Hard deck, is the only way never to crash?

[Ace Dude]

Originally Posted by extrapilot

This is a law of fluid dynamics, it has nothing to do with the sensor type.

The thing is, if you find a place where static pressure = ambient for most of the flight, you are in much better shape. Full scale, a lot of effort goes in to finding the best compromise for that location- but assumptions are made (i.e. fixed wing assumes no lateral flight, no rearwards flight, etc).

But there is no single spot on a frame where that works for all angles of flight (ie an install on the bottom of the machine may be fine for FFF and lateral, but not for vertical punches, etc). There are tricks you can use where there are multiple inlets and a T from them, etc, so that you end up with a good average static pressure. But there will always be some bias in motion for some orientations.

And even if this were not a problem (it is), there is still the issue of latency. That is specific to a setup type, but it doesn’t change with airspeed- even though lead time required for bailout does. That is something more complex systems can manage better- since a typical 300ms lag in that sensor->telemetry->TX script->virtual switch->RX->bailout init loop means 20 feet travelled at 50mph before anything even begins to happen.

You think 300ms is reasonable? Seems like when I turn off my RX it's about 1s (possibly more) before it's reflected on TX display. At 60mph that 1s is 88ft traveled.

[extrapilot]

it depends on a bunch of things. some native telemetry units send a packet as often as one is received. Some, every 4th or less. And then there is processing time at the TX, etc. But it all depends on which hardware. I used 300ms because it is about as good as I get with a Taranis, which is what was used by a couple of people involved in the 'this is a game changer' hype. 300ms sounds like nothing (it is nothing to humans- 3 x the blink of an eye), but as you rightfully mention- it is a LONG ways travel at 50-70mph when you are dealing with a 30ft max window. Then, just look at pitch rate. If you have 360deg/sec dialed in, then you are talking about 250ms to pitch nose down to level (90deg pitch). Yea, that is worst case, and yea, you are decelerating as the thing pitches IF collective is still positive AOA. But the point is still relevant- that this all takes time, and how much time depends on the travel direction (vector) and the machine attitude when it happens. Since a TX-based solution doesnt know those things, you have to assume worst case, and you end up, as has been mentioned here a bunch of times- with a 50ft hard deck at best.

Again, nothing wrong with a simple/cheap solution like this as long as you understand this is not a substitute for a proper auto-bailout system. But that is what people are claiming- the biggest fanboy was suggesting 5-10ft in hard 3D was possible with a baro sensor and a baseline Taranis... Yea.

[phatburner]

@extra pilot

You didn't answer any of my questions. What kind of setup have you built? What are you using right now?

[extrapilot]

Everything from off the shelf FrSky stuff to proprietary hardware that costs more than a typical new car. That is the point- no one here has the hardware I am flying- but that doesn’t change the rules of fluids. I can log stuff at millions of samples/second, and demonstrate how pressure changes with airspeed. It is a waste of time- because it is anecdotal; we are not all flying the same setups- most are nothing like each other, and it changes with flight style and speed regardless.

[ridge-runner]

Originally Posted by extrapilot

This is a law of fluid dynamics, it has nothing to do with the sensor type.

The thing is, if you find a place where static pressure = ambient for most of the flight, you are in much better shape. Full scale, a lot of effort goes in to finding the best compromise for that location- but assumptions are made (i.e. fixed wing assumes no lateral flight, no rearwards flight, etc).

But there is no single spot on a frame where that works for all angles of flight (ie an install on the bottom of the machine may be fine for FFF and lateral, but not for vertical punches, etc). There are tricks you can use where there are multiple inlets and a T from them, etc, so that you end up with a good average static pressure. But there will always be some bias in motion for some orientations.

And even if this were not a problem (it is), there is still the issue of latency. That is specific to a setup type, but it doesn’t change with airspeed- even though lead time required for bailout does. That is something more complex systems can manage better- since a typical 300ms lag in that sensor->telemetry->TX script->virtual switch->RX->bailout init loop means 20 feet travelled at 50mph before anything even begins to happen.

Way back when in fluid mechanics class I was taught that total pressure equals static pressure plus dynamic pressure. A pitot tube airspeed indicator uses the difference between two pressures: 1.) the total pressure at the stagnation point (i.e. the tip of the spear) and 2.) the static pressure from a (carefully located) port positioned on a surface that is parallel to the airflow (a hole drilled into the shaft of the spear). The dynamic pressure is simply the stagnation pressure minus the static pressure and this difference is what the (mechanical) pitot tube airspeed indicator uses.

Ptotal = Pstatic + Pdynamic

Pdynamic = Ptotal - Pstatic

Airspeed = f(Pdynamic, density)

I can see where an improperly placed static port could see pressures lower than static, but if you aim that static port facing into the wind even just a little bit, then it will be reading pressures higher than static. On a moving heli, (which has a surrounding airflow field that is... lets just say complex - and I'm sure you could school us all on the depths of that statement!), I would surmise that a randomly placed absolute pressure gauge would typically read higher than an identical gauge sitting still in the air nearby at the same altitude. Yes, there are likely locations where the pressure is lower but it seems that most locations (anywhere facing the relative wind) would result in pressure readings higher than static (especially when the canopy/fuselage is removed). In this case, the altitude reported by the sensor would be reading erroneously lower, and thus the threshold altitude rescue would be triggered prematurely - i.e. too high, which would be a good thing.

Don't know / not sure!

p.s. The default (programmable) data rate for the FrSky vario sensor is 10Hz (the maximum available), so I'm guessing that typical end-toend latency would be well under 200ms. Although, the sensor itself may be doing some averaging so the readings may have built-in lag time which would increase the latency.

p.s.2. Note that I am not advocating the use of this sort of rescue system - (I'm a simple "Throttle Hold" kind of guy), I just wanted to pic a nit regarding one of your points.

[CZorzella]

Originally Posted by phatburner

@extra pilot

You didn't answer any of my questions. What kind of setup have you built? What are you using right now?

Originally Posted by extrapilot

Everything from off the shelf FrSky stuff to proprietary hardware that costs more than a typical new car. That is the point- no one here has the hardware I am flying- but that doesn’t change the rules of fluids. I can log stuff at millions of samples/second, and demonstrate how pressure changes with airspeed. It is a waste of time- because it is anecdotal; we are not all flying the same setups- most are nothing like each other, and it changes with flight style and speed regardless.

With all due respect, EP, but credibility comes with facts. If you can't answer a simple question and provide us with details of the setup you're using in a RC model, then all this theory behind the rescue / HD technology may not be as relevant or practical for the rest of us mortals.

Carlos

[extrapilot]

I don’t have a single setup. I don’t have a single machine. I don’t use toys to log data. The data acquisition systems I run for real analysis are not available to you. What I have posted is raw and graphical data from an FrSky system. I just ran a test that anyone can run. A stock XR4SB with an FrSky GPS and an FrSky baro. The data was captured on a Taranis Plus with no filtering, and is shown in OpenTX. You guys don’t like it- find your own way. Im not jumping through hoops to explain to you why the laws of fluids apply to you.

Look at the plots. I drove around my block- flat terrain. The system was inside my car (on the dash so the GPS has a view of the sky), with a window open to emulate a canopy with an open section.

You see any trend here between ground speed and indicated altitude? You see the pressure going up with increasing airspeed? No. It indicates increasing altitude (decreasing pressure), in keeping with the laws already referenced.

Then, for fun, I held the sensor, and extended it out the window briefly and back inside, several times. Did the air stagnate and increase the pressure with that sensor out in the free stream, dropping the indicated altitude? No. The pressure dropped rapidly, showing altitude increases of an additional 30-40ft.

OK?

[ShaneJourdan]

^^^
That's pretty telling data.

[ridge-runner]

Originally Posted by extrapilot

... OK?

Yes, and thank you for empirical data. I stand corrected.

I think I may go perform a similar experiment and place the sensor on the front grille. I'm curious to know if the pressure trends will be similar in that location.

Thanks again.

[CZorzella]

Originally Posted by extrapilot

Look at the plots.

Thank you, EP. That's very good insight. 👍

Carlos

[phatburner]

Well what I was trying to ask is have you actually tried this with a toy helicopter? Or are you just talking based on what you think or learned in a fluids lab? You still have not said. With respect, the dash of your car is not a helicopter either, so although it must be nice to drive around with the windows down in January it doesn't actually shed any light on what the data looks like on the actual helicopters people on this thread are in fact flying. What about the people who have posted on this thread and said that they actually have flown it and it works for them? Are you saying that they are liars? I will give a personal anecdote on how this has been a boon for me.

When I got my first CP heli (a trex450 pro) after playing with a 120SR till it wasn't fun anymore, it was a big step both financially and as a challenge to my ability. I practiced a lot on the simulator until I felt ready. The first time I flew it I just hovered and did some slow piros. Stoked, I changed the battery and lifted off for a new adventure. When I did the first circuit I promptly crashed because it turned out to be not like the sim, my eyeball didn't zoom in like the camera and the heli got small and before I knew it, it slid out of a poorly coordinated turn and I got out my wallet. I gather some people navigate the learning curve that way but I can't afford it. Then I heard about this trick with the altimeter and it has definitely been a game changer for me. Now I have the confidence to try things out at the field that I would never have tried before and am learning flips and inverted piros and I haven't crashed once since setting up this deck thing. This saves not only money, but down time and results in going home with a smile rather than a frown. More flying = more learning and more fun too. I think anyone can agree, that is certainly a 'game changer'. I know a guy who gave up on helis because of the cost and the repairs he considered a pain. I'm quite certain that story is not unique. Another fellow doesn't want to expand his skills beyond FFF because of the fear/cost factor. For people like me I think this deck idea is nothing but good. If it starts to become a problem then I'll think otherwise but when will that be? I'm starting to learn to coordinate backwards turns now. So suppose the pressure sensor reads high when I fly backwards what's the worst that could happen? The rescue triggers at a higher altitude. Ok. At least it isn't going to crash. I've had the heli going pretty fast (the length of our runway in about 3-4 seconds) in forward flight and could see it bouncing off the deck level the odd time and it clearly wasn't 30 feet lower than I'd set it. It looked about the same height as when flying slow actually. How do you explain that? Do I have a magic setup that defies the laws of fluid dynamics? I guess so according to what I've been reading here. All I know is that it works.


If someone is already capable of doing hard core 3D I would imagine there would be no benefit and they might likely want to be able to fly low enough to mow the lawn with the blades anyways. That guy can scoff at it all (s)he wants. But I would hazard a guess based on what I see around me at the flying site that there are more people like me in the hobby than the former type. Maybe not right here, on this forum but I'm talking in general. And one more thing, just because someone gets old or struggles with learning that doesn't mean anyone should tell them this hobby isn't for you anymore. I don't know about anybody else but I thought that statement was audacious and rude.

[sutty]

Originally Posted by extrapilot

I don’t have a single setup. I don’t have a single machine. I don’t use toys to log data. The data acquisition systems I run for real analysis are not available to you. What I have posted is raw and graphical data from an FrSky system. I just ran a test that anyone can run. A stock XR4SB with an FrSky GPS and an FrSky baro. The data was captured on a Taranis Plus with no filtering, and is shown in OpenTX. You guys don’t like it- find your own way. Im not jumping through hoops to explain to you why the laws of fluids apply to you.

Look at the plots. I drove around my block- flat terrain. The system was inside my car (on the dash so the GPS has a view of the sky), with a window open to emulate a canopy with an open section.

You see any trend here between ground speed and indicated altitude? You see the pressure going up with increasing airspeed? No. It indicates increasing altitude (decreasing pressure), in keeping with the laws already referenced.

Then, for fun, I held the sensor, and extended it out the window briefly and back inside, several times. Did the air stagnate and increase the pressure with that sensor out in the free stream, dropping the indicated altitude? No. The pressure dropped rapidly, showing altitude increases of an additional 30-40ft.

OK?

Could be interesting perhaps to repeat the out of the window test with one of those static streamers that the FMA CoPilot II uses on their baro sensor, to see how well it smooths down the graph.

I could have guessed that cracking the window part down would have a big effect on the internally mounted sensor You can even pop your ears in the car when doing that, or at least I can. Maybe my ears are sensitive.

As for if that is representative of the best placed sensor on a heli, I have no idea, but I can imagine if you were inside the canopy it would be fairly similar to that, except that it would likely be worse due to increases in pressure in backwards flight and similar, and you can't easily go sideways with the window down to replicate that, lol.

[peteski]

Originally Posted by phatburner

Well what I was trying to ask is have you actually tried this with a toy helicopter? Or are you just talking based on what you think or learned in a fluids lab? You still have not said. With respect, the dash of your car is not a helicopter either, so although it must be nice to drive around with the windows down in January it doesn't actually shed any light on what the data looks like on the actual helicopters people on this thread are in fact flying. What about the people who have posted on this thread and said that they actually have flown it and it works for them? Are you saying that they are liars? I will give a personal anecdote on how this has been a boon for me.

When I got my first CP heli (a trex450 pro) after playing with a 120SR till it wasn't fun anymore, it was a big step both financially and as a challenge to my ability. I practiced a lot on the simulator until I felt ready. The first time I flew it I just hovered and did some slow piros. Stoked, I changed the battery and lifted off for a new adventure. When I did the first circuit I promptly crashed because it turned out to be not like the sim, my eyeball didn't zoom in like the camera and the heli got small and before I knew it, it slid out of a poorly coordinated turn and I got out my wallet. I gather some people navigate the learning curve that way but I can't afford it. Then I heard about this trick with the altimeter and it has definitely been a game changer for me. Now I have the confidence to try things out at the field that I would never have tried before and am learning flips and inverted piros and I haven't crashed once since setting up this deck thing. This saves not only money, but down time and results in going home with a smile rather than a frown. More flying = more learning and more fun too. I think anyone can agree, that is certainly a 'game changer'. I know a guy who gave up on helis because of the cost and the repairs he considered a pain. I'm quite certain that story is not unique. Another fellow doesn't want to expand his skills beyond FFF because of the fear/cost factor. For people like me I think this deck idea is nothing but good. If it starts to become a problem then I'll think otherwise but when will that be? I'm starting to learn to coordinate backwards turns now. So suppose the pressure sensor reads high when I fly backwards what's the worst that could happen? The rescue triggers at a higher altitude. Ok. At least it isn't going to crash. I've had the heli going pretty fast (the length of our runway in about 3-4 seconds) in forward flight and could see it bouncing off the deck level the odd time and it clearly wasn't 30 feet lower than I'd set it. It looked about the same height as when flying slow actually. How do you explain that? Do I have a magic setup that defies the laws of fluid dynamics? I guess so according to what I've been reading here. All I know is that it works.


If someone is already capable of doing hard core 3D I would imagine there would be no benefit and they might likely want to be able to fly low enough to mow the lawn with the blades anyways. That guy can scoff at it all (s)he wants. But I would hazard a guess based on what I see around me at the flying site that there are more people like me in the hobby than the former type. Maybe not right here, on this forum but I'm talking in general. And one more thing, just because someone gets old or struggles with learning that doesn't mean anyone should tell them this hobby isn't for you anymore. I don't know about anybody else but I thought that statement was audacious and rude.

Correct me if I'm wrong, but I didn't think anyone had tried to argue that a hard deck doesn't work for gentle upright flying. It's more the minimum realistic height some people were claiming and whether or not it could cope with more aggressive rescue scenarios.

[nightflyr]

Originally Posted by peteski

Correct me if I'm wrong, but I didn't think anyone had tried to argue that a hard deck doesn't work for gentle upright flying. It's more the minimum realistic height some people were claiming and whether or not it could cope with more aggressive rescue scenarios.

Ummmm.... this seems to work fairly well.

Banging the Hard Deck with CoPilot II with HD module. (3 min 10 sec)

FMA Direct / REVOLECTRIX CPII Hard Deck Module on Airplane (4 min 21 sec)

[peteski]

They both appear to have the hard deck set in space

[WI_Hedgehog]

nightflyr and extrapilot know exactly what they're talking about.

Hard deck may work, but as Skookum points out there are a lot of variables to consider, such as dive rate, angle, head speed, power-to-weight ratio, wind, the flight maneuvers, etc. They've said there are times when the minimum bailout of 30' is not enough, and soft-deck is a much better option.

I seem to remember a thread from a few years ago where a guy said the SK720 bailout was crap because his heli slammed into the ground during his very first bailout test flight. The Skookum engineer asked why he tried a hard dive his very first test, why he set a hard deck and at only 30', and why he did it at something like 70MPH at a steep bank, after a transition maneuver which decreased the pressure making the heli look like it was climbing instead of diving. The reality is that if the guy tried to pull out of that dive he'd never make it, FBL and bailout notwithstanding--it was physically impossible. A soft-deck of 50' probably would have saved him from himself.

Regarding Skookum and vibrations, all bailouts fail under excessive vibrations. SK expects if you're going to pay $1,000 (or $600 currently) for a system you're also going to balance your blades, use round gears and good bearings, etc. For that you get to fly on the edge of insanity as if the model were on rails. A more forgiving FBL won't do that, it'll have more slop, and the owners won't notice nor care, which is the balance we look for in the market.

[nightflyr]

Originally Posted by peteski

They both appear to have the hard deck set in space

Now down around 30 feet (9.1 M )
But keep in mind this is designed as a training tool.
Emphasis on TRAINING


Anyone foolish enough to believe that a hard deck set at 30 feet or lower will stop an intentional breach of the deck at some ridiculous speed and angle is in need of some basic Jr high school math and physics classes.

The hard deck was designed (in the case of FMA Copilot II ) as a automatic safety net to help during practice.

Those doing hard smack 10 feet or less off the deck would never require the need of this type of platform.
Nor was it ever designed or tested to function under those parameters.

As to if a telemetry based configuration could work ...
Providing the same basic guide lines are followed, perhaps.
Not my area of expertise.
Again, attempting a kamikaze to mother earth and expecting these platforms to save your bacon.
I do not believe that was ever their intent.

[peteski]

Originally Posted by nightflyr

Now down around 30 feet (9.1 M )
But keep in mind this is designed as a training tool.
Emphasis on TRAINING


Anyone foolish enough to believe that a hard deck set at 30 feet or lower will stop an intentional breach of the deck at some ridiculous speed and angle is in need of some basic Jr high school math and physics classes.

The hard deck was designed (in the case of FMA Copilot II ) as a automatic safety net to help during practice.

Those doing hard smack 10 feet or less off the deck would never require the need of this type of platform.
Nor was it ever designed or tested to function under those parameters.

As to if a telemetry based configuration could work ...
Providing the same basic guide lines are followed, perhaps.
Not my area of expertise.
Again, attempting a kamikaze to mother earth and expecting these platforms to save your bacon.
I do not believe that was ever their intent.

No argument there from anyone that I've read here so far. But some people seem to be claiming considerably more ambitious targets. To be honest I think I'm well beyond the point where a basic 30 ft hard deck would be of any practical use as I can trigger rescue manually well below that if required. But I can see the point for some.

If there really was a viable 10 ft hard deck that could cope with moderate 3D flying I'd be all over it, but sadly it doesn't appear to exist and especially not in a "home made" system.

[gunny_jeeves]

This thread popped up again? Ok... I will share what I noticed and learned again.

Step 1 - Is it possible to make a hard deck work? (Depends on how you define "work".)
Work="Provide a functioning training tool" - Yes!. Not only works, but works well, and is flexible enough to work using a rescue or a self-level option only. Both work, and both work really well. (Even for 3D flying.) You need a hard deck of 35-50 feet.
Work="Make my heli uncrashable?" - No. Intentional breeches of the hard deck at speed will result in a possible "lawn dart" result. ONLY Occasionally will the timing work out so the heli is saved. (Which this is an accomplishment of sorts as without the hard deck, EVERY dive and breach will result in lawn darting, not just "most of them".)

Step 2 - To get the best use out of your hard deck, you must understand the relationships and workings of your set of tools to make it work.
FBL Rescue --> Know how this works. (Typically, "Enabled" = "arrest descent, flip horizontal, and climb until disabled".) With Self Level only, you will need to flip first then climb. (But admittedly, flipping and climbing at the same time works too, it just makes a hook shape recovery.)
Latency --> Just know that there is latency in every step. For the telemetry to work and trigger, the altimeter gets read as "Below Deck". Last reading was fine... this reading, some number of milliseconds later, is not... so it sends the new reading back to the TX. (It travels to the TX... waiting another 10ms maybe?... ) Once picked up, it calculates (does the math to figure out... YEP. We need to rescue. This is fast but not instant... Few more ms) And it finally sends the rescue signal to the heli. As soon as it travels to the heli, is processed and decoded, it is executed and rescue starts. (Just know stuff adds up, and heli is falling during this time.)
Delays / Durations --> Know your TX and FBL's definitions of delays and durations. (ESP as they pertain to rescue.) Dummy mode works awesome on a Brain... Enable rescue while under HD... Disable it while over it. (No delay, no duration in the TX, so it stays in rescue mode as long as the signal is sent, no guesswork.) Others might need some hard coded in the FBL durations to guess the right time to trigger rescue AND travel back above the deck.
Triggers --> More than anything else, what will determine how much you need to worry about duration are how your triggers work. Think "Fire Hose" vs "Semi-Automatic Gun". GUN --> Squeeze trigger --> BANG! for one instant, one bullet is triggered, even if you hold it down for 30 min. (VBar / Neo is like this) You will likely need to set durations manually in the FBL as it will trigger instantly once, then not trigger again until the heli has registered a height ABOVE the deck (releasing the trigger to be reday for the second pull). Fire Hose --> You see fire, aim, pull trigger. "WOOOOOOOOSSsshhh!!!!" Rescue water starts coming out of the hose and continues to do so until you release the trigger. (Brain is typically like this.) You need no delays or durations as your rescue starts when below deck... and you keep climbing till you are above deck, and release the trigger.

Once all the math and crap is done... For up to an established intermediate pilot learning 3D, the hard deck's sloth is mute as it can correctly react much faster than a beginner, faster than most intermediate pilots, and arguably can SOMETIMES beat even ANY PILOT if they loose orientation, loose sight of the model, or incorrectly read something. So it could conceivably save someone in the expert side a crash under the right situation.

Like it or not, this works a whole lot better than it should, and only is even possible on a Jeti or Open TX radio. (Taranis, 9XR-Pro, or Horus) Of all the crazy facts? That's the one that burns my butt up. Why do we let folks like Futaba (Who sell TX's incapable of this for $3000, more than double a Jeti that can.) and Spektrum / JR / Others... (who in the main stream don't offer anything close to this capability) earn our business?

That's my big gripe, but anyway the Hard deck works way better than it should. Used as a normal tool, it will save your heli in about any case you were not trying to lawn dart it intentionally.

[peteski]

Originally Posted by gunny_jeeves

Why do we let folks like Futaba (Who sell TX's incapable of this for $3000, more than double a Jeti that can.) and Spektrum / JR / Others... (who in the main stream don't offer anything close to this capability) earn our business?

I think it's because the actual demand for a 35-50 ft hard deck is very small. Most people would prefer a manual rescue trigger - given that a 10 ft hard deck is unrealistic. I would go insane if I had to fly above 35 ft all the time!

I realise it's just a training aid but even my grandma could trigger rescue manually at that kind of altitude! I'm now actually wondering what real point there is in having such a high hard deck, given that it doesn't even guarantee stopping a crash if you hit it hard? If you are flying gently and lose orientation, wouldn't you simply press the rescue button manually?