Deep learning strategies for critical heat flux detection in pool boiling

Rassoulinejad-Mousavi, Seyed Moein; Al-Hindawi, Firas; Soori, Tejaswi; Rokoni, Arif; Yoon, Hyunsoo; Hu, Hai‐Jie; Wu, Teresa; Sun, Ying

doi:10.1016/j.applthermaleng.2021.116849

Cited by 27 publications

(4 citation statements)

References 56 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Recently, more studies have focused on developing deep learning based models using visual data for pool boiling heat ux estimation. [290][291][292] One such example was data collection from pool boiling experiments using a commercialized DSLR camera and training of a convolutional neural network (CNN) to predict heat ux in nucleate boiling regime. 290 The CNN model was able to capture bubble morphology which encoded extensive information about the heat ux.…”

Section: Machine Learning Methods For Boiling and Condensationmentioning

confidence: 99%

Advances in micro and nanoengineered surfaces for enhancing boiling and condensation heat transfer: a review

et al. 2023

View full text Add to dashboard Cite

show abstract

Section: Machine Learning Methods For Boiling and Condensationmentioning

confidence: 99%

Advances in micro and nanoengineered surfaces for enhancing boiling and condensation heat transfer: a review

et al. 2023

View full text Add to dashboard Cite

show abstract

“…The principle of the thermoelectric effect method, on the other hand, is that materials with anisotropic thermal conductivity generate an electric field with a transverse component in the main axis of the material when heat passes through it due to the Seebeck effect, thus enabling the heat flux to be obtained by detecting the electrical signal, which allows for the ultra-fast response and is suitable for transient heat flux measurements [113]. With the continuous development of Machine Learning, the image [114,115] and acoustic signals [116] of boiling are detected in order to develop a boiling heat flux measurement system with the aid of the Convolutional Neural Networks (CNNs) [117] and Multilayer Perceptron Neural Networks (MLPNNs) [118].…”

Section: Boiling-heat-transfer Coefficient H and Heat Flux Qmentioning

confidence: 99%

A Review of Pool-Boiling Processes Based on Bubble-Dynamics Parameters

Xiao,

Zhuang,

et al. 2023

Applied Sciences

View full text Add to dashboard Cite

Immersion cooling is widely used for thermal management of servers. The two-phase immersion cooling, which transfers heat by boiling, possesses efficient temperature control ability under intensive heat generation. In the process of temperature control through boiling, the generation and transportation of bubbles play a crucial role in calculating the heat-transfer capacity. Therefore, it holds immense significance to obtain a profound understanding of the mechanisms underlying bubble formation and detachment. Currently, numerous mechanistic explanations and empirical correlations have been proposed to elucidate the various parameters of bubbles during the boiling process. These findings were considered to be valuable references when selecting appropriate boiling media and designing efficient heating surfaces. To comprehensively present the progress of bubble formation and heat transfer in the boiling system, the forces exerted on the bubbles are highlighted in this article. A meticulous review of bubble-force analysis and correlation formulae pertaining to various relevant parameters (e.g., nucleation sites density, bubble growth rate, bubble growth period, and detachment frequency) was conducted. This review article was also expected to provide a novel foundation for further exploration of enhanced boiling heat transfer.

show abstract

“…Classification algorithms have been a rapidly growing field in recent years, particularly with the significant advancements in deep learning and computer vision technologies. These advancements enabled classification algorithms to become highly efficient and accurate, making them applicable to a wide range of applications in various domains Alhindawi & Altarazi (2018); Altarazi et al (2019); Rassoulinejad-Mousavi et al (2021); Padmapriya & Sasilatha (2023); Omeroglu et al (2023). The success of most classification models is contingent on 1) having access to large, labeled datasets that allow for the training of models with high accuracy and 2) that the model is applied (tested) on data coming from the same domain that the model is trained on and is familiar with.…”

Section: Introductionmentioning

confidence: 99%

Pseudo Supervised Metrics: Evaluating Unsupervised Image to Image Translation Models In Unsupervised Cross-Domain Classification Frameworks

Al-Hindawi¹,

Siddiquee²,

Wu³

et al. 2023

Preprint

View full text Add to dashboard Cite

Highlights• New metric for evaluating UI2I translation models using pseudo supervised metrics that is designed specifically to support cross-domain classification.• The metric not only outperforms unsupervised metrics such as the FID, but is also highly correlated with the true supervised metrics, robust, and explainable.• We demonstrate that this metric can be used for future research in this field by applying it to a critical real-world problem (the boiling crisis problem).

show abstract

Deep learning strategies for critical heat flux detection in pool boiling

Cited by 27 publications

References 56 publications

Advances in micro and nanoengineered surfaces for enhancing boiling and condensation heat transfer: a review

Advances in micro and nanoengineered surfaces for enhancing boiling and condensation heat transfer: a review

A Review of Pool-Boiling Processes Based on Bubble-Dynamics Parameters

Pseudo Supervised Metrics: Evaluating Unsupervised Image to Image Translation Models In Unsupervised Cross-Domain Classification Frameworks

Contact Info

Product

Resources

About