Application of fault-tree analysis to troubleshooting the NASA GRC icing research tunnel

摘要

In 1997, the NASA Glenn Research Center in Cleveland Ohio conducted an extensive troubleshooting effort on its icing research wind tunnel (IRT). This effort utilized fault tree analysis as an analytical troubleshooting tool in order to "attack" a serious nonrepeatability problem with IRT icing cloud parameters. The analysis team investigated many problem areas that were indicated by the fault tree. The fault tree analysis was an important aid in the troubleshooting effort which pulled in organizations needed to investigate the problem and caused a number of simultaneous corrective actions to occur. These actions were critical in restoring the repeatability of the icing cloud parameters even though no single root cause of the problem could be identified. In addition, the analysis team found 9 system level considerations which were brought forward to improve tunnel operations. The possibility of implementing the recommendations was investigated by the IRT staff. We learned that the effectiveness of the analysis was greatly enhanced by a team approach, and by a critical team review of all fault tree sections constructed. Personnel investigated the possibility that certain fault events could exist even when it appeared that the likelihood of an event was small. Intermediate events or potential primary faults were not removed from the fault tree unless sufficient evidence was presented to do so. The IRT management also realized that the fault tree analysis report provided a baseline and history for the investigation of this nonrepeatability problem.