EFFECT OF LOADING RATE ON DUCTILE-BRITTLE TRANSITION TEMPERATURE BY REFERENCE TEMPERATURE (T_0) AND MASTER CURVE APPROACH FOR A MODIFIED 9Cr-1Mo STEEL

摘要

Towards evaluation of the ductile to brittle transition temperature,(DBTT) of ferritic steels, the ASTM E 1921 based reference temperature (T_0) and the associate Master Curve approach have now been widely recognized, however, till date the standard restricts itself to static/quasi-static loading rates. It is well recognized that the flow stress of rate sensitive material increases with the strain rate and thus it is imperative that the increase in loading rate would lead to limited plasticity induced brittleness, reflected in higher T_0. There have been efforts in literature for developing empirical correlations to derive T_0 at higher loading rates from T_0 at quasi static loading rates or vice versa. The loading rate is often expressed in terms of stress intensity factor rate (dK/dt), with unit MPam~(1/2) s~(-1). dK/dt is approximately equal to 10~4-10~5 MPam~(1/2) s~(-1) for low velocity tests at 1-2 ms~(-1) and dK/dt ～ 10~6 Mpam~(1/2) s~(-1) for normal velocity impact tests at ～5 ms~1. There is a need to experimentally validate the extent of change in To with higher loading rates, especially for the 9Cr-1Mo family of steels, proposed to be used as wrapper material in the upcoming commercial liquid sodium cooled fast breeder reactors. The present study is directed towards determining T_0 for modified 9Cr-1Mo steel at loading rates of 1.12, 3 and 5 m/s. The material under dynamic loading actually is subjected to a series of forward and reflective stress waves, at least in the initial part of the loading, observed as high frequency oscillations in the load-time or load-displacement traces. Thus for determining the dynamic fracture toughness of materials the inertia effect on the fracture load determination is very important in data analysis, especially for higher loading rates and/or for brittle materials. The amplitudes of the consecutive oscillations are expected to die down with time after the impact and considered negligible after a period of 2τ -3τ, where τ is the period of inertial oscillation, a function of specimen type and geometry and speed of sound through this material. The time period, τ, is taken as 27 μs for the present campaign, implying that after 3τ i.e. 81 μs, all the measured fracture loads are free from any inertial effects. However, at a lower loading rate, i.e ～ 1 m/s, the amplitude of the oscillations are negligible. Thus the time to fracture (t_f) with respect to the 2τ -3τ time span is an important parameter, especially for the loading rates of 3 and 5 m/s, in ensuring the accuracy of the fracture load determination for evaluating the dynamic fracture toughness. It is observed that the time to fracture (t_f) for the loading rate of 3 m/s has exceeded the 3τ time span for all the specimens, the minimum being 100 μs, where as a few specimens for the loading rate of 5 m/s have failed to meet this criteria. Here it is observed that out of eight specimens tested at different temperatures ranging from -25 to 30°C, four specimens are at the border line of 2τ-3τ regime and one specimen with t_f= 10 μs fails to qualify even for consideration. An alternative analytical technique has been used to derive the fracture load for those specimens. T_0 values determined were 8.8 and 11.2°C for the loading rates of 1.12 and 3 m/s respectively, and increases drastically to 26.2°C for a loading rate of 5 m/s. This shift is little higher than that predicted empirically (～14°C) for conventional pressure vessel steels.