Specimens with intentionally embedded weld defects or flaws can be employed for training, development and research into procedures for mechanical property evaluation and structural integrity assessment. It is critical that the artificial defects are a realistic representation of the flaws produced by welding. Cylindrical holes, which are usually machined after welding, are not realistic enough for our purposes as it is known that they are easier to detect than the naturally occurring imperfections and cracks. Furthermore, it is usually impractical to machine a defect in a location similar to where the real weld defects are found. For example, electro-discharge machining can produce a through hole (cylindrical reflector) which neither represents the weld porosity (spherical voids) nor the weld crack (planar thin voids). In this study, the aim is to embed reflectors inside the weld intentionally, and then locate them using ultrasonic phased array imaging. The specimen is an 8 mm thick 080A15 Bright Drawn Steel plate of length 300 mm. Tungsten rods (ø2.4-3.2 mm & length 20-25 mm) and tungsten carbide balls (ø4 mm) will be used to serve as reflectors simulating defects within the weld itself. This study is aligned to a larger research project investigating multi-layer metal NDE found in many multi-pass welding and wire arc additive manufacturing (WAAM) applications and as such, there is no joint preparation as the first layer is deposited over the plate surface directly and subsequent layers contribute to the specimen build profile, similar to the WAAM samples. A tungsten inert gas welding torch mounted on a KUKA robot is used to deposit four layers for each weld, with our process using nine passes for the first layer, down to six passes for the last layer. During this procedure, the tungsten artificial reflectors are embedded in the weld, between the existing layers. The sample is then inspected by a 10 MHz ultrasonic phased array in direct contact with the sample surface using both conventional and total focusing method (TFM) imaging techniques. A phased array aperture of 32 elements has been used. The phased array controller is FIToolbox (Diagnostic Sonar, UK). Firstly, a focused B-scan has been performed with a range of settings for the transmit focal depth. Secondly, a full-aperture TFM method has been processed. All the reflectors of interest were detected successfully using this combination of B-scan and TFM imaging approaches.

Javadi Y., Vasilev M., Macleod C.N., Pierce S.G., Su R., Mineo C., et al. (2019). Intentional weld defect process: From manufacturing by robotic welding machine to inspection using TFM phased array. In AIP Conference Proceedings. 2 HUNTINGTON QUADRANGLE, STE 1NO1, MELVILLE, NY 11747-4501 USA : American Institute of Physics Inc. [10.1063/1.5099761].

Intentional weld defect process: From manufacturing by robotic welding machine to inspection using TFM phased array

Mineo C.;
2019-01-01

Abstract

Specimens with intentionally embedded weld defects or flaws can be employed for training, development and research into procedures for mechanical property evaluation and structural integrity assessment. It is critical that the artificial defects are a realistic representation of the flaws produced by welding. Cylindrical holes, which are usually machined after welding, are not realistic enough for our purposes as it is known that they are easier to detect than the naturally occurring imperfections and cracks. Furthermore, it is usually impractical to machine a defect in a location similar to where the real weld defects are found. For example, electro-discharge machining can produce a through hole (cylindrical reflector) which neither represents the weld porosity (spherical voids) nor the weld crack (planar thin voids). In this study, the aim is to embed reflectors inside the weld intentionally, and then locate them using ultrasonic phased array imaging. The specimen is an 8 mm thick 080A15 Bright Drawn Steel plate of length 300 mm. Tungsten rods (ø2.4-3.2 mm & length 20-25 mm) and tungsten carbide balls (ø4 mm) will be used to serve as reflectors simulating defects within the weld itself. This study is aligned to a larger research project investigating multi-layer metal NDE found in many multi-pass welding and wire arc additive manufacturing (WAAM) applications and as such, there is no joint preparation as the first layer is deposited over the plate surface directly and subsequent layers contribute to the specimen build profile, similar to the WAAM samples. A tungsten inert gas welding torch mounted on a KUKA robot is used to deposit four layers for each weld, with our process using nine passes for the first layer, down to six passes for the last layer. During this procedure, the tungsten artificial reflectors are embedded in the weld, between the existing layers. The sample is then inspected by a 10 MHz ultrasonic phased array in direct contact with the sample surface using both conventional and total focusing method (TFM) imaging techniques. A phased array aperture of 32 elements has been used. The phased array controller is FIToolbox (Diagnostic Sonar, UK). Firstly, a focused B-scan has been performed with a range of settings for the transmit focal depth. Secondly, a full-aperture TFM method has been processed. All the reflectors of interest were detected successfully using this combination of B-scan and TFM imaging approaches.
2019
978-073541832-5
Javadi Y., Vasilev M., Macleod C.N., Pierce S.G., Su R., Mineo C., et al. (2019). Intentional weld defect process: From manufacturing by robotic welding machine to inspection using TFM phased array. In AIP Conference Proceedings. 2 HUNTINGTON QUADRANGLE, STE 1NO1, MELVILLE, NY 11747-4501 USA : American Institute of Physics Inc. [10.1063/1.5099761].
File in questo prodotto:
File Dimensione Formato  
Javadi_etal_QNDE_2018_Intentional_weld_defect_process_from_manufacturing_by_robotic_welding_machine.pdf

accesso aperto

Tipologia: Post-print
Dimensione 1.13 MB
Formato Adobe PDF
1.13 MB Adobe PDF Visualizza/Apri
1.5099761.pdf

accesso aperto

Descrizione: Articolo completo
Tipologia: Versione Editoriale
Dimensione 4.85 MB
Formato Adobe PDF
4.85 MB Adobe PDF Visualizza/Apri

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/10447/425481
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 6
  • ???jsp.display-item.citation.isi??? 8
social impact