Read the forum code of contact
By: - 30th January 2009 at 21:29 Permalink - Edited 1st January 1970 at 01:00
Question: Does anybody have any idea how close an actual production aircraft comes to being exactly like its dimensional specifications state is should be.For example would a fictional fighter that on paper is 15.29 metres long and 12.78 metres wide actully work out exactly that or could it give or take a bit?
That would depend on the aircraft and the part. And that data isn't likely going to be out in the open anywhere as it's proprietary anyway. A good manufacturing engineer could probably tell you what's possible but that's not necessarily going to tell you anything because using the highest possible tolerance when it's not necessary is going to drive the cost up for nothing. I did hear though that on the B-2 the span is within a millimeter or two of nominal. Of course that depends on what temperature you're doing the measurement at. ;)
By: - 30th January 2009 at 21:41 Permalink - Edited 1st January 1970 at 01:00
Question: Does anybody have any idea how close an actual production aircraft comes to being exactly like its dimensional specifications state is should be.For example would a fictional fighter that on paper is 15.29 metres long and 12.78 metres wide actully work out exactly that or could it give or take a bit?
I have no idea when you use meters but aircraft blue prints have a section called the tolerance block and the standard tolerances are as follows:
Dimensions written in hundredths of an inch, .XX the tolerance is +/- .030".
Dimensions written in thousandths of an inch, .XXX the tolerance is +/- .010".
By: - 30th January 2009 at 22:09 Permalink - Edited 1st January 1970 at 01:00
I have no idea when you use meters but aircraft blue prints have a section called the tolerance block and the standard tolerances are as follows:Dimensions written in hundredths of an inch, .XX the tolerance is +/- .030".
Dimensions written in thousandths of an inch, .XXX the tolerance is +/- .010".
Yeah but that's pretty much meaningless. Every part, hell, every feature of every part is evaluated and a tolerance determined. And then of course you do your tolerance stackups and maybe loosen some and tighten others. There is no one-size-fits-all even on a single part.
By: - 30th January 2009 at 22:38 Permalink - Edited 1st January 1970 at 01:00
Tolerances for aircraft dimensions can be surprisingly generous.
I used to do heavy maintenance on C-130's and before and after the input a full symmetry check would be carried out.
For long distances like wing tips, the tolerances were over an inch, but with less for smaller parts of the aircraft.
The tolerances were more for position, than say span, more like distance from a centre line.
Only if the tolerances were way out would an investigation begin, anything near the limits were entered in the aircraft log book, and were usually rigged out on airtest.
One aircraft for instance, had a twist in the fuselage from when it had plugs fitted and was stretched, this meant you couldn't get it to fly straight with the maximum allowed 4 degrees of adjustment on the rudder tab, so an entry was put in the logbook that it was allowed to have 5 degrees of rudder tab to fly straight.
By: - 30th January 2009 at 22:45 Permalink - Edited 1st January 1970 at 01:00
Tolerances for aircraft dimensions can be surprisingly generous.I used to do heavy maintenance on C-130's and before and after the input a full symmetry check would be carried out.
For long distances like wing tips, the tolerances were over an inch, but with less for smaller parts of the aircraft.
The tolerances were more for position, than say span, more like distance from a centre line.
Only if the tolerances were way out would an investigation begin, anything near the limits were entered in the aircraft log book, and were usually rigged out on airtest.
One aircraft for instance, had a twist in the fuselage from when it had plugs fitted and was stretched, this meant you couldn't get it to fly straight with the maximum allowed 4 degrees of adjustment on the rudder tab, so an entry was put in the logbook that it was allowed to have 5 degrees of rudder tab to fly straight.
I remember reading about a "bent" F-105 that didn't like to fly straight back during the Vietnam war. :)
By: - 31st January 2009 at 00:21 Permalink - Edited 1st January 1970 at 01:00
You can get a lot of thermal elongation/contraction based on diurnal heating/cooling that could account for several tenths of an inch over a long wingspan. For example, a 16 degree C swing in temperature will cause a C-130's wingspan to grow/shrink by about 20 mm.
That said, the surface tolerances for steps, gaps and waviness are measured in hundredths of an inch for VLO airplanes.
By: - 31st January 2009 at 09:47 Permalink - Edited 1st January 1970 at 01:00
BAE systems have a manufacturing floor that is self leveling, the engineers noticed that every month the floor recorded a ~1mm shift up and down after some investigation it was found to be due to the moons orbit over the factory.
Some of the tolerances for the Eurofighter are fantastically small (Microns), and the JSF is using the same alignment technology to get it right first time - every time.
Source http://www.telegraph.co.uk/finance/2791900/BAE-shoots-for-the-moon-to-deliver-the-Typhoon.html
Cheers
By: - 2nd February 2009 at 00:38 Permalink - Edited 1st January 1970 at 01:00
BAE systems have a manufacturing floor that is self leveling, the engineers noticed that every month the floor recorded a ~1mm shift up and down after some investigation it was found to be due to the moons orbit over the factory.Some of the tolerances for the Eurofighter are fantastically small (Microns), and the JSF is using the same alignment technology to get it right first time - every time.
Source http://www.telegraph.co.uk/finance/2791900/BAE-shoots-for-the-moon-to-deliver-the-Typhoon.html
Cheers
Not quite.
I cannot tell you the process, (Cannot means not allowed) but the right first time for JSF is not, nothing like, the right first time you will find in a car plant. The parts will be a perfect fit, but it is not the traditional right first time.
The floor movements at Samlesbury/ Warton fall within the natural movements of the planets crust. The Warton site also suffers from the tide twice a day. It's always been there, somebody found a way of measuring it to look smart. That's why all sub-assemblies should be taken from their Assembly jig and dropped into an ICY jig (Interchangeability) to see if it fits to a consistant datum. The modern version of an ICY jig is to digitise everything with a robotic stylus. It's not quicker, and gives you a lot of info to analyse thus adding yet more cost.
Aeroplane tolerances are wide to say the least. Most parts have EOP dimension of +-0.010", which to an engineer is wide enough to drive a bus through.
I came into aerospace from real engineering, where we machined stuff to tenths of a thousandth of an inch. When I came to Aerospace in 1979, I didn't think I'd need a micrometer! You can read a steel rule to 0.010" (Well you could when the rules were made in Sheffield from real material and not the imitation Stainless Steel that is around now). Some rules were actually marked at 0.010" and you could read those to 0.003". You couldn't really see more accurate than a third of a mark on a rule.
Obviously, holes for fitted bolts are very accurate, distance between specific holes are dead accurate. Mechanism's are usually a -+0.005". Critical "Tolerance stack" areas are also tightly controlled, but only at the points of contact. Eg a Wing stringer passes over "Mouse holes" in the Ribs but the rib flange contacts the stringer. The thickness of the Rib flange and the contact Pad of the stringer is controlled as a total of +0.00-0.008" IIRC, otherwise the skin (Covers) would not sit correctly on them all. They dont anyway!......liquid shim or ally shims are used. The body of the stringer between the ribs can be -+0.010". The wings are weighed as a pair, and if there is an imbalance, small counterweights are added to the lighter wing. Because you can't take bits off the heavier wing. They were once made of lead, but it's not PC anymore so tungsten is used because of it's density (Pack a lot of weight in a small place). Or even steel if there is enough room.
Also, the components at the extremes of an aircraft are only aerodynamic shapes, and do not carry such high tolerances. Nose cone. Usually fibreglass and made in a mould. Once the mould is signed off, all parts will be same. So long as the attachment diameter is correct, and the thickness all over is correct, and obviously no cracks then it will be a good part.
Tailcones / Tailcones with EPU's in them. Bent tin items. So long as they fit, an extra half inch long does not matter a jot. Again, as this is so far aft of the CoG, the parts are weighed and if necessary a weight is added to the front. Not usually necessary as I believe each A/c is trimmed for itself. Pilots correct me there please.
Wing tips and aileron/fin tips. SPF formed ally or rubber pressed halves welded together. Again, so long as they fit on, the tolerance to the actual tip is +- 0.030".
A small airliner (A320 ish) length will vary by approximately +-0.5", that's a 1 inch range.
Somebody mentioned air temperature of the assembly, well, it is law for factories and workplaces to be minimum 20deg C (There is a max but I canny remember it) but if temperature is a factor, such as plastics then a temperature controlled environment is vital.
And, don't forget that the parts grow and shrink back again with flight (Pressure) and with temperature. Concorde was 14 inches longer at mach 2 than it was on the ground. You could actually see gaps between the cupboards of the galley when in flight (All calculated for of course).
By: - 2nd February 2009 at 01:01 Permalink - Edited 1st January 1970 at 01:00
Not quite.
I cannot tell you the process, (Cannot means not allowed) but the right first time for JSF is not, nothing like, the right first time you will find in a car plant. The parts will be a perfect fit, but it is not the traditional right first time.
The floor movements at Samlesbury/ Warton fall within the natural movements of the planets crust. The Warton site also suffers from the tide twice a day. It's always been there, somebody found a way of measuring it to look smart. That's why all sub-assemblies should be taken from their Assembly jig and dropped into an ICY jig (Interchangeability) to see if it fits to a consistant datum. The modern version of an ICY jig is to digitise everything with a robotic stylus. It's not quicker, and gives you a lot of info to analyse thus adding yet more cost.
Aeroplane tolerances are wide to say the least. Most parts have EOP dimension of +-0.010", which to an engineer is wide enough to drive a bus through.
I came into aerospace from real engineering, where we machined stuff to tenths of a thousandth of an inch.
What were you doing before aerospace, making guage blocks? :eek:
By: - 2nd February 2009 at 11:01 Permalink - Edited 1st January 1970 at 01:00
Lots of interesting stuff on this subject and far more complex then I first imagined too. :) :eek:
By: - 2nd February 2009 at 11:03 Permalink - Edited 1st January 1970 at 01:00
What were you doing before aerospace, making guage blocks? :eek:
Yes this is quite intriguing, nuclear industry or something equally hazardous where tolerances have to be exact? (though thats what i thought for aerospace more or less)
By: - 2nd February 2009 at 13:48 Permalink - Edited 1st January 1970 at 01:00
Yes this is quite intriguing, nuclear industry or something equally hazardous where tolerances have to be exact? (though thats what i thought for aerospace more or less)
Or it could be engine components. The spinning parts are generally pretty high in certain tolerances. Especially if they're spinning REALLY fast like a small turbine or a high performance piston engine.
Posts: 385
By: Barry Scott - 30th January 2009 at 21:17
Question: Does anybody have any idea how close an actual production aircraft comes to being exactly like its dimensional specifications state is should be.
For example would a fictional fighter that on paper is 15.29 metres long and 12.78 metres wide actully work out exactly that or could it give or take a bit?