An AI-enabled smart ultrasonic contact drying system for nonthermal drying of biopolymeric materials: experimental and theoretical studies

The U.S. government has been committed to achieving the target of net-zero emissions operations by 2050, including a 65% emissions reduction by 2030 via implementing a number of decarbonization strategies that will contribute to improving technological development, innovations, and energy efficiency. The food industry in the U.S., one of the largest in the world and 5th highest consumer of energy in the manufacturing industry (EPA, 2007), plays a crucial role in ensuring that this aim is realized. Drying is an essential technique of food preservation, reducing costs in packaging, transportation, waste reduction, and increasing commodity value of final products. Latent heat of vaporization causes a significant amount of energy to be lost during thermal drying, making it the second most energy-intensive food processing process after distillation. Energy needed in manufacturing is commonly provided by fossil fuels, most of which is dissipated and lost via a release of greenhouse gases (GHG) such as nitrous oxide (N2O), methane (CH4), and carbon dioxide (CO2). Thus, lowering energy consumption and modifying energy input by process electrification would benefit the economy as well as contribute to development of a less carbon-intensive industry. This thesis reported the first attempt to develop a novel sub-pilot multi-frequency, multimode, and modulated (MMM) ultrasonic dryer that was used in dehydration of wet distillers grain (WDG) and apple slices with improved nutrient quality. This unique direct-contact MMM ultrasonic system can provide a uniform, repeatable, and homogenous distribution of high intensity vibrations on the drying surface while avoiding the creation of stationary/standing waves by synchronously exciting multiple vibrating modes. The MMM direct-contact ultrasonic drying has showed promising results in improving quality, reducing energy consumption and drying time, and promoting sustainability. Furthermore, a closed loop hot air drying chamber with a rectangular cross section was designed to develop a hot air & ultrasound hybrid batchwise dryer that utilizes the synergistic effects of conventional hot air drying and ultrasound to conserve energy and improve food quality. In addition to technological improvements, control and optimization strategies could significantly reduce energy consumption, waste, and greenhouse gas emissions, as well as increase sustainability at a lower cost. Multiobjective optimization, however, poses a unique challenge in decision making strategies, which traditionally has focused on singleobjective strategies. It is therefore important to optimize the drying process in accordance with product quality and energy consumption requirements. This thesis also aimed to develop a strategy applying dynamic programming and reinforcement learning to optimize the ultrasonic drying process to reduce energy consumption and improve product quality. This was done with problem formulation and developing appropriate data-driven model for prediction of drying kinetics and color changes of apple slices based on response surface methodology (RSM), and finding optimal operating condition by regulating process controls in sequence for minimizing energy consumption and achieving the desired quality of final product. It is shown that the developed technique can be used to further reduce energy consumption compared to traditional optimization techniques. The outcomes of this work will impact significantly the design and operation of drying systems in terms of production cost, energy savings, time efficiency, quality, and simplification/automation. For the purpose of providing data for development of optimization algorithms in the future to deal with uncertainty and online decision making, as well as to facilitate process monitoring during ultrasonic drying, this study investigated the feasibility of integrating portable near-infrared spectroscopy (NIRS) with multivariate analysis to measure moisture content and hardness of apple slices in real time. Consequently, an efficient and rapid method for determining hardness and moisture content of materials during ultrasound contact drying was developed, and a closed-form equation was introduced for rapid and real-time monitoring. Finally, to provide an understanding of transport phenomena in ultrasonic drying processes, as well as facilitating physics-based simulations, an experimental setup was developed, and a procedure was introduced to calculate permeability during drying. Permeability is one of the key properties of porous media. It is the internal resistance of a porous medium against a pressure-driven flow and signifies capability of the porous medium to allow flow through its microstructures. However, in drying of food products, measuring permeability involved distinct complications. An important issue is fluid involved in a drying process often contains three phases of water, air, and water vapor and the biopolymeric material undergoes shrinkage. The experimental setup and approach allowed intrinsic and relative gas permeability of apple chips to be measured during ultrasonic and hot air drying. Permeability of apple tissues was correlated with moisture content and porosity in a suitable manner.

Name thesis:degree_name

Ph.D.

Level thesis:degree_level

Dissertation

Discipline thesis:degree_discipline

Agricultural & Biological Engr

Grantor

University of Illinois at Urbana-Champaign

Year dc:date

2022

Author dc:creator

• Malvandi, Amir

Contributors dc:contributor

• Feng, Hao
• Rausch, Kent D.
• Kamruzzaman, Mohammed
• Wang, Yi-Cheng

• Ultrasonic Contact Drying
• Smart Drying System
• Dynamic Optimization
• Deep Reinforcement Learning
• Non-thermal Drying
• Mass Transfer Enhancement

Statement dc:rights

• Copyright 2022 Amir Malvandi

Language dc:language

en, eng

Handle dc:identifier

https://hdl.handle.net/2142/117583

OAI record oai:identifier

https://www.ideals.illinois.edu/oai-pmh?verb=GetRecord&metadataPrefix=etdms&identifier=oai%3Awww.ideals.illinois.edu%3A2142%2F117583