Twin Mysteries

A Beech Duke

Usually the cause of a general aviation accident is apparent, or at least can be determined through investigation. But recently the NTSB posted probable causes in two accidents involving piston twins where the Board essentially says we can see what happened but we don’t have any idea why it happened.

The first accident involved a Piper Aztec. The sturdy Piper twin was maintained to the appropriate standards and had about 6,000 hours total time. Not a lot of time for an Aztec.

The weather was a typical Texas summer day with temperature at 90 degrees, scattered clouds at 5,000 feet, wind about 10 knots and visibility of 10 or more miles.

The commercial pilot also held a CFI and had more than 900 hours of total experience. He had recently earned his multi rating and had logged just over 39 hours in twins, but since both engines operated normally during the accident multi experience doesn’t seem to be an issue.

There was absolutely nothing remarkable about the conditions for the VFR flight from Houston Hooks to Tyler. But as the Aztec flew over Lake Palestine near Bullard people on the ground saw the piston twin roll into “several turns” while flying thousands of feet above the surface. The report doesn’t mention how steeply banked the turns were, but the witnesses then saw the Aztec pitch nose up and then head straight down.

The really odd aspect of the accident is that the Aztec was heading straight down without its wings. Investigators found that both wings failed symmetrically in positive overload. There was no evidence of metal fatigue, cracks or corrosion that could have caused the failure. Somehow the pilot had literally pulled the wings off the Aztec in good VFR weather with no reason to believe there was significant turbulence.

The Aztec is certified in the normal category so its structural limit load is 3.8 Gs positive. During certification the Aztec was tested to that loading and showed no deformation of the structure. It was then tested to ultimate load of 5.7 Gs and probably suffered some skin wrinkling or other stress but did not break.

That’s a lot of Gs. A level 60 degree bank turn only loads an airframe to about 2 Gs. Unless the Aztec had accelerated well beyond its design maneuvering airspeed (Va) it would probably be impossible to load more than 5.7 Gs without stalling it. But witness did not report seeing the airplane in a dive before it pitched up and the wings failed.

Control flutter was a possibility but the elevator, stabilizer, fin and rudder all reached the ground in one piece and showed no evidence of any flutter or other failure. The NTSB’s probable cause finding is that the pilot exceeded the structural limits of the airplane which is as obvious and meaningful as also pointing out that the crash was caused by the airplane hitting the ground.

The other piston twin accident mystery involved a Beech Duke departing Sedona in the Arizona Red Rock country. Sedona’s runway is on a plateau with an elevation of 4,830 feet. It was a warm day creating a density altitude of 7,100 feet.

Anyone who has ever flown the sleek turbocharged and pressurized Duke knows that it looks fast just sitting still, but is pretty much of a slug on takeoff with modest at best initial acceleration. But when the NTSB worked the performance numbers it found that the Duke should have lifted off after using 2,805 feet of the Sedona runway, despite the high density altitude. The POH also showed that the Duke could have accelerated to rotation speed, and then aborted the takeoff and stopped in a total distance of 4,900 feet. The Sedona runway is 5,132 feet long.

For some reason the Duke never rotated, and the pilot made no apparent attempt to abort the takeoff. The Duke, with engines sounding at full power, went off the end of the runway dropping off into a deep gully where it caught fire and killed the three people onboard.

Several witnesses, mostly pilots, watched the accident unfold. All agreed the engines sounded normal and like they were making full power. Some said the airplane was moving fast, others said acceleration slowed about half way down the runway. None saw any attempt by the pilot to raise the nose.

The private pilot was not highly experienced but had more than 600 hours total, more than 100 hours of multi time, and more than 60 hours in the Duke. He had attended initial and recurrent simulator training for the Duke and his instructors told investigators he had been very attentive, knew the airplane and its systems well, and was a good student.

What comes immediately to my mind is that the gust lock may have been in the controls. But the lock was found in the cabin, not the cockpit, and the holes in the control column where the lock pins the column were not deformed.

Examination and testing of the engines, propellers and other components didn’t reveal any reason the Duke wasn’t making full power confirming witness reports of the sound of the engines.

At least the NTSB is forthright in its probable cause finding on this one. It reported that it just doesn’t know why the pilot didn’t raise the nose of the Duke or abort the takeoff. Both accident causes remain mysteries.

This entry was posted in Mac Clellan's Left Seat Blog. Bookmark the permalink.

33 Responses to Twin Mysteries

  1. Absent any other obvious explanation, it seems to me that we must assume that any accident involving a piston twin was caused by (drum roll) money exhaustion.

    • Mac says:

      Well, Mike, it wasn’t an engine failure that worries so many about flying twins. But pulling the wings off sure seems like an extreme measure to save money.
      Mac Mc

  2. Jeff Welch says:

    Perhaps pilot incapacitation, at exactly the wrong time, regarding the Sedona mystery.

    Tremendous loss no matter the cause.

  3. phil g says:

    Enough with the disaster porn. We get enough from the MSM

  4. Mike Gorno says:

    Mac – on the Duke – any mention of the trim? Full nose down perhaps?

    • Mac says:

      Hi Mike,
      Pitch trim setting is always a question mark in an accident such as this. The post crash fire consumed much of the Duke, but investigators located the trim actuators and found they were in a “neutral” position that should not have prevented the pilot from rotating. The rudder trim actuator was also found to be in a “neutral” position.
      There have been several accidents in jets caused by incorrect trim setting on takeoff, but those involved airplanes with trimable stabilizers. In that case the stabilizer provides the majority of pitch authority while the elevator is quite small and unable to overcome the forces of the stabilizer position even if the pilot deflects the elevator fully. The Duke has trim tabs, as do most piston airplanes, and the tabs change control forces but do little to alter the authority of the elevator. If a trim tab is not properly set the pilot, if he pulls or pushes hard enough, can overpower the force generated by the tab. That’s not true in large airplanes with trimming stabilizers.
      Mac Mc

  5. Howard Riley says:

    Mac, what in Gods name will this information do for the knowledge seeking EAA member? Your continual foray into lala land with technical inaptitude is getting tiresome, and not of value to the thousands of EAA members seeking real education. You certainly are not representing EAA in your meandering, much less properly fulfilling your executive position of Editor In Chief! Get a life-get out of EAA, please.

    • phil g says:

      Too much focusing on GA issues and not enough about home building and experimental aviation. Maybe visit some chapters and report on what people are building or why people aren’t building

    • Chris mayer says:

      EAA is not just homebuilders. I have an O-2A, which is a twin (despite usually having to point out to bystanders where the second engine is.) Warbirds are part of EAA, as is Vintage — and the Aztec could very well have qualified as vintage. Additionally, these accidents do not seem to be related to the number of engines and could, theoretically, apply to any aircraft…even E-AB. There are no lessons to be learned, yet. Just events that needs further study — and fuel for discussion in a chapter safety meeting.

    • Robin Hou says:

      Unless you only build and don’t fly, once you start flying, GA issues are relevant.

  6. dave mckenzie says:

    Has the pilot’s background and personal life been reviewed and what’s his relationship to all of the passengers? Could it possibly be self destruction?

    • Mac says:

      Yes, Dave, the NTSB always interviews people who knew the pilot searching for any unusual behavior, fatigue, illness or other factors that could contribute. None were found in this case.
      Mac Mc

      • Jacques says:

        Not necessarily true in the Duke incident. There is something clearly stated in the NTSB factual report. I don’t want to be disrespectful here, but it is a sad fact that this pilot had 0.013% alcohol in his blood and about the same percentage in his urine. That blood alcohol level is close to one drink’s worth of alcohol. There was also n-propanol detected, which means the alcohol was imbibed. The accident occurred at 8:30 am. The body excretes about one drink’s worth of alcohol per hour. If the pilot left his hotel an hour before the flight to pre-flight and refuel, that means he had about two drinks worth of alcohol in his blood at that time. I will leave it to each of you to see what that means assuming this pilot followed the 8 hours “bottle to throttle” rule. The NTSB most likely did not put this down as a contributing factor since the level did not exceed 0.04%, the legal intoxication level. The AIM, however, states that as little as one drink can have an effect on one’s flying ability. I would imagine this to be particularly true if the alcohol in one’s system is a remnant. Remember also, the pilot was a two-time Olympic runner who was taught to not give up through the end of a race. While a desirable trait for an Olympic athlete, it’s not such a good one for a pilot. Regardless, the pilot should have calculated the T/O distance prior to departure and shut it down at that point if he didn’t lift off.

  7. Cary Alburn says:

    Demographically, there is a large percentage of EAA members who are strictly GA pilots, me included. I have neither the patience nor the talent to build an airplane, although I admire those who have both. So I don’t see any reason not to include GA issues; EAA has long ago become an organization of many facets. That’s just the way it is, like it or not.

    On the issue of Mac’s blog this time, I’m not sure how helpful a pair of “we can’t figure out why” reports are, other than to remind us that there is a risk in flying, and sometimes that risk isn’t quantifiable.


    • phil g says:

      Aviation is dangerous, disaster porn reinforces that perception. I am sure if your wife read this article she would want you to sell your plane. Mission accomplished

  8. Jeff Welch says:

    Mac. Accident discussion is an important part of accident prevention. Keep up the good work.

    The elephant in the room that no one wants to talk about is….. the number of accidents and fatalities happening in experimental aircraft. For the health and viability of the homebuilt industry, we need to reduce accidents and fatalities posthaste. We need to do this ourselves and not with the help of big brother. With freedom comes responsibility.

    I recommend to AOPA, ASF, EAA that they jointly work on a campaign of awareness to reduce (all GA) fatalities by 50% in 2015. It can be done.

  9. Robert says:

    Re. the Duke going off the runway at Sedona. This sounds like a case of attempting a take off with the parking brake partially set. An airplane like the Duke has a good power to weight ratio and it sets in a nose low attitude. Two conditions for this to be a probable cause. First with a good power to weight ratio it would be possible to accelerate with a partially set parking brake, depending of course on the degree of brake pressure. Second in a nose low attitude, the airplane will not fly off unless the nose is rotated to a positive angle of attack. This is like most if not all jets. When the pilot had reached lift off speed and attempted to raise the nose, with the brake pressure applied the downward rotational moment forcing the nose down would overpower the elevators ability to raise the nose. The pilot being faced with the inability to raise the nose at a critical time would probably continue applying elevator pressure until to late for abort. Which is the only thing to do. If you knew the brake was on you could release it, but that is something most pilots are not taught, and for obvious reasons not practiced. I have flown many full motion simulators, no mention or practice for this condition. My personnel feeling is that a warning system should be in place to prevent attempted takeoff with any brake pressure in the system.

    • Tim says:

      Interesting comment & thoughts. Been a Pilot, AP/IA for 13 years. Worked on & flown the Duke. I speculate if this happened, by rotation speed, the brakes would have been smoking red hot & no longer effective anyway.

      • Robert says:

        In answer to Tim re. hot brakes no longer being effective. Quite a number of years ago Kellogg corp’s Falcon 10 went off the runway at Meigs Field for exactly the reason I mentioned. The parking brake was set at an intermediate position. The aircraft reached rotation speed and the pilot was unable to raise the nose. I believe the Captain was killed, there may have been others as well. I don’t remember all the details. I have also heard of a CJ2 that did the same thing. This resulted in the pilot aborting the takeoff, the aircraft did catch fire after stopping, I don’t know if there were any injuries. Just because brakes are “smoking hot” doesn’t mean they won’t work.

    • Mac says:

      Hi Robert,
      The landing gear was torn off the airplane and there is no mention of the parking brake valve in the report. As you know, the parking brake valve traps hydraulic pressure between the valve and the brakes. If only a small amount of pressure is trapped and the valve is not opened before takeoff, the pilot may not notice the drag of the pressure on the brakes until it is too late. It has happened before.
      Mac Mc

    • Greg Broburg says:

      The situation of starting a take off with the parking brake set happened on a 210M at EAU where I was a rear seat passenger. Right seat was an instrument instructor reading the procedure list, left seat was an instrument rated pilot. We had 6 people aboard, I’m not sure of fuel on board, we were heavy but not at gross. The parking brake was set prior to the runup. The procedure list was put away after the runup and the brake was not released. A takeoff was done with the brake pulled on. TO acceleration was slow but did lift off, the aircraft felt like it jumped when the tires lifted off. The locked brake was noted just off of the ground when the pilot noticed it hitting his knee. What is the value of setting any parking brake prior to the runup? For the Duke, where in the procedure is the item to release the parking brake? Is setting the parking brake and its incumbent risk of failing to release it really worth any benefit?

      • Reid Sayre says:

        This is the reason I have a set of custom checklists. There are actually quite a few. I have a checklist for day before the flight, before leaving home to go to the airport, before starting the engine (before custom checklists, I forgot to remove the chocks (twice) and had to stop the engine and get out and remove them), and some others. During my private training, the checklist supplied with the rental plane omitted “remove left tiedown.” I was not the only student that started the engine with the left tiedown still attached.

        So, I have some long checklists, but a lot of the items are on there because I forgot the item at least once. And some items are not required for every flight, but I would rather have it on there and not need it, today, rather than not have it on there (because it’s an odd case) and forget it.

  10. Robert says:

    Mac, just a note of encouragement. Some of the respondents seem to believe they can only learn from events that pertain to experimental aircraft. When we can learn from other peoples mistakes why would we pass up such an opportunity? I realize the scenario I outlined does not apply to all general aviation aircraft, but it certainly could apply to some general aviation and experimental types. On a personnel note, during a flying career that has spanned 55 yrs. the single one thing that has given me the most problems are brakes, yet training and practice for brake problems is practically nonexistent. The thing about brakes is they can fail either on or off. Depending on your, situation either mode can be equally bad.

  11. Dr. Jimmie L. Valentine says:

    Mac, I love your writing keep up the good work. You have added so much to the EAA magazine since you came aboard.

    Could the young Aztec pilot been attempting a barrel roll?

    • Mac says:

      Hi Dr. Valentine,
      The Aztec pilot did not have a lot of experience, but he was 51, not young in the overall scheme, but certainly younger than many of us who fly. And it is certainly possible that our of boredom or whatever he may have attempted some sort of an aerobatic maneuver that went wrong.
      A few years ago the NTSB determined that the pilot of a Baron tried to roll it and lost control and crashed. That finding was based on the remarks the pilot had made to other pilots. And that pilot was also flying home after attending an air show.
      In the case of the Aztec breakup the NTSB didn’t report any reason to believe the pilot would try some sort of maneuver he nor the airplane was capable of.
      Mac Mc

  12. Robert says:

    Thanks Greg for the input. I believe there are more such incidents than are reported. As far as setting the parking brake for runup, some pilots won’t use the parking brake prior to takeoff. However, there are times when it would be convinent to use it . Say when you get a long delay before takeoff, when you are parked on a slope, or flying a jet where there is enough idle thrust to start the plane rolling without the brake set. Often times on an instrument flight you want to review the clearance and perhaps consult charts prior to takeoff. In a heads down situation it is good to have assurance the plane will not start rolling. You could be behind someone else. I still believe a warning system would be justified. I know many pilots feel that idiot lights are unnecessary, and most of the time they are because we are on top of our game and wouldn’t miss something like taking the parking brake off. But, when the stakes are so high for this one thing, maybe we should reconsider.

    • Greg Broburg says:

      Good points all Robert. I am also remembering an accident in Russia where the right seat pilot had his feet on the brakes during TO. The result was that it failed to get airborne and the entire Lokomotiv hockey team died. Another brakes on incident but not related to a parking brake being set. So the overall concept really needs to include parking brake and foot brakes off. I am also reminded of a Beech 18 accident about 25 years ago near Cambridge MN where the pilot/mechanic had just installed new brake pads made of modern pad material. The braking efficiency of the new pads was so much better than the original that the airplane ended up on its back, on fire, and destroyed. Pilot was killed. Greg

  13. Thankѕ for finallү ԝrіting аƄߋut >
    &#5026ԝіn Myѕteгіеѕ | &#5290&#6513ft Sеаt < Loved it!

  14. My family members all the time say that I am wasting my time here at net, except I know I am getting knowledge all the time by reading thes good posts.|

  15. Chase Burnett says:

    Excellent commentary & blog.Flying mag’s loss.

  16. whoah this blog is great i like studying your articles. Stay up the good work! You already know, many individuals are looking round for this info, you can aid them greatly. |

Leave a Reply

Your email address will not be published. Required fields are marked *


You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>