Category: Flying

Here are some thoughts after flying a factory-new $700,000 Robinson R44 Raven II from Los Angeles to Washington, D.C. Our machine is the gold one in the foreground on the ramp at KTOA:

(One of the teachers in the safety course, regarding a photo of the R22, R44, and R66 lined up, said “They’re not going to win any beauty contests.”)

The $25,000 air conditioning is a great luxury and worth the 33 lb. payload penalty. The lack of a sliding window for taking pictures is a continued disappointment. If you want a good aerial photo you need to land, remove a door, take off again, take the photo, land, put the door back on, etc. Bell and Airbus manage to include simple mechanisms for temporarily obtaining an unobstructed camera position. Why can’t Robinson do this?

The $57,000 autopilot should have been a game-changer and it kind of was, but usually not in a good way. I have trained some Blackhawk pilots who were transitioning down to the R44 and they struggled initially because they always flew the Blackhawk with the stability augmentation system (SAS) enabled. The Genesys autopilot’s core mode is SAS, in which it seems to return the helicopter to whatever attitude was preset. The system is hyperactive, however. If there are small attitude changes from extremely light turbulence or just the vibration of the helicopter, the autopilot will fight those small changes. A human pilot would do nothing and assume that the attitude bumps will average out. In straight and level flight, the SAS system is applying a control input literally every second. It is like having the world’s most nervous copilot on the controls with you. It probably should try to figure out if a human pilot is on the controls and, if so, do nothing until there has been at least a 2-degree disturbance in attitude. We were able to get the autopilot to fly a NAV course, go down a glide path on an LPV approach, etc. It more or less works just like an autopilot in an airplane, but because the helicopter shakes so much more it is not as confidence-inspiring. Is it worth $57,000 (the pre-Biden price of an airworthy certified IFR-capable 4-seat airplane)? Maybe! Despite the annoyance factor, I think it is worth enabling SAS for night flights and any time visibility is reduced or there is a chance of inadvertent instrument meteorological conditions (i.e., going into a cloud).

Robinson has limped into the glass cockpit era, but not gracefully. The Garmin G500 is a primary flight display, but not an integrated flight deck like the G1000 (see the Bell 505). So the pilot needs to look at the warning lights above the PFD to see if anything dramatic is wrong with the machine. Although the helicopter is certified for visual flying only and there is nothing on the PFD that is necessary for continued flight and landing, the FAA requires that Robinson install a full set of backup steam gauges. These are much larger and easier to read than what’s on the Garmin glass (the vertical speed indicator, for example, is at least 5X the size).

The G500 does not display a lot of the information that you’d expect it would. For example, it shows the outside air temperature and the aircraft’s altitude. Even without access to the manifold pressure sensor, it could look up in a table to find out what the current maximum continuous manifold pressure should be (corresponds to a horsepower limit) and display that to the pilot. It becomes the pilot’s job to use a table on the checklist or in fine print on the cyclic. Robinson continues to have separate fuel gauges for the main and aux tanks so it becomes the pilot’s job to do a bit of arithmetic and calculate flight time remaining.

Our second radio was a Garmin 225B. It was entirely unreadable/unusable at night due to being so much dimmer than the G500H and the GTN 750 (the NAV/COM1).

I arrived in Los Angeles for the Robinson Helicopter Company’s safety course on November 13. The engineer who founded the company, Frank Robinson, died the day before at age 92 (AVweb). Frank democratized helicopter flying with his reasonably priced machines that were simple to operate and maintain. I had a few conversations with him over the years and he was always generous in sharing his time and direct in sharing his point of view.

Frank’s son Kurt Robinson now runs the company and he welcomed us to the class: “We want you to know what we know.” He stressed how important the safety course was to his father, a point later backed up by Bob Muse, a legendary LA helicopter pilot and teacher. “Frank missed fewer than 5 safety courses over the years,” Bob noted. “He would reschedule vacations and business meetings so that he could welcome every class of pilots. Frank loved aviation and the company. It was never about money.”

Bell and Airbus (“Eurocopter”) pilots enjoy heaping scorn on Robinsons, but we learned from Bob and Tim Tucker about how the pilots and engineers at all of these companies cooperate and fly each other’s machines. Robinson has been a leader in some safety areas, e.g., crash-resistant fuel bladders and standardizing on the Vuichard technique for recovery from vortex ring state (most likely encountered during a downwind steep approach to an off-airport landing zone).

Some of Bob’s points:

• we overemphasize autorotations in training; it is rare to see accidents that are caused by something that would require an auto
• seek recurrent training every 6 months, which is what the most experienced pilots will get
• look at NTSB Safety Alerts
• take phones away from mechanics; interruption by phone call is a common reason for a procedure step to be skipped

The previous generation’s aircraft mechanics have been retiring and are being replaced by younger less intelligent less conscientious Americans. Maintenance-related crashes are nearly twice as high a percentage of the total (still less than 10%, however) compared to 10-20 years ago. The biggest causes of Robinson accidents are wire strikes and weather, each contributing roughly 30 percent.

A huge number of safety-related initiatives and FAA regulation updates that would improve safety have been delayed by two years or more due to coronapanic. Everything that was on track to be approved in 2020 is still pending. One big change would be to revoke SFAR 73 and update the Robinson POH to add similar requirements, e.g., 20 hours dual before going solo, to the limitations. The requirements would then apply to international customers as well (currently about 80 percent of Robinson’s production is exported).

The latest Robinsons all come with dome light cameras (see Time for a robot assistant up in the dome light of the cockpit? for what I think it should do, but of course it doesn’t!). These have been very helpful in investigating accidents. (As with seemingly everything else in aviation, it was already obsolete when installed. The limit on memory card size is 128 GB, which is good for 10-15 hours. There is also an internal 16-hour memory that the pilots can’t access and that is recorded to even when the camera and audio are switched “off”) The dome light camera also provides some interesting cautionary videos. In one video a Bell 407 pilot, who previously did a lot of flying in Robinsons, is getting current again in the R44. She pulls the mixture, thus shutting off the engine, instead of the carb heat. She then immediately pushes the mixture back down, but the engine quits anyway. If they’d crashed, it would have been due to an “unexplained power loss.” (As it happens, the instructor in the left seat pushed the collective down and did a nice autorotation to the side of a railroad track. We then see him frantically pulling on the rotor brake. It turns out that a train was coming!)

My favorite video featured Julie Link on a sightseeing tour in Hawaii. The R44’s engine quits (the mounting block for the magnetos failed; apparently they’re both on the same piece of metal) and she does an autorotation to a field with two tourists in the back who don’t seem to be aware that things have become perilous. After they land, we hear her say “The engine stopped. It happens. It happens to me a lot.” (She previously did a heroic autorotation in an R22 to a street in Honolulu (Daily Mail).)

Once established in an auto, if the low RPM (97%) horn goes on, I like to take out half of the collective check that is in. Bob says to push it all the way down so as to build that reflex and then pull it back up slightly after the RPM is back to 100 percent.

Bob recommended watching a U.S. Army video on mast bumping.

Robinson now offers a polycarbonate windshield that will ruin a bird’s day, but not yours. We talked to a guy who owns 13 R66 (the turbine-powered Robinson) helicopters and he said that the view is distorted (he also said that he’s had disabling engine problems with 4 out of 13 Rolls-Royce turboshaft engines!). Robinson says that they scratch just as easily as the standard acrylic windshields, but the scratches cannot be repaired.

Those are some of the things that Frank Robinson might have wanted you to know! It is sad that he is gone, but he did pack a lot of achievement into his 92 years. He is the only person in world history who built a sustainable piston helicopter business.

After the class, I joined a former student from MIT who was picking up his new R44 Raven II helicopter. We flew from Los Angeles to Washington, D.C. together and the only squawk was that the left front door became a little tougher to latch after about 15 hours (we picked it up with 4 hours on the collective Hobbs). It is tough to think of another aircraft manufacturer that delivers this kind of quality, especially down in the piston ghetto. I think it is reasonable to say that the more than 13,000 helicopters Robinson has built are Frank Robinson’s true memorial.

Friends in the D.C. area have been texting me tonight regarding a Mooney that landed in some powerlines near KGAI. From CBS:

A small plane crashed into power lines in Maryland, leaving two people dangling about 100 feet in the air, officials said. The crash knocked out power to thousands of people in the area.

A single-engine Mooney M20J crashed into wires near Montgomery County Airpark in Gaithersburg, Maryland, around 5:40 p.m. local time Sunday, according to the Federal Aviation Administration. The plane appeared to be intact as it was caught in a web of power lines about 100 feet in the air.

As of roughly 7:50 p.m., Montgomery County Fire and Rescue Service spokesperson Pete Piringer said the people aboard the plane were uninured. Earlier he said that they remained in a “very precarious situation,” as the wires were still energized. Utility company Pepco said that as of around 8:30 p.m. the lines had been deenergized.

How could this happen? Mooney pilots are usually pretty good! The plane departed from Westchester County, NY, according to ABC. So the task at Gaithersburg was an attempt to land. Where are these powerlines? Let’s check The Google:

The only tall high-voltage lines that I can see are about a mile from the runway, top left of the above image, rather near Giant Food. That’s quite a distance from the runway to be so low to the ground.

Let’s check the weather. 5:40 pm is 17:40 EST then add 5 hours and we get 2240Z. Here’s are the METARS:

KGAI 272156Z AUTO VRB04KT 1 1/4SM BR OVC002 11/11 A2945 RMK AO2 SLP987 T01060106
KGAI 272128Z AUTO 18003KT 1 1/4SM BR OVC002 11/11 A2945 RMK AO2
KGAI 272121Z AUTO 14006KT 3/4SM BR OVC002 10/10 A2945 RMK AO2
KGAI 272110Z AUTO 14007KT 1 1/4SM BR OVC002 11/10 A2946 RMK AO2
KGAI 272056Z AUTO 15004KT 2SM BR OVC002 11/10 A2947 RMK AO2 PRESFR SLP995 60003 T01060100 56050

Why isn’t there a METAR closer to 2240Z? It seems that the 2256Z observation would have come after the massive power failure. Here’s an observation from nearby Dulles Airport:

KIAD 272252Z 21011KT 1 1/2SM BR OVC003 11/11 A2945 RMK AO2 SFC VIS 9 SLP971 T01110106

Could part of the problem be that the altimeter in the Mooney wasn’t set correctly and, therefore, the instruments were showing the plane to be higher than it was? This is an irrelevant factor for an approach with a glide path, which KGAI has. They would have gotten fresh altimeter settings during the entire flight from Air Traffic Control. The departure airport, KHPN, had an altimeter setting of 29.64″ about one hour before accident. So if they’d taken off from HPN and never touched the altimeter they would have been (29.64-29.45)*1000 = 190′ lower than indicated. This kind of mistake resulted in an accident at KBDL to American Airlines 1572 (The NTSB report says that they were on a VOR 15 non-precision (no glideslope) approach.)

The weather was not improving dramatically after 2156Z (4:56 pm). Is it reasonable to land with only 1.25 miles of visibility and 200′ overcast? If you’re a two-pilot crew going into Dulles Airport with its idiot-proof runways and powerful approach lighting system that will cut through the clouds and mist… yes. In fact, the minimums for the ILS 1R at KIAD are 200′ ceiling and 1800′ of visibility.

What about at KGAI? Assuming that this airplane was trying to land, they’d be using the RNAV 14. With the latest and greatest WAAS-capable avionics, the weather minimums are 269′ and 1 mile of visibility. Touchdown zone elevation is 520:

So this would have been an attempt to do the an approach that the weather report suggested could not be accomplished legally. This is actually legal for Part 91 (private) flying, but is not permitted for airlines (Part 121).

On the third hand, had the pilot flown the approach correctly, it should have resulted in a safe landing on the runway even with weather below minimums. Based on my experience, the virtual glide path created by the WAAS GPS box is valid all the way to the ground. (And this was part of our airline training at a Delta subsidiary for ILS approaches; the glide slope (projected via radio waves from the ground) could be used as a reference even after breaking out from the clouds and being within 50′ or 100′ from the runway.) In fact, an autopilot could take the plane down all the way to the runway (would be a rough landing without a flare, but nothing that would be hazardous to the people inside the plane).

So the crash remains something of a mystery at this point. I would want to know if the Mooney (potentially ancient) had a modern WAAS-capable GPS for the RNAV 14 approach. If not, the weather minimum is 400′ ceiling and 1.5 miles of visibility and the challenge for the pilot is substantially greater.

This would be perfect for a Godzilla movie. Godzilla loves powerlines and he is often well-disposed to humanity so he could simply pull the airplane out of the powerlines and set it down on the runway.

Update: I found some relevant Air Traffic Control audio at LiveATC.net. Controllers often speak on multiple frequencies and I found half of a conversation on 128.7 in the archive block starting at 2200Z on 11/27 (so the media reports of a crash at 5:40 pm are probably incorrect; it would have been around 5:28 pm). LiveATC.net captured Potomac Approach, but not the Mooney or other aircraft (so it is likely that the exchanges were happening on a different frequency associated with Potomac Approach; a merged recording of 126.75, 125.52, and 133.85 contains some fuzzy responses from the planes/pilots). The Mooney’s tail number seems to end with “1RF” (One Romeo Foxtrot). At 9 minutes in, the controller suggests (gently) that the 1RF has not followed a previous instruction correctly. At 14:45, the controller says “if you’re able to land at Gaithersburg you can report cancellation…” (i.e., it was foreseen that the weather could be below minimums). At 15:55, a Pilatus PC-12 (“Kronos”) is also going to GAI. At 16:45, the controller mentions, probably to the Pilatus, that a Cheyenne went missed from GAI and was diverting to FDK. At 17:45, the controller acknowledges a communication from 1RF. At 18:30, the controller tells 1RF about traffic and tells 1RF to join the approach. At 20:15, 1RF is cleared for the approach (allowed to descend from previously assigned altitude). At 23:15, 1RF is told to switch to the advisory frequency at the untowered KGAI airport (“CTAF”; used for pilot-to-pilot communication). Just before 27:00 the controller uses the full call sign: N201RF, which is the registration for a 1977 Mooney. Unfortunately, this is to issue a low altitude alert. If indeed the Mooney had switched to the CTAF, the message would not have been heard. Just before 30:00, the controller says “it looks like they made it” regarding the Mooney, but tells the Pilatus that he is waiting to get a cancellation from the Mooney (can’t have two airplanes doing instrument approaches at the same untowered airport at the same time).

My best guess is that the conversation was on 126.75, but the receiver that feeds LiveATC.net is not well-positioned to hear low-altitude planes near KGAI.

Second Update: presidentpicker in the comments below gives us an overlay of the Flightaware ADS-B 3D track and the approach. The tracklog ends at 5:27:20 pm and sadly infers an arrival:

Pilot readers: Before we escaped the Land of Lockdown for the Florida Free State, KGAI was my most common destination airport for trips into D.C. My personal rule was that if the weather wasn’t great I would land at Dulles instead. “Not great” meant winds of more than about 20 knots unless straight down the runway, ceiling below 400′, or visibility of less than 3 miles. I would also go into Dulles if I were arriving after the KGAI FBO closed and I wanted a rental car. The fees at Jet Aviation Dulles or Signature for a single-engine piston aircraft were actually quite reasonable. It’s great to be a skilled pilot, but it is even better to set up a situation in which skills are not required. And there’s nothing wrong with doing an approach to minimums, especially if you’ve set up a two-pilot crew in your little Cirrus, but it doesn’t make sense when one of the world’s largest airports is only 5-10 minutes farther from your ground destination.

The Dulles photo above was taken just before landing at KGAI last week:

We departed Torrance, California in the Robinson R44 Raven II shown above and landed at Gaithersburg on November 21, 2022. There were a few stops in between…

Related:

• Virginia Democrats take over Dulles Airport

One of the long-held dreams of people in general aviation is that a car company would come in and fix all of our woes (high costs, low volumes, intensive maintenance requirements, dumb-as-bricks systems). If Honda, for example, could make an airplane that is as comfortable and reliable as a Honda Odyssey minivan, mass-produced at a reasonable price, life would be awesome.

Well, Honda actually did go into the airplane business! And it took them way longer to push the plane out the door than it would have taken Cessna or Embraer. And an operator of the first-generation plane at NBAA 2022 gave the airplane low marks. (I wrote a review of the plane in 2016.) The airplane is fueled from a single point in the tail, which requires a ladder, and can take nearly 30 minutes for a line guy (this desirable job working in the cold or heat is almost always done by those who identify as “men”) to fill. During this time there will be periodic overflows that will cover the line guy in Jet A. When finished, the plane was never able to hold the advertised maximum capacity. “We were always 100 lbs. light.” The lav is externally service, but in a non-standard way that results in some bad outcomes. “Ten percent of the Gen 1 airplanes went off the runway,” noted the operator. “They’ve maybe fixed that in the newer ones by limiting nosewheel travel depending on speed.”

The plane itself did not end up having way better specs or a lower price than the very light jet/light jet competition. Honda announced a variety of Gen 3 features at NBAA. There is an extra fuel tank under the tail, which increases the ability to accept fuel, extends range slightly. A light next to the fuel filler comes on when the massive overflow spray is imminent:

The cockpit is more or less unchanged. It is a clean Garmin G3000, with no overhead panel and a general lack of clutter:

What will be new in the cockpit are autothrottles and a big button for the amazing Garmin Autoland system.

Performance:

How many Bidies for this wonderful device? About $7 million, which sadly means that all of Honda’s manufacturing and engineering expertise aren’t doing anything to bring the price of new aircraft down (it’s cheaper than the Embraer Phenom 300, but not if you adjust for size, seats, range, etc. (and the per-hour operating cost may be similar); it is more expensive than the Embraer Phenom 100 and Cessna M2 jets and provides some additional performance and cabin size).

Related:

• “Jet It Pivots Away From HondaJet to Phenom 300” (Flying, 11/23/2022): Jet It’s CEO accused the OEM of poor service and said the company had absorbed more than $20 million in off-fleet expenses since 2020.