Tim Furniss/LONDON

The failure of the first Proton M booster last month, was caused by a fire in the turbopump of the second stage engine. The fire was started by a stray particle of aluminium, says Russia's Khrunichev.

At T+277s, the fire started in the Voronezh-built No 3 engine, because of the particle - which was either left by a worker, or was the result of a faulty weld - and propelled into the turbopump during the engine operation.

The Russian company says the turbopump will be redesigned to include a filter, while plans to improve the engine's fuel-injection system will be accelerated.

Proton launches may resume at the end of this month, with the flight of the LM-1 communications satellite. A successful Proton lift-off will clear the way for the launch on 12 November of the Zvezda International Space Station module.

Meanwhile, Lockheed Martin and ILS International Launch Services hope to resume launches in late summer of the Atlas fleet, using the Pratt & Whitney RL-10 engine following the investigation into the explosion of the engine on a Boeing Delta III booster in May.

Boeing says that the cause of the Delta III failure was a new manufacturing process used to solder parts of the combustion chamber. The four sections of the chamber are covered with strengthening brazed seams using silver wire, but flawed brazing left air pockets and allowed a joint to split under the stress of the launch. To eliminate the problem, the brazed seams will be plated.

In addition, the US Air Force investigation into the failure of the RL-10 on a Titan IV launch in April points to a computer software failure rather than an RL-10 problem, with improperly developed software causing an attitude control failure, says the USAF.

Source: Flight International