On 26th May 2003, a British Midland AIRBUS A-321 suffered severe damage from hail en route near Vienna.
This is the synopsis from the official report into the incident published by the UK AAIB.
"The aircraft was in the cruise at FL340 approximately 70 nm south-east of Vienna when it encountered an area of severe turbulence and hail. Some of the flight deck windows became crazed and other areas of the airframe suffered extensive damage although this was not apparent to the crew. The aircraft made a precautionary descent to FL230, in accordance with the required abnormal procedures, and continued the flight to its destination of Manchester. The crew had no indication or warning that the aircraft was about to enter an area of severe turbulence, associated with the upper levels of a Cumulonimbus cloud. When they had been using the weather radar to check the route ahead of the aircraft, sometime before the encounter, the radar returns appeared benign."
Here is an extract from the flight history which details what happened when the aircraft encountered the Cb:
"The radar was set to a scale of 160 nm and with no significant returns ahead and no thunderstorm activity forecast the radar was switched OFF. The aircraft had been in clear skies above towering Cu for most of the flight and, in accordance with normal procedures, the radar had only been turned on when required."
"As the flight progressed the aircraft entered some high Cirrus cloud. The FO, anticipating the possibility of turbulence, switched on the 'seat belt' signs and made a short public address (PA) informing the passengers and cabin crew that this was a precautionary measure. Shortly after the announcement the aircraft entered what the crew described as an area of 'light innocuous turbulence'. Approximately 20 seconds later however, the turbulence increased through moderate to become severe. The autopilot (AP), which was selected ON in the 'Navigation Mode' at a speed of Mach 0.78, disconnected and the aircraft climbed rapidly above its assigned level. Intense hail then began to impact the aircraft. Both flight crew noted the master warning light illuminate as the autopilot disconnected but neither pilot heard the associated audio warning due to the noise of the hail. The FO flew the aircraft manually, selected engine ignition ON, set the speed to Mach 0.76 for the turbulence and turned on the cockpit dome light. The commander changed the range selector on Navigation Display (ND) to 40 nm to check for conflicting traffic on the Traffic Collision Avoidance System (TCAS), monitored the aircraft's speed on his Primary Flight Display (PFD), monitored the first officer's side stick inputs and cancelled the master caution light. Throughout, the PF attempted to regain FL340 and maintain track. The aircraft however, deviated 1,300 feet above to 300 feet below its assigned cruising level, rolling to angles of bank not exceeding 18°. Indications on the Vertical Speed Indicator (VSI) confirmed that on at least one occasion the rates of climb or descent exceeded 5,900 ft/min."
"A Boeing 757 was approximately 25 nm behind G-MIDJ on the same track. The commander of the 757 had his radar selected ON and he could not only see the weather radar returns on his ND but also G-MIDJ displayed by his TCAS. He thought that G-MIDJ had been heading for the gap between two lines of thunderstorms displayed on his radar but realised the gap was closing as the storms were building. He continued using his radar and noted the rapid increase in altitude of G-MIDJ on TCAS as it entered the storm. Initially, in the absence of any radio traffic he initially assumed that G-MIDJ Airbus A321-231, was attempting to climb over the storm. He requested a turn to the right to avoid the weather, which was approved and passed safely clear of the storm activity."
"Moments later the commander of G-MIDJ transmitted to Budapest ATC informing them that they were unable to maintain FL340 due to severe turbulence. He was unable to hear the reply because of the hail. This also prevented the pilots from hearing each other for, although they were wearing headsets, it is normal practice for the intercom to be selected OFF and cross-cockpit conversation to be conducted without the use of intercom."
The report recommended that:
"Present guidance material not only suggests that, in areas of thunderstorm activity, readjusting the radar tilt frequently is the only way to monitor storm development but also that when the upper limit of the storm cell is determined it should be avoided vertically by at least 5,000 feet. The inability of weather radar to detect certain types of precipitation, associated with storm cells, in the upper levels of the atmosphere above 30,000 feet however make it impossible to determine with any accuracy the upper limit of a cell when its vertical development exceeds 30,000 feet. Calculations to determine the aircraft's clearance above the upper limit of a cell can therefore be inaccurate resulting in an aircraft entering the active element of a storm cell whilst attempting to safety over-fly it. It is therefore recommended that…The Civil Aviation Authority should consider reviewing their guidance material concerning the use and interpretation of airborne weather radar, with a view to highlighting the potential for displayed data to be unreliable when used for calculating the safe vertical clearance for overflight of active storm cells."