Beyond the Basics: Multi-Objective Optimization for Design of Heat Exchangers Webinar
May 21, 2025 | 9:30 AM CDT (UTC -05:00)
Commercial heat exchangers in the process industries are designed according to standards from entities such as TEMA or API. The designer’s goal is to maximize the heat transfer coefficient and minimize cost, area, and pressure drop. However, these objectives often conflict. For example, changing a geometric design parameter such as tube diameter may simultaneously cause a favorable change in one performance aspect and an unfavorable change in another. Heat exchanger design thus falls in the domain of multi-objective optimization (MOO).
In this webinar, we review what MOO is, how it differs from single-objective optimization, and how and why a heat exchanger design problem is a MOO problem. We illustrate how to solve the problem using readily available tools in Python®. A heat exchanger designer may have to change many parameters (often one at a time) to identify a single “best” design. This webinar illuminates how applying MOO in an automated process may allow the designer to choose among a class of best designs instead of derivig only one.
Join this webinar to learn more.
Duration: Approximately 60 minutes
Registration deadline: May 16, 2025 (11:59 PM US Central Time)
Facilitated by Parimah Kazemi
Numerical Analyst, earned a BS in Mathematics and Chemistry and a PhD in Mathematics from the University of North Texas, Denton, Texas, USA. Following graduation, she worked as a Postdoctoral, Visiting Assistant Professor, and Senior Analyst before joining HTRI. She has authored scientific publications in refereed journals on scientific computation, physics, and analysis, as well as given presentations both nationally and internationally. At HTRI, her expertise in mathematical modeling, numerical analysis, and scientific computing is focused on the solution of heat exchange problems using state-of-the-art and proven mathematical and numerical techniques. Her skills in data science—specifically, machine learning and artificial neural networks—have increased the usability and value of HTRI's extensive and growing empirical data sets.