東華大學圖書館 |

Language: English

Help

回圖書館首頁

手機版館藏查詢

Back

Switch To: Labeled | MARC Mode | ISBD

Experimental and Machine Learning In...

Qiu, Yue,

Linked to FindBook

Google Book

Amazon

博客來

Experimental and Machine Learning Investigation on Mini/Microchannel for High Heat Flux Applications /

Record Type:	Electronic resources : Monograph/item
Title/Author:	Experimental and Machine Learning Investigation on Mini/Microchannel for High Heat Flux Applications // Yue Qiu.
Author:	Qiu, Yue,
Description:	1 electronic resource (230 pages)
Notes:	Source: Dissertations Abstracts International, Volume: 85-08, Section: B.
Contained By:	Dissertations Abstracts International85-08B.
Subject:	Mechanical engineering. -
Online resource:	https://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=31095916
ISBN:	9798381462609

Experimental and Machine Learning Investigation on Mini/Microchannel for High Heat Flux Applications /
Qiu, Yue,

Experimental and Machine Learning Investigation on Mini/Microchannel for High Heat Flux Applications /Yue Qiu. - 1 electronic resource (230 pages)

Source: Dissertations Abstracts International, Volume: 85-08, Section: B.

Flow boiling in mini/micro-channels proves effective for addressing high dissipating requirements in thermal management systems, yet faces two key challenges: 1) Significant pressure drop leading to low efficiency. 2) The difficulty in accurately predicting heat transfer coefficients and pressure drop due to complex fluid and thermal behavior in these systems.In the first part of the study, copper microchannel manifold (MMC) heat sinks using R1233zd(E) were experimentally investigated. Seven test samples, varying in manifold and microchannel geometry parameters, were tested with mass flow rates from 5 to 15 g/s and subcooled inlet temperatures from 3 to 10 K. Results indicate a maximum heat dissipation capacity of 356 W/cm2 at 15 g/s and 3 K, with manifolds reducing pressure drop significantly. However, trapped bubbles limit heat transfer performance at high heat flux.The second part focuses on improving prediction of two-phase flow boiling heat transfer coefficients and pressure drops in mini/microchannels utilizing machine learning models. Using consolidated experiment databases from 50 sources, an Artificial Neural Network (ANN) predicts saturated flow boiling heat transfer coefficients with a mean absolute error (MAE) of 14.3%. Given the randomness in parameter selection for machine learning tools, a systematic approach is suggested for optimizing the ANN model for predicting flow boiling heat transfer coefficients. Initial input features are selected from literature correlations, and additional features are added based on effect size statistics and mutual information. A comprehensive grid search tunes hyperparameters, resulting in a final optimized model with 17 input features and network hidden layers (10, 20, 50, 100, 200, 400), achieving an MAE of 8.48%.ANN together with other three machine learning models, including K-Nearest neighbors regression (KNN), Extreme gradient boosting (XGBoost) and Light gradient boosting machine (LightGBM), are used in flow boiling pressure drop prediction. The models with dimensionless input features predict the test data with MAEs of 9.58%, 10.38%, 13.52%, and 14.49%, respectively. Machine learning proved to be an extremely useful tool when predicting the flow boiling heat transfer coefficient and pressure drop across a variety of different fluids and conditions.

English

ISBN: 9798381462609Subjects--Topical Terms:

649730
Mechanical engineering.
Subjects--Index Terms:

Artificial Neural Network

Experimental and Machine Learning Investigation on Mini/Microchannel for High Heat Flux Applications /
LDR:03868nmm a22004333i 4500 001 2428446
005 20250410083501.5
006 m o d
007 cr|nu||||||||
008 261011s2024 miu||||||m |||||||eng d
020 $a 9798381462609
035 $a (MiAaPQD)AAI31095916
035 $a (MiAaPQD)OhioLINKcase1702141226482346
035 $a AAI31095916
040 $a MiAaPQD $b eng $c MiAaPQD $e rda
100 1 $a Qiu, Yue, $e author. $3 3815662
245 1 0 $a Experimental and Machine Learning Investigation on Mini/Microchannel for High Heat Flux Applications / $c Yue Qiu.
264 1 $a Ann Arbor : $b ProQuest Dissertations & Theses, $c 2024
300 $a 1 electronic resource (230 pages)
336 $a text $b txt $2 rdacontent
337 $a computer $b c $2 rdamedia
338 $a online resource $b cr $2 rdacarrier
500 $a Source: Dissertations Abstracts International, Volume: 85-08, Section: B.
500 $a Advisors: Kharangate, Chirag Committee members: Liao, Ya-Ting; Carter, Jennifer; McGuffin-Cawley, James.
502 $b Ph.D. $c Case Western Reserve University $d 2024.
520 $a Flow boiling in mini/micro-channels proves effective for addressing high dissipating requirements in thermal management systems, yet faces two key challenges: 1) Significant pressure drop leading to low efficiency. 2) The difficulty in accurately predicting heat transfer coefficients and pressure drop due to complex fluid and thermal behavior in these systems.In the first part of the study, copper microchannel manifold (MMC) heat sinks using R1233zd(E) were experimentally investigated. Seven test samples, varying in manifold and microchannel geometry parameters, were tested with mass flow rates from 5 to 15 g/s and subcooled inlet temperatures from 3 to 10 K. Results indicate a maximum heat dissipation capacity of 356 W/cm2 at 15 g/s and 3 K, with manifolds reducing pressure drop significantly. However, trapped bubbles limit heat transfer performance at high heat flux.The second part focuses on improving prediction of two-phase flow boiling heat transfer coefficients and pressure drops in mini/microchannels utilizing machine learning models. Using consolidated experiment databases from 50 sources, an Artificial Neural Network (ANN) predicts saturated flow boiling heat transfer coefficients with a mean absolute error (MAE) of 14.3%. Given the randomness in parameter selection for machine learning tools, a systematic approach is suggested for optimizing the ANN model for predicting flow boiling heat transfer coefficients. Initial input features are selected from literature correlations, and additional features are added based on effect size statistics and mutual information. A comprehensive grid search tunes hyperparameters, resulting in a final optimized model with 17 input features and network hidden layers (10, 20, 50, 100, 200, 400), achieving an MAE of 8.48%.ANN together with other three machine learning models, including K-Nearest neighbors regression (KNN), Extreme gradient boosting (XGBoost) and Light gradient boosting machine (LightGBM), are used in flow boiling pressure drop prediction. The models with dimensionless input features predict the test data with MAEs of 9.58%, 10.38%, 13.52%, and 14.49%, respectively. Machine learning proved to be an extremely useful tool when predicting the flow boiling heat transfer coefficient and pressure drop across a variety of different fluids and conditions.
546 $a English
590 $a School code: 0042
650 4 $a Mechanical engineering. $3 649730
650 4 $a Thermodynamics. $3 517304
653 $a Artificial Neural Network
653 $a Mean absolute error
653 $a Thermal management systems
653 $a High heat flux
690 $a 0548
690 $a 0348
690 $a 0800
710 2 $a Case Western Reserve University. $b EMC - Aerospace Engineering. $e degree granting institution. $3 3815663
720 1 $a Kharangate, Chirag $e degree supervisor.
773 0 $t Dissertations Abstracts International $g 85-08B.
790 $a 0042
791 $a Ph.D.
792 $a 2024
856 4 0 $u https://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=31095916