En
期刊文章
文章精选
当期目录
过刊浏览
高被引
高下载
特邀文章
专刊合集
待刊文章
Volume 116
您当前的位置：首页 > 期刊文章 > 当期目录 > Volume 116
Efficient prediction method for erosion-corrosion damage in elbows of coal indirect liquefaction industrial pipelines
Liyun Zhu a, Yongchao Li a, Sheng Chen b, Meng He b, Jiarui Shi a, Zhongwei Liu a
a China University of Petroleum (East China), Qingdao, 266580, China
b Technology Innovation Center of Risk Prevention and Control of Refining and Chemical Equipment, State Administration for Market Regulation, China Special Equipment Inspection and Research Institute, Beijing, 100029, China
10.1016/j.partic.2026.06.021
Volume 116, September 2026, Pages 152-168
Received 18 January 2026, Revised 6 June 2026, Accepted 10 June 2026, Available online 27 June 2026, Version of Record 3 July 2026.
E-mail: chensheng_csei@163.com
Highlights

• Euler-Lagrange with EMMS drag and erosion-corrosion models agrees well with industrial data.

• Artificial Neural Network predicts maximum erosion rate and position for elbows under varying parameters.

• Inlet vorticity critically affects erosion rate and position, dictating flow range and damage shape.


Abstract

The elbow of indirect coal liquefaction pipelines is a high-risk area for coupled erosion-corrosion damage from liquid-solid two-phase flow, significantly impacting long-term safe operation. This study investigated coupled mechanisms under various spatial arrangements, structures, and operating parameters using an Euler-Lagrange CFD method incorporating an EMMS drag and corrosion factor correction model. Additionally, four neural models were developed for rapid prediction of maximum erosion rate and location based on extensive simulation data. Results indicate that erosion-affected zone distributions remained consistent across three spatial arrangements with increasing inlet velocity, particle concentration, and corrosion factor, while maximum erosion rate increased. As bend angle increased, the most severe erosion shifted farther from the outlet, with 45° elbows showing lower maximum erosion rates than 90° and 180° configurations. Fluid vorticity elevated maximum erosion rates and transformed erosion patterns from strip-like to cluster-like. The RIME-BP model achieved superior prediction accuracy for maximum erosion rate (R2 ≈ 0.91, RPD >3) and location (R2 > 0.99, 95% residuals within ±2) compared to BP, PSO-BP, and SSA-BP models. This neural network-based method enables rapid damage assessment while considerably improving erosion prediction, monitoring, inspection, and risk prevention efficiency.

Graphical abstract
Keywords
Liquid-solid two-phase flow; Coupled erosion-corrosion model; EMMS; RIME-BP
文章导读