ICF10B

-0,5 0,0 0,5 1,0 1,5 2,0 2,5 3,0 3,5 4,0 0,0 0,2 0,4 0,6 0,8 1,0 t/tf Deflection (mm) Vt Ve Vc V=1µm/h V=5µm/h V=10µm/h Fmax b) a) 0,0 0,5 1,0 1,5 2,0 2,5 3,0 3,5 0,0 0,2 0,4 0,6 0,8 1,0 t/tf Deflection (mm) Vt Vc Ve Fig.6.Load line deflection (Vc: creep component, Ve: elastic component, Vt: total load-line deflection) as a function of normalised test time, t/tf, for HIPed + HT1300°C TiAlCr tested at 700°C a) under constant load, F=10kN, b) displacement rate control, Vrate=1,5&10µm/h. Crack lengths determined from DCPD were compared with those determined from fracture surfaces. The difference was as high as 50 per cent, with those measured on fracture surfaces were higher, which also did not satisfy the standard requirement. In determining the crack lengths on fracture surfaces the unbroken shear ligaments were ignored, whereas crack lengths from the DCPD were erronous because of surface contacts and unbroken ligaments. The measurement of crack length in toughness testing of TiAlCr requires further study. Crack lengths measured on the side surfaces of specimens were smaller than those measured on fracture surfaces by up to 30 per cent. The crack fronts were uneven and the crack growth was non-uniform by more than 20 per cent or 0.15 mm that is not satisfying the standard test requirements [11]. Crack tunnelling and unbroken ligaments on fracture surfaces contribute to the discrepancy in crack length measurements. Crack branching and bifurcation lead to violation of the validity criteria. Large grains and lamellae particularly promote the out-of-plane cracking. Out-of-plane cracking limits need to be specified for titanium aluminides similar to the one in Fig 8 of ASTM E647 [13]. The crack deflection at an excess of 20° may call for mixed-mode stress analyses to compute K as it is also suggested in ASTM E647. The complete set of creep crack growth data is correlated with both C*(t) and K (Figs.1,2), followed by the reduced data (Figs.3,4) to shed some light on the applicability of the concepts of K and C*(t) in these materials. The data from transition range correlate only with K. This is because immidiately after loading, an elastic, or elastic-plastic stress distribution is generated ahead of a crack tip prior to the onset of creep. The criteria was set in the standard [11] to ensure the extensive creep stress distribution will be produced ahead of the crack for the data to correlate with C*(t). The complete CCG data satisfied transition time ttest > tT , and deflection rate ratio >0.9, although the material is brittle. The variation of deflection rate ratio in Fig.5 does not follow the predicted behaviour [14]. Similar discrepancy is also reported for creep brittle Al and Ti alloys [15]. As numerical work was not done in the present study, it may be noted that this aspect need further clarification particularly if the assessment of component made of TiAl is to be done and the crack tip parameter C cV /V & & t t [14] is considered. The stationary stress state is achieved in the second part of constant load test (Fig.4). The decreasing part of the constant load test crack growth rate correlation is attributed to the transition of crack growth mode of a fatigue precracked specimen where initial crack growth follows the transgranular fatigue crack path that changes to creep crack mode with accumulation of grain boundary damage with time. This behaviour in TiAl is also reported by Fuji et al [9]. It is important to note that this transition occurs (Fig.5.a) over a period of finite crack extension, as large as 50% of the total test time, t/tf=0.5. The data from displacement rate controlled tests exhibited typical transition behaviour over a range of crack growth caused by change of stress state via changing applied load under displacement rate control in creep brittle material. This discussion is supported by the partitioned deflection in Figure 6, where creep component of deflection is 0 for a long period of time, t/tf>0.5. It was followed by constant Ve and

RkJQdWJsaXNoZXIy MjM0NDE=