Introduction: Dissimilar metal welds contain complex microstructures at the fusion boundary which are not present in the parent materials. These complex microstructures arise from the differences between material properties and the physical / chemical reactions resulting form these differences in properties and are often unavoidable. These deserve further attention due to their participation in the Hydrogen Assisted Cracking (HAC) phenomenon studied at the Ohio State university (OSU) using the Delayed Hydrogen Cracking Test (DHCT). It is hypothesized that these features act as stress concentrators, altering the local stress state significantly which affects numerous aspects of the HAC failure mechanism. Using Cross Weld Tensile Testing (CWTT) and Digital Image Correlation (DIC), the local strain state can be determined which can yield valuable information on the stress strain state in local regions. The fusion boundary (and its complex microstructures) exists in a small region, thus the traditional macro CWTT DIC testing method used previously is not applicable. However, using a micro sized sample, and micro tensile stage loaded in a Scanning Electron Microscope (SEM), the study of mechanical behavior at these reduced length scales becomes possible. Using SEM in-situ DIC, the local stress-strain relationship at the fusion boundary will be studied and determined for typical microstructural features common to nickel base filler metals to low alloy steel DMWs.
Experimental Procedure: The DIC tensile test procedure resembles a typical CWTT but requires a specific sample preparation. Samples are prepared for the SEM as if they were to be imaged. This is done by grinding and polishing samples in a typical format. Once polished, samples are patterned with a speckle pattern suitable for SEM DIC. This is typically done using platinum sputtering, but the research group is also looking into other means to accomplish patterning such as etching samples (and using microstructural contrast as a speckle pattern). The sample will be tested to failure and imaged in the SEM at different loads for comparison at no load. DIC data will be obtained using Correlated Solutions VIC 2D software with the appropriate algorithm inputs for SEM-DIC. Using tools available in the software, local strain of regions of interest can be measured and compared. Combining this data with typical plotting and data analysis programs such as MATLAB and Excel yields comparisons which can be used to determine best practices for future weldments. The most important output of this test is the micro-level strain distribution at typical microstructural constituents in the fusion boundary area of DMWs, including partially melted CGHAZ grain, weld metal penetrations along grain boundaries of partially melted CGHAZ grains, continuous and discontinuous partially mixed zones, and the planar growth region.
Results and Discussion: Two dissimilar metal welds will be tested using in-situ SEM DIC. One weld (F22-Alloy 625 weld metal) is resistant to HAC when tested in the DHCT. The other weld (F65-alloy 625 weld metal) is susceptible to. During the micro-level DIC testing, microstructural constituents of interest will be identified for future studies. The aim is to identify microstructural constituents related to HAC nucleation, through the study of the local stress strain interactions in the fusion boundary area, and if possible, rank these in terms of their potential susceptibility to HAC .
Conclusions: The conventional CWTT has a significant shortcoming when testing DMWs. The CWTT only outputs the global properties of the weldment that reflect the region of ultimate failure, where most strain will accumulate throughout the test. In DMWs, there are localized differences that can greatly impact performance. DIC aided CWTT can overcome these shortcomings by giving more localized straining results, which better represent the gradient in mechanical behavior across the weld related to the local variations in composition and microstructure.