Performance characterization Don clausing C Don Clausing 1998 16881 Fig. 1
Performance characterization Don Clausing © Don Clausing 1998 16.881 Fig. 1
Failure modes Noises lead to failure modes (Fm) One set of noise values leads to fm Opposite set of noise values leads to FM, Simple problem solving chases the problem from Fm, to Fm, and back again, but does not avoid both fms with the same set of design values-endless cycles of build/test/fix (B/T/F) C Don Clausing 1998 16881 Fig 2
Failure modes • Noises lead to failure modes (FM) • One set of noise values leads to FM1 • Opposite set of noise values leads to FM2 • Simple problem solving chases the problem from FM1 to FM2 and back again, but does not avoid both FMs with the same set of design values – endless cycles of build/test/fix (B/T/F) © Don Clausing 1998 16.881 Fig. 2
Failure modes occurrence Failure modes occur when the essential performance characteristic deviates excessively from its ideal value Some failure modes are simply values that deviate so far that output is not useful; e.g voltage from power supply is too low Some failure modes are catastrophic; e.g paper that arrives too soon jams in copier C Don Clausing 1998 16881 Fig. 3
Failure modes occurrence • Failure modes occur when the essential performance characteristic deviates excessively from its ideal value • Some failure modes are simply values that deviate so far that output is not useful; e.g., voltage from power supply is too low • Some failure modes are catastrophic; e.g., paper that arrives too soon jams in copier © Don Clausing 1998 16.881 Fig. 3
Performance characteristic What is good performance characteristic to use when reducing the occurrence of failure modes? Dont merely count occurrence of failure modes Can't distinguish between following two cases C Don Clausing 1998 16881 Fig 4
Performance characteristic • What is good performance characteristic to use when reducing the occurrence of failure modes? • Don’t merely count occurrence of failure modes • Can’t distinguish between following two cases © Don Clausing 1998 16.881 Fig. 4
Case 1 -easy to fix Occurrence of failure mode FM FM C Don Clausing 1998 16881 Fig 5
Case 1 – easy to fix Occurrence of failure mode FM1 FM2 © Don Clausing 1998 16.881 Fig. 5
Case 2-difficult to fix Occurrence of failure mode FM FM C Don Clausing 1998 16881 Fig 6
Case 2 – difficult to fix Occurrence of failure mode FM1 FM2 © Don Clausing 1998 16.881 Fig. 6
Case 1 and Case 2 Both cases have same failure rate But situations are very different Counting failure rate is very weak approach to the reduction of failure rate Concentrate on ideal function What is the system supposed to do? Then make system do it all of the time C Don Clausing 1998 16881 Fig. 7
Case 1 and Case 2 • Both cases have same failure rate • But situations are very different • Counting failure rate is very weak approach to the reduction of failure rate • Concentrate on ideal function – What is the system supposed to do? • Then make system do it all of the time © Don Clausing 1998 16.881 Fig. 7
The engineered system N olse Signal System Response Control factors C Don Clausing 1998 16881 Fig 8
The engineered system Noise Signal System Response Control factors © Don Clausing 1998 16.881 Fig. 8
Ideal function Ideal function 00.0u2 (response) SIGNAL, M C Don Clausing 1998 16881 FI
Ideal function Ideal function (response) RESPONSE SIGNAL, M © Don Clausing 1998 16.881 Fig. 9
Actual response Ideal response 00.0u2 Effect of noises 1 SIGNAL M2 C Don Clausing 1998 16881 Fig 10
Actual response Ideal response Effect of noises RESPONSE M1 SIGNAL M2 © Don Clausing 1998 16.881 Fig. 10