I have previously addressed some of the causal factors in the error chain leading up to accidents that are broadly grouped under "loss of control" or LOC. I have previously written about and conducted seminars and webinars on how maneuvering flight errors can lead to a LOC accident. Likewise, I have addressed spatial disorientation through articles, seminars, and webinars. But maybe it is time to take a step back and look at the bigger picture.

The concept of losing control of an airplane seems foreign to most general aviation pilots. The airplanes that we fly are, for the most part, pretty stable and forgiving. Yet the accident records abound with reports of LOC accidents in some of the most common, user friendly, airplanes. We have no data on how many loss of control events occur with the pilot being able to recover. We only know how many end in accidents. I would guess that a rather high percentage of LOC events end in LOC accidents. These accidents are particularly bad news because they are likely to be fatal.

So I think we need to approach the problem from two angles. First, we need to have a good understanding of the situations that set us up for a loss of control so that we can avoid those situations. That involves a solid review of aerodynamics, particularly the conditions that produce stalls and spins, the effect of load factors on stall speed, and the importance of maintaining coordinated flight. Second, we need to make sure that we are proficient in our aircraft control near the edges of the performance envelope. This is best accomplished by taking some recurrent training from a competent flight instructor. The instructor must be experienced in the type of airplane to be flown and be competent and comfortable in pressing the envelope. We must repeat this for the various kinds of airplanes that we fly. I am not saying that if we are flying very similar airplanes such as a Warrior and an Archer we need to train separately in each. But if we fly a Cessna 172 and a Bonanza, then we must address each airplane separately because they handle quite differently near the edges of the envelope.

Many people would advocate a third step, upset training. I think that upset training can be valuable, but I would rather have a pilot concentrate on avoiding a loss of control than in how to recover from it. Read more about that in my "Commentary" section.

I can't help with the actual flying, but I can briefly discuss some of the reasons pilots end up losing control of healthy airplanes. This is not a complete list, but it is a place to start.

• Maneuvering Flight Error
• Spatial Disorientation
• Distraction
• Thunderstorms, Icing, or Turbulence (Including wake turbulence)
• Lack of Proficiency (IFR)

Some of these factors may overlap. Or, two or more may appear as separate links in the error chain.

The Maneuvering Flight topic has some sub-topics. One of the more common is the classic skidding turn from base to final. This is brought about by the pilot realizing that the turn is going to overshoot the runway so rudder is applied to bring the nose around. This results in a skid which messes up the aerodynamics and induces a stall well above the normal stalling speed. The stall during uncoordinated flight causes a spin. Recovery at the relatively low altitude is unlikely. The stall/spin LOC accident that results from attempting to clear an obstacle on takeoff or trying to stretch a glide also comes under Maneuvering Flight topic. Also, let's not forget the accidents that happen when a pilot tries to be an airshow performer to show off for family or friends. The LOC accident there would also be a sub-topic under Maneuvering Flight.

Spatial disorientation is next. I did a webinar on that subject in October of 2010. I consider three risk factors for spatial disorientation, reduced visibility (night or instrument conditions), lack of proficiency, and physiology. The first two are probably self-explanatory, but physiology might need some explaining. If we are ill, taking medications, or fatigued, our risk of becoming spatially disoriented increases substantially. Over-the-counter cold and allergy medications are particularly dangerous. I am seeing more and more fatal accident reports in which the toxicology report lists one or more OTC medications. Unless the concentration of the medication is very high, impairment is not listed in the probable cause finding because it is impossible to know whether or not the medication played a role. But we all know that many of these medications have adverse effects. The slight slowing of reaction time or the clouding of decision making ability while using one of these medications just might have made the difference between living and dying.

A distraction can be the first link in the error chain for a LOC accident. This is more likely when flying on instruments, but I have a number of documented cases in which something as simple as a door popping open after takeoff caused the pilot to lose control of the airplane in VFR conditions. Programming the GPS or autopilot, a sick passenger, or any kind of minor malfunction can be the distraction that causes the world to turn upside down. The simple remedy is to fly the airplane regardless of what else is going on.

Flying into weather that exceeds the capabilities of either the airplane or the pilot is another causal factor in LOC accidents. This one usually creates an intermediate link in the error chain. A lack of thorough preflight planning is frequently found as an earlier link in the chain.

Finally, for the instrument pilots, a lack of proficiency in flying solely by reference to instruments shows up as a causal factor in many LOC accidents. A typical scenario is the instrument rated who hasn't stayed on top of the game since earning the rating. Then an important trip comes along and the weather is down. Sometimes this is the return trip after a weekend getaway with an important meeting on Monday morning. The pilot isn't particularly comfortable, but thinks, "I have an instrument rating so I might as well use it." But the instrument scan is slow and the procedures are a bit faded in the memory. Task saturation leads to confusion and then to loss of control. In the human factors world, the pilot's decision to go would be considered an early link in the error chain that allowed external factors to affect the decision making process.

This is only a brief introduction to the subject of loss of control. One could easily fill a book about the subject. I am creating a course on the subject and I hope to have it finished by the end of April. I will let everyone know when it is available.

We will look at two LOC accidents, one that happened in VFR conditions and one that happened in IFR conditions.

The pilot lost his life in this Cirrus SR22 accident that occurred in South Carolina in January 2009.

NTSB photo Click to enlarge

According to the NTSB accident report, "About 6 minutes after departing in visual meteorological conditions, the pilot reported fumes in the cockpit to air traffic control, and advised them that he planned to divert. The pilot subsequently began descending toward a nearby airport, and entered the traffic pattern. As the airplane neared the runway threshold, on what appeared to be a "normal" approach, it suddenly pitched nose down and impacted the ground about 900 feet short of the runway threshold. A witness who approached the airplane immediately following the accident reported no fire, smoke, or abnormal fumes. Post accident review of the airplane's maintenance records revealed that about 3 weeks prior to the accident, the pilot had reported an odor of "burnt electronics" in the cabin, and subsequently had the airplane inspected by a maintenance facility. No anomalies were found during the inspection and all of the airplane's systems continued to function normally thereafter. A post-accident examination of the wreckage revealed no evidence of any pre-impact mechanical malfunctions, failures, or fire in the accident airplane. A detailed inspection of the airplane's electrical system and avionic components revealed a single damaged capacitor located in the primary flight display that displayed evidence of thermal distress. Functional testing of the circuit card containing the damaged capacitor suggested normal operation."

The NTSB listed the probable cause of the accident as follows: "A loss of aircraft control during the landing approach for undetermined reasons."

It seems that there probably was an odor of hot electronics in the cabin. That is certainly cause for concern and the pilot's decision to get the airplane on the ground at a suitable airport cannot be faulted. The problem was probably not imminently dangerous, but the pilot had no way of knowing that. But it appears that somewhere between the time the pilot decided to get to an airport and the few moments before the accident, he allowed the distraction to cause him to lose control of the airplane. No matter what else is going on, we have to fly the airplane. A loss of control accident is almost always fatal. Even if we cannot get to an airport and things are really bad, an off-airport landing while under control has a much greater chance of being survivable than does impacting terrain while out of control.

Click here to read the full accident report on the NTSB website.

January 2009 was a bad month for pilots flying Cirrus airplanes. This accident involved a Cirrus SR20 that crashed following a loss of control accident in Menomonie, Wisconsin in January of 2009

NTSB photo Click to enlarge

The NTSB accident report says, "The private pilot, who had received his instrument rating about one month prior to the accident, departed at night on an instrument flight rules (IFR) flight plan. During the weather briefing, the pilot was advised that conditions would be conducive for icing with increasing cloud cover. The pilot departed and climbed to 6,000 feet mean sea level (msl), but about one hour after departure he requested 7,000 feet to get above the cloud tops. On arrival, the pilot was cleared to descend to 4,000 feet msl and to fly direct to the initial approach fix for the destination airport. During the descent, the airplane encountered instrument meteorological conditions. The cloud bases were at 1,100 feet above ground level (agl) near the accident site.

NTSB photo Click to enlarge

Air traffic control (ATC) asked if the airplane was picking up ice, and the pilot reported that it was not. Radar track data indicated that the airplane started a right turn, and about one minute later the airplane was lost from radar contact. The airplane was destroyed by impact with terrain. Witnesses nearby reported hearing the engine running at a high power setting. The inspection of the airplane and engine did not reveal any pre-impact anomalies. The airplane was not certified for flight into known icing conditions. The airplane’s parachute system was not deployed, and the parachutes' safety pin with the red colored "Remove Before Flight" tag was found in the activation handle, still in the handle holder. The preflight checklist calls for the safety pin to be removed prior to flight."

The NTSB Probable Cause of the accident stated, "The pilot's failure to maintain control of the airplane while flying at night in instrument meteorological conditions."

The probable cause as stated by the NTSB is of course correct. The pilot lost control of the airplane. But that doesn't tell us much that will help other pilots avoid the same fate. The limited information available does not allow us to isolate a single factor that led to the accident, but we can look at several areas to see what might have been done differently.

There is one glaring error on behalf of the pilot. The safety pin was left in the deployment handle of the ballistic parachute system. Removing the pin before flight would not have prevented the loss of airplane control, but would have allowed the pilot to activate the system and might have saved the lives of those aboard.

It appears that the pilot was legally qualified to conduct the flight. In fact, he had renewed his medical certificate on the day of the accident. He had received his instrument rating just forty-one days prior to the accident, so he met recent experience requirements for instrument flight. But legal doesn't always mean safe. He had some actual instrument experience. The NTSB report shows about 19 hours in the 13 months preceding the accident, but it is not clear how much of that was with an instructor. He also logged about 16 hours of night flight in the same period. Again, there is no indication how much of that was PIC. He had completed the Cirrus factory training program within the four months immediately preceding the accident, but only the VFR course. That raises a question about his familiarity with operating the technically advanced systems in the IFR environment. Since the accident occurred as the pilot was descending for the approach, perhaps he was programming the autopilot at the time he lost control of the airplane. If that was the case, it would not be the first time that had happened.

Finally, we have to look at the possibility that the airplane was picking up ice. The weather conditions were certainly conducive to structural icing and there had been some pilot reports of ice in the general area. The pilot was aware of that and he had been asked twice by ATC if he was encountering ice. He responded that he was not. The pilot did not seem to be under any unusual stress during his last communication with ATC. If he was having any sort of problem, he did not reveal it either through his words or his tone. Click here to listen to the ATC communications. But detecting ice on an airplane with white wings at night is not always easy. There is no indication as to whether the pilot had ever experienced structural icing in the past. If this was his first encounter he might not have been aware that ice was forming. He perhaps had the autopilot flying the airplane. There are plenty of documented cases in which the autopilot attempted to maintain the altitude that it was programmed to fly, but allowed the airspeed to deteriorate until a stall developed. The stall speed would be increased if there was ice on the airplane and the stall might have been sudden and severe. We will never know exactly what happened in that final couple of minutes before impact.

The pilot was aware of the possibility of structural icing before he departed. Legalities aside, one has to question his decision to depart. But from my own experience flying in northern latitudes in the winter, I have made that same decision many times and I have gotten away with it. Except for the one time that I almost paid the price. My close call caused me to change my thinking about flying when structural icing is a possibility in an airplane that is not properly equipped to handle it. It is a tough decision when the forecast almost always calls for ice three months out of the year. There has been much discussion about what is legal in this regard and I do not want to get into that. But legal or not, flying in structural icing conditions in an airplane not approved for flying in known icing can get us killed. I had a close call in May of 2000. Had I found myself in the same situation within my first thousand hours of flying, I would probably not be here today. It took every bit of my skill, knowledge, and experience plus a whole lot of divine intervention to get me to a runway while still flying and under control. Click here to read my ice story.

Click here to read the full accident report on the NTSB website.

Many people advocate that all pilots take upset training. No, that isn't training to learn how to deal with your spouse after you have spent too much time at the airport. (Though now that I think about it, I have great experience in that area and could probably create a valuable course.) Upset training refers to a course with the objective of learning how to control the airplane at the edges of the performance envelope and how to recover if control is lost. I am in the minority, but I think that the value of upset training is over-rated in preventing LOC accidents. Here's why. First, most general aviation LOC accidents happen at relatively low altitude, particularly when maneuvering in the pattern. If control is lost unexpectedly at pattern altitude or below, it is unlikely that a recovery can be made before the ground is contacted. Second, most upset training is conducted in a fully aerobatic airplane that handles quite differently from the "family sedan" that most pilots fly. I have done enough aerobatic flying in enough different kinds of airplanes to know that even though the procedures are similar, recovering from an upset in a Citabria or Extra 300 is quite different from wrestling the Baron back upright. And third, I have heard pilots boast that they took this great upset training course and now are much more confident in their ability. The problem is that they took the course eight years ago. If the skills acquired in the training have faded, we are left with an overconfident pilot.

Please don't get me wrong. I am not opposed to upset training and I believe that it can be very valuable. In fact, I think that any kind of training is valuable. But for pilots with limited financial resources, I would rather see those finite resources spent with a good, competent instructor exploring the edges of the envelope in the airplane or airplanes that are routinely flown. Then I would like to see similar training happen again, at least annually. If funds are readily available, by all means, run to a reputable provider of upset training. A better pilot will emerge. Taking upset training at least semi-annually would be beneficial. But if the budget is tight, spend a fraction of the cost of upset training on becoming better at avoiding the upset.