定位销
焊接
接头(建筑物)
结构工程
多元统计
转速
预测建模
对接接头
材料科学
机械工程
计算机科学
工程类
复合材料
机器学习
作者
Yun Xu,Danying Gao,Aleksandr Grigor’yevich Chernykh
出处
期刊:Holzforschung
[De Gruyter]
日期:2022-03-10
卷期号:76 (6): 544-555
被引量:4
摘要
Abstract Since rotational wood welding was patented in 1997, this technology has been successfully used in the field of timber structures and furniture. A growing body of literature has investigated the mechanical behaviour of wood-dowel welding joints in tension to study the effect of various factors on welding strength. However, up to now, only two predictive approaches were proposed by researchers, and due to the lack of small samples, they were applied in a limited scope. Therefore, this paper aimed to find a way to establish a multivariate model to compute their carrying capacity. A pull-out test database including 1338 trials from previous literature was used as the basis of modelling. Due to the similarity of joints with glued-in-rod, after evaluating its model validation, the modelling method was applied to the welded joints. The data analysis was carried out to determine the primary influences (such as: rotational speed, predrilled hole diameter, welded length and substrate density). Based on the non-linear regression analysis, a multivariate prediction model was proposed, and results showed positive accuracy with measured values. This study provides a new comprehensive assessment of the withdrawal strength of welded joints, and can help identify future technical guidelines within the wood industry.
科研通智能强力驱动
Strongly Powered by AbleSci AI