Authors:

Ruey-Bin Tsai

Date Created:

Not specified

Course Title:

Professor:

Not specified

About Paper:

Quenching is the process of rapidly cooling a material in a medium with a lower temperature. Controlling quenching conditions, such as quench speed, medium viscosity, and material orientation, is crucial for metals processing, as deviations can lead to drastically different microstructures and properties. While conventional metal quench oils, such as mineral oils and other petroleum derivatives, are effective, they are often expensive and unsustainable. Therefore, identifying sustainable quench alternatives, such as vegetable oils, and understanding their quench properties is essential. This study focuses on measuring the cooling characteristics of over 70 commercially available oils following the ASTM D6200 standard. These characteristics are utilized in a Gaussian process machine learning model, along with other quench-related properties such as boiling point, degradation point, heat capacity, and viscosity, to predict cooling curves. The aging of the oils during quenching is also investigated through accelerated aging and rheometry. The constructed machine learning model predicts cooling curve characteristics with high accuracy. Pearson correlation plots reveal strong connections between the cooling rate of the quenched metal and the boiling point and viscosity of the quenchant. Additionally, accelerated aging of high-oleic soybean oil demonstrates an exponential relationship between aging time and oil viscosity. Although further quench tests and refinements of the machine learning model are necessary, this work provides a basis for understanding the quench-related properties of commercial oils and identifying sustainable quench oil alternatives. Keywords: Quenchants; Vegetable Oils; Machine Learning; Thermal Aging

Source:

Purdue University / 2024

Topics:

No topics listed

Co-authors:

Ruey-Bin Tsai