The service life of high temperature furnace tubes varies significantly in
engineering practice. Predicting the service life of the tubes has thus lon
g been a concern. Due to the difficulties in defining variability of creep
data, previous research has mainly concentrated on deterministic creep dama
ge models. In order to study the random nature of service life, a new stoch
astic creep damage model is proposed in this paper. A comparison with resul
ts calculated by use of the Monte Carlo method verifies the creep damage mo
del. The randomness of the creep damage is demonstrated with a calculation
on HK-40 furnace tubes which provides an effective means to assess the reli
ability of the furnace tubes. (C) 2001 Elsevier Science Ltd. All rights res
erved.