Detection of faults in laser beam welds by near-infrared camera observation

Dorsch, Friedhelm; Braun, Holger; Keßler, Steffen; Pfitzner, Dieter; Rominger, Volker

doi:10.2351/1.5062445

Cited by 7 publications

(7 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Fig. 10 [71]- [72] developed a coaxial monitoring system integrating VIS and NIR-camera without auxiliary illumination, and a real-time image processing system analyzes the camera images regarding welding irregularities and delivers information to characterize the weld process and its result. In laser lap welding, Lapido et al [73] presented a novel approach for real-time monitoring evolution of the melt pool under several welding procedures by utilizing uncooled PbSe image sensors in the mid-wavelength infrared range.…”

Section: Co-axial Visual Sensingmentioning

confidence: 99%

Progress and perspectives of in-situ optical monitoring in laser beam welding: Sensing, characterization and modeling

Zhang

et al. 2022

Journal of Manufacturing Processes

View full text Add to dashboard Cite

Section: Co-axial Visual Sensingmentioning

confidence: 99%

Progress and perspectives of in-situ optical monitoring in laser beam welding: Sensing, characterization and modeling

Zhang

et al. 2022

Journal of Manufacturing Processes

View full text Add to dashboard Cite

“…The threshold 𝐶 in (7) for the calculation of geometrical quantities in ( 8)-( 10) sensibly influences results: the bigger the threshold, the smaller the melt pool and vice versa. Accordingly, the estimates of area, length and width of the melt pool are function of C. Setting the threshold is therefore related to the selection of isotherm of the temperature that allows the distinction of the melt pool from the surrounding powder.…”

Section: Calibration Of Process Emission Estimatesmentioning

confidence: 99%

“…estimates from DBSCAN algorithm provides a width which is slightly underestimated; 2. process emission at the melt pool tail (i.e. at the left) has a slow decay [7]. Estimates from process emission images tends to be overestimated because of the emission of already solidified material.…”

Section: A Threshold Calibrationmentioning

confidence: 99%

“…To overcome this issue, additional elements (e.g. neutral density filters) can be inserted in the optical chain to damp process emission in the laser irradiated zone [7]; this would result in higher values of threshold values and lower sensitivity of the HTB algorithm to its variation. It is hypothized that estimation performances would increase especially for width estimation.…”

Section: A Threshold Calibrationmentioning

confidence: 99%

“…• externally illuminated images are taken as reference as in [6][7] but quantitative estimation of geometrical features is performed; • an optimization is run to make estimates coming from externally illuminated images and process emission ones match in average sense.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Estimation of melt pool size by complementary use of external illumination and process emission in coaxial monitoring of selective laser melting

Pacher

Mazzoleni

Caprio

et al. 2019

Journal of Laser Applications

View full text Add to dashboard Cite

Coaxial imaging of melt pool dynamics provides several advantages over other monitoring methods in SLM. The ability to track the processing zone ensures the possibility to observe defect formation dynamics mainly related to melting and solidification. Commonly, the melt pool dynamics are observed by means of process emission. In process emission images, geometrical information of the melt pool are not directly available and their extraction would require the use of a calibrated sensor in order to measure the temperature levels; as a consequence commonly an arbitrary threshold is applied to the image. The use of external illumination for monitoring purposes allows for suppressing the process emission and observing the melt pool geometry by means of the reflected light. On the other hand, the obtained images show lower contrast and can be difficult to process by means of image processing algorithms. Accordingly, this work proposes the complementary use of external illumination and process emission for characterizing the melt pool dynamics in SLM. For the purpose, an open SLM platform with an inhouse designed coaxial monitoring module is used. Images with external illumination were used to estimate the melt pool size for AISI 316L. The information was used to set a threshold value for determining the melt pool size observed at the near infrared emission band. The proposed strategy proved promising for real time monitoring and control applications and can represent a feasible solution for industrial systems.

show abstract