Investigating the processing of steel

Karisa Steffans, Tech Editor

Ever had an interest in learning more about events that affect steel processing? On Tuesday, Feb. 11 at 11:00 a.m., Bryan Webler will be giving a lecture on one such phenomenon. He will focus specifically on non-metallic inclusions, or small oxide, sulfide, or nitride particles that can affect the properties and processing of steel. His presentation will last for about an hour, and will be taking place in room 610 of the Minerals and Materials Engineering Building.

Steel processing is a necessary aspect of manufacturing for many products. As a result, it is important to have a strong understanding of things that can affect this process, so we can predict and model how steel will behave once it has been processed. As a result, it is important to understand what these non-metallic inclusions do within this process, and how to model them. These inclusions form during liquid steel processing. As a result, it is important to be able to control them so that the steel can be refined.

One major challenge that comes from attempting to do this stems from the complex shapes the inclusions form. They also consist of complex chemicals and can come in various sizes. However, given the environment they come from, perhaps this is not a surprise. These inclusions arise when liquid steel is processed, so they form in an environment of high heat with a variety of chemical reactions and fluid flows influencing their development. As a result, it is difficult to describe and model these inclusions.

A potential way of assisting in forming a model is to study and gather data about these inclusions. Webler intends on discussing the various approaches to studying and modeling their behavior, both by using a data-driven model and by using a mechanistic approach. In other words, he will be discussing observing these inclusions through numbers and comparing it to observing them through more physical observations.

Within the mechanistic approach, researchers focus on physics and chemistry. By doing so, they can describe how the reactions within the metal cause magnesium and calcium modification of the inclusion compositions. Experiments have been done to study these reactions, so that the results could be compared to a kinetic model that had been made. Webler will be discussing the results of this experiment, and what they say about the model that had been originally used to model this reaction.

In comparison, the data-driven approach focuses on the information contained in images of the inclusions. By using machine learning methods, they’ve built a model to predict inclusion composition. Computers were also used to differentiate between inclusions and non-inclusions such as pores or contamination.

Both of these approaches can provide insights into these inclusions. By understanding them better, we can develop better control methods for these in the future. Webler will be discussing this in more depth on Tuesday for anyone interested in learning more.