<p>This study investigates the production of boss parts for thick-walled ultra-high pressure vessels (99 MPa) used in hydrogen storage, utilizing a hot-spinning process to accommodate an optimally designed plug. The minimum permissible thickness for the dome and boss parts, made from SA-372 Grade J Class 70 material, was determined to be 28 mm using ASME Sec. 8 Div 3 calculations. Finite element method (FEM) and machine learning were employed to optimize the plug design, with sensitivity analysis identifying thread length and major diameter as key factors affecting equivalent stress and fatigue life. The final design, with a 65 mm thread length, 30.17 mm major diameter, and 8UN thread pitch, achieved a fatigue life of 1×10<sup>8</sup> cycles. Simulation and experimental results confirmed the successful formation of boss parts, with a roller stop timing of 3.0 seconds identified as a critical parameter for controlling hole size. Despite material buildup, all measured thicknesses exceeded 28 mm, ensuring structural safety. A comparison between FEM predictions and field measurements showed good agreement in critical geometrical parameters, with the experimental boss length of 118.1 mm differing by only 5.5 % from the simulation results, confirming the reliability of the modeling approach. From an application perspective, the conservative prediction of a smaller hole diameter ensures sufficient allowance for threading, reducing leakage risks in high-pressure environments. This study demonstrates the feasibility of using hot-spinning for the safe and efficient production of boss parts in hydrogen storage vessels.</p>

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Advanced plug design optimization using machine learning and hot-spinning for boss parts in high-pressure hydrogen vessels (99 MPa)

  • Rivaldo Mersis Brilianto,
  • Gunyoung Park,
  • Chul Kim

摘要

This study investigates the production of boss parts for thick-walled ultra-high pressure vessels (99 MPa) used in hydrogen storage, utilizing a hot-spinning process to accommodate an optimally designed plug. The minimum permissible thickness for the dome and boss parts, made from SA-372 Grade J Class 70 material, was determined to be 28 mm using ASME Sec. 8 Div 3 calculations. Finite element method (FEM) and machine learning were employed to optimize the plug design, with sensitivity analysis identifying thread length and major diameter as key factors affecting equivalent stress and fatigue life. The final design, with a 65 mm thread length, 30.17 mm major diameter, and 8UN thread pitch, achieved a fatigue life of 1×108 cycles. Simulation and experimental results confirmed the successful formation of boss parts, with a roller stop timing of 3.0 seconds identified as a critical parameter for controlling hole size. Despite material buildup, all measured thicknesses exceeded 28 mm, ensuring structural safety. A comparison between FEM predictions and field measurements showed good agreement in critical geometrical parameters, with the experimental boss length of 118.1 mm differing by only 5.5 % from the simulation results, confirming the reliability of the modeling approach. From an application perspective, the conservative prediction of a smaller hole diameter ensures sufficient allowance for threading, reducing leakage risks in high-pressure environments. This study demonstrates the feasibility of using hot-spinning for the safe and efficient production of boss parts in hydrogen storage vessels.