The early stage of fatigue crack propagation in al 2048

  • PDF / 3,715,617 Bytes
  • 8 Pages / 612 x 792 pts (letter) Page_size
  • 49 Downloads / 242 Views

DOWNLOAD

REPORT


THE

author has had a continuing interest in the e a r l y stage of fatigue c r a c k growth in aluminum alloys, in particular in noncrystallographic c r a c k propagation a f t e r initiation at s m a l l intermetallic surface sites. T h e r e has been considerable r e c e n t activity in this a r e a , concentrated principally on initiation mechanisms or on formulating expressions for the c r a c k growth r a t e .1-5 One f a c e t of the problem that has been heretofore untreated due t o the associated experimental difficulties is the assessment of the c r a c k closure load associated with microcrack propagation a f t e r initiation. The bulk of c r a c k closure experimentation t o date has involved the study of macrocrack development in l a r g e fracture mechanics test specimens. 6-17 Because of the deformation mechanism required to produce c r a c k closure, it is not unreasonable t o expect that for the case of a s m a l l sharp m i c r o c r a c k in an initially undeformed medium the closure load will be zero, and that a nonzero closure load will subsequently be developed as the c r a c k propagates. Exactly this effect has been predicted by Newman.18 Remaining to be investigated is what affect nonideal conditions and ambient environment have upon the closure load of a developing microcrack. Ultimately such insight must be had before m i c r o c r a c k propagation r a t e s as affected by interaction with alloy microstructure and environment can be understood quantatively. In the course of studying the e a r l y s t a g e of fatigue c r a c k development in an A1 2219 alloy it was r e c o g nized that surface microcrakcs propagated in low relative humidity (RH) air were e a s i e r t o resolve with Scanning Electron Microscopy (SEM) because the p o s t test residual c r a c k openings were l a r g e r . ~ That is, c r a c k s propagated at h i g h e r humidity were more nearly closed under zero load. Residual c r a c k openings in e x c e s s of 1 gm were measured for 400 p m long c r a c k s propagated at low RH. Additionally, it was noticed that over the period of a y e a r , the residual c r a c k openings r e l a x e d slowly back t o nearly z e r o . It was concluded that for a g i v e n c r a c k length, a l a r g e residual c r a c k opening meant that the associated c r a c k closure load was l a r g e . The closure load at a m i n i mum had t o be l a r g e r than the load required to e l a s -

tically open a nondeformed c r a c k to the same residual c r a c k opening. This can be seen schematically in Fig. 1. The parameter 5 is the c r a c k opening m e a s u r e d at the surface as a function of load and 5(0) is the r e s i dual opening at zero load. The low residual c r a c k opening observed at h i g h e r RH is consistent with low closure load values obtained in macroscopic measurements at elevated humidity.'6 The residual c r a c k opening relaxation is also consistent with closure load relaxation observed on macroscopic c r a c k s . B a s e d on this, a s e r i e s of experiments have