Weld Metal Toughness: Sources of Variation

Abstract

Listen

Translate article

Many material properties are statistical in nature. If one measures the same property of the same material repeatedly, ideally the result is a normally distributed “bell” curve about a mean value. This ideal case does not necessarily hold true for all mechanical properties of interest in steel weld metals. Tensile strength measurements tend to exhibit normal behavior for a given weld metal chemical composition deposited using a reasonable consistent welding procedure, Figure 1a. However, toughness measurements are not nearly as well-behaved or predictable. In a tensile test, assuming a defect free weld, the strength measurement is based on the bulk response of the material throughout the gage length. In a Charpy V-notch (CVN) impact test, again assuming a defect free weld, the toughness measurements are controlled largely by the very local response of the material at the point of highest stress where fracture initiates just below the notch. This paper presents a detailed assessment of a C-Mn weld metal and explains how CVN toughness can vary from less than 20 ft-lbf to over 200 ft-lbf in the same weld, often with test specimens located adjacent to one another in the test weld, Figure 1b. The much localized microstructure features that give rise to this degree of variation are a combined result of chemical composition, welding procedure, pass sequence, and individual welder technique. The evidence suggests that retained austenite in coarse grained regions of the as-deposited weld metal transform to martensite at the CVN test temperature, effectively creating local brittle zones in the weld metal. This example provides basis for examination of a broader range of microstructural discontinuities in steel weld metals and their potential influence on toughness measurement.