Introduction: Re-work and fit-up issues due to welding-induced distortion are significant contributors to hull fabrication cost in the shipbuilding industry. Integrated Computational Materials Engineering (ICME) tools such as finite element modeling (FEM) can be used to quantify distortions associated with fabrication of complex stiffened panels and other lightweight structures. However, these ICME tools require detailed, reliable databases of temperature-dependent material properties to maximize the accuracy of calculated distortions and residual stresses. Of highest importance to the fidelity of such models are the thermo-physical and thermo-mechanical properties of the material(s) being joined. While some of these data exist for steels commonly used in marine construction, their suitability for use in high-fidelity welding models is low because: 1) the methods of data generation are inconsistent across time and research institution, 2) significant gaps exist in the data at elevated temperatures, and 3) reported values are subject to variations in material chemistry and processing practices across the available studies. This presentation reports the results of a Lightweight Innovations for Tomorrow (LIFT)-sponsored program that sought to develop pedigreed temperature-dependent material property databases for navy-relevant steels. It focuses on data associated with DH36, the first of six investigated materials.
Experimental Procedure: Material from a single 4.76 mm (3/16 in.) thick DH36 steel plate was analyzed in this study. The chemical composition of the material was measured using glow discharge atomic emission spectrometry. Thermal diffusivity (α), specific heat (Cp), and density were measured as functions of temperature. Thermal conductivity was calculated from α, Cp, and the room temperature density of the material. Solidus and liquidus temperatures were evaluated using a single-sensor differential thermal analysis (SS-DTA) apparatus, where a 6 g (0.2 oz) button of material underwent induction levitation melting and was dropped into a copper crucible while temperature data was collected with a Type C thermocouple.
Heat affected zone (HAZ) phase transformations were assessed via dilatometry using a Gleeble. First, heating rates between 100 and 2000 °C/s (180 and 3600 °F/s) were used to study variations in on-heating austenitization temperature (Ac1 and Ac3). Then, continuous cooling transformation (CCT) diagrams were developed for four regions of the HAZ that are subject to microstructural transformation during welding: the intercritical region (ICHAZ); the low-temperature, fine-grained region (FGHAZ1); the high-temperature, fine-grained region (FGHAZ2); and the coarse-grained region (CGHAZ). Optical and scanning electron microscopy (SEM) were used to identify transformation products. The CCT diagrams were then used to inform elevated temperature mechanical property testing.
On-heating tension tests were performed at an external laboratory. On-cooling mechanical properties were measured in-house using a Gleeble. Prior to mechanical testing, specimens were thermally cycled to generate different initial microstructures based on the results of the CCT diagram development. A peak temperature of 1350 °C (2462 °F) and cooling rates of 1, 10, and 100 °C/s were utilized. All specimens were then cooled to room temperature, a contact longitudinal extensometer was affixed, and the specimens were heated to the desired test temperature before being tested to fracture.
Results and Discussion: Experimentally measured solidus and liquidus temperatures were 1453 and 1483 °C (2647 and 2701 °F), respectively. Measured thermal expansion coefficients were 15.6 x 10-6 ± 0.6 x 10-6 °C-1 (8.7 x 10-6 ± 0.3 x 10-6 °F-1) for the untransformed base metal below 650 °C (1202 °F) and 20.9 x 10-6 ± 2.1 x 10-6 °C-1 (12.0 x 10-6 ± 1.2 x 10-6 °F-1) for austenite above 950 °C (1742 °F). Measured Ac1 temperatures varied between 725 and 775 °C (1337 and 1427 °F) over the range of experimental heating rates, and this observation was attributed to the interplay between thermal and mass diffusion during heating. Ac3 temperatures were relatively constant with changes in heating rate.
HAZ dilation data indicated combinations of ferrite, pearlite, and bainite transformations during cooling. Ferrite transformations generally began at temperatures between 650 and 750 °C (1202 and 1382 °F), with the start temperature decreasing slightly with increasing peak temperature and cooling rate. When observed, pearlite transformations typically began 100 to 150 °C (180 to 270 °F) below the ferrite start temperature and followed the same trends as the ferrite transformation with respect to peak temperature and cooling rate. Pearlite transformations were not observed for cooling rates above 10 °C/s (18 °F/s). The onset of bainite transformations occurred between 500 to 600 °C (932 to 1112 °F) and was only observed at cooling rates of 25 °C/s (45 °F/s) and greater. Dilation data and microstructural analysis revealed an increase in bainite content with increasing peak temperature.
Elevated temperature yield strengths of specimens cycled to a peak temperature of 1350 °C (2462 °F) varied between 150 and 620 MPa (22 and 90 ksi). The observed variations with cooling rate and test temperature were found to be consistent with the changes in microstructure demonstrated by the CCT diagrams.
Conclusions: The data generated in this study are important for accuracy of naval structural steel weldment FEM tools. Temperature-dependent thermophysical properties govern the extent of heat flow and the rate at which expansion and contraction stresses may develop. Heating rate-, cooling rate-, and peak temperature-dependent transformation behavior must be available to enable appropriate selection of temperature dependent mechanical property data. Finally, temperature-dependent mechanical properties themselves must be understood because they will ultimately determine the magnitude of residual stresses and distortion that can develop in welded structures. The work performed in this study has measured all of these material properties for DH36 steel, laid the groundwork for increased FEM fidelity, and defined a repeatable procedure that has been applied to a range of other naval structural steels to develop similar databases.