Go Back  PPRuNe Forums > Flight Deck Forums > Tech Log
Reload this Page >

Haptic Feedback Discussion Migrated From Max Thread

Wikiposts
Search
Tech Log The very best in practical technical discussion on the web

Haptic Feedback Discussion Migrated From Max Thread

Thread Tools
 
Search this Thread
 
Old 16th Sep 2019, 08:34
  #1 (permalink)  

Only half a speed-brake
Thread Starter
 
Join Date: Apr 2003
Location: Commuting not home
Age: 46
Posts: 4,321
Received 3 Likes on 3 Posts
Haptic Feedback Discussion Migrated From Max Thread

Originally Posted by fdr
The 'buses have no feedback, they really are nintendo devices, but they do have a nice manoeuvre demand logic. The natural control feel of a FBW system is about the same as the control panel of a microwave oven for the SSC systems. If the designer wants to give feedback, that is an artificial feel, fed back by servos to a moveable control,
Utterly wrong. Although perfectly irrelevant to the thread, the danger of complete misunderstanding of the ‘bus haptic feedback necessitates my remark, apologies.
FlightDetent is offline  
Old 16th Sep 2019, 09:19
  #2 (permalink)  
 
Join Date: Dec 2013
Location: Weltschmerz-By-The-Sea, Queensland, Australia
Posts: 1,366
Received 80 Likes on 37 Posts
Hmmm. This will be interesting. What is this haptic feedback to which you refer? The Airbus I flew had exactly the same sidestick force in all manoeuvres and throughout the speed range. It really is like a Nintendo in that you are limited to visual cues as the tactile ones provide no data.
Australopithecus is offline  
Old 16th Sep 2019, 13:55
  #3 (permalink)  
fdr
 
Join Date: Jun 2001
Location: 3rd Rock, #29B
Posts: 2,956
Received 861 Likes on 257 Posts
Originally Posted by FlightDetent
Utterly wrong. Although perfectly irrelevant to the thread, the danger of complete misunderstanding of the ‘bus haptic feedback necessitates my remark, apologies.
Are you referring to limit protection at alpha prot through to alpha floor as a feed back? (FCTM 2.60.6)

Possibly for the current Airbus, haven't flown one for a couple of years, however haptic control feedback was patented (assigned to Honeywell who were involved in the bus FCS) in 2006 but appears to have lapsed. TU Delft ran studies on the benefits of that as a proposed enhancement on an A330 in 2016, all following Billings work in 1996. The bus SSC's I have used had an SSC resistive force proportional to stick deflection and velocity from passive mechanical spring/dampers. If they were smarter than that, I missed that in the testing. The bus provides an increase in resistance to the SSC order to over bank the aircraft, that is an advisory of the envelope protection. That is a cueing of a limit rather than a feedback of the flight vehicles condition. If you had asymmetry in lift between the wings, the system is going to absorb that and the flight crew will be given the normal passive constraints per a normal condition. Thats roll... for pitch, if you are off speed, well, you never are, the aircraft is doing its own thing, and there is no cue that the aircraft is off speed (aoa) as a consequence. That leaves pitch to be either passive in nature until a limit condition occurs e.g., alpha prot/ alpha floor. The G500/G600 SSCs incorporating tactile feedback with BAE Systems’ “Active Inceptor System,” was headlined as the first civil aircraft use of that concept, around 2015. The G's also were synchronous SSC's which seems to make some sense, would have been handy for AF447 at least. Crew flying a 330 into a fun strip have the Captain comment to the FO on the taxi in that he could have used some more pitch in the flare. The FO knew that he had full back stick on at the time, Captain only knew when the FO 'fessed up on the taxi in. Synchronous controls would have given a bit more info.

If AI had previously applied haptic feedback would appreciate a reference to that effect. I can't find any reference to a haptic feedback system incorporated in the 320.330 or 340 FCOMs or FCTM of mine (alpha prot/alpha floor notwithstanding).

Note: the evaluation of the A330 FCS adapted to having haptic feedback was interesting and should not be surprising. For the primary tasks that were evaluated, there was either no observable difference or negligible difference in tracking accuracy, or workload, that would be expected in pitch considering the flight path demand logic of the airbus, but it was interesting in the roll side, where little benefit was observed. The test series was constrained in scope, which may explain the observed lack of benefit in roll.

Now going to synchronous SSC design, and even on the ATS would be interesting to test near limit conditions for S.A. Suspect any bus driver flying the G's SSC would note the difference.


P.S.: The one time that feed back would be quite handy in the Airbus system and could help in a tracking task would be in direct law, as possibly such feed back would assist in trim input to the pilot. However, as direct law comes about following degradations which have potential sources in the sensing that would be needed for the feedback system, CAS, AOA etc... then sometimes it just sucks to have an electrojet, and you get to work with a trim system to a control that has performance outcome as its feedback rather than force.
______

Billings,C.E.(1996). Aviation Automation:The Search for a Human-Centered Approach. CRC Press,1st edition.

Ellerbroeck, J., Rodriguez Martin, M.J.M., Lombaerts, T., van Paassen, M.M., & Mulder, M.; (2016) Design and evaluation of a Flight Envelope Protection haptic feedback system. Paper, IFAC (International Federation of Automatic Control). https://doi.org/10.1016/j.ifacol.2016.10.481

EP1918196A1, EP1918196B1, Abel, S.G., Hanion, C., Pilot Flight Deck Control Stick Haptic Feedback System and Method.

US9156546B2 Irwin J.G. III., Einthoven P.G., Miller, D.G., Spano, M.S., Active-inceptor tactile-cueing hands-off rate-limit

Last edited by fdr; 16th Sep 2019 at 16:14. Reason: added P.S.:
fdr is offline  
Old 16th Sep 2019, 18:01
  #4 (permalink)  
 
Join Date: Jun 2009
Location: florida
Age: 81
Posts: 1,610
Received 55 Likes on 16 Posts
Salute!
At the risk of getting back to the rationale for the actual system causing two crashes, I'll make at least one more post about that and about not all the management, politics and questionable engineering we have seen the last year.
@ FDR and Tom , Australo , et al,
ER: feedback. And help me understand "haptic"
Being up at my mountain retreat now, I do not have my technical reports from 35 years ago that show the actual plots of stick force versus "commanded" aircraft response.for the YF-16 and the FSD version. I soloed in the original Block 1 ( less than 100 delivered, and was lucky to fly serial numbers like 001, 007. 013 and so forth), and instructed in the Block 5, 10 and 15.
The stick did not move in the Block 1 - it was like a pipe in concrete. We flew with it using wrist pressure. The 1/8 inch movement in the Block 5 and later was not apparent to this old fart, but it made the brass happy and might have helped a small amount in formation flying, although breakout force was the same - about a pound or slightly less for both pitch and roll
Our feedback was visual and seat of pants gee and body rates from our innner ear gizmos. No problems. The forces were fairly linear after breakout, tho the roll had one "bend" in the plot. So 16 or 17 pounds of roll command resulted in max of 300 degrees per second and 30-34 pounds of aft stick commanded 9 gees biased by AoA.higher than 15 degrees.
As wth the 'bus, we did not have any problems with old or new pilots, and like the 'bus we could not see or feel what the other dude was doing with the stick. Our pressure versus command was damned near a straight line, so 16 pounds aft stick was about a 4 gee command.
I do not like the hybrid systems such as those in the Boeings, as I see too many "connections" between the electronics and the mechanical components to ascertain whether it is HAL or Otto or a jammed cable or control surface that is commanding full nose trim, much less an undocumented system I have not been told about ( GASP!!!!)
In any case, I revert to my comment about SA being aware of what HAL should be doing compared to what you would be doing manually. Your gees and body rates should be what you expect, and the only way to burn those inputs into your physiological memory is to fly manual as much as you can early in the program to establish a baseline.

Oh well,Boeing and the FAA dug this hole and it will be interesting to see them climb out of it.

Gums sends...
gums is offline  
Old 16th Sep 2019, 22:31
  #5 (permalink)  
fdr
 
Join Date: Jun 2001
Location: 3rd Rock, #29B
Posts: 2,956
Received 861 Likes on 257 Posts
Originally Posted by gums
Salute!
...I do not have my technical reports from 35 years ago that show the actual plots of stick force versus "commanded" aircraft response.for the YF-16 and the FSD version. I soloed in the original Block 1 ( less than 100 delivered, and was lucky to fly serial numbers like 001, 007. 013 and so forth), and instructed in the Block 5, 10 and 15.
The stick did not move in the Block 1 - it was like a pipe in concrete. We flew with it using wrist pressure. The 1/8 inch movement in the Block 5 and later was not apparent to this old fart, but it made the brass happy and might have helped a small amount in formation flying, although breakout force was the same - about a pound or slightly less for both pitch and roll
Our feedback was visual and seat of pants gee and body rates from our innner ear gizmos. No problems. The forces were fairly linear after breakout, tho the roll had one "bend" in the plot. So 16 or 17 pounds of roll command resulted in max of 300 degrees per second and 30-34 pounds of aft stick commanded 9 gees biased by AoA.higher than 15 degrees.
As wth the 'bus, we did not have any problems with old or new pilots, and like the 'bus we could not see or feel what the other dude was doing with the stick. Our pressure versus command was damned near a straight line, so 16 pounds aft stick was about a 4 gee command. Gums sends...
Your memory seems to be doing better than your teeth there Gums. At some point a 2 step pitch demand was added to the DFCS (15DFCS004), along with an S bend like 3 step lateral demand (16DFCS006). No idea at what block. The Airbus gains were of a 2 step form and varying with configuration, with envelope protection and landing phase varying that, but that is going back about 15 years. Force-rate values for the 16 were per your memory.

This may be off topic or it is right in the middle, the question was whether collectively the industry is losing skills through the use of automation, and lack of hand flying, or whether that really makes much difference in FBW aircraft. The response to the MCAS bingles appears to include regulator calls for more hands on flying, which may be the magic solution, or it may paradoxically just add more loss of S.A. events. Your comment that the cues that you are using are not from the FCS, but are from the instrumentation and performance of the aircraft, which is notable. Dropping out the AP and using the SSC puts the driver into the control loop irrespective of force feel of the control systems, which would then suggest that the in loop vs out of loop differences are related to the cognitive tracking task that the driver gets to do, being active or passive in the loop, feel being of less importance. Then the issue is whether the cognitive task demand of the tracking task overall detracts from the bandwidth available for S.A. more or less than it benefits it. That is not a trivial concern, as the industry plants aircraft boneyards off the end of runways routinely from the crew being so preoccupied with getting on the ground with hand flying that they end up landing 7000' down a 10,000' runway. ET and JT feature in that sort of deal, but so do US, French, German aircraft, so loss of S.A. is agnostic or just endemic.

cheers.
fdr is offline  
Old 16th Sep 2019, 23:49
  #6 (permalink)  
 
Join Date: Jun 2009
Location: florida
Age: 81
Posts: 1,610
Received 55 Likes on 16 Posts
Salute!

Thanks, fdr, I needed that.

I do not feel we are getting too far off topic. The mods shut off this on the Tech log for some reason.

Were I on one of the legal team experts for the forthcoming trials I would make a big deal about what we have been discussing lately here. The management and certification process is a different aspect, IMHO.

Gums sends......


gums is offline  
Old 17th Sep 2019, 03:14
  #7 (permalink)  
 
Join Date: Mar 2003
Location: The Dirty South
Posts: 449
Received 22 Likes on 6 Posts
Originally Posted by gums
Salute!

Thanks, fdr, I needed that.

I do not feel we are getting too far off topic. The mods shut off this on the Tech log for some reason.

Were I on one of the legal team experts for the forthcoming trials I would make a big deal about what we have been discussing lately here. The management and certification process is a different aspect, IMHO.

Gums sends......


I don’t know why they shut if down either. Perhaps the over abundance of ‘enthusiasts’ ?

Re. FDR link. - Dickson will do as he’s told.


JPJP is offline  
Old 17th Sep 2019, 16:56
  #8 (permalink)  

Only half a speed-brake
Thread Starter
 
Join Date: Apr 2003
Location: Commuting not home
Age: 46
Posts: 4,321
Received 3 Likes on 3 Posts
Mods: thanks for moving this. I'll need to clear some mess caused by the overly agitated choice of the term "haptic feedback".

Not even "tacile reactivity" would do, for what it is worth the bus has a dull, spring-loaded stick with force proportional to its displacement from neutral. Which is still full eons ahead from the claimed nintendo / microwave similarities.

Public thanks to fdr for the PM received, thrilling knowledge and shared experiences again.
FlightDetent is offline  
Old 17th Sep 2019, 18:50
  #9 (permalink)  
 
Join Date: Jun 2009
Location: florida
Age: 81
Posts: 1,610
Received 55 Likes on 16 Posts
Salute!

Guess we can bitch, whine and moan over on a new thread about "feedback" for control inputs versus aircraft response and aero surface displacement, huh?

I finally looked up the "haptic" entry on Wiki:
Haptic technology, also known as kinaesthetic communication or 3D touch,[1] refers to any technology that can create an experience of touch by applying forces, vibrations, or motions to the user.[2]
I really like that term "kinaesthetic communication: so I'll use that from now on to impress the newbies and wannabees ( lOL, jez kidding)

Brief background ++++++++++ I learned to fly in planes such as the Champ,Luscombe and Taylorcraf, that all had piure, basic control systems.Then to T-37 and T-33. I could feel every pressure and even "buzz/burble" in some portions of the envelope due to impending stall and/or high mach flow. Moved on to the Century series and irreversible hydaulic controls where the ailerons and elevator could not provide any feedback to the stick. We used springs like the 'bus for aileron, plus bellows to make the stick stiffer in pitch at higher dynamic pressure. Retired using a stick with zero feedback and 1/8 inch of movement to the 35 pound pitch limit and 17 pound roll limit.

I take some small issue with Detent's view, in that there are several control systems for personal flight simulators that employ "force feedback" and not simply springs. Early others used the same spring implementation as the 'bus and the VooDoo and SLUF I flew - the more you move the stick, the higher the force reuired. You know, the basic spring equation where the force required is proportional to the square of the displacement and the spring constant. The biggest advantage the 'bus has over the Viper is the physical movement adds another "kinaesthetic communication" channel ( LOL).

To be honest, I prefer the small movement, if any, and primarily force transducers to command the plane's comtrol surfaces. Most of the precise profiles, tracking scenarios and timy corrections flying close formation experiences I had did not involve moving the stick hardly at all, and it was all pressure on the stick. As technology for hydraulic valves advanced, I could see the difference between the T-33 aileron boost and the A-7D system. The Viper was different, as the stick pressures I used did not directly move a hydraulic valve component.

The visual and kinetic feedback we get nowadays in the pure FBW planes has been shown to be more than enough to maintain aircraft control and execute very precise maneuvers. If the regulatory agencies wish to require control forces or "feel" related to AoA and gee and roll commands, my feeling is we have to go back and come up with new requirements for certification. I point to the Airbus 320, and we have had 30 years of FBW for that plane to examine. That's my story and I am sticking to it.

Gums sends...
gums is offline  
Old 18th Sep 2019, 15:29
  #10 (permalink)  
fdr
 
Join Date: Jun 2001
Location: 3rd Rock, #29B
Posts: 2,956
Received 861 Likes on 257 Posts
The TU Delft findings should not have been a surprise to anyone who looks at the control architecture of the Airbus that was evaluated. It has a PI control loop running C*. so that is pretty straight forward, and indicates that the aircraft has no speed stability in the normal control. Boeing uses a C*U program, where the "U" is the speed stability component. The Boeing driver sets the reference speed with the trim switches, analogous to a conventional aircraft in outcome, the plane will pitch to the trimmed speed (AoA). As the Airbus is neutrally stable at all times (it doesn't really care about speed as a control target, it is driving a path determined by g or pitch rate... it goes where you ask it to) the only feedback that is needed is a suggestion to the pilot as to the fact that he has a force above breakout applied to the SSC. Airbus also provides envelope protection cues in low speed/high AOA cases, for longitudinal, and with an over bank case for roll channels.

Feedback for being "off speed" is inconsistent with the g/pitch rate of the C* system. The asynchronous controllers however leave the other pilot out of the loop as to what inputs are being made, other than on the ground where the ADI cues are used. Gulfstream has crossed the streams by applying synchronous controllers, with haptic feedback. I fly a much older G that is fly by cable, so am not aware what control logic was used, but haptic input by G would only make operational sense in normal flight for a C*U type system using SSC's, or by the C* logic of Airbus for envelope protection.

Personal experience on the 320/330/340-300 and 600 vs more experience on the B777 and 787 makes me appreciate the Airbus system in normal flight conditions. In wind shear, crosswinds, turbulence or with control degradation, Boeing is more comfortable to my tastes. The Boeing C*U has a higher tracking demand for the pilot, but there is feel for the dynamic response of the aircraft, whereas with the Airbus, intervention in perturbations to the flight path can lead towards APC problems (PIO's by another name). Both design architectures can give wild rides when they go weird from sensor failures or any other anomaly that goes outside of the expected failure modes, the 772ER/RR ex Perth was nearly as unpleasant as Kev S's A330 oddity with QF072. Both made a mess of the interior, heads, etc and gave the crews interesting problems to deal with.

Aside from the active SSC on the G500/600's the Dassault FBW offers a slight variant to the Airbus system, which is system behaviour feedback such as vibrations, as cues.
fdr is offline  
Old 18th Sep 2019, 18:45
  #11 (permalink)  
 
Join Date: Jun 2009
Location: florida
Age: 81
Posts: 1,610
Received 55 Likes on 16 Posts
Salute!

@FDR, et al....
I would bet that the 'bus is statically stable in pitch in the direct "law". Not "speed/AoA neutral" stable In other words, take out the "protections" and have the stick command control surface movements directly. Relaxing the stick to neutral returns the surfaces to whatever "trim" posiiton existed before moving the stick. Gotta look back at the FCOM for that mode/law WRT trim, but think it was control surface position as we had in planes since Glenn Curtis, Boeing brothers and Kelly.

The Viper also appeared neutral WRT speed/AoA when close to one gee flight, but like the 'bus, that was due to the FBW control law and not basic aero characteristics, pitch moments, etc. Using our pitch override function when in a deep stall resulted in "direct", proportional movement of the stab, just like the 'bus. In short, we could tell HAL to go take a leap until AoA below 30 degrees.

Gums sends...
gums is offline  
Old 19th Sep 2019, 06:12
  #12 (permalink)  
fdr
 
Join Date: Jun 2001
Location: 3rd Rock, #29B
Posts: 2,956
Received 861 Likes on 257 Posts
Originally Posted by gums
Salute!

@FDR, et al....
I would bet that the 'bus is statically stable in pitch in the direct "law". Not "speed/AoA neutral" stable In other words, take out the "protections" and have the stick command control surface movements directly. Relaxing the stick to neutral returns the surfaces to whatever "trim" posiiton existed before moving the stick. Gotta look back at the FCOM for that mode/law WRT trim, but think it was control surface position as we had in planes since Glenn Curtis, Boeing brothers and Kelly.

The Viper also appeared neutral WRT speed/AoA when close to one gee flight, but like the 'bus, that was due to the FBW control law and not basic aero characteristics, pitch moments, etc. Using our pitch override function when in a deep stall resulted in "direct", proportional movement of the stab, just like the 'bus. In short, we could tell HAL to go take a leap until AoA below 30 degrees.

Gums sends...
It is statically stable, with a mild phugoid around the speed trimmed by the driver on the THS wheely thingy, being dynamically stable. Thats in direct law. By design, the C* doesn't trim to any speed as such, so it is neutral. Boeing (and your Viper I guess) using a speed trim component, C*U are statically stable in normal mode. The Boeing without confusers is still stable, your Viper, has the single use alternate landing mechanism.

To get to be unstable statically (and dynamically too...) apparently takes a large loading error such as the positioning of an MRAP through the aft pressure bulkhead of a B744F. Once you get to the neutral point, it is all pretty interesting, but not necessarily for very long. Investigated a 744F incident that went back more than 10% aft of the envelope, it was still just forward of the neutral point but dang if those elevators didn't get a workout. On landing the aircraft 4 pointed its way onto the grass, NLG was about 6' shy of the concrete. On the takeoff the nose gear came off the ground without control input at 56KCAS... and the plane rotated off the ground by itself before V1, should have sold tickets for the airshow. The plane had a bank on all by itself which was why the tail didn't strike the ground, it was jacked up by pivoting around the wing gear. What the crew did then was a story unto itself.

I have been out the back of the envelope but also out the front, the forward one ended up having to split the flaps to be able to gain a trim for landing. 6.5 tons from the aft cargo got put in the fwd cargo, which doesn't sound like much... rotate took about an extra 3,000' of pavement which happened to be available, with full elevator and continuous ANU trim as soon as the rotate didn't work as advertised.

Off topic: Came across the photo of your Viper with the "roll enhancing feature", that is up there with the IDF's F15 clipped wing for a surprise happy ending.
fdr is offline  
Old 19th Sep 2019, 17:09
  #13 (permalink)  
 
Join Date: Jun 2009
Location: florida
Age: 81
Posts: 1,610
Received 55 Likes on 16 Posts
Salute!

@fdr, et al

Vipers, until late 90's, and with gear up ( normal law) had no AoA bias on pitch gee command until above 15 degrees, and then gee was limited by increasing AoA until 27 degrees or so and gee was limited to 1 gee. So in essence, we had neutral speed stability. However, the basic plane in "direct" law had negative static stability until about 0.95 M

I could not find a lotta difference in the Airbus laws WRT pitch except the pitch attitude WRT local level and the AoA limiting such as we had. If we climbed at 30 deg and had trimmed for 1 gee, then that is what we felt, and the nose would rise to preserve that 1 gee. Hence, steep climbs could help you getting into a deep stall when elevator ran outta authority to get nose down.

The mechanical uber alles folks should take a ride with me and then a good debrief to unnerstan.

Gums sends....






gums is offline  
Old 19th Sep 2019, 17:20
  #14 (permalink)  

Avoid imitations
 
Join Date: Nov 2000
Location: Wandering the FIR and cyberspace often at highly unsociable times
Posts: 14,576
Received 425 Likes on 224 Posts
There was I, thinking that haptic feedback was a product review by a someone representing a religious sect....
ShyTorque is online now  

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off



Contact Us - Archive - Advertising - Cookie Policy - Privacy Statement - Terms of Service

Copyright © 2024 MH Sub I, LLC dba Internet Brands. All rights reserved. Use of this site indicates your consent to the Terms of Use.