Boeing's latest delay for the 787 Dreamliner was blamed on a 57-day strike by machinists, but the program's schedule already faces additional pressure by new disclosures about improperly installed fasteners.According to sources across the program, the number of fasteners needing replacement range from about 2,500 to 5,000 per aircraft or shipset. Boeing publicly estimates that less than 3% of fasteners installed to date will have to be removed and reinstalled.
Boeing underscores that, "the issue is with installation of the fasteners, not the fasteners themselves."
Significant engineering and machinist resources across the program are being devoted to solving this problem inside Boeing and structural partner facilities as quickly as possible.
The challenge to the programme schedule centres on getting fasteners removed and reinstalled, as well as the potential damage to the composite material that could occur.
"The risk involved is that some of the fastener holes will need to be oversized. This is a common practice on in-production repairs," said one veteran engineer.
"Fastened structure is designed to allow for future reworkability, primarily for in-service repairs."
Boeing faced time consuming repairs on Dreamliner One following the July 2007 rollout when temporary fasteners caused damage after being removed to make way for permanent ones.
Boeing is re-training all 787 machinists in its Everett facility on new fastener installation procedures. Compounding the problem, sources say, is the slow pace of workers returning following the conclusion of the IAM strike. Machinists have until November 10 to return to work, according to the strike resolution. Only machinists who have completed the re-training are permitted to work on the aircraft again.
Those familiar with the fastener situation tell FlightBlogger that the problem originated in two separate types of fastener installation on the four flight test and two ground test aircraft, as well as the more than a dozen shipsets currently at supplier partners.
The first problem stems from the holes drilled to affix titanium and carbon fibre together. When holes are drilled into titanium, a burr is often left on the edge of the entry side of the hole. Because of the extraordinary strength of titanium, when a fastener is installed in the hole, the head will sit on the burr rather than flush against the surface.
With the head of the fastener resting on the burr, the loads will be distributed on that one spot rather than evenly across the surface. In addition, in the event of high side-to-side shear loads, in a worst-case scenario, the high-strength titanium burr could cut the fastener undermining structural integrity.
Titanium is used in key structural areas of the aircraft such as the joined sections in the fuselage and horizontal stabilizer.
Sources say the fastener problem was first discovered on the engine pylons on the static test airframe. The pylons have been removed from all aircraft in Everett and returned to Spirit AeroSystems in Wichita, Kansas for repair.
All major structure partners, with the exception of the wings supplied by Mitsubishi, are impacted by this problem, including Vought, Global Aeronautica and Alenia.
Fastener Anatomy
The re-training of Boeing staff covers preparation of the holes after they are drilled prior to fastener installation. Typically, after a hole is drilled the edges of the hole are slightly sloped to remove any possible debris or irregularities and creating a symmetrical sloping surface at edge of the hole.
There are generally four methods of preparation, all essentially identical with increasing size. In size order from smallest to largest a hole can have a deburr, fillet relief, chamfer or countersink to accommodate a fastener.
According to several sources in Everett, WA and Charleston, SC, in many cases, holes were prepared using one of these four methods, but the specification for installation was often unclear and size of the required slope was insufficient.
The second fastener problem finds its roots in the fastener shortage that challenged Boeing early in the program. Fasteners of varying lengths were installed improperly creating the possibility of structural instability.
The fastener tolerances are known as 'protrusion maximum' and 'base minimum.' Protrusion maximum addresses the amount of allowable thread on the bolt to emerge out of the nut to avoid any threading, or bottoming out, of the shank or non-threaded area of a fastener inside the hole.
In some situations where the fastener is too large, too much force can damage the fastener during installation by threading the shank rather than sitting flush against the surface undermining the entire fastener system.
Using longer fasteners is perfectly acceptable if washers are used.
The drawback for using larger fasteners, another veteran engineer says, is inevitable weight gain that comes with using larger parts than originally intended.
The base minimum keeps from having any threads in the hole of the fastened material.
If the shank of a fastener is too short then there is a chance the threaded part of the bolt is inside the hole and not enough thread is running through the nut. In this case, the result would be a compression of the two surfaces rather than using the whole fastener for stability.
Ideally a nut will sit snugly at the end of the bolt's thread allowing the entire shank to be entirely inside the fastened material creating a structurally sound connection.
The remedy: remove the incorrectly sized fastener and replace it with the proper fastener.
Though the timeline to solve the problem is unclear, Boeing will likely be hard pressed to release guidance for first flight and delivery until this situation is well in hand and production staff are back at work.
The re-training of Boeing staff covers preparation of the holes after they are drilled prior to fastener installation. Typically, after a hole is drilled the edges of the hole are slightly sloped to remove any possible debris or irregularities and creating a symmetrical sloping surface at edge of the hole.
There are generally four methods of preparation, all essentially identical with increasing size. In size order from smallest to largest a hole can have a deburr, fillet relief, chamfer or countersink to accommodate a fastener.
According to several sources in Everett, WA and Charleston, SC, in many cases, holes were prepared using one of these four methods, but the specification for installation was often unclear and size of the required slope was insufficient.The second fastener problem finds its roots in the fastener shortage that challenged Boeing early in the program. Fasteners of varying lengths were installed improperly creating the possibility of structural instability.
The fastener tolerances are known as 'protrusion maximum' and 'base minimum.' Protrusion maximum addresses the amount of allowable thread on the bolt to emerge out of the nut to avoid any threading, or bottoming out, of the shank or non-threaded area of a fastener inside the hole.
In some situations where the fastener is too large, too much force can damage the fastener during installation by threading the shank rather than sitting flush against the surface undermining the entire fastener system.
Using longer fasteners is perfectly acceptable if washers are used.
The drawback for using larger fasteners, another veteran engineer says, is inevitable weight gain that comes with using larger parts than originally intended.
The base minimum keeps from having any threads in the hole of the fastened material.
If the shank of a fastener is too short then there is a chance the threaded part of the bolt is inside the hole and not enough thread is running through the nut. In this case, the result would be a compression of the two surfaces rather than using the whole fastener for stability.
Ideally a nut will sit snugly at the end of the bolt's thread allowing the entire shank to be entirely inside the fastened material creating a structurally sound connection.
The remedy: remove the incorrectly sized fastener and replace it with the proper fastener.
Though the timeline to solve the problem is unclear, Boeing will likely be hard pressed to release guidance for first flight and delivery until this situation is well in hand and production staff are back at work.
on November 7, 2008 8:26 PM | Reply
Jon,
An excellent description thank you.
Its a very basic engineering oversight and really no excuse.
How many months I wonder to go back and chamfer every titanium hole in each airframe?
Its fair to assume that a new chamfering operation will need to be introduced into the assembly procedure, but frankly one of the first lessons I learned as an apprentice was to remove burr.
on November 8, 2008 8:50 PM | Reply
As Andrew says, basic, and one of the first apprentice lessons. Perhaps a factor was use of unskilled/untrained labor? Though not stated directly, it appears Boeing must remove/replace ALL fasteners (at least of certain types or in certain installations) to confirm/achieve absence of burrs (though it's interesting that Mitsubishi wing parts apparently are OK). The problem appears connected with the global partnership plan: suppliers might have left Boeing assemblers to assure holes were deburred while assembly planners might have expected holes to have been deburred before final inspection at source and did not write deburring into operations instructions. Reference to large fasteners is unfortunate, since Jon's notes imply only incorrect shank/thread lengths rather than diameters. Finally, it's surely not part of the explanation that washers which may be used during assembly in conjunction with longer-shanked fasteners can be called 'burrs'...?
on November 9, 2008 8:25 AM | Reply
I've heard that in aerospace assemblies, a hole can cost more than the fastener (drill, ream, deburr, inspect). One would think that there would be a single "owner" for the hole. If your team begins the process of making a hole, then you finish it and do all the steps to make it "per spec". Indeed, deburr is so fundamental, it's no wonder that it was assumed as being done correctly. I wonder if this is another "gotcha" with outsourcing. Seems hard to believe that Boeing employees would miss this.
on November 10, 2008 3:58 PM | Reply
Sounds like the crews at some of the plants were doing it right due to "common sense", and the contractors, lacking experience were not. Still saving money on that cheap labor?
... in many cases, holes were prepared using one of these four methods, but the specification for installation was often unclear and size of the required slope was insufficient.
on November 11, 2008 10:05 AM | Reply
Very interesting and informative piece.
But how can Boeing say '3%' with any certainty?
Don't they all have to come out, just to make sure?
I'd like to see a photo of a fastener, to complete the description.
on November 11, 2008 4:32 PM | Reply
I really don't know why Boeing are still bothering with all these old technology nuts and bolts.
A better solution would be to use Intevia Intelligent Fasteners (Intevia.com) which use SMA technology and can communicate with aircraft safety systems and on the ground. Plus they are lighter than conventional fasteners/solenoids so pay for themselves in fuel savings.
As a patriotic yankee, I'm a big fan of Boeing design & engineering, but I'm worried that they will be left behind by Airbus. I've heard that Airbus are using Intevia Fasteners on their A350.
on November 12, 2008 6:07 PM | Reply
Jon,
What about lightning strike implications of badly fitted bolts?
For example, the below linked article says how "Boeing will install each fastener precisely and seal it on the inside to ensure a snug, spark-free fit"
http://seattletimes.nwsource.com/html/businesstechnology/2002844619_boeing05.html
on November 13, 2008 3:07 AM | Reply
Deburr:
could it be that most of the holes are done
on CNC machines that get their programming
via CAD-> CAM integrated exports?
If the designer did not think of drawing
a chamfer no deburring will be done.
Except you have an experienced engineer
enhancing the CNC code a bit at the
contractor site.
Overall Boeing seems to have abysmal
QA. How do they manage to not have
their convetional aircraft fall out
of the sky ;-)
OK, probably all done by oldschool
engineers. But those are dying out fast.
Intevia Fasteners:
Having a network accessible releasable
bolt would just be the right thing to
have.
Visualise the plane coming "unscrewed"
at just the right moment.
uwe
on November 16, 2008 12:47 AM | Reply
Prepping a fastener hole, especially for a shear joint, is much more involved than just drill, deburr and install.
In metal substrates the hole prep involves drill, deburr, ream, inspect, coin, etc. Then the selection of the fastener with a proper grip length is also critical. There should be no threads loaded in shear, but the grip length should also not cause the threads to bottom out in the mating nut/collar/nutplate before the parts are clamped up. It is usually accepted practice to stack up 2 or 3 flat washers to correct this. Or you can use counterbored nutplates for blind installations.
Lastly, there is usually a requirement to have at least two threads visually protruding beyond the locking element of the nut/collar/nutplate.
Safety and reliability in aircraft manufacture is all about careful control of process and procedure. Boeing should know better than to let this happen.
on November 18, 2008 8:31 AM | Reply
As a non-engineer i am astonished that a company like Boeing could get into such problems with what would seem a very basic - though critical - task. As third parties are designing and building new technology stuff quality control is critical.
on December 6, 2008 3:16 PM | Reply
I have worked in aerospace engineering for 45 years. In my first six months I was taught how to prepare a hole - it's the very basics of learning to work on aircraft, one of the fundamentals of the safe manufacture and maintenance of all aircraft, C150 to B747 to A380 and through the military.
Now, if subcontractors are not yet smart enough to realise this, then they are cetainly not smart enought to manufacture sub-assemblies.
It is my view that Boeing have very serious problems - not only only on the B787 (which are severe), but issues that extend over all their build programmes. I would look at a massive re-training exercise, and look closely at management and their decisions.