Fretting fatigue crack analysis in Ti–6Al–4V |
| |
| Authors: | Alisha Hutson Ted Nicholas Reji John |
| |
| Affiliation: | aUniversity of Dayton Research Institute, Dayton, OH 45469, USA;bAir Force Institute of Technology (AFIT/ENY), Wright-Patterson AFB, OH 45433, USA;cAir Force Research Laboratory, Materials and Manufacturing Directorate (AFRL/MLLMN), Wright-Patterson AFB, OH 45433, USA |
| |
| Abstract: | A study was conducted to verify the efficacy of a fracture mechanics methodology to model the crack growth behavior of fretting fatigue-nucleated cracks obtained under test conditions similar to those found in turbine engine blade attachments. Experiments were performed to produce cracked samples, and fretting fatigue crack propagation lives were calculated for each sample. Cracks were generated at 106 cycles (10%-of-life) under applied stress conditions previously identified as the fretting fatigue limit conditions for a 107 cycle fatigue life. Resulting cracks, ranging in size from 30 to 1200 μm, were identified and measured using scanning electron microscopy. Uniaxial fatigue limit stresses were determined experimentally for the fretting fatigue-cracked samples, using a step loading technique, for R=0.5 at 300 Hz. Fracture surfaces were inspected to characterize the fretting fatigue crack front indicated by heat tinting. The shape and size of the crack front were then used in calculating ΔKth values for each crack. The resulting uniaxial fatigue limit and ΔKth values compared favorably with the baseline fatigue strength (660 MPa) for this material and the ΔKth value (2.9 MPa√m) for naturally initiated cracks tested at R=0.5 on a Kitagawa diagram.Crack propagation lives were calculated using stress results of FEM analysis of the contact conditions and a weight function method for determination of ΔK. Resulting lives were compared with the nine million-cycle propagation life that would have been expected in the experiments, if the contact conditions had not been removed. Scatter in the experimental results for fatigue limit stresses and fatigue lives had to be considered as part of an explanation why the fatigue life calculations were unable to match the experiments that were modeled. Analytical life prediction results for the case where propagation life is observed to be very short experimentally were most accurate when using a coefficient of friction, μ=1.0, rather than for the calculations using μ=0.3 |
| |
| Keywords: | Fretting fatigue Stress intensity Crack growth Titanium |
| 本文献已被 ScienceDirect 等数据库收录！ |
|
