To evaluate the susceptibility to stress corrosion cracking (SCC) of stainless steel and nickel based alloy components, it is important to know the degree of plastic strain. Indeed, SCC is enhanced by the strain hardening induced by manufacturing and welding. We evaluate the residual plastic strain in a thick tube made of Ni-based alloy 600, generated by the welding operation with alloy 182 deposited metal. By using these data we can predict the crack growth rate and improve structural integrity assessments of components.

Hardness measurements or numerical stress analysis are usually applied to approximate plastic strain. Here, we use an innovative method by correlating deformation with mechanical characteristics from Electron Backscattered Diffraction (EBSD) [1] acquired during in-situ tensile test. We use a reference sample to calibrate and quantify the equivalent plastic deformation as a function of the average macroscopic deformation. Then, measurements are done on a sample after a welding operation and compared to the calibration data. The reference sample and the mock-up were manufactured with the same alloy 600. Tensile test specimen was prepared using Electrical Discharge Machining (EDM) and mechanical polishing to obtain an EBSD-quality finish. Micro-lithography was done to deposit markers on the surface.

In-situ experiment was done using a Tescan Mira 3 Scanning Electron Microscope (SEM) and a Micromecha tensile test machine allowing uniaxial loading at EBSD position.

Regarding [2], a preliminary work was done to optimize acquisition parameters and to check the statistical representativeness of the data (500 x 500 µm areas, step size = 250 nm). Data treatment consists in calculating for each point of the EBSD map the Kernel Average Misorientation (KAM). KAM gives local information about the plastic deformation [3]. In order to make the data quantitative, the KAM distribution is plotted. The shape of the distribution is very sensitive to the microstructure.

Results of the in-situ mechanical test are presented in figure 1. We found a logical and monotonic evolution of the misorientation distribution: the average, the position of the maximum and the width of the distribution are going up with the increase of the deformation (measured using markers), in accordance with [1]. We use those data to calibrate equivalent plastic deformation for the mock-up.

Fifteen measurements were performed on a weld mock-up with exactly the same SEM and EBSD acquisition parameters as used for the calibration experiment [2]. Location of the measurements were chosen in order to draw vertical and horizontal profiles regarding the welding interface.

We found least square fit method with a degree of freedom on the abscissa (up to 0,04°) as the best way to compare data from the mock-up and from the calibrated distributions. An example of a fit is given in figure 2. A map of the sample with the deduced values of equivalent plastic deformation is presented in figure 3.

From a material point of view, results show low levels of equivalent deformation. The maximum is 6,8%. Logical and monotonic decrease is observed going away from the front of the heat affected zone (HAZ, defined by the limit between large and small grains), the equivalent strain is 0 at 3 mm from the HAZ. Areas 4 mm away from the HAZ show an equivalent strain lower than the tensile test specimen in its initial state. This may be due to stress relief during heating by the welding operation, a problem of representativeness of the uniaxial tensile test compared to the welding operation or unexpected deformation of the reference test piece.

[1] Kamaya M. et al. (2015) Nucl. Eng. Des. 235, 173.

[2] Wright S. et al. (2011) Microsc. Microanal. 17, 316.

[3] Britton T. B. et al. (2010) Scripta Mater. 62, 639.

Figures:

Figure 1: Distribution of the KAM values (number of iteration vs misorientation angle), 1st neighbors, for different steps of uniaxial deformation. We observe a logical and monotonic evolution of the distribution with the deformation.

Figure 2: Superposition of a calibration curve (4.8% deformation) with data acquired on the mock-up. The fit is good between the two distributions.

Figure 3: Values of equivalent deformation measured on the mock-up. Areas of acquisition are represented by red squares. There is a monotonic evolution of the values showing gradient of deformation parallel to the HAZ (dashed line).

« Back to The 16th European Microscopy Congress 2016

EMC Abstracts - https://emc-proceedings.com/abstract/study-of-strain-fields-induced-by-welding-in-nickel-alloy-600-using-in-situ-mechanical-tensile-test-approach-and-standard-ebsd/