Third article only partial copy of the interesting part....
Following Clues To Solve The Icing Puzzle | Technology content from Aviation Week
"The focus of many of the 1990s’ thrust-loss studies had been on the AlliedSignal (later Honeywell) LF502-powered Bae 146, including a 1992 incident in which an Ansett-operated aircraft had lost power on all four engines over Western Australia. However, it was a 2002 event in the U.S. involving a McDonnell Douglas MD-82 in which the aircraft descended to 17,000 ft. before being able to restart its engines that produced one of the biggest clues. Although not caused by ice building up in the core of the aircraft’s Pratt & Whitney JT8D-217 engines, the evidence showed ice particles had blocked the inlet of a pressure sensor which sent an erroneous message to the autothrottle. The MD-82 was also equipped with SLD ice detectors, but because these did not trigger, the event became a turning point in the understanding that engine failure was more likely linked to ice particles.
“It was a new discovery, but in fact it wasn’t quite new,” says Strapp. “They knew about it in the 1950s because they had a problem with the Bristol Britannia and flameouts in its Proteus engines.” The issue began in April 1956 when two of the four turboprops on a BOAC Britannia flamed out at 20,000 ft. over Africa on a route proving flight to Nairobi. Kenya. The event was a mystery as the engine had successfully passed through intense ice- certification tests in Ottowa and, just as in recent events, no airframe icing was present. The only clue to the presence of ice particles was a thin white “witness line” along the null point on the leading edges.
As a result, Bristol, Rolls-Royce and the certification authorities “did a lot of work back in the 1950s characterizing the atmosphere. It was work that, in essence, we repeated. But it was no longer traceable and we didn’t know how accurate it was,” says Strapp. “However, they knew a lot about ICI [ice crystal icing] and by time we got onto HIWC in 2004 this was not common knowledge. We didn’t think you could get icing from just dry ice crystals. You can get the same conditions at turboprop altitudes if you are flying in the tropics, and they were,” he adds. Part of the reason the lessons were forgotten was the unusual reverse-flow configuration of the Proteus and the fact that the more popular pitot-style engines that succeeded earlier generations were not susceptible to ICI. “It went off people’s minds,” says Strapp."