On the 9th of August 2023, a VX4 crashed at Cotswold Airport in Gloucestershire during a test flight over the runway.
The VX4 is a prototype air taxi, designed and manufactured by Vertical Aerospace Ltd based in Bristol in the west of England. There’s some confusion (at least, I’m confused) about the naming conventions. The aircraft registered as G-EVTL is referred to in the accident report as a VA-1X (not to be confused with their very first prototype, a quadcopter called VA-X1). In company documentation from 2020, the prototype aircraft is initially referred to as a VA-X4. In the media and the current website, it’s called the VX4. In any event, G-EVTL is the fullscale prototype of an electric high-wing aircraft with eight propellers which is capable of vertical take-off and landings (eVTOL). The five-seater aircraft (pilot and four passengers) is pitched as an urban air taxi, with plenty of room for luggage. Vertical Aerospace say that it is 30 times quieter than the equivalent helicopter with zero operating emissions. The prototype completed its first tethered flight in 2022 and aims to be the first certified winged all-electric Vertical Take-Off and Landing aircraft. They already have orders for up to 1,000 aircraft, with customers including American Airlines, Virgin Atlantic and AirAsia.
The prototype is designed as a piloted aircraft but the initial test flight in July 2023 was unmanned. The prototype lifted off, hovered and flew at a speed of 40 knots before landing. Aviation photographer Michael Stappen was at Cotswold Airport and got a photograph of that first flight.
Photograph courtesy of Michael Stappen
The goal is to offer safe and energy-efficient city transport with cruise speeds of 150 mph and a range of up to 100 miles. Vertical Aerospace stated at the time that this marked their next phase of testing: crewed flight missions. Luckily, the next flight was also unmanned.
The flight on the 9th of August was a test of the aircraft’s performance in the case of an engine failure during hover. G-EVTL has eight electric propulsion units (EPUs) powered by lithium-ion battery subpacks. Each EPU consists of a three-phase motor, an inverter and a thermal management system and drives one propeller. The eight carbon-composite propellers are placed on the wings: four on the leading edge and four on the trailing edge. The forward five-bladed propellers tilt in an angle range of 0° (straight ahead) to 100° (slightly past vertically upwards), allowing the aircraft to shift between vertical lift and forward propulsion. The rear four-bladed propellers are fixed into place to point vertically upwards.
For the unmanned test, the pilot controlled the aircraft from a remote cockpit with a curved screen offering a panoramic view of the instruments. A second pilot maintained visual contact with the aircraft. Additional team members monitored the aircraft status and were able to quickly speak to the pilot as needed.
G-EVTL lifted off with all EPUs operating. The front tilting propellers are numbered 1 – 4 from left to right. Once stable in ground effect, the pilot shut down EPU1 (left-side outboard) to simulate the performance in the case of an engine failure. Then the pilot carefully climbed to 30 feet above the runway.
The system response was what they hoped for: the prototype remained in a stable hover. The pilot kept the aircraft hovering at 30 feet above the runway for 10 seconds. The pilot then increased acceleration with a target of seven knots ground speed. But as it passed through 2 to 4 knots, there was a loud POP. A propeller blade flew off of EPU3. The right-inboard pylon (pylon 3) fractured and the aircraft fell at 19.5 feet per second to crash into the ground. The right wing broke off and the nose gear collapsed on impact. In a second and a half, the flight test had unexpectedly failed.
From the accident report: G-EVTL after the accident
The team carried out their emergency response plan. The airfield Rescue and Fire Fighting Service quickly arrived and used a thermal camera for any evidence of the lithium-ion batteries overheating. The second pilot put on protective gear and approached the aircraft with a “high-voltage trained hook man” to shut down the electrical systems. When they turned off the high voltage system, the battery contactors opened, disconnecting the system from the batteries. Then they connected an ethernet cable to the aircraft so that they could use their laptop to troubleshoot the system.
Once the battery pack voltages and temperatures remained normal over three hours, they proceeded to recover the aircraft. Well, the report says “recover” but I’m not sure G-EVTL would feel the same, as they had to cut off the broken right wing.
Vertical Aerospace confirmed that there was no additional vibration or loads on the blade when it detached. There was no sign of a foreign object in the video footage. They quickly released a preliminary report and worked closely with the AAIB. In the days after the crash, the CEO announced that “…all the lessons that we’ve got that will affect passenger safety and aircraft design that we think the whole industry can benefit from, we’re very keen to share.”
The forward EPU propeller blades have an external sheath adhesively bonded (glued) to a carbon fibre spar which is connected to the propeller hub. The detached sheath of the EPU3 blade had most of the adhesive on the internal surface, with very little adhesive remaining on the blade spar.
From the accident report: Released propeller blade including failed spar-to-sheath adhesive bond (lower three images, viewed in direction ‘A’) (courtesy of manufacturer)
Once the adhesive bond failed to hold the blade sheath to the spar, the bending load on the blade spar fractured.
Investigators were happy that the aircraft continued flying after the EPU had been shut down and that this had not caused the blade to detach. They focused on what happened after the blade separated. The propeller, spinning at 1,200 rpm, caused pylon 3 to twist up while the propeller kept rotating. When the pylon broke, it damaged the wiring harness. The EPU3 motor wiring pulled out of the connectors. The video footage showed a bright spark of electrical arcing between the high voltage power cables and the connector body. This led to EPU4 losing its tilting function; as a result the flight computer spooled it down, as designed for a tilt fault. The rear propeller directly behind EPU3, EPU7, also shut down, likely caused by an inverter resetting.
As a result, there was no longer enough vertical thrust, although, to be fair, the prototype remained level as it descended towards the ground. A vertical descent rate of 19.5 feet per second is twice the maximum descent speed and it collapsed under the impact.
The adhesive bond seemed to have failed as a result of progressive degradation. Vertical Aerospace found two similar propeller blades, meant to be used as spares, and inspected them using CT scanning (computed tomography, basically a series of x-rays). Both blades also showed large unbonded areas as well as variations in the shape of the blade spar. They concluded that the blade structural design and the quality assurance processes had been a contributing factor to the blade detaching.
The AAIB concludes:
The blade released from EPU3 was caused by a failure of the adhesive bond between the propeller blade sheath and spar. It is likely that defects introduced in the bond when the blade was manufactured grew progressively larger during the blade’s operational service to the point that the remaining bond area was insufficient to retain the blade under normal operating loads.
Large out-of-balance loads generated by the blade release caused structural failure of the right inboard pylon, resulting in damage to the aircraft’s wiring harnesses. This caused a loss of thrust from motors 4 and 7. Whilst the aircraft’s flight control system was able to maintain a level attitude, the high rate of descent caused by the loss of vertical thrust resulted in substantial damage to the aircraft when it struck the ground.
All of the remaining blades, referred to by Vertical Aerospace as “Generation 1” have been withdrawn from use. Vertical Aerospace was already in the process of introducing Generation 2 blades which do not suffer from the same risk of bonding failure. In total, they identified 36 product and process improvements as a result of the accident. Vertical confirmed yesterday that the Generation 1 blade would not be used for any further prototypes and that they are no longer using the same supplier.
The CEO said that this would not impact their goal of having the VX4 certified by 2026 and that he believed investors would continue their support of the aircraft.