Guided wave propagation and scattering in pipeworks comprising elbows: Theoretical and experimental results

• Published in: Journal of physics. Conference series Vol. 581; no. 1; pp. 12011 - 12
• Main Authors: Bakkali, M El, Lhémery, A, Baronian, V, Chapuis, B
• Format: Journal Article
• Language: English
• Published: Bristol IOP Publishing 29.01.2015; IOP Science
• Subjects: Computer simulation; Elbows; Finite element method; Inspection; Matching; Mathematical analysis; Mathematical models; Numerical prediction; Physics; Pipe; Pipe bends; Pipework; Propagation; S matrix theory; Scattering; Spherical coordinates; Wave propagation
• ISSN: 1742-6596; 1742-6588; 1742-6596; 1742-6588
• DOI: 10.1088/1742-6596/581/1/012011

Elastic guided waves (GW) are used to inspect pipeworks in various industries. Modelling tools for simulating GW inspection are necessary to understand complex scattering phenomena occurring at specific features (welds, elbows, junctions...). In pipeworks, straight pipes coexist with elbows. GW propagation in the former cases is well-known, but is less documented in the latter. Their scattering at junction of straight and curved pipes constitutes a complex phenomenon. When a curved part is joined to two straight parts, these phenomena couple and give rise to even more complex wave structures. In a previous work, the SemiAnalytic Finite Element method extended to curvilinear coordinates was used to handle GW propagation in elbows, combined with a mode matching method to predict their scattering at the junction with a straight pipe. Here, a pipework comprising an arbitrary number of elbows of finite length and of different curvature linking straight pipes is considered. A modal scattering matrix is built by cascading local scattering and propagation matrices. The overall formulation only requires meshing the pipe section to compute both the modal solutions and the integrals resulting from the mode-matching method for computing local scattering matrices. Numerical predictions using this approach are studied and compared to experiments.