Screening datasets for laser welded steel-copper lap joints

V1:

The data show a definitve screening design to evaluate the the influence of six factors (laser beam power (W), welding speed (m/min), angular position in welding direction (°), focal positon (mm), gas flow rate (l/min), material thickness of the steel sheet (mm)) in three levels and 18 parameter combinations on the weld depth and the geometrical dimensions of the weld metal in laser welded steel-copper joints in the lap configuration with steel on the top side. Every parameter combination was repeated 5 times and every sheet was cuttet 4 times to overall generate 360 cross sections. Every line in the dataset stands for a cross section which was evaluated regarding the dimensions of the weld metal. Additionally, there was a dichotomous data column added for cracking in the weld metal (yes/no).
The dataset is not suitable for modelling a precise predicting model of weld depth in the copper sheet, but shows a correlation between cracking (yes/no) and the weld depth in the copper sheet. This can be discribed very well in a binominal logistic regression.
Moreover the average crack lenght and count of cracks was added in Versions V1.1

V2:

The data show the same screening experiments but with adjusted laser power levels to garantee sufficient weld depth for all parameter combinations (which was lacking in V1). The target is the average weld depth in the bottom copper sheet for every parameter combination. The data can be used to calculate the parameter effect strenght of the different factors on the weld depth and to build a simple linear model of the coherences between the parameters and the resulting weld depth.
V2.1 ist basically the same dataset but with the resulting copper dilution instead of the weld depth.

V1:
本数据集呈现了一套确定性筛选设计（screening design），用于评估6个因素在3个水平下共18组参数组合对搭接构型（钢侧在上）的激光焊接钢-铜接头的熔深及熔敷金属几何尺寸的影响，6个因素分别为：激光束功率（W）、焊接速度（m/min）、焊接方向角位置（°）、焦点位置（mm）、气体流量（l/min）、钢板材料厚度（mm）。每组参数组合重复5次实验，每块板材切割4次，最终共获得360个焊接截面。数据集中每一行对应一个被评估熔敷金属尺寸的焊接截面，此外还增设了熔敷金属开裂情况的二分类数据列（开裂/未开裂）。
本数据集不适用于构建铜板熔深的精准预测模型，但可揭示熔深与开裂情况（开裂/未开裂）之间的相关性，该相关性可通过二项逻辑回归（binominal logistic regression）进行良好拟合。
此外，V1.1版本新增了平均裂纹长度及裂纹数量字段。

V2:
本数据集包含相同的筛选实验，但调整了激光功率水平，以确保所有参数组合均可获得足够的熔深（该问题在V1版本中存在）。本数据集的目标变量为每组参数组合下底部铜板的平均熔深，可用于计算各因素对熔深的参数效应强度，并构建参数与熔深之间相关性的简单线性模型。
V2.1：V2.1版本为同一数据集，但将目标变量替换为铜元素稀释率而非熔深。