Fouling modeling and prediction approach for heat exchangers using deep learning

Sundar, Sreenath; Rajagopal, Manjunath C.; Zhao, Hanyang; Kuntumalla, Gowtham; Meng, Yuquan; Chang, Ho Chan; Shao, Chenhui; Ferreira, Placid; Miljkovic, Nenad; Sinha, Sanjiv; Salapaka, Srinivasa M.

doi:10.1016/j.ijheatmasstransfer.2020.120112

Cited by 66 publications

(19 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…[50]- [52], implantable biomedical devices [4], [53], etc., and for concurrent temperature and fouling measurements in industrial pipelines [34], especially in low-temperature heat exchangers [54]- [56] to ensure profitable heat recovery. Overall, new or existing magnetostrictive sensor packages can be suitably adapted to measure temperatures remotely, using either the TCF-or TCVbased technique.…”

Section: Discussionmentioning

confidence: 99%

Design and analysis of magnetostrictive sensors for wireless temperature sensing

Rajagopal

Sinha

2021

Review of Scientific Instruments

Self Cite

View full text Add to dashboard Cite

Magnetostrictive transducers are commonly used as actuators, sonar transducers, and in remote non-destructive evaluation. Their use in wireless thermometry is relatively unexplored.Since magnetostriction-based sensors are passive, they could potentially enable long-term nearfield thermometry. While the temperature sensitivity of resonance frequency in magnetostrictive transducers has been reported in previous studies, the origin of the temperature sensitivity has however not been elucidated. Here, we identify material properties that determine temperature sensitivity, and identify ways to improve sensitivity as well as the detection technique. Using a combination of analytical and computational methods, we systematically identify the material properties that directly influence the temperature coefficient of resonance frequency (TCF). We first experimentally measure the shift in resonance frequency due to temperature changes in a Metglas strip to be 0.03%K -1 . Using insights from theory, we then experimentally demonstrate a 5-fold improvement to the TCF by using Terfenol in place of Metglas as the magnetostrictive sensor material. We further demonstrate an alternate temperature sensing technique that does not require measuring the resonance frequency, consequently reducing instrument complexity. This work provides a general framework to analyze magnetostrictive materials and the sensing scheme for near-field wireless thermometry.

show abstract

Section: Discussionmentioning

confidence: 99%

Design and analysis of magnetostrictive sensors for wireless temperature sensing

Rajagopal

Sinha

2021

Review of Scientific Instruments

Self Cite

View full text Add to dashboard Cite

show abstract

“…Nikoo et al [27] used extended Derjaguin, Landau, Verwey, and Overbeek (DLVO) theory to predict the CaSO 4 fouling propensity and compared the results obtained from this technique with experimental data in their investigation, and found 75% of all data consistent with the proposed fouling propensity indicator (FPI) criterion. Sundar et al [28] developed a generalized and scalable model for accurately predicting fouling resistance based on deep learning for cross-flow heat exchangers. The heat exchangers' fouling and heat transfer physics have been demonstrated through a neural network framework.…”

Section: Literature Surveymentioning

confidence: 99%

CFD analysis of Gypsum crystallization fouling in 2D plate heat exchangers

2022

IJATEE

View full text Add to dashboard Cite

The most common equipment used in processes industries are heat exchangers. These devices are used for exchanging heat between two process fluid streams. The heat exchanger is an essential modern tool in the process industry. The objectives of heat exchangers in the process industries are to meet boiling, condensation, bulk heating, and cooling or evaporation requirements.*Author for correspondence Heat exchangers' performance and efficiency can be measured by measuring the amount of heat transferred using the minimum area of heat transfer and the pressure drop. The main problem faced during the working of heat exchangers is fouling, which results in their poor efficiency. Fouling is the deposition and buildup of undesired elements in process fluids, such as scale, suspended particles, insoluble salts, and so on, on the interior or external surfaces of processing equipment such as boilers and heat exchangers. The most significant consequence of fouling is the reduction in the thermal efficiency of a

show abstract

“…PHM-XAI works associated with anomaly detection and failure prognostic are summarized. In the order of presentation: (i) interpretable model [40,41]; (ii) extraction-based approach [42]; (iii) decision rules and knowledge-based explanation [43]; (iv) attention mechanism [44]; (v) model agnostic [45]; and (vi) visual explanation technique [46].…”

Section: Related Workmentioning

confidence: 99%

Abnormality Detection and Failure Prediction Using Explainable Bayesian Deep Learning: Methodology and Case Study of Real-World Gas Turbine Anomalies

Nor¹,

Pedapati²,

Muhammad³

et al. 2022

Preprint

View full text Add to dashboard Cite

: Mistrust, amplified by numerous artificial intelligence (AI) related incidents, has caused the energy and industrial sectors to be amongst the slowest adopter of AI methods. Central to this issue is the black-box problem of AI, which impedes investments and fast becoming a legal hazard for users. Explainable AI (XAI) is a recent paradigm to tackle this challenge. Being the backbone of the industry, the prognostic and health management (PHM) domain has recently been introduced to XAI. However, many deficiencies, particularly lack of explanation assessment methods and uncertainty quantification, plague this young field. In this paper, we elaborate a framework on explainable anomaly detection and failure prognostic employing a Bayesian deep learning model to generate local and global explanations from the PHM tasks. An uncertainty measure of the Bayesian model is utilized as marker for anomalies expanding the prognostic explanation scope to include model’s confidence. Also, the global explanation is used to improve prognostic performance, an aspect neglected from the handful of PHM-XAI publications. The quality of the explanation is finally examined employing local accuracy and consistency properties. The method is tested on real-world gas turbine anomalies and synthetic turbofan data failure prediction. Seven out of eight of the tested anomalies were successfully identified. Additionally, the prognostic outcome showed 19% improvement in statistical terms and achieved the highest prognostic score amongst best published results on the topic.

show abstract

Fouling modeling and prediction approach for heat exchangers using deep learning

Cited by 66 publications

References 32 publications

Design and analysis of magnetostrictive sensors for wireless temperature sensing

Design and analysis of magnetostrictive sensors for wireless temperature sensing

CFD analysis of Gypsum crystallization fouling in 2D plate heat exchangers

Abnormality Detection and Failure Prediction Using Explainable Bayesian Deep Learning: Methodology and Case Study of Real-World Gas Turbine Anomalies

Contact Info

Product

Resources

About