If you replace conventional controls with a side stick or similar 4-axis controller, and then need two pilots to fly some manoeuvres - how is that an advance?

Early Sikorsky FBW development
 

Good question Sycamore. There were two of us that did most of the front seat flying. The learning curve for flying the 4 axis system was very flat. By the end of each flight we were pretty accomplished flying basic tasks. Little of that was retained between flights. More complex tasks, ie. descending decelerating climbs and turns, quick stops, precision and confined area landings were more than difficult. To say the least we were operating in the lower right hand corner of the Cooper Harper Scale!

The 3 plus one configuration was very similar to a basic helicopter control system. Similar to flying in the front seat of an AH-1 Cobra. Having yaw control on the side arm controller resulted in unwanted yaw control input while performing more complex maneuvers. Having the ability to make small trim changes in collective from the cyclic was a plus.

I spoke with Bill Dana of X-15 and F-16 AFTI fame. He believed that most flying tasks could be achieved by ensuring that the aircrafts basic handling qualities were optimized.

Hope this answers some of your questions.

Was in a discussion with a person closely connected with the MHP qualification testing and who noted that one of the previous posts included an error as to the configuration of the Comanche controller vs the MHP controller. The Comanche had a three axis ( pitch/roll/yaw controller while the Controller in the MHP has only pitch and roll, with the yaw axis controlled by the pedals.

My recollection is a little hazy but I thought everything after Comanche was 2 axis cyclic. We had an old 3 axis controller we used for the X2 but we didn't use the yaw axis, just conventional pedals.

Back to this thread subject for a moment, IFMU, but several of the media reports say that the Canadian military authorities are getting the 92 MHP ships back flying having changed nothing but some procedural information in the operators manual etc. That tends to indicate to me that there isn’t any question of whether the pilot can overcome any Flight Director/Autopilot inputs. Am I missing something?

I don't think so. Manipulating the controls while the FD is coupled, in any type of aircraft, is a recipe for surprises.

I dont remember the exact wording but from memory the information sad ”we know exactly what happend and why”. If they withdraw the operational pause I’m positive Sikorsky has issued a AD/EAD or something like that telling aircrew how to operate to not get in the same trouble. Most likely it would say ”Do not use fly through with F/D /Upper modes coupled” or maybe with a combination of maximum pitch/bank attitudes(?) as the information had an part of ”not tested flight regime” or something like that.
All this should be followed by updated FBW software later on, I guess ?

I agree with 212man, when manipulating controls with the FD coupled, it can be scary. The S92, with standard control have already limitations for when engaging the FD in correlation of speed, height and flight conditions. It is also the same for other manufacturer. Many incident reports have proven that already. So maybe is not only related to the FBW system, but a lack of knowledge.

There is substantial difference between analog and digital FBW.
 

While you comment on analog and digital flight control systems, my point is there is difference between analog and digital FBW. While the analog FBW transmit signals at different voltages a digital FBW transmit signals by binary values (0 and 1). The first could be exposed the disturbance from outside while the latter is non effected by external sources as packages of information would have control logarithm established.

My understanding is that it was the A320 in 1987 and B777 in 1994 thar where the first commercial aircraft to utilize a digital FBW.

The NH90 for example has a FBW system, although it has two digital lanes and two analog lanes, so it can only do what an analog computer can do (basic damping and such), as Nick Lappos mentioned in an earlier post.

And further:

“2nd Aug 2006, 02:56
(...) The basic architecture of the NH 90's FBW is that all computations are made in a pair of analog computers that match the outputs of a pair of digital computers. Any miscompare shuts down that lane. This means that the FBW on the NH90 is incapable of doing anything that can't be done with resisters and capacitors. It is therefore relegated to damping and making normal helo flight controls, basically 1975 flight controls, in spite of the fact that it is a FBW.”

So FBW is divided into Analog FBW and Digital FBW. It is not the same and represents two very different ways of FBW.

No, thats not correct

NH90 has four FBW-computers. Two digital and two analog. The two digital are the ones providing normal handling and upper modes. Digital#1 is the one in charge normally with Digital#2 in stand by, with no degradation in performance if #1 “goes sick”.

Analog #1 and #2 is pure backup, with reduced handling qualities, more or less no decoupling and no upper modes.

The four lanes isnt connected to a specific computer but is handled by a actuator control computer (two
ACC’s actually) that selects what signals to send to the actuators.

= NH90 have digital FBW with analog backup.

Analog FBW computers as main FBW Computers should be a long time since put in
Aircrafts.
Im sure all three mentioned have digital FBW.

Thx - with simple redundancy (two systems) on the digital FBW incorporated into the NH90 could that aircraft be certified to the latest requirements with regards to FCS?

Other FBW aircraft are certified along latest requirements with triple redundancy (four non degraded parallel systems).

With a strongly degraded backup solution on the NH90 compared to a triple redundant digital solution I would still claim there are differences between an analog and digital FBW of which NH90 might be considered a hybrid.

This FBW system takes around fifteen minutes to make. It matures for about twenty years with no costs to the operator. When ready it takes about £30,000 to programme it to fly your helicopter. It then requires between £30k to £80k to maintain it but it can be upgraded to your new helicopter at no cost apart from administration. It will, with care, last about forty years against replacing an existing system costing many millions every ten years.

It's called a pilot.

15 minutes? Show off.

Pitchlink, please put aside for a moment how the control laws are processed in the FCCs. This is immaterial to my point. My concerns the Cyclone are with with the cyclic inceptor configuration.

Imagine a conventional mechanical flight control system, but with the autopilot actuators in series with the linkage, versus in parallel as is conventionally incorporated.

When the autopilot would then command the aircraft, the cyclic stick would not move. Not only would the pilot have no tactile or visual indication of the auto pilot inputs form the cyclic, the pilot would also not have any indication of how much control authority was remaining.

This is how the Cyclone autopilot architecture functions. Digital versus analog, makes no difference.

I was closely associated with the S92FBW and CH148 flight control development and would like to weigh in on a couple of speculations.

It would be very difficult for control margins to be an issue. First, more than adequate control margins were demonstrated through the full flight envelope, which included both steady state trim conditions and extreme maneuvering.

While the controllers are indeed passive unique trim design, the CH148 incorporates an envelope cueing system, which provides both aural and visual cues to control margin encroachment. The most common reason for activating these cues would be if the aircraft was being operated outside of the operational CG envelope.

Some argue that proportional control position feedback is essential to safely operate, but there have been frustratingly too many accidents where pilots found themselves outside of CG but didn't recognize control positions being near limits. For that reason, cueing would be an improvement. Upon weight on wheels, the FBW system controller converts to a proportional controller - valuable during slope landings, but the cueing system excelled in warning the pilots nearing limits so they can have enough control remaining to recover off an excessive slope.

The autopilot uses very little control envelope, so providing proportional feedback is really of little value except to show it is doing something. Further, I would defy anyone to be able to simply look down at the cyclic position and tell me how much control is remaining. You know it when you hit it, but you can't really tell when you're 10% from the stop.

A couple other tidbits:
The controller configuration is 2 axis center mounted unique trim cyclic, floor mounted unique trim pedals, and conventional, trimmable, displacement collective.

FCC architecture is "dual-dual-triplex". There are 3 FCCs, each FCC has two lanes (either one can operate the servos). Each lane has dual processors, and they incorporate dissimilar software to address common mode failures.

RVGuy, please take this response as a desire for an open discussion of different system architectures, not an attempt to tie unique trim cyclic to the cause of this accident.

The S92FBW development preceded the Bell 525 by over two years. So when developing the 525 FBW architecture, Bell studied it in depth. A unique trim cyclic is obviously the lightest and least expensive configuration. But is it optimal for pilot cues?

During informal discussions with the FAA, Bell was advised that the FAA would not certify a unique trim cyclic configuration for a helicopter. Primarily due to the lack of tactile cuing between pilot and copilot. But also due to the lack of auto pilot tactile cuing. This is why the 525 while being a FBW side stick, still mimics the function of a conventional mechanical cyclic.

The USMC also had their reservations regarding unique trim cyclics. This is why they demanded installation of active BAE cyclics sticks that could mimic the function of mechanical mechanical controls in the CH53K.

This sounds very much like the Airbus throttles debate!

I forgot to mention, the Leonardo 609 and Bell 525 share the same FCC architecture as the S92FBW. Not surprising considering BAE provides the FCCs for all three aircraft.

Sounds worthy of its own thread. Would be interesting to find out the direct cause of the accident in this thread.

FAA and Controls
 

CTR wrote: "During informal discussions with the FAA, Bell was advised that the FAA would not certify a unique trim cyclic configuration for a helicopter."

Provokes a question. The 525 certification process was via " Special Conditions ", I think, as Part 29 and the attendant Advisory Circular do not yet address the totality of the FBW implications.If that is correct, then CTR is proposing ( I think? ) that everyone else is bound by a Special Condition agreed upon by Bell and the FAA, but no one else, that is not yet published as 14 CFR Part 29 and therefore is not law?

John,

Could you please rephrase your question? After reading it, I felt like I was in a “Who’s on first” Abbott and Costello routine. ;-)

My response was indeed to dispel your attempt to link lack of active control margin feedback to the accident. It is important for current operators of that aircraft to know there is more than adequate control margin throughout the operational flight envelope.

For those not familiar with the term, “unique trim” refers to a controller which resembles a joy stick. When relaxed, it is spring/detent centered. Inputs move away from the center detent by a small displacement, typically one to three inches.

Now to your other comments. I was a presenter at the early certification meetings to educate the FAA as to the Sikorsky FBW approach. There is nothing in the FARs that would prevent implementation of a unique trim controller. The main issues with the non-traditional controller approach were in demonstration of FARs such as longitudinal static stability. With automatic trim follow up (a fundamental element of unique trim), the stick is always in the same position regardless of trimmed speed. So we came up with an alternate method to demonstrate stability.

If Bell decided against a unique trim approach, that was their decision based on an assessment of the effort required to meet the certification challenges, not an FAA mandate. BTW, all the publicly accessible photos of the 525 cockpit show a sidestick with very little space for proportional control movement, at least with acceptable sensitivity. Makes one wonder, if it isn’t unique trim, what is it?

There are indeed benefits and challenges to the unique trim approach, and I have presented a couple of papers on the subject. The challenges, however, do not generally have to be mitigated by incorporation of active feedback. The control strategy with a unique trim stick is quite different than a proportional controller. If you look at time histories of a flight with a unique trim stick, you’ll notice the stick is in detent about 80% of the time. The stick moves in and out of detent at a relatively high frequency. Thus requiring the non-flying stick mirror the movement of the flying pilot is of questionable value.

Autopilot feedback? Think about what happens when you are coupled to a flight director airspeed and drag off speed with stick input. When you release the stick, do you respond to the subsequent pitch movement because the stick is moving forward or the nose is pitching down? I suggest it’s the latter.

I have considerable time in experimental helicopters with both active and passive unique trim sticks and, comparing the two, I didn’t find the active stick to provide a demonstrable improvement. And there are some nasty failure modes embedded in the active stick architecture that need to be addressed in emergency procedures.

Lastly, your contention the USMC “demanded installation of active BAE cyclics sticks” is false. Sikorsky convinced the Marines that there was potential in the active stick to incorporate tactile cueing, and if they didn’t invest in the technology now they might not be able to retrofit it later. The aircraft could handle the approximately 30 lbs weight penalty and the cost was just lost in the scope of the entire program. They are unique trim sticks and do not “mimic” the functions of mechanical controls.

CTR-sorry-it was a longer,involved sentence:

“Provokes a question. The 525 certification process was via " Special Conditions ", I think, as Part 29 and the attendant Advisory Circular do not yet address the totality of the FBW implications.If that is correct, then CTR is proposing ( I think? ) that everyone else is bound by a Special Condition agreed upon by Bell and the FAA, but no one else, that is not yet published as 14 CFR Part 29 and therefore is not law?”

The point I was getting at was:

1. The FAA told Bell they would not approve a unique trim cyclic.
2. Therefore Bell did not provide one.
3. The clear implication was that a unique trim cyclic wasn’’t certifiable, i.e., if Bell couldn’t do it neither could anyone else.
4. But remember, this is a Special Condition applying to the 525.
5. Bell’s Special Conditions are not in 14 CFR Part 29, thus are not law as far as say Boeing or Sikorsky proposing a unique trim controller.

CTR, allow me to go further. In thinking about my response above I can see where it might be misinterpreted.

The Bell folks certainly have had meaningful experience with FBW technology. When the FAA advised them that they wouldn’t be certifying a unique trim controller, Bell could very well have responded ( if they so desired to ) that they were wrong about that and proceeded to explain why, and how their system would answer all the safety and human factors considerations etc. They have the experienced and knowledgeable engineering talent to back that up whereas in the rotary wing/vertical lift area, the FAA apparently does not. I write “ apparently “ advisedly, because if they did, three decades after FBW appeared in vertical lift, they would have amended the relevant certification requirements and qualification procedures, and they have not.

Response to RVGuy’s Comments
 

RVGuy,

Immediately responding properly to your extensive comments required time I did not have. I finally have some free time. Sorry for the delay.

To try and maintain clarity and continuity in my response, I have pasted in multiple sections of your postings in quotes.

“My response was indeed to dispel your attempt to link lack of active control margin feedback to the accident. It is important for current operators of that aircraft to know there is more than adequate control margin throughout the operational flight envelope.”

Please re-read my posts. I specifically state I am not attempting to tie unique trim cyclic to the cause of this accident.

“Now to your other comments. I was a presenter at the early certification meetings to educate the FAA as to the Sikorsky FBW approach. There is nothing in the FARs that would prevent implementation of a unique trim controller. The main issues with the non-traditional controller approach were in demonstration of FARs such as longitudinal static stability. With automatic trim follow up (a fundamental element of unique trim), the stick is always in the same position regardless of trimmed speed. So we came up with an alternate method to demonstrate stability.”

Prior to your discussions with the SW FAA Office to certify the S92FBW, I was working with the same FAA personnel on the certification of the FBW flight control system for the Leonardo 609.

Just prior to Sikorsky giving up obtaining FAA certification for the S92FBW, I was involved with developing the certificate basis for FBW system on the 525. By an interesting chain of events, in between working on both of these programs I turned down an offer from Sikorsky to work on the certification of the S92FBW. So while I do not question any of your statements validity, I believe they are based on your impressions with relatively brief meetings with the FAA. My statements are based on the sum total of over 20 years working to civil certify FBW rotorcraft. I have the mental scars to prove it ;-).

You are correct there is nothing in the FARs that specifically prevents the use of unique trim cyclics. But then again, there is next to nothing in the FARs for Part 29 aircraft regarding Fly by Wire. As you well know, the FARs are vague, and their interpretation is up to the discretion of the FAA. Recently, I have witnessed two identical Issue Papers submitted to the FAA for approval on FBW issues. The one that was sent to the NY office was approved, the one that went to the SW office was turned down.

RVGuy, you may have walked away from meetings with the FAA thinking that they had accepted unique trim as being acceptable for certification. However, prior to the 525 program we specifically asked the SW FAA if unique trim would be acceptable for Part 29 (without mentioning the S92FBW), and were told no. This is not secondhand, I asked the question myself.

“If Bell decided against a unique trim approach, that was their decision based on an assessment of the effort required to meet the certification challenges, not an FAA mandate. BTW, all the publicly accessible photos of the 525 cockpit show a sidestick with very little space for proportional control movement, at least with acceptable sensitivity. Makes one wonder, if it isn’t unique trim, what is it?”

While neither the 609 or 525 are yet FAA certified, they are both are much closer to achieving this milestone than the S92FBW ever got. This is not to say that unique trim is inadequate. But it has limitations, especially when transitioning pilots familiar with conventional mechanical controls.

You need to get the opportunity to sit in a 525 mock up or aircraft. You’ll be surprised at how much displacement the cyclic actually has. It was modeled initially on the Zulu Cobra side stick displacement. The Z is all mechanical with SCAS. With all that linkage and resulting lost motion, a very short stick would’ve resulted in an unacceptable dead band. So yes, the 525 mimics conventional mechanical controls, with proportional movement. This is best displayed on aircraft start up. Prior to rotors turning the 525 does not have hydraulic pressure to the swashplate actuators. Therefore, upon start up the cyclic stick moves to synchronize up with the swashplate position. The same goes for the 609.

As I mentioned before, the biggest concern with dual control unique trim the FAA expressed was with pilot to copilot coordination. When the FAA was asked why unique trim was acceptable on large commercial airliners but not for helicopters, pilot to copilot coordination when operating close to obstacles was the first concern. This is why although the 525 has side sticks, the two-sided sticks are mechanically linked.

Thanks to Nick Lappos’s brief stay at Bell, the 525 incorporates some advanced control features similar to what he pushed for on the S92FBW. But the FAA moves very slowly and excepting new technology. So many of the original features had to be stripped out in order to achieve certification. But it’s now only a software update to bring them back.

“There are indeed benefits and challenges to the unique trim approach, and I have presented a couple of papers on the subject. The challenges, however, do not generally have to be mitigated by incorporation of active feedback. The control strategy with a unique trim stick is quite different than a proportional controller. If you look at time histories of a flight with a unique trim stick, you’ll notice the stick is in detent about 80% of the time. The stick moves in and out of detent at a relatively high frequency. Thus requiring the non-flying stick mirror the movement of the flying pilot is of questionable value.

“I have considerable time in experimental helicopters with both active and passive unique trim sticks and, comparing the two, I didn’t find the active stick to provide a demonstrable improvement. And there are some nasty failure modes embedded in the active stick architecture that need to be addressed in emergency procedures.”

I agree concerning the potential nasty failure modes with many active stick architecture’s. This is especially a concern with dual controls, where force feel must also replicate pilot to copilot interaction. This is another reason why the 525 cyclics are mechanically linked. The single actuator that provides force feel for each axis can be kept to a very low load and rate, similar to conventional mechanical controls.

“Lastly, your contention the USMC “demanded installation of active BAE cyclics sticks” is false. Sikorsky convinced the Marines that there was potential in the active stick to incorporate tactile cueing, and if they didn’t invest in the technology now they might not be able to retrofit it later. The aircraft could handle the approximately 30 lbs weight penalty and the cost was just lost in the scope of the entire program. They are unique trim sticks and do not “mimic” the functions of mechanical controls.”

I had the opportunity to visit be BAE and get a demonstration of an earlier generation of the CH 53K active sticks. Additionally, I presented a paper at the SAE A6 conference, when NAVAIR present a paper on the CH-53K active sticks. My statement was based on what I was stated by both BAE and NAVAIR representatives. Also if you reread my statement you will see that I stated that active sticks were incorporated so that the cyclic controls ‘COULD’ mimic conventional mechanical controls, not would mimic. As was the case with the 525 certification, being able to mimic conventional mechanical controls reduced potential program risk.

Finally, I believe you were being kind by saying it was only a 30 lb hit in weight For active sticks on the CH 53K. For both cyclics and associated electronics, I believe it’s more than twice that amount.

Stay healthy,

CTR

PS John Dixon, I will respond to your posting as soon as I get some more time.

Good post, and I look forward to the response from the RVguy. You two are clearing the air a bit re FBW, and there was one issue touched on that has been a problem to me for decades-the “out of dateness“ of the longitudinal static stability requirement.

The idea behind the requirement sounds fine: it should require forward control input to go faster and vice versa. The associated “ stability “ part-that the stick position vs speed must show a positive/negative slope sounds fine, until one thinks about it. The assumption is that if the pilot doesn’t move the control, the aircraft will stay right there, on speed. Its stable..All is well. The reality is that when a change in speed is demanded, the control is moved forward ( for example ) then as the ship gets to the desired speed, the pilot makes the adjustments to stop the acceleration and fine trim the ship to the new speed. Its “ nice” if that measured position is forward of the original, but what matters is if the ship stays there.

In the modern fleets of machines there are some that stand out as having to add equipment ( sensors, actuators etc ) to meet the positive stick position slope: the CH-47 series, the UH-60, the S-76 and the 53E. DASH, Pitch Bias Actuators ( PBA ) etc. The UH-60 case stands out in my mind. There was a speed area at max aft CG where the stick position measurement in smooth air etc was flat to slightly negative. We had to add all the claptrap associated with the PBA. We tried to reason with the Army that the pilot would never know the difference ( if he didn’t have the PBA )because the AFCS had airspeed hold automatically locking on after 12 seconds ( i.e., the ship would stay where the pilot put it, speed wise ). Lost that argument. Scene shifts forward 3-4 years and the PBA subsystem was a frequent cause of maintenance actions. Had another meeting with the Army, who now knew that the pilots couldn’t determine any change in the handling and the PBA was removed.

My point is that if the aircraft stays where the pilot trims it, the stick position slope argument with modern control systems is baseless-the modern control system solves the original intent with modern technology. Solution on the other side of the discussion is to modernize the regulations.

CTR - Check your PMs please

Grizz

Grizz,

See response.

Stay healthy,

CTR

https://atlantic.ctvnews.ca/canadian...2020-1.5485958

Canadian Forces pilots not warned about autopilot before deadly Cyclone crash in 2020
By Michael TuttonThe Canadian Press Staff

Contact

Published Friday, June 25, 2021 4:34PM ADT Last Updated Friday, June 25, 2021 4:40PM ADT

https://www.ctvnews.ca/polopoly_fs/1...020/image.jpeg

From top left: Sub-Lieutenant Abbigail Cowbrough, a Marine Systems Engineering Officer; Sub-Lieutenant Matthew Pyke, Naval Warfare Officer; Master Corporal Matthew Cousins, Airborne Electronic Sensor Operator; Captain Maxime Miron-Morin, Air Combat Systems Officer; Captain Kevin Hagen, Pilot; Captain Brenden Ian MacDonald, Pilot.
HALIFAX -- As a pilot guided one of Canada's navy helicopters up into a tight turn, neither his training nor cockpit indicators warned of how a built-in autopilot would take control and plunge the Cyclone into the Ionian Sea, a military report has concluded.

All six Canadian Forces members on board died in the crash on April 29, 2020.

According to a board of inquiry report obtained by The Canadian Press, when the pilot was flying the turn, commonly called a "return to target," he had pointed the nose up and used his feet to turn the helicopter's tail, overriding the autopilot to complete the manoeuvre of less than 20 seconds.

The report, however, said testing wasn't done during the aircraft's certification to identify what would happen if a pilot overrode the autopilot more than "momentarily" and in certain complex situations. "The automation principles and philosophy that governed the Cyclone's design never intended for the (autopilot) to be overridden for extended periods of time, and therefore this was never tested," it said.

This was the case even though -- as the report stated -- pilots are known on occasion to override the autopilot system without manually pressing a button on their control stick, called the cyclic.

The report said that at the time of the crash, the autopilot -- referred to as the flight director -- was set to an air speed of about 260 kilometres per hour before one of the pilots pitched the aircraft's nose upward for the turn.

It was supposed to fly back over HMCS Fredericton and practice hoisting people onto the deck. Instead, the frigate's CH-148 Cyclone helicopter crashed off the coast of Greece while returning from a NATO training mission. That crash caused the worst single-day loss of life for the Canadian Armed Forces since six soldiers were killed in a roadside bombing in Afghanistan on July 4, 2007.

The report indicated the crash might have been averted if the pilot had manually chosen to turn off the autopilot during the turn. But it also stated that it wasn't unusual for pilots to override the autopilot and there were no explicit instructions in the manuals on the necessity to manually turn off the flight director.

In addition, the report said the pilot appeared unaware the computer would attempt to regain control near the end of the turn.

When the helicopter flipped around, the report said, the pilot pulled back as far as he could on the cyclic, attempting to right the aircraft that the computer was flying into the sea. Within seconds, the helicopter hit the ocean at massive force.

The board of inquiry said it found no evidence the flying pilot recognized he had lost control of the aircraft until it was too late.

Critical to the crash, the report said, was the aircraft's software, which was certified by the military. If the autopilot is overridden, the computer accumulates digital commands, referred to as "command bias accumulation." The more commands a pilot sends manually to the computer while the aircraft is coupled with the autopilot, the more this bias accumulation occurs, the report said.

'PILOT'S ABILITY TO CONTROL AIRCRAFT WILL BE REDUCED OR LOST'

After a pilot overrides the air speed set by the autopilot, a "feed forward look" occurs, the report said, adding that in some situations, "the pilot's ability to control the aircraft will be reduced or lost."

The board of inquiry said the pilots' training didn't cover "with sufficient detail" certain risks of flying the aircraft, leaving the flyers unaware the autopilot would seek to keep control of the helicopter.

The return-to-target manoeuvre, which led to the crash, was being flown by others in the maritime helicopter community, the report said. That manoeuvre has been disallowed since the crash.

The report makes six recommendations, five of which involve better training for pilots to make them aware of the potential problems that could occur if they override the autopilot. It recommended creating special cockpit signals pilots could use to warn each other about overriding flight directors for extended periods of time.

The report also recommended the military consider an engineering change "to automatically disengage the flight director under certain conditions, such as when the flight director is overridden in multiple axes, or for an extended period of time."

According to a senior military source, that recommendation is not shared in a second, independent report by the military's Directorate of Flight Safety, expected to be released next week.

The second report said pilots must be well trained to almost instantly press a single button on their control stick to disengage the autopilot if they're not getting the response they want out of their controls. It said, however, that automatically disengaging the autopilot might pose its own risks in some situations, especially when a pilot believes the autopilot will keep functioning.

The second report instead argued that the software -- and its "bias accumulation" -- needs to be addressed by American aviation company Sikorsky Aircraft, the manufacturer of the Cyclone.

https://atlantic.ctvnews.ca/polopoly..._960/image.jpg

"We need to look at that software and see if we can eliminate this from the software altogether, being careful to understand when you make any changes like that you may introduce a butterfly effect and cause problems elsewhere," the source said.

The board of inquiry report, signed by three members of the panel on Nov. 20, 2020, concluded the pilots were not distracted and the crew "flew well together." It added that the aircraft captain had a strong command of the helicopter and the co-pilot showed "good situational awareness" throughout the mission.

The second report is expected to provide further analysis on the factors behind the crash.

The military source said the recommendations of the two reports must be meshed into a single set of findings for consideration by senior Royal Canadian Air Force officers.

A spokeswoman for Sikorsky referred all questions on the report to the Canadian Forces.

This report by The Canadian Press was first published June 25, 2021.