Session: Construction & Industrialization

Paper Number: 164952

164952 - Parametric Modelling and Robotic Trajectory Generation for Inspection of Welded Tubular Joints 

Abstract: 

Introduction

The inspection of welded tubular joints in offshore jacket-type structures is a critical process to ensure structural safety, as these structures are exposed to extreme conditions such as marine corrosion and constant dynamic loads. In this sense, structural integrity depends significantly on the quality of their welded tubular joints.

Non-Destructive Testing (NDT), particularly Phased Array Ultrasonic Testing (PAUT), is a widely used technique for detecting welding defects before commissioning. However, manual inspection presents challenges in terms of accuracy, repeatability, and operator ergonomics. In this context, robotic automation offers a solution that enhances precision, standardisation, and traceability by recording sensor positions and orientations throughout data acquisition. Moreover, minimising direct manual intervention reduces ergonomic risks, improving working conditions in industrial environments.

Contribution

This study proposes a parametric modelling and robotic trajectory generation methodology for inspection of welded tubular joints. Unlike conventional manual inspection, this approach enables automated trajectory planning, significantly reducing the need for manual adjustments. By integrating robotic automation and PAUT technology, this method ensures repeatability, traceability, and improved inspection accuracy, contributing to safer and more efficient structural assessments in offshore environments.

Methodology

A KUKA KR8 2100 robotic arm with six DOF is used, along with a turntable as the seventh axis. A Phased Array Panther II ultrasonic sensor, mounted on the robot flange, moves along the entire weld, scanning the joint in its entirety and recording data as it progresses.

Preliminary results/conclusions

The developed methodology allows adaptation to Y, X, and K joint types, significantly reducing reprogramming time and associated costs by facilitating the parametrisation of new geometries without the need for extensive manual adjustments.

Presenting Author: Micaela Gomez Coronel Universidade da Coruña

Presenting Author Biography: Micaela is a civil engineer with a master's degree in robotics and industrial informatics. After completing her master's, she started working on a shipyard automation project, where she collaborates on the robotic automation of industrial processes such as welding and weld inspection. Her work involves integrating industrial robots and optical sensors to improve precision, efficiency, and quality control in shipbuilding and offshore structural applications.

Authors:

Micaela Gomez Coronel Universidade da Coruña
Alberto Ramil Universidade da Coruña
Armando Yáñez Universidade da Coruña
Javier Montero Universidade da Coruña