Variation in build length

[18-Wheeler]

Have been chatting on another forum and a guy that's sort-of involved in aviation says he has a friend who worked for Boeing and on one type of airliner, due to the large tolerances in the parts, made for up to 75 millimetres difference in length of the fuselage.

Sounds excessive to me, but I'd like to know how much variation there is in a large airliner fuselage length from new? I can't imagine it'd be more than 5mm or so .... ?

[rudderrudderrat]

Hi 18-Wheeler,

The thermal coefficient of expansion for Aluminium is about 23 * 10 to the minus 6 / degree C.

My spread sheet says that a 20 deg change in temp on a 747 fuselage (70 metres) amounts to 32 mm. So I can quite easily believe your figure of 75 mm.

[18-Wheeler]

Thanks, but I'm not after thermal variation, only the difference due to building one.

[rudderrudderrat]

Hi 18-Wheeler,

I was just trying to say that 75 mm in a hull length of 70 metres is to an accuracy of about 1 thousandth. Normal thermal expansion in 20 degs C change, (just sitting on the tarmac), is about half the build tolerance.

Sounds about right to me.

[onetrack]

I don't believe that statement. Engineers can build bridges in two sections that are pinned in the middle, and after extending the lengthy structures to meet in the middle, they can then pin or bolt the sections together without any mismatch or alignment problems.
Aircraft are built to exacting standards and alignments, and I suspect the story is more BS than truth. 7mm I could believe, but not 75mm.

[galaxy flyer]

OTOH, the book that told the story of the B777, said that, having been designed on CATIA, the first fuselage was remarkably accurate to the "drawings". There was mention that the typical B747 carried about a 1,000 pounds of shims to make everything fit. 741 or 744 wasn't mentioned.

GF

[muduckace]

Have been chatting on another forum and a guy that's sort-of involved in aviation says he has a friend who worked for Boeing

My sisters boyfriend's best friend told me that... Aircraft are not bridges. 75 mm does not in any way surprise me as 3 inches overthe 231 feet and 10 inches the 744 shoots for.

Thermal expansion is a factor as the production length is at 70 deg f, the length that the person who told he person who told you could vary that much depending on the temperature it was above or below 70 deg when said aircraft was measured.

Not for nothing but the guys who are putting these aircraft together are not the finest of scholars, most of them have one little section of the aircraft that they hammer the same rivits into day in day out.

From personal experience with thermal expansion, I have done my fair share of rigging and building aircraft cables. Fact is they are often a bear to deal with on long runs as the aluminum contracts/expands at a different rate than the steel cables do. There are calculations for tension based on temperature and the job is allmost impossible to get done right unless the aircraft is at a controlled temperature for a decent length of time.

3 inches could easily be made up for over several turnbuckle connections, it happens on ocassion that cables have to be custom made.

Bottom line yes an aircraft could be 75 inches off spec and no the manufacturer and operator probably do not give a damb.

[Juliet Sierra Papa]

Bottom line yes an aircraft could be 75 inches off spec and no the manufacturer and operator probably do not give a damb.

Classic Met/Imp blunder which can be so disasterous.

[muduckace]

Oops, good thing I am not Boeing, might have to modify gates and hangars worldwide.

[Ivan aromer]

When aircraft were "bespoke" there was a large variation in the length/span.
Not usually a problem. That is until some bright spark decides to modify all the Nimrods (basically a Comet 4) up to a new standard; first aircraft in the hanger, all the jigs made and set up, second aircraft in NOTHING fits anywhere. Probably goes a long way to explain the huge cost over run and very very long delay. Or is that me being cynical again

[onetrack]

The design of the 747 fuselage is semi-monocoque. The fuselage is built in sections, which are then placed in a jig so they can then all be joined. The jig is the one of the most expensive parts of the whole process. A jig is designed to ensure that all finished products that come from that jig are identical - within set limits of tolerances. How you could produce two fuselages, from a fuselage jig, with a 75mm variation in length, is something I'd like explained to me in detail. A tolerance limit of 75mm appears to me, to be impossible. I understand that the assembly factory is temperature controlled, so wide temperature variations in the assembly process can be discounted. IMO, the original poster of the 75mm variation is getting his wires crossed, and the 75mm figure is the likely maximum fuselage length variation between major extremes of temperature and extremes of imposed flying forces.

[18-Wheeler]

The design of the 747 fuselage is semi-monocoque. The fuselage is built in sections, which are then placed in a jig so they can then all be joined. The jig is the one of the most expensive parts of the whole process. A jig is designed to ensure that all finished products that come from that jig are identical - within set limits of tolerances. How you could produce two fuselages, from a fuselage jig, with a 75mm variation in length, is something I'd like explained to me in detail. A tolerance limit of 75mm appears to me, to be impossible. I understand that the assembly factory is temperature controlled, so wide temperature variations in the assembly process can be discounted. IMO, the original poster of the 75mm variation is getting his wires crossed, and the 75mm figure is the likely maximum fuselage length variation between major extremes of temperature and extremes of imposed flying forces.

That's what I thought, OT.
I would just like some confirmation of it - or not, as the case may be.

[Brian Abraham]

Find the proposition hard to believe. Remember when the first 777 was put together Boeing were talking that with the new jigging methods in place, that one wing tip measured up exactly, where as the other was two thousandths of an inch out. (location of wingtip as measured and in relation to the complete aircraft)

Brit aircraft were something else again. Small volume, and largely hand built. There is a thread on the Belfast and its ability to handle out size loads, depended on which airframe, as all the doors differed markedly in size.

[rmm]

Having worked quite a bit with Boeing production drawings to manufacture parts, they all have tolerances. E.g you might make a stringer that the drawing says cut to 18.5 ft +/- .200". So I think it's probably possibly if all parts ended up on the high side.

In relation to the jigs ask yourself the question as to why Boeing never supply pre-drilled structural parts. At best you get whats called pilot drilled holes (these holes might be drilled to .062 but final size on installation is 0.375) to allow for misalignment. When you buy a cargo door skin from Boeing it's bigger than the hole it goes into. The maintenance manual says trim it till it fits.

I personally have removed a pax window forging from one 733 to fit to another 733 at the exact same location and the holes didn't even come close to lining up. Even after opening up the holes to 2nd oversize (that's a .030" increase on the diameter) the holes were still half a hole out.

Rmm

[Dan Winterland]

I used to fly the Victor. Every one was hand bulilt and our engineers used to know that there was little point in robbing things like panels or control surfaces for another aircraft as they were just not going to fit.

I suspect Airbus is pretty hot on its' tolerances. The components are built in several different factories. There can't be much lattitude.

[dixi188]

I used to work for British Aircraft Corporation building BAC 1-11s. The standard tolerance on parts was + or - 0.010" (10 thou.) In theory if you added all the tolerances of fuselage skins together you might end up with a variation in length of about 0.2", (fuselage about 100ft long) but since the sections were joined in a jig I doubt there was much difference in overall length.
I don't think Boeing do things much differently, so to get an error of 3 ins seems unlikely.
I guess the B787 might be built a little differently.

Holes are often pilot drilled in skins and left blank in the ribs and stringers to enable adjustments to be made to the fit before final drilling and riveting.
Some clever person at Edgely tried to final drill all parts for the Optica before assembly. CNC accuracy? I believe the fitters had to do a lot of fiddling to get the thing assembled. Then they went bust.