Fatal - MTOSport 5Y-KWV, Kenya 19 NOV 2020

I would not draw anything from small numbers, Phil. Accidents happen randomly, and focusing on a small subset has next-to-no statistical power.

e.g. I launch my new gyro brand, and the first 9 fly happily. Woohoo! 100% safety record. Then the tenth one crashes. A wholly unacceptable fatal rate of 10%.... Obviously, neither figure is/was the truth. We must wait till 100 machines, or a thousand, to get closer to the truth...

Only the big picture can tell us anything, and even then we must be cautious, because rates can change over time, and different countries have vastly different environments, regulatory, and even cultural.

Do France and Spain have a high accident rate because they are flying a lot of ELAs? (which seem to have a bad accident rate)

or

Do ELAs have a bad accident rate because lots of them are flying in France and Spain? (and none in the UK)
 
Last edited:
I don't really understand the rationale for ½ a figure for a collision with another gyro. It seems to me that's 1 each, even if they happened simultaneously and in the same incident.
There's no easy way to do it.
If I didn't halve them I would end up with one more accident than actually occurred. Halving seemed to me to do the least harm.
Either way, it would only alter a fraction of a percent in an analysis that isn't that perfect to begin with...
 
Last edited:
I would not draw anything from small numbers, Phil. Accidents happen randomly, and focusing on a small subset has next-to-no statistical power.

e.g. I launch my new gyro brand, and the first 9 fly happily. Woohoo! 100% safety record. Then the tenth one crashes. A wholly unacceptable fatal rate of 10%.... Obviously, neither figure is/was the truth. We must wait till 100 machines, or a thousand, to get closer to the truth...

Only the big picture can tell us anything, and even then we must be cautious, because rates can change over time, and different countries have vastly different environments, regulatory, and even cultural.

Do France and Spain have a high accident rate because they are flying a lot of ELAs? (which seem to have a bad accident rate)

or

Do ELAs have a bad accident rate because lots of them are flying in France and Spain? (and none in the UK)

Yes I see the general point which is agreed with but that shouldn’t excuse a deeper drill into the data and the detail of individual accidents- which in my view is partly why we are in the current situation of repeating many common errors (all be they aren’t fatal).

First of all if you look at data in territories that have robust (or should that be more robust) accident reporting and investigation then at least if you want to drill down you can. So we are both UK based and one might argue our system is mature and robust. So our numbers on Cavalon are very interesting because we have 30ish airframes (I say ish because of Cavalon Pro, CKYV being re-built as CLGG and one heading out to the USA, but regardless around 30 for 7 accidents - so about 23%, which is consistent with the data from the USA all be it over a longer timeframe.

So there are some similarities in the metrics and similarities in the accident types - take off accidents? Tick. Landing accidents? Actually no - not to my knowledge. Then some pre-rotation SNAFUs, some bad airmanship and some as yet unexplained - but with low time/inexperienced pilots.

So again a pattern. The take off/pre-rotation are training / technique issues and we might have to wait a while for the explanations on the current unknowns but already the situation on the 2 unknown fatals (one in USA, one UK) raise half an eyebrow on a number of points - namely the rebuild in USA and the suggestion of 2nd solo in the UK and the disruptive nature of the virus throughout 2020.

We will of course see but I’ll bet my mortgage that nothing in any of the final reports becomes so surprising and when you look at Cavalon accidents to date in UK or USA nothing has pointed toward anything that more rigor and a more disciplined process wouldn’t go some way to fixing. Even medical events... well gee we relaxed the medical requirements and we find people having medical problems that we didn’t foresee... funny old world.
 
Someone once said, IIRC

"70% of dead pilots were dead before they left the ground. They just didn't know it...."

Insufficient training
IMSAFE
Poor ADM for the sortie
Poor maintenance / poor checks
Etc.
 
Last edited:
I think you are doing a good job with your numbers and analysis.

I find your numbers interesting TyroGyro and your methodology thoughtful.

Thank you for your effort.
 
Actually the data would be better if accidents like mid air collision where It has nothing to do with being a gyro should be taken out.
 
There are only 3 mid-airs (all in France), amounting to less than 4% of all global fatals.

The only other candidate for removal is the UK incident where an angry civilian confronted a taking-off gyro, with terminal consequences for him.

The irony of your post is not lost.

It is apparent from trawling the accidents that the vast majority are not the fault of the gyro.

Someone put it rather pithily: "Gyros don't kill people; people kill gyros"
 
On the subject of checks, a wise old instructor told me:- "the best chance of finding a fault is when cleaning your aircraft. So clean it often." and "in addition to your checklists, sometimes just 'stare' at the aircraft."

Well, I have probably the most comprehensive gyro "A" checklist in the UK. Two pages, and it takes about 20 minutes to complete.

A couple of weeks ago, I was just admiring my gyro in the hangar (it had just been re-registered with new marks). I was 'staring' at it.

Then... I noticed that a nut and washer were missing from the port-side carb ! [the one at mid left in the pic]

rotax-912-carb-overhaul-bing-64-01-uncleaned-1-LRG.jpg


How had I missed that? How long had it been in that state? Had it departed in flight?

I was shocked that I had no answers.

The checklist says check the three springs on the carb. The problem is tunnel vision. You only see what you are looking for, even though something else should be patently obvious!

Checklists are certainly required, but I am grateful for the wisdom of an instructor.

Clean your aircraft, and from time to time just stare at it.
 
Last edited:
Familiarity breads contempt, I have seen many high hours excellent pilots do a cursory pre-flight, I am sure I am not alone in this, it is part of the human condition.
 
  • Like
Reactions: PJB
On the subject of checks, a wise old instructor told me:- "the best chance of finding a fault is when cleaning your aircraft. So clean it often." and "in addition to your checklists, sometimes just 'stare' at the aircraft."

Well, I have probably the most comprehensive gyro "A" checklist in the UK. Two pages, and it takes about 20 minutes to complete.
My checklist fits on half a page (actually three columns of small type on an 8x5 index card), but 20 minutes would be a fast pre-flight for me.
Here's an incipient problem I found by just staring, though...
 

Attachments

  • hose chafe.jpg
    hose chafe.jpg
    17.8 KB · Views: 33
The only other candidate for removal is the UK incident where an angry civilian confronted a taking-off gyro, with terminal consequences for him.
Wow, I had not heard about that one. By "civilian" I take it you mean a non-aviator?
 
I vaguely recall discussion on another thread long ago, involving a protest about fox hunting, some overflying that hunters found objectionable, anger on both sides, an encounter on the ground where the gyro was operating, and a pedestrian struck by turning blades. That may be the same incident.

 
It is apparent from trawling the accidents that the vast majority are not the fault of the gyro.
I think that may change if the trend of fully enclosed gyros continues, unless there is some methodology to insure the gyro handles predictably with the large increase of area in front of the mast that is often not offset by an increase in tail volume behind the mast. How about testing the gyro by hanging it from the rotorhead directly sideways to the wind and see how strong the tendency to point into the wind it has, or if it even does ?
 
I think that may change if the trend of fully enclosed gyros continues, unless there is some methodology to insure the gyro handles predictably with the large increase of area in front of the mast that is often not offset by an increase in tail volume behind the mast. How about testing the gyro by hanging it from the rotorhead directly sideways to the wind and see how strong the tendency to point into the wind it has, or if it even does ?


That is not how aerodynamics work. The tail has airfoils most of the time. Airfoil creates pressures and different airfoils create different pressures and though placement further back or having a larger side area is valid, it isn't the whole story. Its napkin design. Just as in airplanes design of the tail requires knowledge of airfoils used and expected pressure they will create (on either side) to try and keep it straight. You are looking for neutral to slightly positive stability in yaw to allow time for the pilot to put in the right input for the expected cruise speed range. Pilot has to still be trained in using all three axis. There isn't any substitute for training. May be yaw dampers like airliners or new SR22 use.
 
Last edited:
There are only 3 mid-airs (all in France), amounting to less than 4% of all global fatals.

The only other candidate for removal is the UK incident where an angry civilian confronted a taking-off gyro, with terminal consequences for him.

The irony of your post is not lost.

It is apparent from trawling the accidents that the vast majority are not the fault of the gyro.

Someone put it rather pithily: "Gyros don't kill people; people kill gyros"
Thanks for all your efforts compiling those statistics , not an easy task.

Years ago I took a course on statistics and below was an example they used

I decided to post it here because of your (true) comment : .... not the fault of the gyro.

The following is intended as educational

=============================
Aircraft #1 has ..... 100 fatalities per 100,000 flight hours
Aircraft #2 has ........ 15 fatalities per 100,000 flight hours

QUESTION: Which is the safer aircraft ?

Answer is below

============================
Turns out aircraft #1 is the safest

Reason given below

================================
Aircraft #1 was one commuter plane with 100 people on board
Aircraft #2 were 15 light aircraft with one person on board.
================================

But that is not really true either

Because all crashes were attributed 100% to pilot error (not the aircraft)

Thus we could correctly state that both aircraft have a good safety record

=================================
The true statistics (in these examples) should read ....

In 100,000 flying hours 16 pilots made errors that caused the deaths of 115 people and destroyed 16 perfectly good aircraft
==================================
 
It's a knotty problem...

For Commercial, with a varying proportion of survivors per fatal crash, and essentially no correlation with flight length, the Q-Statistic seems the best.

For gyros, with, from the pilot's point-of-view, an invariant number of POB, fatal accidents per million flight hours is more accessible, and calculable.

For those interested, you can find the Q-Statistic for various Commercial airliners here [the "Rate"]
 
Last edited:
That is not how aerodynamics work. The tail has airfoils most of the time. Airfoil creates pressures and different airfoils create different pressures and though placement further back or having a larger side area is valid, it isn't the whole story. Its napkin design. Just as in airplanes design of the tail requires knowledge of airfoils used and expected pressure they will create (on either side) to try and keep it straight. You are looking for neutral to slightly positive stability in yaw to allow time for the pilot to put in the right input for the expected cruise speed range. Pilot has to still be trained in using all three axis. There isn't any substitute for training. May be yaw dampers like airliners or new SR22 use.
I'm not claiming it to be a perfect test, but it is relatively easy and inexpensive to do and I think it might give an indication of potential instability in a hard sideslip or slow flight with a strong crosswind, etc. Do you have a better test?
 
I think part of why that's probably not a great test is that by the time you are "directly sideways" to the relative wind, things have already gone completely sideways (so to speak). The side pressures on the front (big fuselage with canopy) and empennage (slim fins) are going to look pretty different at the various intermediate angles between 0º and 90º.
 
I can’t speak for other countries but here in Australia the gyro hours flown are not a true representation of the current situation due to the reluctance of some certified pilots to record their hours.
Then there is the ‘mustering pilots‘ who on average would do up to 1000 hours a year but unfortunately don’t officially record their hours due to the legality surrounding commercial gyro operations.
The significance of accidents per hours flown appears to be misunderstood by some people, but is extremely important to insurers when calculating the exorbitant premiums we are now faced with.
 
I think part of why that's probably not a great test is that by the time you are "directly sideways" to the relative wind, things have already gone completely sideways (so to speak). The side pressures on the front (big fuselage with canopy) and empennage (slim fins) are going to look pretty different at the various intermediate angles between 0º and 90º.
I suppose you could have the gyro hang facing the relative wind and see how far off axis it remains stable and tries to head back into the wind. How many degrees off axis if or before it becomes unstable then. Pretty much the same test. Do you think Tyger, that the addition of full enclosures is making the underlying gyro safer, about the same, or less safe? Do you think they are generally well engineered conversions or do you think they are being treated as benign add ons?
 
Top