| BlueSky Business Aviation News | |||||||||||||||||||||||||
| . | |||||||||||||||||||||||||
As you are all probably aware, last Friday (April 1st, so inappropriate) Southwest Airlines Flight 812 lost pressurization while in climb when a five foot long crack at a lap joint opened up. This, of course, comes on top of a crack in the crown skin of a 757 (due to a manufacturing error by the supplier) and another crack in the crown skin near the tail on another Southwest 737, and the media can't really explain the differences. Let me see what I can do.
As-built not quite the same as-designed Of course, the airplane as built differs slightly from the airplane as designed. The little details that get screwed up can have a significant impact on fatigue life. Oversized holes, double-drilled holes, holes too close to the edge of the part, crooked holes, over-driven and work hardened rivets, well, the list seems endless, really. The manufacturers work very hard to minimize those issues, and they do quite a good job. Fatigue is different in different parts Some parts of the airplane are worse for fatigue than others. Usually the skin in the crown (top skins) forward of the wing are the ones most affected. Crown skins see the highest loads, generally, and forward of the wing sees those loads reverse every cycle. You may already know this, but fatigue that simply takes a load further in that same direction isn't as bad as one that reverses direction. So, a tension load that goes up and down, but still remains tension is less bad than a load that reverses and goes from tension to compression and back to tension, even when the delta load is exactly the same. Going from 100lb tension to 500lb tension is less problematic than going from 200lb tension to 200lb compression. The skins in the crown aft of the wing see mostly tension. The crown skins forward of the wing see tension in flight and compression on the ground. The lap joints where we rivet one skin to another sees bending (along with other loads), strangely enough, and that also causes fatigue problems. Fatigue cracking near a lap joint is the problem the Southwest plane had. Scribe lines A key thing to keep in mind about lap joints is scribe lines. This became a real issue five years ago, or thereabouts. A scribe line is a mark on the skin made by a sharp metal tool sometimes used to remove the line of sealant on the low side of the lap. A lap, by the way, is where one skin panel overlaps the next one so we can attach one to the other. We run three lines of rivets down a lap to hold it together. Think of it as a seam. We even call them lap seams, sometimes. The upper skin is always lapped over the lower skin to minimize problems with moisture. Further, we run a bead of sealant down the lower side of the lap to help keep the water out. This is called a fillet seal. To repaint the airplane, that fillet seal has to come off. In the early days, people would use steel putty knives or even razor knives to scrape or cut away that sealant, sometimes leaving a small, but very sharp scratch in the skin. Sadly, that scratch is at the very worst place possible on this joint. I know, a scratch doesn't sound like a big deal, but a very sharp scratch like a scribe line acts as a stress multiplier. It's like a really fresh banana, one that you can't break open with your fingers, but if you put a small cut in the skin, it cracks open easily. Same principle. Hoop, longitudinal,and bending stresses at the lap A lap joint takes hoop stresses and longitudinal stresses from holding in the cabin pressurization. An aircraft cabin is a pressure vessel, really. All pressure vessels see at least hoop loads, most see both hoop and longitudinal. Hoop loads are tension loads running around the circumference of the cabin like a circle (hence the name, "hoop"). Longitudinal loads are orthogonal or at right angles to the hoop loads; they run fore and aft (hence the name, righ?). The way the math works out, longitudinal loads are exactly half of the hoop loads. The hoop loads want to run through the exact center of the joint and this forces the lap to straighten out a bit, causing the skin on either side - top and bottom - to bend a bit; bending adds tension loads on one side and compression loads on the other. So, the outside of skin on the lower side of the lap joint has additional tension loads from bending stresses in it that fall away a few inches farther down. Unfortunately, where the scribe lines are found is generally that part where the bending stresses are highest - think green banana skin.There is an AD (Airworthiness Directive) out on scribe lines and we have added the proper procedures to the normal maintenance programs. However, there was a lot of confusion early on about how to handle them. Characteristic cracks I should also mention that every airplane cracks in characteristic places. A 737 will eventually crack in the window belt covering the riser duct at Body Station 512. Every single one will. It's just a matter of time. And they have to be repaired. A 727 will crack the frame web at the stringer pass-through in a group of frames aft of the wing, close to the floor (usually stringer 15). All of them. It's just a matter of time - cycles, really. Airbus aircraft have their characteristic cracks, as do Douglas aircraft and all other airplanes, for that matter. The key is to know the characteristic cracks and inspect for them every chance you get to stay ahead of them. The aircraft that Southwest operates have fatigue cracking in all the usual spots, and we know the usual spots; there are very few surprises anymore, really, and those are almost all airframe-specific (e.g. a truck ran into tail number 724). The crack on flight 812 is a new one; we have never had a fatigue crack in that part of the airplane before--just aft of the rear spar. Given our extensive history with the 737, this seems very different, and we do not yet know why it happened, and there could be a number of explanations, but right now I would look very closely at scribe lines. All airplanes have cracks One of the conventional wisdoms we learn early on is that every airplane in service has a crack somewhere (I know, not what you wanted to hear, is it?). I can't swear to the truth of it, since I have never seen a study to that effect, but some airplanes do come in for maintenance without any cracks being noted for us in Engineering to repair. Of course, some cracks are quite minor and are repaired per the procedures in the Structural Repair Manual, which does not require notifying Engineering. But, because of the way we design airplanes, those cracks aren't much of a threat, usually. We do, in fact, design with cracked structure in mind, so, it's not as bad as it might sound. However, having residual capability is not the same as having full capability. Cracked structure is weaker than structure that is not cracked, obviously. It can get complicated. Design for fatigue cracks The fatigue cracks around the lap joints are a known issue, have been for decades, literally. There are no good solutions other than trimming out the cracked structure when you find it and splicing in new structure, which takes time and money, but is nothing like rocket science. We do this sort of thing every day. It isn't magic. The thing that keeps us relatively safe with all this is that we design the skins to come apart gracefully with tear straps every twenty inches to hopefully keep a crack in one structural bay, though it doesn't always happen that way. Boeing pioneered this technology n the 707 in part as a reaction to the Comet (long story). Generally speaking, a crack opening up in one location, causing a decompression event is not that big of a deal. Generally speaking. We have designed the airplane to handle that sort of event. Of course, if MANY small cracks come apart in a cascading failure (like the Aloha Air 737), then we have serious, life-threatening problems, or if we have structure cracking through in turbulence or turbulence in bad weather with icing, etc. However, the stringers are there to pick up the load in case the skin cracks. The stringers by themselves cannot take the full certification load, obviously, but they do provide the necessary redundant load path to keep the airplane in one piece long enough for the crew to get the airplane safely down on the ground. Doing it right Given what we know right now, there are things that need to be done, and from all reports, Southwest has been doing them. I.e., they grounded all aircaft of the same type and inspected all 79, finding five with cracks. This is exactly how it is supposed to work. Having found the damaged aircraft, they can rapidly repair them and return them to service without a problem. In the meantime, technical experts will be doing microscopic examinations of the damaged part that was cut out of the airplane (not as big a deal as it sounds since we would have had to cut it out anyway to do the repair), leading to a better understanding of exactly what happened, which should lead to better maintenance procedures. If this all seems very bad for Southwest, please keep in mind that they operate some 169 Classic 737s, and their nearest competition is Lufthansa with 33. Statistically speaking, if we are going to find something new on the 737 fleet, we will find it on a Southwest airplane. They operate more 737s than anyone and they operate them on shorter routes than anyone else, making Southwest Airlines the King of Cycles in the 737 world. My personal experience with Southwest as an engineer is very positive. They are as fine a group of people as I have ever worked with, and I would not hesitate to fly them. They know their business and that gives me confidence. Terry Drinkard is a Contract Structural Engineer based in Jacksonville, Florida whose interests and desire are being involved in cool developments around airplanes and in the aviation industry. He has held senior positions with Boeing and Gulfstream Aerospace and has years of experience at MROs designing structural repairs. Terry’s areas of specialty are aircraft design, development, manufacturing, maintenance, and modification; lean manufacturing; Six-sigma; worker-directed teams; project management; organization development and start-ups. Terry welcomes your comments, questions or feedback. You may contact him via editor@blueskynews.aero Other recent articles by Terry Drinkard:
|
|||||||||||||||||||||||||
