The 1.5-meter-long gash that opened up in the upper cabin of Friday's Southwest Airlines Flight 812 from Phoenix to Sacramento will have a deep impact on the nature and frequency of commercial aircraft maintenance. The Federal Aviation Administration (FAA) issued a directive on Tuesday ordering about 175 Boeing 737 aircraft—80 of which are registered in the U.S., most of those operated by Southwest—to be inspected using an electromagnetic device that can identify metal fatigue.

The FAA is targeting Boeing 737 series 300, 400 and 500 aircraft that have accumulated more than 30,000 flight cycles (takeoffs and landings) in order to prevent a repeat of the April 1 incident. The fuselage of a 15-year-old Southwest Boeing 737–300 ruptured 18 minutes into the flight at an altitude of about 10,670 meters, forcing the pilots to make an emergency landing at Arizona's Marine Corps Air Station Yuma.

The National Transportation Safety Board (NTSB) says its investigators have found cracks in portions of the lap joint running on two lines of riveted joints covering the length of the fuselage of the aircraft involved in the incident. Subsequent Southwest inspections turned up cracks in the lap joints on five other aircraft, grounding them as well. The electromagnetic eddy-current test being performed uses a probe to send high- and low-frequency signals down into the skin of the aircraft. The probe is moved from one rivet to the next. Any crack in the metal changes the current's signal and tips off inspectors to a potential problem.

The riveted joints that failed on Flight 812 were not extensively checked because they were thought not to be susceptible to fatigue, according to the NTSB. "What we saw with Flight 812 was a new and unknown issue," Mike Van de Ven, Southwest's executive vice president and chief operating officer, said in a press release.

Southwest, the largest U.S. domestic carrier with more than 3,400 flights daily, follows a business model that relies exclusively on Boeing 737 aircraft, which for the most part make frequent flights along heavily trafficked regional routes, although the airline has expanded to the Midwest and east coast in recent years. This approach, along with bare-bones service, saves Southwest money but also puts a lot of cycles on its airplanes.

Scientific American spoke with Snorri Gudmundsson, an assistant aerospace engineering professor at Embry–Riddle Aeronautical University in Daytona Beach, Fla., about what causes cracks such as those that may have led to the fuselage rupture, what Flight 812 passengers experienced when their airplane opened up, and how neural networks might be able to someday detect cracks before they become a problem. Prior to joining Embry–Riddle, Gudmundsson worked for 15 years as a flight test engineer, structural engineer and the chief aerodynamicist at Cirrus Aircraft in Duluth, Minn.

[An edited transcript of the interview follows.]

What are some of the reasons that cracks might appear in an aircraft's outer aluminum skin? What may have caused the actual rupture?
In order to provide comfort or actually make it possible for a passenger to live at the altitude where it's efficient to run a jet engine—between 30,000 and 40,000 feet [9,150 and 12,200 meters]—you have to pressurize the cabin, so that the pressure inside the cabin is the same as it is at sea level.* There's an analogy with a balloon—if you blow up a balloon, the pressure inside the balloon is higher than the outside pressure, which is why it expands. On every flight the airplane takes off, flies to those altitudes, and pressurizes the fuselage. When it descends, the fuselage is depressurized. And then you do it again and again and again for subsequent flights. Each of these events is called a cycle. You're basically putting force into the aircraft's aluminum and [then] relieving it. Eventually, the aluminum begins to give in, and that phenomenon is called fatigue. When you pressurize an aircraft tens of thousands of times, the material's properties change—and one day it's flying and just cannot take the next cycle.

How common are the cracks that were found in the aircraft's fuselage?
Cracks like these are common in aluminum. The longer the aircraft is in operation, the more frequently they begin to appear. Where the crack appears on the aircraft determines whether it is a nuisance or a serious thing. The people who design these aircraft know where the most critical areas are, and they tell the operator which areas to inspect extremely well and which areas to inspect maybe less. The older the aircraft is the more prevalent these cracks are and the harder it is to keep track of them. If you don't have mechanics inspect these locations carefully enough, one or two or three may slip under the radar and something like this may happen.

*Editor's Note (4/08/11): Gudmundsson later clarified that cabin pressure ranges anywhere from seal level to about 1,500 meters.

The Boeing 737–300 in question has been in service for 15 years. Is that a long time for this type of aircraft?
It is really not a question of age in years as much as it is a question of how many cycles the airplane has in 15 years. It appears that the business model that Southwest has is one in which some of their airplanes are already old and already have thousands of cycles when they acquire them, and they operate them so rapidly that they go up to this critical number of cycles faster perhaps than they would for a different airline. To give you an idea, a Boeing 737 might be designed for 70,000 cycles, something that might happen over 20 years of operation normally. Aloha Airlines Flight 243, which experienced explosive decompression in flight in 1988 that caused a piece of the roof to rupture, killing a flight attendant, was a Boeing 737–200 that had been through about 90,000 cycles when that incident happened.

What was it like for the passengers on board Southwest flight 812 experience when the cabin ruptured?
In these situations people who were sitting in an atmosphere that corresponds to about 5,000 feet [1,500 meters] above sea level all of a sudden are sitting in an atmosphere that corresponds to 30,000 or 35,000 feet [9,150 or 10,650 meters]. At that point, air in the body starts to escape, but the biggest terror for people would probably be the popping noise associated with the rupture, followed by a very rapid accumulation of condensation on the windows—which rapidly goes away.* Oxygen masks come down. It's an immediate procedure at that point for pilots, when they realize there's a rapid decompression in the airplane, to dive down to 14,000 feet [4,250 meters], because below that altitude is where almost any human being is capable of breathing. This type of dive occurs normally at about 4,000 feet [1,200 meters] per minute, and I guess that would be a terrifying experience also, because most of passengers are not going to realize that this is actually being done to save their lives. The pilot is quickly taking the airplane someplace where there is oxygen-rich air.

Would a five-foot by one-foot rupture such as the one Southwest flight 812 experienced greatly destabilize the aircraft for the pilot?
This was way too small a crack to be an issue in terms of the stability of the airplane.

The aircraft had been given its last "heavy check" in March 2010. How might the cause of the rupture have been detected ahead of time?
There are several techniques that are used, including the eddy-current technique and x-rays. These cracks do not just appear out of nowhere. It takes years before such cracks would cause a panel to fail. I cannot tell you why Southwest did not detect these cracks in their last major overhaul. An aircraft will fly for maybe 15,000 cycles before they start to inspect for cracks. Then, they will do it every 3,000 cycles, or something like that. Depending upon how quickly an aircraft racks up cycles, it may be anywhere from two years to six years between overhauls where they are actually looking for cracks. I have to say that I am surprised that they didn't catch these cracks in that particular airplane. Why? They will have to answer that.

What are they looking for during a normal preflight maintenance check?
During a preflight check, typically the co-pilot will do a walk around the airplane to check the airplane's wheels, sensors and external controls to make sure nothing is blocking them. A walk-around is not designed to catch cracks because these are usually microscopic.

*Editor's Note (4/08/11): Gudmundsson later clarified that there would have been a very rapid accumulation of condensation of humidity in the airplane's cabin that would appear like a fog (not clouding the windows, however) and then would quickly dissipate.

Is this a case where there is a problem specifically with the Boeing 737 aircraft or would closer inspections across all different commercial aircraft yield similar fuselage problems?
All airplanes are subject to metal fatigue. The only way to catch it is by proper maintenance procedures. Every day, many airplanes have patches and skin plates put on them to prevent further fatigue, and we never hear about it. It only becomes a problem if the maintenance is inadequate.

Southwest has replaced the aluminum skin on many of its 737–300 airplanes in recent years, according to a spokeswoman. The planes that the company has grounded over the past few days had not had their skin replaced. What does this tell you?
They are aware of this problem and are trying to prevent it from becoming too serious. I don't want to say anything that's unfair, and I don't know how they operate their maintenance program, but it is not unreasonable for the flying public to question two incidents such as this occurring from at the same airline. [On July 13, 2009, Southwest Flight 2294 from Nashville to Baltimore was forced to divert to West Virginia after a hole formed on the top of the Boeing 737's fuselage near the tail, resulting in depressurization of the cabin and deployment of the oxygen masks.] Southwest is known for short flights, which means their airplanes accumulate a lot of cycles over a short period of time. Perhaps because of that they should change their maintenance procedures so they have a better chance of catching these cracks before they become failures.

Does this encourage the airline industry to look into new types of composites and other materials that might be used to build their airplanes?
The industry has been moving toward the use of composites for a while. But it doesn't matter whether it's composites or aluminum—all materials have their shortcomings. Aluminum is a fantastic material. The problem with aluminum, though, is that it doesn't have what we call an endurance limit. Steel, for example, has a known stress endurance limit. This means that as long as the stress levels in the material are kept below a certain value, you can cycle it endlessly. In the case of aluminum it doesn't matter if you apply high stresses or low stresses, eventually you're going to break the material. Of course it will take you a lot longer if your stress levels are low. Moving away from aluminum and toward other lesser-known materials, however, might be opening a different can of worms.

Instead, it would be better to incorporate into the airplane a system that would monitor crack growth. One way of doing this that is being developed at Embry–Riddle is to put a microphone on the skin of the airplane that picks up noise. You would then use a neural network, basically artificial intelligence to break the sound into constituent components and identify the sources of different types of sounds. For example, you create a mathematical model that can estimate the characteristics of a crack generating a particular sound. This system would be on every flight, and when it determines too much noise is coming from a particular direction it would warn the pilot.