The ductile-brittle transition of a zirconium alloy due to hydrogen

Abstract

Zircaloy-4 is generally used as fuel element cladding and an in-core structure c'm~cn..~t in light water reactors. Its mechanical properties are degraded by the presence of hydrid ~a wh ch are primarily formed by the absorption of excess hydrogen from the corrosion reaction: Zr + 2H20 ZrO 2+ 2H 2 [1]. Lin et el.[2] and Baiet el. [3-5] studied the mechanical properties of hydri.l~ Z£rcaloy-4 alloys independently by tensile testing the smooth specimens. Both of them obser,ed a room temperature ductile-brlttle on the reduction of area, when the hydrogen content in the specimen is higher than some critical value. Their deflnltion of transition is the abrupt change of the value of reduction of area, which is somewhat different from the conventional definition determined from impact tests. Since the effect of hydrogen on reduction of area 18 more distinct than on elongation, the reduction of area has been used as a parameter to exhibit hydrogen effect. Although reduction of area Is not a standard parameter for a notched specimen geometry, it can be used as a relative criterion for judging the ductility of a specSmen being tensile teated. In thl8 paper we followed the definition of Lin and Bai. The observations of ductile-brittle transitions mentlonedsboveweremostlyperformedusingmmooth tensile specimens in standard tensile tests. Since notched specimens provide a triax£al state of stress, which will enhance brittle fracture, they are more suitable to be used to study the ductile-brittle of metals than smooth tensile specimens. In the present study, notch tensile tests were performed on Zircaloy-4 alloys at various temperatures up to 300C in hydrided conditions to investigate the change of ductile-brittle due to notch effe.:t. Additionally, no previous study has been done on Zircaloy-4 alloy to determine whether -he ductile-brittle is existed in a hydrogen gas environment. To determine this phenomenon, notched, uncharged Zircaloy-4 specimens were tensile tested in a hydrogen ~as environment at various temperatures up to 200°C. We found that the ductile-brittle oxlsts not only for the hydrided Zircaloy-4 but also for Zircaloy-4 tested in the hydrogen gas environment. The specimen materials were commercial Zircaloy-4 purchased in plate form. The chemical compositions are listed in Table I. All specimens were used in the as-recelved condition, annealed at 760C for one hour and air cooled. The grain shape was equiaxed and the grain sizes wore 15 ~m and i0 pm for specimens A and B, respectively. The yield strengths for specimens A and B were 484 MPa and 435 MPa, respectively. Specimens A (2 mm thickness) with dimensions 150 mm x 25 nm x 2 on were hydrided and then tensile tested in air. Specimens B (1.6 mm thickness) with dlmenslons 75 mm x 25 ~ x 1.6 ~ were tensile tested in hydrogen gas. All the specimens wore out transverse to the rolling direction. A saw cut (about 7 mm deep, notch root radi .s abf,ut 0.125 mm) was made in the middle of one edge of each specimen. For the hydrlded apeclmL~ns, the material was gaseously hydrided at 350C in an autoclave at various pressures for different durations. The hydrlding procedures are described elsewhere [6]. After charging, the hydride morphology was examLned using an optical microscope. For each specimen, after testing, sectluns near the fracture surface were cut off to determine the hydrogen content using the inert gas fusion method at 1800°C. For those tests performed in hydrogen gas, specimens were sealed in an autoclave. The autoclave system was evacuated down to 6.67 Pa using a mechanical pump, then flushed with argon gas to ambient pressure, and evacuated again. The autoclave was then at least twice flushed with hydrogen to ambient pressure end evacuated. Finally, hydrogen was admitted up to experimental pressure. The hydrogen pressures used Ln this study were 101, 1010 and 2020 kPa. High purity hydrogen with H 2>99.9995% was used. The tensile tests wore carrled out at room temperature (25C), 100, 200 ° , and 300c (only for hydrided specimens) on a constant extension rate testing machine (CERT) at a cross-head speed of 3 x 10-'m/sec. The time for a single test lasted from seven hours to twelve hours,depending on the hydrogen content, hydrogen pressure or experimental temperature. In general, as the hydrogen content, hydrogen pressure, or exp~-rlmental