首页|大曲率弯扭叶片涡轮产品注射模具设计关键技术的研究

大曲率弯扭叶片涡轮产品注射模具设计关键技术的研究

扫码查看
由于大曲率弯扭叶片涡轮产品具有空间结构的复杂性,局部曲面存在多法向交错的特征.为了解决涡轮叶片变截曲面结构带来的空间成型、分型与抽芯困难,研究了增压涡轮产品注射模具设计的关键技术.利用Unigraphics NX平台进行三角网格离散化,采用Python语言,对叶片空间离散点法向进行高斯映射,建立了叶片密集点云与高斯球面法向点域的映射关系.基于高斯球面点域的K-means++法向聚类和可视性分析的方法,将分类结果映射到叶片,通过切割形成了模具镶块的初始复合结构.据此设计了多层瓣合式镶件组合机构用于涡轮产品的注射模具成型与抽芯,将传统的三板式模具改进为四次分型结构,提高了涡轮产品注射模具的设计质量以及自动化生产效率.
Study on Key Technology of Injection Mold Design for Turbine Products with Large Curvature Bowed-twisted Blades
Because of the complicated spatial structure of big curvature twist blade turbine products,the local surfaces feature multi-normal intersection properties.The essential technologies of injection mold design for supercharged turbine product in order to tackle the issues of spatial molding,parting and core extraction induced by the varied truncated surface structure of turbine blades is analyzed.The Unigraphics NX platform was used firstly for triangular mesh discretization,and then the Python language was adopted for Gaussian mapping of the discrete point normal to the blade space,establishing a mapping relationship between the dense point cloud of the blade and the Gaussian spherical normal point domain.A K-means++normal clustering and visibility analysis for Gaussian spherical data were suggested.The classification results were transferred to the blade,and cutting was employed to construct the initial composite mold inserts.As a consequence,a multi-layer flap-fitted insert assembly mechanism for injection mold and core extraction of turbine products was developed,transforming the typical three-plate mold into a four-part structure that improves design quality and automated production efficiency.

triangular discretizationgaussian mappingnormal clusteringvisibility analysis

曹钱、李光明、陈磊、杜国源、岳俊、蒙川

展开 >

西南科技大学 制造科学与工程学院,四川绵阳 621010

航空工业成都飞机工业有限责任公司,成都 610073

四川航天燎原科技有限公司,成都 610100

三角离散化 高斯映射 法向聚类 可视性分析

国家自然科学基金项目四川省国际科技合作项目四川省大学生创新创业项目

619014002021YFH0089S202110619121

2024

机械科学与技术
西北工业大学

机械科学与技术

CSTPCD北大核心
影响因子:0.565
ISSN:1003-8728
年,卷(期):2024.43(10)