Creep-fatigue life assessment of cruciform weldments using the linear matching method

Yevgen Gorash, Haofeng Chen

Research output: Contribution to journalArticlepeer-review

35 Citations (Scopus)
316 Downloads (Pure)


This paper presents a creep-fatigue life assessment of a cruciform weldment made of the steel AISI type 316N(L) and subjected to reversed bending and cyclic dwells at 550C using the Linear Matching Method (LMM) and considering different weld zones. The design limits are estimated by the shakedown analysis using the LMM and elastic-perfectly-plastic material model. The creep fatigue analysis is implemented using the following material models: 1) Ramberg-Osgood model for plastic strains under saturated cyclic conditions; 2) power-law model in “time hardening” form for creep strains during primary creep stage. The number of cycles to failure N? under creep-fatigue interaction is defined by: a) relation for cycles to fatigue failure N dependent on numerical total strain range "tot for the fatigue damage !f ; b) long-term strength relation for the time to creep rupture t dependent on numerical average stress ¯ during dwell t for the creep damage !cr; c) non-linear creep-fatigue interaction diagram for the total damage. Numerically estimated N? for different t and "tot shows good quantitative agreement with experiments. A parametric study of different dwell times t is used to formulate the functions for N? and residual life L? dependent on t and normalised bending moment ˜M , and the corresponding contour plot intended for design applications is created.
Original languageEnglish
Pages (from-to)1-13
Number of pages14
JournalInternational Journal of Pressure Vessels and Piping
Early online date18 Jan 2013
Publication statusPublished - 1 Apr 2013


  • creep
  • damage
  • finite element analysis
  • FSRF
  • low cycle fatigue
  • type 316 steel
  • weldment


Dive into the research topics of 'Creep-fatigue life assessment of cruciform weldments using the linear matching method'. Together they form a unique fingerprint.

Cite this