The present project aims to develop a reliable numerical model to predict the extent and severity of macrosegregation in a 40MT (metric ton) medium-carbon high strength steel ingot as a function of casting parameters with the view to minimize the occurrence of the macro-scale defect. For this purpose, the commercial finite element modeling (FEM) code Thercast® was at first determined to be used for the simulation work. This basis of such a decision rested on the critical literature review on the evolution of solidification modeling technique in the past 60 years, the powerful computing capability of the software and its performance of the three-dimensional simulations of mold filling and solidification based on a volume-averaged solid-liquid two-phase model.
The factors influencing model accuracies were eliminated one by one. For instance, an axisymmetric model in the dimension of 1/24 of the actual casting system was determined to be representative enough for the necessary simulation work. An average general cell size of 35 mm was identified to be the discretization parameter for its realization of the compromise between computational accuracy and efficiency based on series of sensitivity studies of solidification to mesh size. Modeling boundary conditions were set based on industrial practice, which took into account the local friction and the thermal exchange between the steel and the mold components, as well as the mutual exchange between mold components themselves in the casting system. Input temperature-dependent parameters of all concerned materials, reflecting their thermodynamic, thermomechanic and thermodiffusion behaviors, were identified by means of combined approaches of computational thermodynamics program Thermo-Calc® software, literature reports, published empirical equations, Thercast® software material database, industry sources, theoretic calculations, as well as room and high temperature tension tests in laboratory. The applicability of the used softwares and models was verified before applying them to the investigated steel. The experimental planning was strictly formulated based on simulation tests. Lever rule microsegregation model was ascertained to be the most reasonable depiction of the physical behavior of the steel in study in large-size cast ingot. This selection was made after comparative examinations of various microsegregation schemes against experimental measurements, together with subsequent adoption to the regions formed in different stages of the solidification process. A thermomechanical model was finally deemed as the baseline formulation for all the parametric studies in terms of the impact study of thermomechanical shrinkage on macrosegregation and its reproductions of experimental findings.
The dependence of the solidification behavior on the variations of filling rate, melt superheat and initial mold temperature was then investigated, respectively, using the established reliable model. The three casting control parameters were selected due to the shortness or confliction of relevant data in the literature, as well as the complication, high-cost or impossibility of the employment of other macrosegregation control techniques. The studied casting variables included the filling rates of 0.084, 0.107 and 0.145 m/min (with filling times of 38, 30 and 22 min), the initial melt superheats of 75, 65 and 55 °C, the initial mold temperatures of 50, 120 and 250 °C. These casting parameters represented the conditions the most commonly encountered in industry during casting of large size ingot. The results indicated that a quicker filling operation, a lower melt superheat and a colder mold condition tended to decrease the macrosegregation intensity in the upper section of the casting, along the centerline, and in the mid-radius solute-enriched bands. The alleviation effect was associated with the changes in flows induced by pouring jet, thermosolutal convection, thermomechanical deformation of the phases, along with the variations of the temperature gradient in the casting and the solidification speed. The predictions of the studied cases were verified with experimental characterizations on the macro-/micro-structures, solutal distributions, dimensional change of the casting and the temperature variations on mold outside surface.
The findings in the dissertation allow for a better understanding of the underlying mechanisms responsible for the occurrence of macrosegregation in ingot casting process. They should also be helpful for the casting process design of a given ingot and could be used in industry to improve the quality of large size ingot production and the productivity of high value-added steels or other alloys which are prone to macrosegregation.
| Date | 21 Feb 2020 |
|---|
| Original language | American English |
|---|
| Awarding Institution | - École de technologie supérieure
|
|---|
| Supervisor | Mohammad Jahazi (Supervisor) |
|---|
Zhang, C. (Author),
Jahazi (Supervisor),
21 Feb 2020Student thesis: Doctoral thesis › Doctorate in Engineering: Engineering