A Hybrid Modeling Approach for Performance Prediction of Fouled Spiral Fin-Tube Heat Exchanger

Abstract

Spiral finned tube heat exchangers are extensively used in petrochemical, power electronics, and metallurgical industries due to their high efficiency and compact design. However, fouling accumulation during operation significantly reduces heat transfer efficiency and increases pressure loss. This study develops a hybrid approach integrating discrete element method (DEM), finite element analysis (FEA), and HTRI Xchanger Suite 7 software to correlate fouling thickness with thermal performance and establish a prediction model for tube-side outlet temperature under varying conditions. DEM simulations analyze dust deposition patterns and determine equivalent fouling thickness distribution. A fouling-integrated FE model then evaluates how fouling thickness affects both heat transfer and flow resistance coefficients. Through orthogonal experimental design considering fouling thickness, ambient temperature, and inlet air velocity, thermal resistance values calculated from FEA are imported into HTRI to predict outlet temperature. A random forest algorithm is subsequently employed to develop a multivariable prediction model. Validation conducted on a spiral finned tube heat exchanger at Chongqing Xiangguosi Underground Gas Storage Co., Ltd. (Chongqing, China) confirmed close agreement between simulated and actual fouling patterns. The maximum relative error of the predicted outlet temperatures on the testing dataset was 0.1869%, demonstrating the proposed method’s potential to support performance evaluation and operational optimization of fouled heat exchangers.