condensation is a well-known phase change phenomenon which can be found in a wide range of commercial applications such as power generation, water desalination, thermal management, refrigeration, and air conditioning.  In these applications, filmwise condensation (FWC) is the dominant condensation mode. However, dropwise condensation (DWC) has attracted significant attention in recent years due to enhanced heat transfer performance such as a much higher heat fluxes compared with FWC.  Numerous experimental researches have been reported in identifying the effects of contact angle, advancing and receding contact angle, sub-cooling temperature, saturation vapor temperature and coating layer thickness on DWC heat transfer performance and some theorical models have been developed to calculate heat transfer performance.  However, there is a lack of an accurate prediction method/tool which can correlate these key parameters with DWC heat transfer performance for DWC system design and guidance purposes.

Artificial neural networks (ANNs) is a specific type of artificial neural algorithms which is vastly used in predicting without complete understanding of physical problem. For training ANNs, different algorithms have been developed. This paper reports a preliminary study of predicting heat flux of a vertical DWC surface in a wide range of contact angle (CA), advancing and receding contact angle (CA~adv~, CA~rec~), sub-cooling temperature (T~sub~), saturation vapor temperature (T~sat~) and coating layer resistance (R~coat~) with an ANN method which is optimized by genetic algorithm (GA). The inputs of the network are T~sub~, T~sat~, CA, CA~adv~, CA~rec~ and R~coat~ of DWC and the target is heat flux. Feed forward network with genetic training algorithm is used. For GA, the input is weights and biases of the network and the output is optimized weights and biases. Root Mean Square Error (RMSE) was used to show the error between predicting and experimental heat flux from literature. The transfer functions that were used in hidden layers and the output layer are sigmoid and linear act functions, respectively. A total of 700 data points was utilized in the ANN-GA.

A good agreement was obtained between the predicted heat fluxes of DWC surfaces and experimental results reported by literature. The developed machine learning model shows good potential as an effective tool in predicting thermal performance of DWC on hydrophobic surfaces under a wide range of contact angle, contact angle advancing and receding, sub-cooling and saturation temperatures, coating layer resistance. 


A Preliminary Study of Dropwise Condensation Using Artificial Neural Network Method

Division / Sponsors: 
Aerospace, Bioengineering, Applied Mechanics, Design Engineering Division, Advanced Energy Systems, Manufacturing Engineering Division, Noise Control and Acoustics, Nondestructive Evaluation, Diagnosis, and Prognosis (NDPD), Nuclear Engineering Division, Process Industries Division, Fluids Engineering, Heat Transfer, Materials, Microelectromechanical Systems, Safety Engineering and Risk Analysis, Computers & Information in Engineering


Paper Proceedings: https://asmedigitalcollection.asme.org/IMECE/IMECE2020/volume/84478

IMECE2020 - Virtual Conference

Supporting the need among researchers worldwide for additional online and multimedia content, ASME now offers the ASME Conference Video Collection. This platform will collect and provide long-term access to conference presentation videos, encompassing all technology areas within ASME's technical conference and events program. About 30,000-40,000 engineers across all mechanical engineering and related disciplines participate in ASME conferences each year. This includes authors of approximately 6,000 to 8,000 technical papers published by ASME, the majority of whom also submit video presentations for our virtual conferences. The video collection will be available as an add-on to existing product packages or as a separate subscription.