· Table of Contents
Change in the elastic constants with thermal embrittlement of duplex stainless steelTetsu ICHITSUBO, Masakazu TANE, Hirotsugu OGI, and Masahiko HIRAO
Graduate School of Engineering Science, Osaka University, Osaka 560-8531
We report the change in elastic constants of duplex stainless steel (JIS-SCS14A) with isothermal aging at 400°C up to 10000h. The elastic constants have been measured by the resonant ultrasound spectroscopy (RUS) method combined with the electromagnetic acoustic resonance (EMAR) method; we have identified the four specific vibration (OX, OY, OZ and OD) groups among eight groups of the free vibrations by EMAR. The elastic constant c11 monotonically increases with the isothermal aging, but c44 remains virtually unchanged, resulting in the increases in Young's modulus and the Bulk modulus. The contactless measurement with EMAR provides intrinsic attenuation coefficients; the coefficients decrease with the isothermal aging. These results indicate that the longitudinal-wave velocity and the attenuation coefficients can be candidates for the non-destructive evaluation of the thermal embrittlement in this material.
Keywords: resonant ultrasonic spectroscopy, electromagnetic acoustic resonance, vibration mode, ultrasound, longitudinal wave velocity
Duplex stainless steels consisting of austenite and ferrite phases form a microstructure that austenite phase particles are distributed in a ferrite phase matrix. The materials possess several superior characters, e.g., corrosion resistance, formability and high strength, and therefore have been used in many products such as pressure vessels, pipes to heat exchangers, etc. Thus, the duplex stainless steels are regarded as high potential industrial materials, but there is a well-known problem of "475°C embrittlement" [1-7]; this phenomena is believed to be caused by the spinodal decomposition of a ferrite phase into a (Fe-rich) and a' (Cr-rich) phases. This embrittlement yields an increase of the hardness of the ferrite phase and a significant decrease of ductility and fracture toughness. The materials are most rapidly embrittled when exposed at temperatures of 400-500°C. The spinodal decomposition is possible to occur even at lower temperatures, of course, even at the practical service temperature, for example, 325°C in the pressurized water reactors.
For a long-period use in safety, we should elucidate the mechanism of the thermal embrittlement and establish a technique of nondestructive evaluation for the remaining life of the materials. If the elastic constants change with the thermal embrittlement, measurements of the ultrasound velocities shall become a quite useful means. We have then studied the changes in the elastic property and ultrasonic attenuation of the polycrystalline duplex stainless steel, JIS-SCS14A, with the thermal aging embrittlement. According to the previous report , pulse-echo method is useless, because the grain size of the austenite phase is of the order 100mm and the scattering of the ultrasound is too strong to measure the ultrasound velocities. To determine the elastic constants in high accuracy, we employ the two resonance methods; the resonant ultrasound spectroscopy (RUS) [9,10] and the electromagnetic acoustic resonance (EMAR) [11,12]. This paper reports mainly the change in the elastic constants of the material.
The test metal was made by centrifugal casting after being melted in an electric furnace. The material was annealed at 1090°C for 5.5h and subsequently quenched into water. Table I shows the chemical composition. The accelerated aging test was performed; the specimens were aged at 400°C higher than the service temperature 325°C for 0, 300, 1000, 3000 and 10000h. These times correspond to 0, 0.24, 0.81, 2.4 and 8.1 years at 325°C. This conversion is subjected to Arrhenius's relation, t2=t1[Q/R(1/T2-1/T1], where t is the time, T is the temperature and Q is the activation energy of diffusion.
|Volume fraction of a phase%||C||Si||Mn||P||S||Ni||Ci||Mo|
|Table 1: Chemical composition (mass%) of the duplex stainless steel (SCS14A)|
Before the acoustic measurements, the degree of thermal embrittlement was evaluated by Charpy-impact and Vickers hardness tests [Figs. 1 (a) and (b)]. We confirmed the decrease of the absorbed energy (< 10%) and the increase of hardness of the ferrite phase (~ 200%) with the increase of the aging time.
|Fig 1: Changes in (a) absorbed energy in Charpy-impact test of the duplex stainless steel and (b) Vickers hardness of ferrite (a) and austenite (g) phases with isothermal aging.|
The rectangular parallelepiped specimens for RUS and EMAR measurements were cut out of the materials aged for respective times, measuring about 5 mm along each edge. The specimen surfaces were mechanically polished with sandpapers.
The RUS method to determine elastic constants was developed by Demarest for cubic specimens  and subsequently was extended to rectangular parallelepiped specimens by Ohno . When the material is of orthorhombic symmetry (including isotropic, cubic, hexagonal, tetragonal symmetries), there are eight independent groups of free-vibration modes: OD (dilatation), EV (torsion), OX, OY, OZ (shear) and EX, EY, EZ (flexure) [9,10]. The resonance frequency of a free vibration mode depends on the density, the dimensions and the elastic constants of the specimen. With the known elastic constants, the resonance frequencies can be calculated using the measured density and dimensions. Inversely, the elastic constants are possible to be determined through the iterative calculations from the measured resonance frequencies (Fig. 2).
|Fig 2: Determination of elastic constants from the measurement and calculation of resonance frequencies by the RUS method.|
The key point here for the reliable determination is to identify the vibration modes; that is, we should know an exact correspondence between the resonance peaks and the vibration modes. Wrong correspondence could lead to erroneous or unrealistic results. This is a possible cause of error with the RUS method, which detects a large number of resonance peaks from all eight groups. However, this is not true for the EMAR method, because it can preferentially excite and detect only one vibrating group . Figure 3 shows the principle of driving an electromagnetic acoustic transducer (EMAT), which consists of a permanent magnet and a solenoidal coil, and the mechanism of excitation of the specific vibration group (OX vibration group in the figure). The other vibration (OY, OZ and OD) groups can be excited by rearranging the geometric configurations of the magnet and the coil. The details of EMAT and the mode-selective EMAR method have been described in the literature .
|Fig 3: Excitation of the OX vibration group by the EMAT. (a) The specimen is inserted into the coil and the magnetic field is applied. (b) The eddy currents on the surface of the specimen are induced by the high frequency current passing in the coil. The Lorentz force is generated by the interaction of the magnetic field and the eddy current. (c) The specimen is deformed by the shear Lorentz forces as shown in the figure. (d) The symmetry of the deformation satisfies only that of the OX vibration group.|
The RUS measurements were carried out within the frequency range of 200-600kHz at 0.01kHz steps. A part of the measured resonance spectrum is shown in Fig. 4 for the unaged specimen. The vibration groups of these resonance peaks were identified by the EMAR method and are shown in Fig. 4. The number in brackets denotes the order in the vibration group. Thus, by employing the EMAR method, the four groups of the vibrations were identified unambiguously. The elastic constants of each specimen have been determined on the assumption that all the specimens are isotropic.
|Fig 4: Resonance spectrum measured by the RUS method. Some of the resonance peaks are identified by the EMAR technique.|
Figure 5(a) shows the elastic constants, c11 (= l + 2m) and c44 (= m), as functions of the aging time, and Fig. 5(b) shows Young's modulus and the bulk modulus calculated using them. The elastic constant c11 increases with the aging time, but c44 remains virtually unchanged. Young's modulus slightly increases, while the increase of the bulk modulus is very remarkable. As seen in Fig. 1(b), the hardness of the ferrite phase increases, but no change is found in the austenite phase. This is probably because the internal structure of the ferrite phase changes due to the spinodal decomposition; the ferrite phase decomposes into two phases made with stronger atomic bonds (i.e., Fe-Fe and Cr-Cr bonds). The elastic constants correspond to the second derivative of the interatomic potential with respect to the atom position. Therefore, supposing that the lattice parameter is unchanged by the spinodal decomposition, the elastic constants of the ferrite phase should increase with the spinodal decomposition. The result of the present measurement is consistent with this view. The stiffening of the duplex stainless steel is caused by the microstructure change in the ferrite phase that occupies only 26% in volume fraction. Further consideration by the micromechanics model is required for a deeper understanding this phenomenon.
|Fig 5: (a) Elastic constants, c11 (= l + 2m) and c44 (= m), as functions of the aging time. (b) Young's modulus and the bulk modulus calculated from c11 and c44.|
The attenuation coefficients, which means the inverse of the decay time constant in the ultrasonic resonance, were also measured by the EMAR method; the details will be addressed elsewhere. The attenuation coefficients tend to decrease with the aging time. In the present frequency range, the origin of ultrasonic attenuation is mainly the inelastic motion of dislocations, i.e., the dislocation dumping. The decrease of the ultrasonic attenuation indicates that the dislocations becomes less mobile with the aging time, which is consistent with the hardness increase of the material.
We have investigated the change in the elastic constants of the duplex stainless steel (JIS-SCS14A) with the thermal embrittlement. The measurements by RUS and EMAR have provided the accurate elastic constants of the material. The elastic constant c11 increases but c44 remains unchanged with the isothermal aging. This result suggests that the longitudinal-wave velocity and the attenuation coefficients are useful for the nondestructive evaluation of the thermal embrittlement in this material.
|© AINDT , created by NDT.net|||Home| |Top||