Category Archives: Aviation Safety

Research on the Stinson 108-3 Flies Through Trees at CML2 Video

Ok, this is a Stinson 108-3, which like;y has a 165hp Franklin engine and a max gross weight of 2400 lbs with a factory empty weight of 1320lbs. Fuel capacity was 50 gallons. Being these aircraft are often heavily modified, its hard to tell how legit those numbers are.

The airport is CML2, which has a field elevation of 130ft MSL and its located just north of Victoria BC. The lone runway is 1800x75ft. Maximum recorded temperatures in the area are under 100F. On the day of this flight, the pilot said it was hot… so giving the benefit of a doubt, lets go with 100F and moderately humid at say 70%. Such would put density altitude at around 3300 ft which will increase the length of the take off roll.

The airplane has 3 souls on board, all of which are pilots, so lets assume each weighs 170 lbs. We don’t know the fuel loading, but since fuel was not available on the airport, the tanks cannot be full. Lets figure around 35 gallons, which works out to be around 210 lbs. Alas, even though its under max gross weight at ~2100 lbs (includes some additions like shoulder belts, avionics, dust/dirt, fabric, dope, etc over the years), it will not perform anywhere near as well as the idealized factory values.

On a positive note, the runway slops downhill, so even though it is an unimproved surface, take off roll will decrease, assuming winds are aligned with the runway.

As this is an older and potentially highly modified aircraft, online performance charts are lacking. Stinson states the take off roll under unknown conditions to be 980ft. Some pilots suggest fully loaded under unknown conditions it is 1500ft. Being the aircraft did achieve flight, despite improper pilot technique, a calculated take off roll under these conditions would likely fall somewhere between the two figures, lets say 1250 ft.

Bottom line, taking off on this 1800 ft runway under the above assumptions provides for a ~45% margin. (1800ft runway-1250ft takeoff roll)/1250ft takeoff roll.  As a result, its pretty unlikely that the other pilots on board questioned the pilot in command as to whether the flight was safe or not.

Alas, this presents an interesting question, one of which many of my students have asked over the years… what sort of margin should I add to my performance figures? In this case, 45% almost resulted in an accident. Being 30-50% is a pretty common answer I’ve gotten back from fight reviews over the years, such is a common buffer for a lot of private pilots, but is it really the right one?

I remember as a young CFI asking an old bush pilot this exact question, having had my own near miss at the end of a runway with too small a margin. His answer, and the one both he and I go with for personal flying… 100%!

More on this in an upcoming post.


Sources

Density Altitude Calculator

Humidity Dewpoint Calculator

Stinson 108-3 Specifications

CML2 Airfield Data from Skyvector

Commentary from the pilot in command in the above video

I really appreciate that the pilot in command provided some commentary on this. Its far too easy to sit an an armchair and quarterback this type of thing as a “this will never happen to me, I’m not stupid” that is, until it does. Bottom line, if NASA with all their safety checks and technical acumen can blow a units conversion and crash a satellite, who are we to think we are immune to calculation and/or judgment errors when it comes to aviating.

CRM Questioning the Senior Pilot

The crash of 2 Boeing 747’s in 1977 was cited by a NASA workshop which initiated CRM (Crew Resource Management). Skipping all the buzz words and hoopla, the big deal was crew coordination and to encourage questioning authority… ie, if a FE or copilot saw things going south, not to sit back and wait for fecal matter to hit the fan.

See a recreation of the disaster in this video.

As a result of the NASA study, United Airlines rolled out CRM training to their crews in 1981. While airline safety improved, CRM didnt really drop right into general aviation operations being most are single pilot ventures… It took a while, more buzzwords needed to be created, things needed to be tailored to fit, and lobbyists needed to be plaacted etc… but after 24 years or so, SRM entered the general aviation scene in 2005.

The thing is, most every chief flight instructor I worked for didn’t need an acronym and fancy FAA promo materials to utilize the human factors issues from CRM, ie situation awareness, workload management, automation management, and aeronautical decision making. Many adopted CRM as part of their 135 ops, and it was a natural thing that many of these concepts rolled into their 61 flight schools… but, the bit about crew coordination, questioning authority etc was for the most part left out for seemingly obvious reasons.

I think this was and is a mistake… The PPL likely wont be flying as SIC anytime soon, especially as most wont go on to 135 or airline careers. Otoh they will be flying with or even for other pilots…

I about bought the farm years ago as I didn’t bug the senior pilot…. with 14,000+ hours, combat experience, and 40+ years as a pilot in all types of aircraft, who was I, a 19 year old n00b to question him? That was a hard lesson, and its one thats stuck with me and something I try very hard to impress upon my students.

Another reason for including the crew aspect early on is despite decades of CRM, crew coordination is still a problem.

Consider the very sad story behind AF447 and TK1951.  Granted, there were other problems that started the  chain, but all along the way CRM weaknesses are popping up. While its easy to arm chair quarterback, had crew operations been included from the earliest days of flight training, one has to wonder if it wouldn’t have made a difference.

Test your own CRM with this scenario simulator from Transport Canada.

http://data.tc.gc.ca/archive/eng/civilaviation/regserv/safetyintelligence-airtaxistudy-simulation-decision_simulator-497.htm

 

Evaluating Safety Management (PTS proposed changes) Part 1

“The goal of the airman certification process is to ensure the applicant is ready to safely manage the risks of flight as pilot-in-command, consistent with the privileges of the certificate or rating exercised.”

In other words… lets add a bunch of extra cost and time to pilot training in the interest of safety management. Evaluate it not only during a written exam, but also take critical time away from from the stick and rudder arena, and then somehow magically pilots will be safer.

In 1977, the FAA had this nifty illustration as concerns teaching in AC-60-14.

A pilot had a hinge on his head, and a CFI would take a pitcher of knowledge and pour it into the student, and then knowledge would come out of the students mouth. It was a comical example of how learning does not occur…. or perhaps better said, of how long term useful learning does not occur. Short term wise, the pitcher thing can sort of work.

There’s a whole subset of aviation instruction specifically built up to make money via leveraging short term memory for the written exam. Such courses have all the test questions, and they go through nearly every one over a 2-3 day period. While doing so, they also present memory tricks and such for a lot of questions which pretty much guarantees someone will pass the exam.

While cram and forget courses seem a questionable sort of practice, there is a need for them by some students. The complexity of the written exam is such that unless one is good with the paper side of things, and/or is a recent student, the written exam can be a barrier. More than a few times over the years I’ve a recommended a student who has difficulty to go to one of those weekend guaranteed to pass the exam courses over the years just to get it out of the way. The really important info we cover in flight and ground school, and does get checked again during the oral, just not as much in depth as the written. Consider how the exam would need to change if the test questions and answers were not published before hand?

And now the FAA is proposing adding more complexity to the oral and practical (and in time, no doubt the written)? Will it really make a difference in the safety domain? I sort of doubt it, but there will be money to be made in 141’s and a new industry of cram and forget oral and practical prep will come into being. Otoh, how many potential pilots will drop out due to increased complexity?

This is not to say Safety Management etc is a bad thing. The proverbial “A superior pilot is one who uses his superior judgment so as not to have to use his superior skill.” is all too true. The challenge is how do we get folks to the superior judgement arena?

Safety by example was promoted, ADM was promoted, FITS was promoted, a subset of airline CPM was promoted, and the accident rates don’t change very much. Peripheral matters like doubling the size of the PTS, and poking at means of evaluation at huge expense to our students is unlikely to make much of a dent in rates either.

 

 

Flight Reviews and Safety

The overall goal of a flight review is safety. Most assuredly such was in the minds of the legislators when 61.56 was written. The thing is, if one just looks to the regs, things dont work out too well, and in some cases, they can be counterproductive.

Safe operations are not guaranteed by the minimum 1 hour of dual and 1 hour of ground every couple years. Rather safety is a function of pilot skill and judgment on each and every flight. The flight review is an opportunity for a pilot to gain greater awareness as to how they might handle future flight ops safely. Most assuredly, even the best CFI in the world is not going to uncover every area of weakness, or hazardous attitude in a couple hours, or even 10 hours for that matter.

If a flight review was a CFI personally guaranteed signoff that one was good to go for the next two years, flight reviews would become very long, and very evil in short order. We’ve all heard of or experienced the checkride from a warm spot near the center of the earth. Ie, where in the pilot being evaluated is put in the pressure cooker, and is exposed to every possible emergency ever, all occurring at the same time, all the while being grilled on everything in the PTS, and every distraction known to man kind is then piled on top for good measure.

Granted, there are times and places where such may be appropriate. While a prep course for a 709 ride as a confidence builder comes to mind, such an approach to a flight review is counter productive. The reason being, is in almost any other case, going too hard core becomes a motivation killer, which then often leads to less recurrent training, which then leads to a less than safe pilot. The same can be said when it comes to sending students down rabbit trails for the part 91 portion of a flight review.

By the same token, an untailored 1 hour ground, 1 hour flight canned approach to a flight review is neither adequate nor appropriate. Sure, it might pick up a weak area here or there… but its like most anything. If you dont have a clue where a problem lies before you get started, its like searching for a needle in a haystack. You likely wont find it, and even if you do, it could well be the wrong one.

The FAA preaches tailoring the flight review based upon an initial interview. Likewise, they suggest the time frames be what they may. Ie, short of an already very proficient pilot, the 1 hour ground/1 hour airwork aspect is too short a time slot. Some pilots on the other hand, often view any attempt to go beyond the minimums as a way to bilk them out of money and a total waste of time.

In the past, when one determined such on the initial interview, it was common practice to advocate for the Wings programs, rather than a flight review. Today, being Wings is proficiency based, rather than time, it still may be more aligned with a students expectations than a flight review, but it is critical to be upfront in that regard. There will always be some pilots who will squak, no matter how much you preach safety or the benefits of training to a given level of proficiency.

Shallow Fog METAR MIFG

Shallow fog is defined by the AMS as the following. “In weather-observing terminology, low-lying fog that does not obstruct horizontal visibility at a level 2 m (6 ft) or more above the surface of the earth. This is, almost invariably, a form of radiation fog.” It is encoded in a METAR via the symbols MIFG. Some great photos of it are available at Everything Looks Better from Above.

One a positive note, it often burns off pretty fast during the day. On a negative note, it can form incredibly fast at night… and worse, it can be really hard to identify at night. In some cases, it may not be apparent until you are in the landing flare, when all of a sudden, visibility goes from near perfect to almost zero, at the worst possible time. In other cases, it can change from shallow fog, to a thick fog bank in a flash.

On another negative note… automated instrument systems often due not detect it, and worse, it usually occurs either right when ATC is going home, or shortly there after. Its interesting to note AMOFSG places it in group D, where sensors are considered difficult, expensive, unproven, or unreliable to utilize. Apart from the technical aspect, their is a political and economic one as well when it comes to sensors. Airliners, corp jets, and many multis engine a/c do not suffer the same adverse affects from MIFG due to their elevated cockpits… ie eye level is almost always above 2 meters, where as the pilots of many ASEL craft are stuck right in the soup. Ultimately, the lack of automation and sensor capabilities, means that if a METAR includes MIFG… it was observed by a human on the field.

Another fun part about MIFG, is that its not very predictable, rather its an issue of probability. Ie, if the dewpoint/temperature spread is wide, its very windy, and the sky is overcast, it is unlikely that any type of radiation fog will form, much less shallow fog. On the other hand, if winds are below 5 knots, temperature/dewpoint spread is under 6 deg F, the sky is clear, and the sun has recently set… it is more likely fog will form, and in many cases, a thick layer of radiation fog is more likely than shallow fog.

To add further insult to injury, shallow fog often randomly forms in agricultural microclimates surrounding a rural airports far from any weather observers. Some crops have much greater moisture retention than others, some have significantly greater total emissivity than others. The net effect is that some airports experienced drastically different radiative cooling properties and others much less so. While such make an accurate prediction of shallow fog difficult, it also meant that if airport A was socked in, more than likely a nearby airport B is clear.

It should also be noted that some airport industrial parks also lend themselves to localized shallow fog development more so than others. Namely decreased dewpoint/temperature spreads due to cooling towers and the like, combined with large surface areas of high emissivity (vacant bituminous parking lots) sets the stage for its development.

With that much gloom and doom… whats a pilot to do?

First, if a METAR indicates MIFG at night, ie it was observed by a human at the field, landing at such a location in a small aircraft is likely to result in a go-around followed by proceeding to ones alternate. In fact, if temperature/dewpoint spread is that narrow, and the winds are light, the dangers of radiation fog forming are very real… and while an approach could be attempted with most MIFG occurrences, a thick fog bank would mean heading direct to ones alternate.

As one who has encountered MIFG at night many a time, the following was my basic procedure when conditions made its formation likely.

Check in with unicom, or any pilots in the area to see what the conditions currently are. Remember, AWOS does not detect it. Granted, often times it forms after everyone has left for the day.

Overfly the field above pattern altitude and look carefully at the runway lights… it they are not bright and clear, it is likely MIFG is present. Then again, I have had the fun of things changing from when I overflew the field to when I was in the landing flare.

If the landing lights appear even the slightest bit obscured or fuzzy, making a low pass is in order. First, such will serve to scare most animals off the runway… and secondly, the proximity to the runway, in addition to the landing light will make its detection easier.

Always assume if conditions are right, MIFG will form, and your visibility just before, or in the flare may go to near zero. Always, always be prepared for a go around and followed by proceding direct to ones alternate.

Ensure the runway lights are on full bright on base… you do not want to be in the flare, encouter MIFG, and at the same time have them shut down or return to low intensity.

The landing light, and strobes should be turned off on short final… there is nothing more disturbing than having the entire cockpit light up in flashing white lights, all the while you are trying to reconfigure the a/c for a go-around.

Parental Responsibility Lap Children

I came across the following article where the NTSB and FAA are at odds over the regs which provide for children under 2 years old to fly on their parents lap. It was interesting to note the comments… pretty much the same old, same old, parents arguing that they didn’t want to pay, and business travelers saying good riddance. One comment sort of stood out though. A parent said, I am responsible for my child, she doesn’t run up and down the aisles, she doesn’t throw tantrums… she is just fine in my lap.

Likely said parent, and the parents so gung ho about lap children have never experienced moderate/severe turbulence, much less CAT (clear air turbulence). I dont care how responsible you are, or how strong you are… if you experience major turbulence, you will not be able to hold unto your child… or in the worst case scenario, in a crash, you wont be able to hold onto your child. Granted, CAT is rare, accidents are rare… and if it was just your child who might be seriously injured or killed, I could sort of see the responsibility aspect. On the other hand, if your child becomes a missile and kills someone else as well as dieing themselves… parental responsibility doesnt go very far.

It was also interesting to note the availability of child harnesses such that a under 2yr old can be secured to the parents safety belt. Which such most certainly reduces the issue of a child becoming a missile… I dont know how well such would work in severe turbulence if the child was on a tether, rather than being tightly secured to the parent.

Tail Stall Detection and Recovery

My FIRC this time around has a section on tail stalls, and being the information is somewhat rare, I thought it appropriate to cover here. Bare in mind, this is generic information for airplanes under 12,500 lbs, the actual flight manual or pilots operating handbook, if updated is always the best source.

Tailplane Stall Detection:

Tailplane stalls most often, but not always occur when flaps are being extended, or power is being added. One should also note, that if wings are picking up ice, the tail is likely to do so as well, in fact perhaps 2-3X as much as the wing… and sadly, its not visible to the pilot. In any case, if the potential exists for tail icing, one should be hand flying, rather than relying on the autopilot. Yoke forces are key.

Symptoms include

  • Lightening loads.
  • Difficulty trimming.
  • Pilot-induced (pitch) oscillations.
  • Buffets in yoke, not airframe. (The yoke pulls forward, sometimes smashing to the stop and can’t be pulled back; forces of more than 100 lbs. can occur.)
  • Very sudden pitch-down, which can be unrecoverable on approach.

Do note, that stall warning systems can also end up iced over, and as such may not function.

Tailplane Stall Recovery

  • If you just added flaps, and symptons of a tail stall occur, retract the flaps.
  • If the yoke slams forward, it may be necessary to apply significant force to pull it back. In either case, as contrasted with a wing stall, the recovery action is the reverse, ie pull back.

In General

  • pull back
  • reduce power
  • reduce flaps

Being tail stall recovery is the reverse of everything one is taught, its imperative to that one detects the yoke forces, rather than the airframe going mushy.

Additional Resources:

Actual Tail Stall Event
http://aircrafticing.grc.nasa.gov/courses/inflight_icing/related/3_2_3f_RI.html

Useful videos and data from Nasa
http://aircrafticing.grc.nasa.gov/resources/related.html

AC No: 23.143-1 “Ice Contaminated Tailplane Stall”

Aircraft Accidents where tail stall occurred

http://www.flightsafety.org/fsd/fsd_jun-sep97.pdf

10mar64 DC4 Slick Airways,.Boston-“Loss of balancing forces on stabalizer, due ice accretion,causing aircraft to pitch nose down too low to effect recovery…

15jan’77 Viscount,Skyline sweden,”ice on leading edge of Stabalizer causing flow seperation and Stabalizer…

06Apr58 Viscount,Capital airlines,Michigan,Undetected ice accretion on stabalizer,in conjunction with specific speed and configuration,caused loss of pitch control…

Airline regulation, the glory days… sort of

Its in the airlines best interest to make money…. and as long as the probability of an accident has not changed appreciably from quarter to quarter, they will continue to reduce costs. Note, the key word appreciably… a 0.1% reduction in safety every quarter does build over time, even if its not noticeable on a quarterly report, much less in practice.

In the 70’s, service was what sold, regulated ticket prices were multifold higher than what they are today. However, the process controls and such to improve safety were still in development. Instead, the airlines relied on highly skilled people in nearly all areas of their operation. And despite that, the accident rates were much higher than today. Case in point, the crew coordination on theEastern Air Lines 401 crash. A highly experienced crew, albeit new in type, focused all their attention on a landing gear annunciator, while their plane slowly descended into a swamp.

747 Piano Bar
747 Piano Bar

 

Today, with a high degree of automation, process controls in design, manufacturing, operations, and service insuring high reliability, the high levels of skill needed in years past, as long as nothing goes wrong, is not really needed, at least not all the time. There in lies the problem, as long as nothing goes wrong. And the problem is exacerbated, by the temptation for the airlines, the government, and the employees, to rely more and more on process, and less so on skill, in order to minimize labor costs, and thereby maximize the revenue stream.

Over the next few blog entries, I’m going to take a look at these different groups, and changes over the years which have created potentially more broken links which can lead to an accident, despite vast improvements in process control. These issues are prevalent, not only in the airline world, but also in general aviation. They are systemic problems, and sadly short of returning to the days of regulation, are unlikely to be solved. The feedback provided by the free market system, while it can work, has such huge time lags, far too many people would die, or be seriously injured, before the constraints of the free market system could kick in, if they would at all due to secondary consequences.

Skill vs Judgement

If I look back to the day I got my private pilots certificate over 20 years ago, I don’t think I was ever as technically proficient. Sure, I couldn’t do the commercial manuevers, or shoot an ILS to ATP standards, but my precision, and finesse in take offs and landings was never better. It was the same with straight and level. I could hold +/-10 feet for extended periods, and it was rare I would deviate much more than +/-50 feet. Of course at the time, I was shooting about 50 landings/week, and my goal was formation flying, so I worked for precision. My judgement at that time, well thats another story.

I often wondered why private and commercial pilots had to jump hoops in order to fly a specific aircraft with insurance whereas the CFI would only need a cursory checkout to teach in such a plane. This excludes of course multi-engine aircraft where we need 5 hours in type. I think its the judgement issue.

Continue reading Skill vs Judgement