de Souza Bruno Mota, Bestard Guillermo Alvarez, Alfaro Sadek C Absi
Postgraduate Program in Mechatronic Systems, University of Brasília, Brasília, Brazil.
Electronic Engineering Group, Faculty of Science and Technology in Engineering, University of Brasília, Brasília, Brazil.
Data Brief. 2025 Jul 10;61:111883. doi: 10.1016/j.dib.2025.111883. eCollection 2025 Aug.
In pursuit of smarter and more efficient welding methods, assessing the quality of welded components remains critical but is often constrained by traditional destructive testing methods for evaluating production batches. This work addresses this challenge by presenting a comprehensive dataset collected during the Gas Metal Arc Welding (GMAW) process, which has been utilised in models for estimating weld bead width, penetration, and reinforcement. The experiments employ thermographic sensors and laser profilometers integrated into a motorised and controlled welding setup. Detailed descriptions of the dataset structure, acquisition techniques, and potential applications are provided. The dataset can be a valuable resource for researchers seeking to develop and test advanced control algorithms and dynamic models of the GMAW welding process. Data from four experiments are available, offering extensive material for training, validation, and further exploration.
在追求更智能、更高效的焊接方法过程中，评估焊接部件的质量仍然至关重要，但往往受到用于评估生产批次的传统破坏性测试方法的限制。这项工作通过展示在气体金属电弧焊（GMAW）过程中收集的综合数据集来应对这一挑战，该数据集已用于估计焊缝宽度、熔深和余高的模型中。实验采用集成在电动和受控焊接装置中的热成像传感器和激光轮廓仪。提供了数据集结构、采集技术和潜在应用的详细描述。该数据集对于寻求开发和测试GMAW焊接过程的先进控制算法和动态模型的研究人员来说可能是一个宝贵的资源。现有来自四个实验的数据，为训练、验证和进一步探索提供了丰富的材料。
