First principles analysis of the 737 Max crashes by a product guy.
29th October 2018 — A Lionair Boeing 737–8 MAX, registration PK-LQP performing flight JT-610 from Jakarta to Pangkal Pinang (Indonesia) with 181 passengers and 8 crew, was climbing out of Jakarta when the aircraft reached a maximum altitude of about 5400 feet, then lost height, radar contact was lost about 35nm northeast of Jakarta over the Java Sea. No survivors were found.
10th March 2019 — An Ethiopian Boeing 737–8 MAX, registration ET-AVJ performing flight ET-302 from Addis Ababa (Ethiopia) to Nairobi (Kenya) with 149 passengers and 8 crew, departed Addis Ababa’s runway 07R and was climbing out of Addis Ababa when the aircraft levelled off at about 9000 feet MSL, radar contact was lost shortly after. The aircraft wreckage was found near Ejere. No survivors were found.
Same aircraft type, two similar occurances. This is my attempt at using science and engineering to understand possible explanations of the reports coming out from the two incidents and the role MCAS system could have played in the incidents. MCAS (Maneuvering Characteristics Augmentation System) was introduced by Boeing into the 737 Max aircraft after the aircraft received newer, higher efficiency engines which; kind of tend to get bigger in diameter because of certain engineering principles.
Bigger Engines
Lets see how jet engines looked some decades back. Below is a picture of the 737–100/200 variant carrying a PW JT8D low bypass gas turbofan engine which is considerably smaller in diamter than the engines you see today.
The low bypass jet engines let much of the incoming air into the engine to be sent into what is called the engine core which is essentially the compressor, combustion chamber and the turbine, all mounted on a single axle. The exhaust air at high temperatures is forced out through the back of the engine to create a reaction force which moves the aircraft forward. Higher the volume of air used for combustion, higher the fuel burn, lower the overall efficiency.
Engine designers in the 60’s figured that allowing for a slower, larger fan to pull air into the engine and then combusting only a small portion of the incoming air in the engine core allowed for high fuel efficiencies. Less reaction force from a smaller engine core, more thrust from the larger fan up front — all leading to engines that are far simpler in design and easier to build and maintain.
Introducing Bypass Ratio — The bypass ratio (BPR) of a turbofan engine is the ratio between the air mass flow rate of the bypass stream to the air mass flow rate entering the core. The JT8D engine has a BPR of around 0.96 which means pretty much all of the air entering the engine is used for cobustion.
Now lets see how engines looked on the later 737–700/800/900 models (some of the most popular commercial aircraft on the planet) which came right before the infamous 737 Max series.
On the high bypass turbofan engines; much of the air entering the engine “bypasses” the engine core thereby allowing for the engines to use much less volumes of air for the actual combustion process and let the fan in the front of the engine create most of the thrust. The fan is driven by the engine core which is fed by the “non-bypass” air. The above image shows the 737–800 with a CFM56 engine which is a high baypass turbofan engine with a bypass ratio of 5.6 which means that only 17% of the air entering the engine goes into the combustion chamber, the rest of the 83% air bypasses the engine core.
On the 737 Max series, the CFM LEAP engines (photo below) are considerably larger in diameter. Guess the bypass ratio on this one? 12.5! Only 8% of the air entering the engine is actually used for combustion, the rest of the air is just pulled in by the rotating fan up front creating most of the thurst. The cores are so small that you can see through the engine (try this next time you board a new aircraft).
For the aviation enthusiasts out there; the turbo prop engines of the likes of the PW100 on the ATR 72 has a BPR of 50–60, although this doesn’t necessarily make them more efficient.
Engine manufactures are trying to make the fans up front create more thrust by designing bigger fans and better composite fan blade designs that would be driven by smaller, more efficient engine cores. Interestingly both of these have their own challenges, the first one being that the fan blade tips become supersonic beyond a certain engine speed & fan diameter. Companies like Pratt & Whitney envisioned the “geared turbofan” engines that allow the fan & the engine core to be mounted on separate shafts that spin at different angular velocities allowing the fan to run at lower speeds than the core. The new PW1000G that powers the A320NEO series aircraft is a geared turbofan engine. These engines have been in news for the wrong reasons due to multiple in flight shutdowns because of the complications resulting from gearbox lubrication issues at higher cruise altitudes. The second challenge is the need for engine core temperatures to be considerably higher to increase core efficiency forcing companies like GE to discover new materials like the Ceramic Matric Composites (CMC) to build engine core components to sustain higher core temperatures in their CFM LEAP engines which power the 737 MAX aircraft.
Design Choices
Now; what does all this boil down to? Bigger engines with large diameter fans. Because the LEAP engines are bigger, and because the 737 sits so low to the ground (a deliberate design choice that let the 737 serve small airports with limited ground equipment back in the 60's), Boeing moved the engines slightly forward and raised them higher on their underwing pylons. (If you place an engine too close to the ground, it can suck in debris on the ground while taxiing).
Airframe problem
I guess a design change and certification for a taller landing gear or a higher wing ( to accomodate for the larger engine) could have been more expensive & time consuming and would have been ruled out as a solution to the the bigger engine problem. A pylon redesign seems like a easier and cheaper fix.
Aerodynamic problem
This change introduced a new challenge in the way the 737 MAX handled in flight especially when the engines were creating thrust; the forward engines in thrust introduces a moment between the center of thrust and center of gravity of the aircraft, this pitch up moment is translated into a “nose light” handling condition which leads to the control system having insufficient elevator & stabilizer authority to pitch the nose of the aircraft downward; critical in the event of a impending stall. In simple terms, the new engines tend to make the aircraft go into a stall in certain flight conditions and pitching the nose down is the fastest solution to a impending stall.
Systems engineering problem
Boeing wanted a simple fix for the handling deficiency caused by the larger engines with minimal engineering changes, minimal flight crew & maintenance crew training. The easiest way was to add new features to the “Elevator Feel Shift System”; a stall protection system already available on the previous 737 model the 737 NG (New Generation) designed to increase the “pich feel forces” on the pilot’s control column by about 4 times the normal feel in case of a stall which effectively discourages the pilots from pulling back on the control column in case the aircraft detects a stall. A classic case of a accident caused by the pilot pulling back on the control yoke while entering a stall is the Air France Airbus A330 crash in the Atlantic. Watch Capt. Sully’s explanation of the accident here.
Enter MCAS — Maneuvering Characteristics Augmentation System
The MCAS is activated without pilot input and commands nose down stabilizer to enhance pitch characterists during steep turns, elevated load factors and during flaps up flight at airspeeds approaching stall. MCAS activates when AoA (Angle of Attack) exceeds a threshold based on airspeed and altitude. MCAS trims or tilts the 737 MAX’s horizontal stabilizer by 2.5 degrees in 10 seconds; the rate of tilt is a factor of airspeed or Mach number. The flight crew is not given visual warnings when MCAS changes flight parameters.
Sensor problem
As per the preliminary investigation into the crash of Lion Air flight 610 in October 2018, the flight crew had struggled with the aircraft’s Air Data Inertial Reference Unit giving unreliable airspeed & AoA (Angle of Attack) indications. The first officer’s IAS (Indicated Airspeed) indicated significantly higher airspeed than the captain’s airspeed indicator. Captain’s AoA incidicator indicated about 20 degrees higher than the first officer’s AoA indicator.
The AoA is measured using the AoA indicator vanes or the alpha vanes. The AoA indicator vanes are simple devices that measure the AoA of a aircraft using the difference between the mean direction of the relative wind striking the wing and the aerodynamic “chord” of the wing. Chord of the wing is a straight line joining the leading and trailing edges of the wing.
As a result of the high angle of attack values coming in from the AoA sensors, the stick shaker on the left side (captain’s side) activated. The stick shaker is part of the aircraft’s stall protection system which shakes the control column and makes a auditory alert. It has a electric motor which is connected to a unbalanced flywheel which shakes the control column alerting the pilots to a impending stall.
Design problem
In the Lion Air flight, on the basis of a non serviceable AoA sensor generated signal, the MCAS system pushed the aircraft’s nose down by trimming the stabilizer nose down soon after takeoff. We are yet to understand if and why the MCAS takes input from just one non-redundant sensor? Why was a conflict of Air data reference on two sides not considered in the design? Interestingly the data going into the flight computers could be corrupted if alpha vane sensors are damaged by hail, hitting the aerobridge at the airport, birds or multiple other factors. The Birgenair flight 301; crashed due to unreliable airspeed indications caused by a wasp nest inside the pitot tubes; equipment was a Boeing 757.
Training problem
Reports indicate that Boeing didn’t warrant airlines to train their 737MAX flight crews on the nuances of flying with a new system like the MCAS. In fact the pilots of Lion Air flight probably never heard about the existence of MCAS on their aircraft. With a failed AoA sensor giving confusing IAS readings in the cockpit, the crew must have been attempting to secure their aircraft from the impending stall or understanding that the sensor had failed and continued flight with the airspeed unreliable procedures they have trained for in the simulators. All the while without knowing that a automated system (MCAS) was triming their aircraft nose down uncommanded.
The MCAS trimmed the aircraft nose down again and the pilots on Lion Air trimmed the stabilizer nose up, however the MCAS system continued to push the nose down.
The pilot’s control column movement controls the elevators while the MCAS system manipulates the horizontal stabilizer pitch. How are they different? Check below.
Overall, on the Lion Air flight the stabilizer trim position increasingly moved towards nose down until it was no longer possible to counter the pitch down moment via the pilot’s control column inputs.
The only way the flight crew can control the stabilizer is through the thumb controlled stabilizer trim on the control column or the hand cranked wheel trim on the center console. The only way the crew could have subverted the MCAS was by cutting out the stabilizer trim cutout switches which takes away MCAS’s ability to electrically adjust the stabilizer trim.
Interestingly; the same problematic flight profile occurred on two previous Lion Air flights on the same aircraft before the ill fated flight but the MCAS system hadn’t caused as much nose down trim because those flight crew had figured out by chance that they had to cut out the stabilizer trim cutout switches to subvert whatever was causing the automated nose down attitude. They could quite possibly have done this because of the training they obtained handling the “runaway stabilizer trim” procedure which essentially means that they thought that they had a problem with the automated stabilizer trim system that is no longer serviceable and needs to be permanently switched off.
The FlightRadar24 data from the image below clearly shows the pilots on the Lion Air flight fighting to keep their aircraft from pitching down.
Stabilizer trim jackscrew
The horizontal stabilizer trim on most of the modern airliners is manipulated using a jackscrew and gearbox mechanism. When pilots turn the trim wheels, the aft cable drum turns a jackscrew through a gearbox. When the jackscrew turns, the stabiliser changes its pitch. During electrical trim, stabiliser trim actuator turns the jackscrew through the stabiliser gearbox using a DC motor. The stab trim cutout switches in the cockpit turns this electrical system off thereby isolating MCAS or any other automated system from the stabilizer trim system. The stabilizer jackscrew on the Ethiopian 737 MAX crash site was found to have the been configured for a nose down position, possible indication that someone or some system had trimmed the aircraft nose down in flight before the crash.
Overall; I am sure with the two crashes and the emergency airworthiness directive and from what we are learning from these accidents, there will be no 737 crew out there that does not clearly understand how the MCAS system works and how they could take full control of their aircraft in case of a emergency.
The above article is a explanation of my understanding of the 737 systems and aircraft systems that I learned as a student pilot and the information I gathered from the internet from both unreliable and reliable sources like my friend MentourPilot & Trevor. I could be wrong, I stand corrected in case of a clarification. That is how science works.
The purpose of this article is to go into the incidents with a slightly deeper technical understanding of the systems involved and shed light on the story behind the headlines.
