/will not see, or acknowledge; the over-worked phrase "The camera never lies" is a lie, as anyone, who's indulged in the hobby of photography, will tell you.

Basically, lenses of a focal length less than the diagonal measurement of the negative make items near the camera seem disproportionately large; photograph a girl close-up, with even a mildly wide-angle lens, and she'll display a nose like Schnozzle Durante/Cyrano de Bergerac. Portraits should always be done with a mildly telephoto lens, to keep items in some sort of proportion; overdo it, though, and perspective gets flattened completely.

When Peter Cooke was verifying the existence of the late-war "bulge," introduced on the upper cowling of the IX/XVI, he photographed VIIIs & IXs from around 100 yards, with an 800mm telephoto, also taking great care to set up at exactly 90 degrees to his subject; this "flattened" the image enough to enable him to concentrate on the outline shape of the nose.

Edgar

This effect is not a lie, nor is it distortion, as many would have you believe.

This effect is not due to the focal length of the lens, but is purely due to the distance from subject to camera- take a photo with a wide-angle lens from far away, and as long as you have enough resolution, when the image is cropped, the image will be the same as if you had taken the image with a longer focal length lens.

In other words lenses cannot alter perspective, only position can.

Edited August 2, 2011 by ben_m

[Andrew Jones]

Adding to what has already been said about the Airfix/Aeroclub seafire nose comparison , I decided to get out my trusty profile gauge and measure the nose profiles, not being willing to rely on the old mark one eyeball to be entirely truthful . The results;;

1/ Measuring the Aeroclub nose profile,

2/ Comparing the Aeroclub profile to the Airfix nose,

To all intents and puurposes they appear identical.

Andrew

Edited August 3, 2011 by Andrew Jones

[TheModeller]

To all intents and puurposes they appear identical.

Which is exactly what a lot of us have been saying for some time, thank you for taking the time to demonstrate how similar they are.

Whatever inconsistencies that exist between the two are insignificant and to my eye barely worth worrying about.

Doubtless certain contributors to this thread will point out some glaring flaw in your logic or measurements and will simply refuse to accept the evidence of thier own eyes despite telling us that drawings are not to be trusted and we should rely on our eyes to prove how fatally-flawed the new kit is...

Of course thats not something I'm going to concern myself with for a moment.

[Jon Kunac-Tabinor]

Just to ask a question here that's always puzzled me:

How much difference could there be in the actual built airframe to the factory engineering plans? I mean I'm assuming that if you measured 50 real Spitfires you'd get slightly differing values from each just because, even allowing for jigs etc, anything that's essentially human assembled will always have some variance.

So I guess what I'm saying is that if you measure a real spit in order to make a plastic model tooling - how do you know that the Spit you are measuring matches the manufacturers plans?. And if there is a difference- which one is "right", the reality of the built machine, or the "ideal perfection" of the plans?

PS are we safe to say He Who Shall Not Be Measured's name - Voldemarty!!!! aargh

Jonners

[TheModeller]

Just to ask a question here that's always puzzled me:

How much difference could there be in the actual built airframe to the factory engineering plans? I mean I'm assuming that if you measured 50 real Spitfires you'd get slightly differing values from each just because, even allowing for jigs etc, anything that's essentially human assembled will always have some variance.

So I guess what I'm saying is that if you measure a real spit in order to make a plastic model tooling - how do you know that the Spit you are measuring matches the manufacturers plans?. And if there is a difference- which one is "right", the reality of the built machine, or the "ideal perfection" of the plans?

I think you answered your own question Jonners, I'd go with the reality of the actual airframe, the 'factory-drawings', which most often aren't something you or I would consider a 'scale plan' represent the ideal perfection you mention, the actual airframe is a tangible real-world object that can be examined, measured, photographed and analysed.

Many years ago I worked as a buyer for a major UK company that were developing new machinery for mass production of consumables for vending machines, the office I was in was more than three quarters filled with drawing-boards and men creating drawings of the required components which I would then take to local engineering shops to have fabricated, not one drawing was 'to-scale', few ever are, and even fewer were created to the same scale as their neighbours, it all depends on the size of the widget being designed, the measurements of the required widget were called out on the drawings.

I can't imagine the factory drawing offices at Supermarine working any differently, there are no such things as 'factory scaled drawings' at least not of a complete airframe. Of course there will be general arrangement drawings, some may well be very detailed, but ultimately they aren't 'scaled-drawings' as you or I would understand them.

PS are we safe to say He Who Shall Not Be Measured's name - Voldemarty!!!! aargh

I'm the wrong person to ask mate, being in my 50's now I've got too old to read childrens books!

Edited August 3, 2011 by TheModeller

[dragonlanceHR]

Thank you Andrew,

that was the missing piece!

Vedran

[Andrew Jones]

I think you answered your own question Jonners, I'd go with the reality of the actual airframe, the 'factory-drawings', which most often aren't something you or I would consider a 'scale plan' represent the ideal perfection you mention, the actual airframe is a tangible real-world object that can be examined, measured, photographed and analysed.

Many years ago I worked as a buyer for a major UK company that were developing new machinery for mass production of consumables for vending machines, the office I was in was more than three quarters filled with drawing-boards and men creating drawings of the required components which I would then take to local engineering shops to have fabricated, not one drawing was 'to-scale', few ever are, and even fewer were created to the same scale as their neighbours, it all depends on the size of the widget being designed, the measurements of the required widget were called out on the drawings.

I can't imagine the factory drawing offices at Supermarine working any differently, there are no such things as 'factory scaled drawings' at least not of a complete airframe. Of course there will be general arrangement drawings, some may well be very detailed, but ultimately they aren't 'scaled-drawings' as you or I would understand them.

You are absolutely correct, as I understand it an engineering or GA drawing is not to scale, but, it is dimensioned , and anything made from this drawing is made by reference to the dimensions and not the drawing itself.

Edited August 3, 2011 by Andrew Jones

[gingerbob]

How much difference could there be in the actual built airframe to the factory engineering plans? I mean I'm assuming that if you measured 50 real Spitfires you'd get slightly differing values from each just because, even allowing for jigs etc, anything that's essentially human assembled will always have some variance.

Less difference than you might think. People like to talk about British equipment being "hand made", but that's not the way it is explained in the various reading I've done. There are no doubt certain things like cowling panels and wing root fairing panels that require fettling when installing a replacement, but the basic chunks would be built in jigs (which had to be proved) and had to meet specs or get approval if they did not in some way. I don't really imagine some squadron or repair unit in North Africa, or India, is going to try a bunch of wing panels to find one that fits, for example. So, maybe one rudder is a sixteenth of an inch longer in chord than another, but over the length or span of the airplane I wouldn't expect a significant variation. Warbirds may be somewhat less reliable, depending on who is doing the work and what they've got to work with (and how many times it's been done to that example before!)

And as for the shop drawings, they aren't intended to be "scale" drawings, they are intended to show the worker how to make the part. As Andrew says, it is the dimensions that matter- the picture is just there to help you visualize.

bob

[TheModeller]

Bobs comments are spot on, the manufacture of many types, but Spitfires in particular, was a distributed operation, my Grandmother worked as a seamstress before the war, she and many of her friends were driven to Andover by works bus to make the fabric coverings for the elevators, tailplanes and rudders of early Spits which were then shipped to assembly workshops for fitting to the airframes.

If tolerances weren't maintained 'out-sourced' components would have been difficult to fit resulting in slowed production rates. So although some variance was possible it was within specified tolerances, I think the point to emphasise is that "hand made" is very different from "hand assembled".

Edited August 3, 2011 by TheModeller

[Naun123]

Less difference than you might think. People like to talk about British equipment being "hand made", but that's not the way it is explained in the various reading I've done. There are no doubt certain things like cowling panels and wing root fairing panels that require fettling when installing a replacement, but the basic chunks would be built in jigs (which had to be proved) and had to meet specs or get approval if they did not in some way. I don't really imagine some squadron or repair unit in North Africa, or India, is going to try a bunch of wing panels to find one that fits, for example. So, maybe one rudder is a sixteenth of an inch longer in chord than another, but over the length or span of the airplane I wouldn't expect a significant variation. Warbirds may be somewhat less reliable, depending on who is doing the work and what they've got to work with (and how many times it's been done to that example before!)

And as for the shop drawings, they aren't intended to be "scale" drawings, they are intended to show the worker how to make the part. As Andrew says, it is the dimensions that matter- the picture is just there to help you visualize.

bob

Depends on the dimensioning and tolerance scheme used. In some cases the tolerances add up, in others they don't. For example, if you're using a tolerance scheme where the tolerances add up, and each tolerance is +/- 1/16 of an inch, then if you've got 16 parts and all of them are high, the assembly will be 1" longer than it should be. If they're all short, then it'll be 1" shorter than it should be.

In some cases, you tolerance from a point on the assembly so that the overall assembly is no longer or shorter than the tolerance, regardless of how much each part of the assembly is off, like when you need a piece of furniture to fit through a doorway. The exact cushion width on a couch doesn't really matter, but the assembled couch needs to fit through the standard door.

http://en.wikipedia.org/wiki/Geometric_dim...and_tolerancing

[Naun123]

Bobs comments are spot on, the manufacture of many types, but Spitfires in particular, was a distributed operation, my Grandmother worked as a seamstress before the war, she and many of her friends were driven to Andover by works bus to make the fabric coverings for the elevators, tailplanes and rudders of early Spits which were then shipped to assembly workshops for fitting to the airframes.

If tolerances weren't maintained 'out-sourced' components would have been difficult to fit resulting in slowed production rates. So although some variance was possible it was within specified tolerances, I think the point to emphasise is that "hand made" is very different from "hand assembled".

Depends on where they are. The width of a spitfire aileron and it's hinge placement probably had to be fairly exact. Other measurements probably not as much.

[Naun123]

Beautiful Seafire!

[Andrew Jones]

Beautiful Seafire!

Thanks Naun, it's the Aeroclub conversion [obviously ] mated to an Airfix C-wing from a mark IX kit, decals by Eagle Strike.

Andrew

[gingerbob]

Depends on the dimensioning and tolerance scheme used. In some cases the tolerances add up, in others they don't.

It doesn't work that way with a Spitfire- the wing panel, or the fuselage from firewall (Station 5) to tail joint, was built in a jig that took the whole thing. Each rib or fuselage station had a location on the jig- no "stacking" could occur. One of the books by someone who worked for Sup went into great detail- when a new fuselage jig was built, a fuselage that had been made in an existing jig would be dropped into it. If it didn't fit, the jig had to be corrected.

Edit: Besides, the most basic components are the wing spar(s), built up to a "D-section" leading edge on a Spitfire, and the longerons/ stringers for the fuselage. You sort-of set the limits of the piece and then fill in all the structure.

bob

Edited August 5, 2011 by gingerbob

[Edgar]

Depends on the dimensioning and tolerance scheme used. In some cases the tolerances add up, in others they don't. For example, if you're using a tolerance scheme where the tolerances add up, and each tolerance is +/- 1/16 of an inch, then if you've got 16 parts and all of them are high, the assembly will be 1" longer than it should be. If they're all short, then it'll be 1" shorter than it should be.

But that's not the way that engineers work. A tolerance is just that, but not every piece will exhibit it, so, to start, the worker will use only parts with zero error, and put them together. When he runs out, he will look for parts with a + error, which are matched by others with a - error, and use them, so that they cancel each other out. Workers were not expected to use parts "Come what may," so, if he ended up with some that he couldn't match, he was entirely likely to put them on the scrap pile, telling the foreman what he'd done.

Depends on where they are. The width of a spitfire aileron and it's hinge placement probably had to be fairly exact. Other measurements probably not as much.

This is a section from a wing drawing; note the 7 holes, in the end of the wingspar, with tolerances of +.001"/-.0002" (that's one thousandth, and two-tenths of a thousandth of an inch.) A drawing of "frame 5" (the fireproof bulkhead) shows a stub double-sided spar, with two lots of 7 matching holes, into which the wingspar slid, and 7 bolts held it in place, with no "tolerance" or sloppiness permitted.

Remember, too, that, for Eastleigh production, wings and fuselages were built miles apart, and brought together for final assembly.

Edgar

P.S. The Academy kits were not based on the Airfix pair, predating them by quite a margin; they were produced from a G.A. (not scale) drawing, taken from a magazine.

Edited August 5, 2011 by Edgar

[Miggers]

Edgar said:

"But that's not the way that engineers work. A tolerance is just that, but not every piece will exhibit it, so, to start, the worker will use only parts with zero error, and put them together. When he runs out, he will look for parts with a + error, which are matched by others with a - error, and use them, so that they cancel each other out. Workers were not expected to use parts "Come what may," so, if he ended up with some that he couldn't match, he was entirely likely to put them on the scrap pile, telling the foreman what he'd done."

Absolutely correct.

That is exactly the way we assembled pumps and pump/motor sets when I worked in that side of the engineering industry.

The only difference was that if a part was + out of tolerance or flat,it could be sent back to the machine shop to be checked

to see if it could be re-opped to bring it back to tolerance.

If something was -out of tolerance,it could be shimmed out to make up the difference,but usually if something

was - out of tolerance,it would be scrapped.

Mark

[Naun123]

But that's not the way that engineers work.

I'm an engineer. We do exactly that. You don't limit tolerances when you don't need to because it adds to manufacturing costs. The couch for example, who cares if the seat cushions are 14 5/8" or 14 3/4"? You do however care whether or not it fits through a standard door frame. You want it to fit the people that are sitting on it, so you won't make it too narrow by feet, but the width is probably the only "do not exceed" tolerance.

Building in adjustability is often cheaper than building to a tight tolerance. The clutch pedal assembly on my car had oval mounting holes so that they didn't have to manufacture the stamped steel bracket to any tight tolerance. The rod that connects the pedal to the clutch's z-bar was threaded so that it could be extended or shortened as needed (several inches). That way neither the frame nor the car's body had to meet specific tolerances.

Aircraft design is the same. I worked on the A-10, and NONE of the pylons were exactly interchangeable between one airframe and another (or even one side and another) even though they were all built to tolerances. Getting all 4 pylon mount bolts and the lollypop bolt up front in their holes at the same time was often an exercise in frustration. Nothing fit "exactly". Especially the sheet metal panels that went around and sealed the pylons to the wings where dihedral created gaps. We were always having to oval out holes or move nut plates to allow things to fit. Other panels were the same (leading edges, engine nacelles, side fuselage panels, etc). You usually couldn't take them one from one jet and bolt it to another without sheet metal shop out there adjusting things for you.

I volunteered in a resto shop for a museum, and they had exactly the same problem with their C-119. Parts from one aircraft did not fit on the one they had without "adjustment". I've also worked on a C-118 (stripping the interior), an A-26, and an Albatross (sanding mostly for refinishing). And I've witnessed similar problems on each.

I've never worked on anything Supermarine, so I can't make any definitive claims about them or their process.

A tolerance is just that, but not every piece will exhibit it, so, to start, the worker will use only parts with zero error, and put them together. When he runs out, he will look for parts with a + error, which are matched by others with a - error, and use them, so that they cancel each other out. Workers were not expected to use parts "Come what may," so, if he ended up with some that he couldn't match, he was entirely likely to put them on the scrap pile, telling the foreman what he'd done.

There's no such thing as a zero error part. Everything is either plus or minus. It's always a question of how far. If the assembler has to take time to match parts, then you've got an extremely inefficient assembly process. It's the Industrial Engineer's job to design the process so that any assembler can take any part and bolt it to any other mating part and have the assembly meet specs. Otherwise, you might wind up with a bunch of - parts, no matching + parts, and either you scrap everything, or your assembly line grinds to a standstill while you wait for someone to create parts with matching tolerances. Statistical process control is used to constrain the variance and defect rate.

[Naun123]

It doesn't work that way with a Spitfire- the wing panel, or the fuselage from firewall (Station 5) to tail joint, was built in a jig that took the whole thing. Each rib or fuselage station had a location on the jig- no "stacking" could occur. One of the books by someone who worked for Sup went into great detail- when a new fuselage jig was built, a fuselage that had been made in an existing jig would be dropped into it. If it didn't fit, the jig had to be corrected.

Edit: Besides, the most basic components are the wing spar(s), built up to a "D-section" leading edge on a Spitfire, and the longerons/ stringers for the fuselage. You sort-of set the limits of the piece and then fill in all the structure.

bob

I have no doubt that there were quite a few parts built to tight tolerances.

What about the cowl, spinner, rudder, wingtips, etc? Any stacking possible there?

I watched a guy re-fabric a set of DC-3 ailerons he built recently (he was redoing everything, but I only saw him working on the ailerons). He had two sets of ailerons that he was using as a pattern for making new ones. Neither of the originals matched each other exactly. Ribs and other interior parts weren't in exactly the same spots because they were built by hand just like the ones this guy was making. But he was still building them to specs.

And even with a jig, some variance is inevitable because the jig won't be exactly the same size at every temperature. You've got to deal with the coefficient of thermal expansion (CTE). Airframes built in one season could be a slightly different size than airframes built in another because they were built with a jig and the size of the jig changed when it was hot in the summer and cold in the winter.

You can really see the CTE in railroads, where it's commonly called "sun kink". We're having a lot of trouble with that during our current heat wave.

http://en.wikipedia.org/wiki/Sun_kink

http://www.trainorders.com/discussion/read.php?1,161542

For example: http://www.abc2news.com/dpp/news/state/csx...ins-due-to-heat

Conversely, the extreme cold can cause enough linear shrinkage to sheer rails too. http://www.fuzzyworld3.com/3um/viewtopic.php?f=5&t=4000

I just think that all of this emphasis we, as modellers, are placing on "knowing" the "exact" anything about any of this stuff is a bunch of malarkey. Fact is, an A-10 sitting in the desert sun is a different size than the gate guard in Alaska. Even on the same flight line at the same temperature on the same day, we couldn't take things from one jet and bolt them directly to another jet made using the same jigs on the same assembly line without at least some fight.

As long as the shape is right and it looks like what it's supposed to be a model of, I'm OK with it. The Academy XIV doesn't look right to me. Neither does the Hobbycraft XV or the Hasegawa IX. The Airfix XII does.

[Andrew Jones]

As an example of the importance of the correct use of tolerances ,is the cause of the crashes of two Mk XIVE's of 17sqn at Madura, India in July 1945, both aircraft broke up in the air causing the death of both pilots, Flt Sgt Gardener and Flt sgt Whiting ,a third aircraft flown by Flt Sgt Holland managed to land with part of a wing missing.

The squadron was immediately grounded and an investigation started, it was discoved that the aircraft all had loose wings, the cause was traced to the spar bolts. These bolts should have been refrigeration fitted, however the aircraft had been delivered by sea in a knocked down form and had been assembled by TATA Airways in Karachi. The company's fitters had found the spar bolts the be "oversize" and not realising they had to be refridgerated to fit, had reduced their size with emery cloth, to enable them to fit.

The remedy for this error was for all the spar holes to be reamed out slightly oversize and new bolts to be specially manufactured and refridgeration fitted.

[Spitfire31]

…You've got to deal with the coefficient of thermal expansion (CTE). Airframes built in one season could be a slightly different size than airframes built in another because they were built with a jig and the size of the jig changed when it was hot in the summer and cold in the winter.

And that is the reason that the Academy Mk XIV is considered by some to be just a teeny bit bulbous. So there!

/Joachim

[gingerbob]

I just think that all of this emphasis we, as modellers, are placing on "knowing" the "exact" anything about any of this stuff is a bunch of malarkey. Fact is, an A-10 sitting in the desert sun is a different size than the gate guard in Alaska.

As long as the shape is right and it looks like what it's supposed to be a model of, I'm OK with it. The Academy XIV doesn't look right to me. Neither does the Hobbycraft XV or the Hasegawa IX. The Airfix XII does.

Well, maybe you aren't looking at the Academy, Hobbycraft, or Hasegawa kits on the right weather-days. Yes, I know about thermal expansion, but scale that down to 1/48 (or even 32) and I don't think it is going to explain the deviation from tolerance of a model. How can you evaluate the fidelity to scale of a kit (remember, they're called scale models) if you don't have some actual numbers to measure against? And why is it a problem if some of us like to have some baseline to start from? Maybe it is just (at least in part) an excuse to learn more about an airplane that we like. The ones who do real research are not generally the ones that start shouting as soon as the latest pictures of a test shot, or even CAD rendering, are posted.

Yes, the Spitfire's cowling panels were hand fitted, and you would be very unlikely to take one off one airplane and fasten it, sans adjustment, onto another, but the cowling panel blanks were built to a spec and would be effectively the same size. Possibly with some slight amount extra around the edges to "trim to fit", though probably not even much of that. It wasn't matched-hole tooling, I don't believe, but the undrilled skins are not going to be varying much in size. And the same is true of your DC-3 ailerons- sure the ribs might shift a little from one to another, but I bet the overall dimensions of the aileron weren't much different. Someone drilling a hole in a slightly different place is not going to change the scale/shape of a model.

bob

[dragonlanceHR]

And why is it a problem if some of us like to have some baseline to start from?

This is a mystery to me ever since I started following r.m.s and the various scale modeling forums. It's not like we're forcing someone to read this or build his models in any way.

Oh well,

Vedran

Edited August 6, 2011 by dragonlanceHR

[Edgar]

And even with a jig, some variance is inevitable because the jig won't be exactly the same size at every temperature. You've got to deal with the coefficient of thermal expansion (CTE). Airframes built in one season could be a slightly different size than airframes built in another because they were built with a jig and the size of the jig changed when it was hot in the summer and cold in the winter.

Which presupposes that the unthinking, uncaring management didn't heat their factory in winter, leaving the staff to freeze, then made them swelter in summer, by keeping all of the windows closed. Remember that the Spitfires were being built, day & night, in 12-hour shifts, not to the more leisurely tempo of today's factories.

The Westland historian told me how any Spitfire, or Seafire, built by themselves, or Supermarine, could have its removable parts thrown in a heap, then put up against any other similarly-built airframe, and they would still fit, perfectly. Castle Bromwich-built airframes didn't have the same finesse, but the parts, kept with it during repair, went back onto the original airframe without fuss.

Edgar

[Spitfires Forever]

Yes, another "Black Hole" we modelers get into. I have a closet full of Spitfires and have co,e to the conclusion that I have a choice, and that is to either ignore the inconsistencies or obsess over them at the risk of just giving up and selling the lot on e-bay so I can start collecting Hummel figures or something less frustrating. I have chosen a middle road in order to preserve my sanity and still enjoy building my models. I stay wash form the worst kits ( Fujimi and Occidental/Italieri then buy corrections like DACO make for kits like the Mk XIV, and then stick with the ICM for Mk's VII thru Mk IX and MK XVI. For the Mk I/II and Vb the Tamiya kit will do just fine. Each kit has it's pluses and minuses, but we must ask ourselves if we want to enjoy modeling or turn it info a chore that we dread like pulling weeds in the yard or painting the house. If it ain't fun then why do it? I would suggest finding a middle ground, and if the later mark kits have features that bother you then just don't display the Hasegawa Mk IX next to the ICM ( the difference in length is pretty noticeable in comparison) and go from there. I hope all the distraught Spit fans out there can find peace in my words. good luck and good modeling.

Cheers