The recent pair of fatal crashes involving Boeing 737 Max 8s in Indonesia and Ethiopia has generated considerable debate about crew training for the type, sparked by the interim reports from the two investigating agencies. According to the reports, both crews failed to intervene successfully to save the aircraft following a technical malfunction.
In the spotlight are the training requirements for extending a pilot’s existing type rating on one variant in a manufacturer’s designated series, such as the 737, to cover an additional variant in the same series – for example, changing from a 737NG to a 737 Max, or an Airbus A320 to an A320neo.
Max 8s parked
Angel Medina G / EPA-EFE / Shutterstock
There is a basic question here that the industry needs to review, crew training experts say. Is the designated regulatory minimum training focused on the differences – between systems, performance, procedures and handling – sufficient to make pilots type-rated on one aircraft in a series competent to operate another in the same series? This is a critical issue, because this “differences training” enables pilots to add the new variant – legally – to their licence under the common type rating scheme.
Senior industry training experts are worried about one particular side-effect of the fierce competition between manufacturers – the battle to keep the costs of training crew to fly new variants as low as possible. Some believe that, although the letter of the law may be met regarding the “operational suitability data” that enables types in a series to be compared for training and operational purposes, the spirit of the regulations may be lost or pushed beyond intended boundaries. An example provided to FlightGlobal by a pilot rated on all the 737 variants is that “the switches [in the different 737 variants] are labelled the same but don’t deliver an identical result”.
There are other examples of “fudges” by manufacturers to make differences appear less significant than they are, and circumstantial evidence is gathering to the effect that the Maneuvering Characteristics Augmentation System (MCAS) – a system unique to the 737 Max – was one of those. Boeing admits that MCAS was designed to make the Max feel and handle like the NG and earlier models when it is being flown manually at low speed with flaps retracted. FlightGlobal’s test pilot Michael Gerzanics confirms the Max does indeed, in normal operation, feel like the NG to fly – although he is critical of the manufacturer’s opacity regarding MCAS.
If Boeing’s intention was to add a system to make the handling feel the same on the Max as it did on the NG, why did it not brief Max trainee pilots fully on MCAS during the “differences training” courses for airlines that took delivery of the first examples of the re-engined narrowbody? And why did it not include MCAS details in the flightcrew operating manual (FCOM)? Following the crashes, Boeing has admitted that it should have done both. The crews in the aircraft that crashed were taken by surprise when MCAS intervened unexpectedly, triggered by incorrect sensor inputs. In both cases, MCAS operated at a point when it was not designed to do so.
There is a significant factor here that has added considerably to the competition pressure Boeing faces. When Airbus went totally fly-by-wire (FBW) on all its new aircraft types with the introduction to service of the A320 series in 1988, it gained an advantage over Boeing in maximising the potential for pilot cross-fleet qualification (CFQ) between all types in its entire fleet. This goes beyond just the common type rating across variants within a series. Boeing now has two fly-by-wire (FBW) aircraft in its fleet – the 777 and 787 – so it is reducing this CFQ disadvantage.
The latest 737’s flight control surfaces, however, are still basically mechanically-controlled, with FBW systems like MCAS added to modify handling of the Max. But the Max also features another little-heralded FBW component called the spoiler control electronic unit (SCEU), which interacts automatically with the air data computer, flight control computer and spoiler lever. So the Max is a hybrid mechanical/FBW aircraft, no longer the simple airborne Jeep it was in its classic versions – yet its many devotees still talk about it as if it were.
Despite the considerable differences wrought by re-engining the Max and installing compensating FBW systems like MCAS and SCEU, until the 737 Max series was grounded on 13 March this year, the designated “differences training” that pilots had to undertake to extend their 737NG rating to the new variant consisted of a few hours of computer-based training (CBT) on a tablet computer, with no time required in any kind of flight simulation training device (FSTD).
The Lion Air 737 Max 8 crash occurred on 29 October 2018 and 11 days later the US Federal Aviation Administration (FAA) issued an emergency airworthiness directive (EAD) related to information emerging from the accident investigation. It stated that, “if an erroneously high angle of attack sensor input is received by the flight-control system, there is a potential for repeated nose-down trim commands of the horizontal stabiliser". This, without naming it specifically, referred to the action of MCAS. The system was not referred to by name in the FCOM, nor by describing its intended function as an automatic system.
The FAA EAD then stated its requirement for action by Boeing and 737 Max operators. This entailed “revising certificate limitations of the airplane flight manual to provide the crew with runaway horizontal stabiliser trim procedure to follow under certain conditions". The FAA was advising crews, in the event of an erroneous triggering of the MCAS, to carry out the runaway horizontal stabiliser trim procedure. For crews, perhaps an “MCAS malfunction” page might have been a more intuitive place to look for in the emergency checklist.
“Under certain conditions”, it transpires, the MCAS can indeed present the crew with symptoms that look momentarily like those of a runaway horizontal stabiliser trim, but in fact the cause is MCAS operation triggered – as designed – by an angle-of-attack reading above a certain value when the flaps are retracted (clean-wing). If that angle of attack were a true indication, it would be warning the MCAS that the aircraft is closing with the “clean-wing” stalling speed. In this small section of the flight envelope, when the aircraft is being manually flown and the angle of attack is registering above the trigger level, the MCAS is programmed to motor the horizontal stabiliser to push the nose down, reducing the angle of attack.
The MCAS is needed because of the significant changes in the Max compared with the NG, including a higher-power, heavier engine with a bigger fan. The size of the CFM International Leap-1Bs meant that, to maintain ground clearance beneath the cowling, its pylon had to site the engine further forward and higher relative to the wing leading edge. The nose undercarriage length also had to be increased by 8in, giving the aircraft a higher nose-up attitude on the ground and during landing. All these effects combined to change the wing/engine cowling aerodynamics and the engine thrust vector.
Without MCAS, the Max would have been vulnerable to stalling when flying under manual control at low speed with flaps up. According to a report filed to the Royal Aeronautical Society (RAeS) Flight Operations Group (FOG), this occurs because, in that small corner of the Max’s flight envelope, the aircraft had a tendency to pitch nose-up due to aerodynamic and power vector differences between the NG and the Max.
The unanswerable question – at this stage – is whether the FAA would have felt able to certificate the 737 Max with the knowledge of this increased stalling risk but without automatic pitch compensation from a system like MCAS. And even if it would have approved it without MCAS, the marked differences in handling characteristics between the 737NG and the Max in parts of the flight envelope might have required a “differences training” regime that demanded expensive pilot time in a FSTD – perhaps even a full flight simulator (FFS).
The report to the RAeS group, by a member fresh from carrying out the Max “differences course”, remarks: “The new system has a number of added (and mainly automatic) functions, such as Maneuver Load Alleviation, Emergency Descent (increased spoiler deflection), Elevator Jam Landing assist (uses spoiler deflection to augment pitch requirements) and Landing Attitude Modifier [which] helps pitch control on the glide slope at low flap settings, and is required to maintain clearance for the nose landing gear’s 8in extra length.” The reporting pilot said there was “no mention” of the MCAS on the course.
The conundrum the industry faces right now is that airlines are always happy for the manufacturers to produce new, more efficient variants of successful aircraft such as the 737, but they do not want to have to pay for their pilots to take time away from the line to train to extend their type rating to the new machine. For that reason the manufacturers work hard to minimise the changes, sometimes applying cosmetic fixes to mask genuine differences, to stay within the limitations that allow the common type rating for the series to be extended to the new machine.
Some influential people at the top of the training and flight safety world now believe that there needs to be a thorough review of existing common type ratings and cross-crew qualifications. It is understood that the International Civil Aviation Organisation is working on the subject, but if that is indeed so, it is not ready yet to talk about it. It would be a touchy subject for both the airlines and manufacturers, because it might highlight the need for more detailed “differences training” or even rule that a particular variant in a series has too many differences and thus demands a separate type rating.
On the other hand, the pressure to limit differences can be a constraint on offering potential technological advances that are ready for deployment, so a reality check might give manufacturers the opportunity to offer better products.
By the time this article appears the FAA will have presented to a gathering of significant national aviation authorities (NAA) a “software fix” for the MCAS, and proposed modifications for the differences training that take the new changes into account. On 16 May Boeing reported that it had completed not just simulator tests on the new systems, but more than 360h of flight tests across 207 sorties.
The FAA will certificate the modified MCAS and associated systems when it believes the USA-based 737 Max fleet can safely fly again. Information released so far indicates that it will involve the MCAS getting its sensor feed from both the aircraft’s angle-of-attack vanes, not just one as before, and a cockpit warning of “AoA disagree” if the two sensor vane readings diverge significantly.
The US Air Line Pilots Association stated on 8 May that it is satisfied with the proposed fix, and ventured its opinion that the 737 Max with the newly-modified MCAS will not need simulator time to conduct a differences course. It will be up to the foreign NAAs to determine whether they are sufficiently convinced of the new system’s integrity to withdraw their grounding orders, and to approve the proposed parameters for the differences training.
EASA type ratings guidance
The European Union Aviation Safety Agency (EASA) sets out generic parameters for judging how to manage common type ratings and decide differences training. “The Operational Suitability Data (OSD) system is part of the certification process," it states. "It evaluates and identifies the difference between a candidate aircraft and a base aircraft. Operators using mixed fleet should have approved Operator Differences Requirements (ODR) Tables that contain the following: ODR tables list all differences between a candidate aircraft and a base aircraft with regard to general characteristics (e.g. dimensions, MTOW, etc.), aircraft systems (e.g. hydraulic, electrical, etc.), and manoeuvres/handling characteristics. Each element is assigned a Difference Level for training, checking and currency. Differences training must address all of these differences at the appropriate Difference Level (e.g. through handout, aided instruction, flat panel trainer, simulator, etc).”