Failsafe- Throttle down but what about the control surfaces?

[Bob Johnstone]

What may suit one Model will not suit all models when trying to get a circling failsafe mode. Indeed some would not be able to do this in practice, as there are variables outwith a preset control. The best we can do is minimize the risk firstly to others and only secondary to model safety. The only constant that I can see is to reduce or cut power, with the risk posed by either being recognised. The flying site or event must be taken into consideration also when chosing.

[Bob Thompson1894]

nice to know Bob, hope others do as well, we have the comments of, well, its not a large model, so i dont need to set it, i have even heard what fail safe? a lot dont know they even have one, and when the heated debate was taking place, it was a while ago, but we where in large model company, small stuff was not on the menu, the comments where aimed at show flying, got quite heated, but as always with Boddo, it ended with a smile all round,

Well I have just been to Rolls-Royce flyin (great day lads, thanks!) and was thinking about the failsafe scenario. I watched a number of largish models and some jets fly by, and thought 'if that goes into failsafe now, what would the result be?' With throttle closed and all surfaces neutral, the worst would be the model hitting the tarmac at a shallow angle. Chances are, the model would come out of failsafe and recover. If the surfaces suddenly went into 'flick roll' mode, in order to create a spin, the the chances are that I would have got a faceful of model doing around 50 mph. (Dont forget jets also are allowed 3 seconds before going into engine cut!) In that 3 seconds, that model could be anywhere on the field, with full throttle for 3 seconds and all surfaces hard over. Its a recipe for disaster. Mad, mad, mad.

[Alan Cantwell 1131]

i dont think, and never have thought, that there is an answer to suit all, it would be a brave man that set rules and regs for all classes, suffice it to say, that the CAA has deemed it a minimum throttle to idle, do for me

[Bob Thompson1894]

Possibly true, Alan, but trying to get a 'circling failsafe mode' as above, is also a recipe for disaster. The aim is to get the model on the ground in as safe a way as possible, as quickly as possible, and being able to judge which direction it is heading in is, to me at least, part of that.

[Alan Cantwell 1131]

Works for me, Bob,

[Chris Bradbury]

No waffle Chris, apart from the set the control surfaces over

Thanks Alan, though I think I may have been misunderstood, as I'm not recommending controls over or neutral or hold as best. In fact I'm happy that the recommendations only say throttle to off, as I think the controls need to be set by model and individual knowledge, as what is right for one model and flying site won't necessarily be right for another.

For example in my fleet, my gliders are all set to banked turns of various degrees to ensure they come down back on the slope, my helicopters are all throttle off and hold, as even if they were on full pitch or cyclic at the moment of loss, the energy would be taken away quickly and my largest Extra is set to full snap roll, as it has massive surfaces and after just one snap has a very predictable slow flopping stall spin. So as you can hopefully see, I've set them all differently based on my personal knowledge of my models.

With the superb choice of models we have, I don't think its possible to specify which way is right or wrong it needs to be down to the pilots choice and their knowledge of their model, as at the end of the day if something awful happened, they're the ones that might need to stand up in an enquiry and justify their settings etc.

[Alan Cantwell 1131]

no they wont, they will state that they have set the model to the CAA guidance, and they will be in the clear, at shows, on the FS demo, the flightline scrutineers only look for throttle shut, for me, it will always be control surfaces level, and throttle to tick over, no flicks, no nothing, with gliders, i can see the exception, i dont know about helis, when they shut down, they are a lead weight, interesting discussion that will go on an on, until i have turned to dust, there are many viewpoints, all legal, not stupid, each to their own, and so it should be

[Dave Hayfield]

As far as the CAA are concerned a failsafe is required to prevent the model from flying into airspace where fullsize may be operating, ie. ' to enable the model to land safely'. This does not mean exactly what we would understand, that the model comes down without damage to anything or anyone. Personally I don't think there can be any hard and fast settings for large model controls in failsafe, there are too many variables. I suppose the only thing to attempt is to slow the model down in as much as possible in failsafe with throttle reduction or cutout and by lowering flaps if fitted, other control surface settings are meaningless unless they are like barn doors and react at low airspeed to offer more drag.

[Jonathan Pelham]

It's an interesting point that the appropriate fail safe can be different depending on where the aircraft is operating and the specific risks.

Military and commercial RPAS are often intended to return to a set point when the link is lost. However I've read numerous reports on military RPAS where on losing link they head for a fail safe rally point to wait for the link to re-establish and unfortunately the aircraft has changed launch and recovery area since the rally point was last set and the aircraft runs out of fuel going to their obsolete rally point as the operator forgot to enter the new rally point.

The safety case for a danger area(or flight display) is always going to be linked to a consistent fail safe policy across all aircraft. The distance between the point of closest approach and the outside of the danger area must be the distance within which the aircraft will land/impact the ground.

Wasn't one of the outcomes of the 1952 Farnborough airshow tragedy that aircraft should not conduct manoeuvres which in the event of a failure would result in them impacting in the crowd. Manoeuvres to be parallel or away from the crowd?

[Clive Hall]

I don't know whether this will help, but here goes:
I did an experiment a long time ago on a small model with a HAL 2000 fitted, and the radio programmed to turn the HAL on only if the radio went into failsafe, and to set the throttle at low idle at the same time. Turning the transmitter off resulted in the model levelling itself from any attitude and going into a steady landing approach. For safety reasons this was tried on only one occasion, in an isolated field on a remote farm.
I think this gives an ideal situation because it results in a slow descent and consequently minimum impact also giving anyone nearby the maximum time to see where the model is going and to get out of the way.
It was some time ago when this was tested. HAL is not the ideal gadget, but any modern self levelling stabiliser such as the Eagle Tree Guardian would do this with ease. All it takes is to fit the thing, then set the failsafe to switch on the appropriate gain and mode. To counter the risk of the model flying away just leave the rudder out of the stabiliser, connected directly to the receiver, and set it for a small deflection, then the levelled aircraft will descend in a circle, or near one allowing for wind effects. The landing is not perfect because the gadget can not manage the flare to ground.
The problem with holding the last command is that it might have just placed the model in a tight turn, which the failsafe would convert into a high speed spiral dive.

[Bob Thompson1894]

Agreed, but of course, none of this is any use if the link between battery(s) and receiver is broken. The model will then carry on at the last setting, with the throttle staying exactly where it was put. Large models with dual systems have redundancy, but smaller models are at risk of this. After having a flyaway with a small model where the switch went faulty, I now intend to set up ALL models of any size with two switches, (split lead from battery and then both plugged in to Rx) to lessen the chances of this scenario. Having seen the inside of a small switch, these things are the weakest link in the chain.

[Clive Hall]

Bob, this is getting away from the main thread, but following experience with small electric foam things I am now inclined to solve the switch problem by doing away with switches and instead plugging the power lead from the receiver to the battery directly using Deans plugs. They don't let go easily and the current capacity should satisfy anything.

My suggestion for auto levelling by failsafe was to solve the conundrum of what is best to do with control surfaces, which pre-supposes that the battery is still in contact with the radio. Clearly if the power supply goes down then usually so does the model.