Friends have been asking me to explain the recent Robinson R44 autorotation off Miami Beach.
The incident, caught on surveillance camera:
If it isn’t an emergency and your machine happens to be equipped with fixed or pop-out floats and you’re practicing, it looks like the following video (throttle is rolled to idle to simulate engine failure and, due to a sprag clutch, the engine isn’t helping to maintain rotor speed).
Here’s one to a hard surface (cheating a little with a slide-reducing headwind that you can hear in the microphone):
Let’s assume that there was an engine failure in the Miami crash, which could be due to a mechanical problem, to running out of fuel, to someone pulling the mixture control inadvertently or turning off the magnetos (I always hate to see keychains on aircraft keys or, for that matter, ignition keys to begin with (jets don’t have them so you can’t turn off a jet with your knee)), etc. In that case, since we see that the rotor blades are spinning, the Miami pilot reacted correctly by lowering collective pitch and, probably, pulling back a little on the cyclic. This preserves rotor speed and enables the blades to windmill as the helicopter descends. The potential energy from being up in the air turns into a source of power to keep the blades turning, but that power can’t be used if the blades are at a steep angle of attack compared to the new relative wind (coming up from the ground).
The airspeed also looks pretty good. It is supposed to be 70 knots in an R44 (POH), but 60 knots is also sufficient for a reasonable flare and landing. What seems to have been missing in the Miami crash is the cyclic flare at the bottom. This maneuver, not that different from flaring a fixed-wing airplane on landing, turns the kinetic energy of the forward airspeed into a climb that cancels out the descent from the glide so that the net vertical speed is close to 0.
(At the end of the flare, if you want to get everything perfect and not damage the tail, you stick forward to level the skids and finally pull the collective to use the energy of blade rotation to cushion the fall from 5′ to the ground.)
Why wouldn’t the pilot flare? One thing that we tell people in training is to begin their flare at “treetop height”. This is tough to put into practice when there aren’t any good vertical references. Even experienced seaplane pilots have a tough time judging height above the water when the water is smooth (“glassy”). One can see from the top video, when witnesses are being interviewed, that there wasn’t a lot of wave action. Aside from the difficulty of judging height above smooth water there is, of course, the difference between training and the real world of surprise and shock that things aren’t going as planned.
Fortunately, nobody was killed in the Miami crash. Counterintuitively, the injuries might have been less severe if the helicopter had contacted pavement. That’s because the skids are designed to absorb much of the downward energy of a crash, but they can’t do this job when the machine smacks down in water. In order to meet FAA and EASA certification standards, the seats themselves also have to absorb downward energy by crushing and that, presumably, is what saved the occupants from being killed by the impact that we observe on the video.
It will be interesting to see what the NTSB can learn…
Related:
- “Teaching Autorotations” (some tips for CFIs)
- HeliTrak Collective Pull Down (reacts immediately even if the pilot doesn’t)
- Cirrus SR20 (you don’t have to be a hero because you can pull the parachute)
Thank you for taking the time to share your thoughts on this. I really appreciate the insight and often read these posts multiple times. Thank you!
Nice post Philip, thanks. Your estimate of 60 knots at the crash seemed high to me at first viewing. But I watched a video on Youtube using .25 speed the URL start feature where you can specify the starting point to the second (allegedly). 60 knots is ~100 feet/second, a Robinson R44 is 30 feet long, it actually seemed possible (to me) it was travelling ~ 3 lengths/second. Anyone have a decent method of estimating air speed from a Youtube video?
Thank you for a great post. The second video (with the Star Trek music lol) was a sweet landing. I didn’t know auto rotation could be so cool.
My takeaway from this, is that if everything I know about science is learned from movies like Contagion and Don’t Look Up, it means I don’t know much about science.
When I saw this on the news yesterday, the first thing I thought was what does Philip think about this? I only fly aircraft with wings that don’t move, but that didn’t look like a very good autorotation to me.
The trick with autorotating into water is undoing the harness & getting out before it sinks. The 2018 crash makes surviving all autorotations into water miraculous.
I’m not an aviator of any sort, but I was at the Grafenwöhr Army Training Center when a cease-fire was called.
One of our AH-1 Cobras had suffered a sudden loss of oil pressure. The pilot attempted RTB by turning and gaining altitude, but there was a huge bang and he lost all power. This was at approximately 75 knots and below 150 feet, so all he could do was yank the controls (sorry, that is the limit of my chopper knowledge) and try to slow the impact, but it was too little, too late, for autorotation.
Both pilot and CPG suffered spinal compression fractures. I believe both had to leave the service.
When the Cobra was hoisted out of the woods, there was one of the infamous “Graf pigs” (wild boar) underneath, killed on impact.
Kevin: 75 knots and 150′ is actually within the “safe” zone for every helicopter https://en.wikipedia.org/wiki/Helicopter_height%E2%80%93velocity_diagram that I’ve ever seen. The pilot should have enough time to (a) do nothing for 1 second, (b) lower collective, (c) glide a bit, (d) flare, (e) level, (f) pull collective.
0 knots and 150′ can also work, actually, as a demonstration exercise. The pilot rolls the engine to idle him/her/zir/theirself and doesn’t leave everything as-is for 1 second the way that the FAA certification standards require.