Archive for the ‘Safety’ Category

What you hear may NOT be what you get

Wednesday, December 9th, 2015
RNAV 21 HXD Mins

RNAV Runway 21 HXD Mins

I’ve learned the hard way that weather forecasts sometimes bear about as much resemblance to reality as political promises before an election. (Conversely, forecast are sometimes true—you can be the judge on the political winds of fortune.) But pilots, like voters, want to believe what we’re told even when indicators are clearly pointed toward a less favorable outcome. Then it’s time for some of that decision-making stuff.

Earlier this fall, I had planned to visit a friend in Hilton Head (KHXD) for lunch. The non-stop flight normally takes about 25 minutes. A lingering stationary front had sort of, kind of, moved off the coast and the forecast was for decreasing clouds and improving visibility. At the departure point (KLRO) near Charleston, South Carolina, the ceiling was 800 broken with unlimited visibility and plenty of sunshine above a thin cloud deck. Good VFR was expected by noon, essentially the same for KHXD.

I launched IFR and had a delightful trip down the coast toward Hilton Head at 4,000 feet. Broken to overcast tops were running about 1,500 feet. The automated weather at KHXD was also reporting 800 broken and 8 miles visibility. That all-important temperature­-dew point spread was within two degrees, but this looked like an easy RNAV approach. With Sol doing the heavy lifting, the deck might even burn off to the point that the approach wouldn’t be loggable for currency purposes.

Switched over to KHXD tower where the controller updated the weather to 1,200 scattered and better than 10 miles. EZ day! Yet the clouds seemed rather solid over the final approach fix, but I fully expected to break out. Just any time now— even though it appeared to be a bit lower than what was just advertised. You can see where this is headed. Somewhere, between the final fix and MDA, the controller mentioned that the clouds looked somewhat thicker over the approach end of the runway, but the airport was wide open. Hmmm…no time to think about what might have, should have been, etc. as the aircraft was now well below 800 feet and only a few hundred feet above minimums.

A slightly delayed breakout

A slightly delayed breakout

In daylight conditions one of the precursors to an impending breakout is that it gets darker beneath the aircraft since ground doesn’t reflect sunlight nearly as well as cloud. None of that was happening today. Solid. Minimums. Drat—time to miss, and no cheating. Written about that too many times. Power, pitch, flaps, gear, and start climbing. Advise ATC that we’re going missed.

Just as the controller gave the missed instructions, the aircraft flew into brilliant sunshine with a perfect view of the airport. Still in alignment with the runway, but a little high. Decide quickly.

Put the gear down again—kind of important—full flaps, and reduced power to settle back into the slot. Advised the tower of my intention to land but apparently he didn’t hear and asked if I’d heard the miss instructions just as we touched down. “We’re rolling out now.” He chuckled, “Welcome to Hilton Head— guess that fog bank up to the north end is pretty low.”

A few take-aways:

  • In retrospect, it’s logical: The ASOS was located midfield on the sunny part of the airport; and because it reports only what is happening right there, weather in the landing zone—a mile or two away—could be much different. Caveat Emptor!
  • Had the minimums on the approach been lower (462 feet, agl) there might not have been time to reconfigure.
  • In a faster and less maneuverable aircraft than a Bonanza, this would have been ill-considered.
  • There was more than adequate runway, and it’s really good form to touch down in the first third of the runway—on short fields it’s essential! This was an easy first third touchdown.
  • It’s so easy to rationalize our behavior when the outcome is what we hoped for.

I am constantly reminded—as one who sits on high writing about this wonderful activity of ours and having lots of time to second-guess many who made headlines—that sometimes we’re good, sometimes we’re lucky, and sometimes it’s both. This was a perfect judgment-training scenario, and the “book” answer would have been to complete the miss. This is a plausible scenario for a gear-up or possible overshoot in a fast aircraft. The book recommendation would have been to go to the alternate unless it was a transient condition or you didn’t get to minimums. (In the latter case, that’s a mandate for more training and practice.) The approach end fog burned off about 10 to 20 minutes later.

Reality, however, is that judgment training and evaluation is much more complex than we, in the sanctity of hindsight bias, will often admit. Regulators, instructors, attorneys, accident investigators, and safety writers just might fall into that mindset occasionally, and the hypothetical arguments are endless. Would love to hear some of your real-world judgment moments where it wasn’t quite as cut and dried as some of our political candidates would have us believe.

Big Watches, Fuel Gauges, and Annunciator Lights

Tuesday, November 24th, 2015

The notion that engines run on air is no more absurd than attempting flight with insufficient fuel in a tank connected to the engine. Yet hardly a week goes by that somewhere a pilot is attempting to go where none have gone before.

The technology of measuring fuel has gotten much better than the wire and cork device used on the Piper Cub. The Cub’s design was stone simple—mount the fuel tank directly in front of the pilot (not such a great idea from a crashworthiness perspective, but only a short hose run to the engine) and attach a cork to the end of a wire rod. The length of rod protruding through the fuel cap indicated how much go-juice remained. Only two things could go wrong: The cork could become saturated and sink, giving the pilot a sinking feeling that no fuel remained, or worse, the rod could become slightly bent and jam, falsely indicating more fuel on board.

We paid little attention to any of that and always used our big pilot watches to determine how long to stay aloft. Time is one of the best indications of what’s left in the tank(s), provided you know what was on board to begin with and are familiar with the engine’s burn rate—both highly recommended procedures.

New pilots are often told that the fuel gauges on light aircraft are unreliable and that the FAA only requires them to be accurate in one condition—empty. Let’s dig into that a bit more. As usual, there seems to be some gray!

Under CAR 3.672 Fuel Quantity Indicator, the Civil Air Regulations by which most legacy aircraft are certificated, “Means shall be provided to indicate to the flight personnel the quantity of fuel in each tank during flight…Fuel quantity indicators shall be calibrated to read zero during level flight when the quantity of fuel remaining in the tank is equal to the unusable fuel supply as defined by § 3.437.” [ital. added]

Under the new regs, FAR 23.1337 Powerplant Instruments Installations—(b) Fuel quantity indication, “There must be a means to indicate to the flightcrew members the quantity of usable fuel in each tank during flight. [ital. added] An indicator calibrated in appropriate units and clearly marked to indicate those units must be used. In addition: (1) Each fuel quantity indicator must be calibrated to read ‘zero’ during level flight when the quantity of fuel remaining in the tank is equal to the unusable fuel supply determined under § 23.959(a)…” It’s not appreciably different from the earlier rule.

It could be reasonably argued that quantity indications should be accurate but sometimes the hardware is suspect. The FAA goes on to say in a safety publication, Time in Your Tanks—P-8740-03, “Fuel gauges are subject to malfunctions and errors. Therefore, unless restricted by the gross weight or center of gravity limits, it is considered good judgment to ‘top off’ the tanks at fuel stops. If the fuel load must be limited, you should endeavor to get an accurate measurement of fuel quantity by using a dipstick calibrated for the aircraft.” Dipsticks are a good measure…IF…they are properly calibrated…AND…the aircraft is sitting level, which allows Murphy’s law to slip in.

With the “rebirth” of GA in the ’90s, most manufacturers added low-fuel annunciator lights to new production aircraft (a.k.a. fuel “idiot lights”—they have been installed on most automobiles and larger aircraft for decades) separate from the gauging system to make it really hard to run the beast out of fuel or a tank dry. In general, it’s worked very well although one young CFI and his student managed to run a new production Cessna, equipped with lights, out of fuel. No need to belabor the point where I’m sure both participants will do better in the future. The fuel accident numbers of new tech aircraft compared to the classics is much better and it’s not because pilots suddenly got smarter.

But suppose you own a classic, and installing annunciator lights is prohibitively expensive. A highly recommended upgrade, second only to installing shoulder harnesses, is to install a fuel flow transducer/totalizer which will accurately measure fuel flow to within a few tenths of a gallon. As for quantity, it must be properly programmed, so it’s not quite as goof-proof as the low fuel lights. Used on every flight, it provides great peace of mind—either more fuel is required or it isn’t. As far as not running a tank dry, a timer or that stopwatch function on your chronograph or some other clock on board should help.

Back to the FARs—Do the gauges need to be accurate or not? In my view, it’s irrelevant. The Air Safety Institute’s “Golden Hour” approach of always landing with one hour of fuel on board elegantly solves the problem.


The science of building/maintaining aircraft

Wednesday, November 11th, 2015

young mechanic (1)

Building and maintaining aircraft is mostly science and perhaps a bit of art. There should be a keen awareness on the tradeoff made between affordability, suitability, and safety. If it’s a homebuilt and you built it, than who knows better than you on how to maintain it? That’s somewhat debatable and while the amateur-built regulations give us plenty of rope with which to hang ourselves, there are some best practices that should be taken as gospel. Many have been learned the hard way.

A few observations:

Hot Seat—Understanding amperage, circuit breakers, and good wiring technique is more than just a good idea. A homebuilt flying the Everglades some years ago in visual conditions asked ATC if they were aware of any fires in the area. ATC knew of nothing, but the pilots soon discovered that the automotive electric seats they had installed had turned into hot seats. Unfortunately, the plunge into the swamp was fatal and the NTSB noted that a circuit breaker supposedly protecting wiring and accessories from a short circuit, didn’t. We can discuss the wisdom of electric seats in homebuilts but wiring practices are really non-negotiable. It was high price to pay for a bit of ignorance.

Hose Clamps—Another fatal accident occurred when a mechanic decided to use an automotive hose clamp to secure a muffler and exhaust pipe on a Piper Cherokee, resulting in an in-flight fire. The aircraft version, while about 15 times more expensive, had a critical detail that the mechanic apparently overlooked or thought unimportant. There is a pin on the clamp, which is inserted into a hole in the muffler and exhaust pipe and holds the two together even if the clamp loosens. Some parts of aircraft really do need to be expensive and well-made.

Plumbing—A recent honest mistake apparently involved the builder of an RV-10 who suffered an engine stoppage due to fuel starvation. There is currently a $35 million lawsuit pending against Vans Aircraft and the manufacturer of the fuel flow transducer that the NTSB reported was plugged by an improperly-used sealant in the system. There was loss of life so the legal system is involved. Sometimes that’s warranted but this has little, if anything, to do with Vans in my view.

There is a long thread on Vans Air Force that looks at the pros and cons. One of the writers thoughtfully points out that there is perhaps not enough guidance on what and how one should specifically assemble various parts and pieces. In my two earlier examples, it seems the guidance was pretty clear and the crux of this suit may hinge on what the builder knew and when he knew it. How much guidance should a manufacturer for a kit or a component provide, and does that incur additional liability? Is it their responsibility to educate builders on everything that a certificated mechanic should know? Guess a jury will sort that one out.

One thread respondent points out that in too many cases there is “tribal wisdom” or the excuse that if the Framistan hasn’t failed in 350 hours it must be done right. Perhaps, but there are some notable examples of certificated aircraft where a rigorous process didn’t uncover a particular fault until years or decades later. Is that evidence of negligence?

I’ve had the honor of meeting with Dick VanGrunsven and you won’t find anyone more concerned about design, safety, pilot proficiency, judgment, ease of kit assembly, and the use of proven methods. There are thousands of Vans kits flying and he’s become one of the most successful manufacturers in the world.

But building and maintaining an aircraft is still a daunting task. The certificated manufacturers have to regularly defend against a variety of suits—most of which are groundless but very expensive. The “gotcha approach” once a suit is filed doesn’t always serve the industry or buyers of aircraft particularly well. An admission that something could be improved, such as more guidance or improving the hardware, is often used as evidence of wrongdoing when that isn’t usually the case.

This trial will be closely watched from many quarters. In the interim let’s insure that our machines are so well built (if we built them) and maintained that we’re willing to put our families into them. Don’t forget that flying part either.