Quality assessment in laser welding: a critical review

Stavridis, John; Papacharalampopoulos, Alexios; Stavropoulos, Panagiotis

doi:10.1007/s00170-017-0461-4

Cited by 182 publications

(90 citation statements)

References 109 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Stavridis et al [7] presented relevant studies regarding quality assessment methods and techniques for inspection on laser welding, namely, image processing, acoustic emissions, x-ray, and eddy current. Furthermore, they mapped the existing modeling approaches used for correlating measured weld characteristics and defects with the process parameters.…”

Section: Introductionmentioning

confidence: 99%

Quality Assessment and Process Management of Welded Joints in Metal Construction—A Review

Péreira

Melo

2020

Metals

View full text Add to dashboard Cite

This paper aims to express the maturity and dynamism of welding in industry. It follows the direct observation of international construction standards and rules, here described as an introduction to the most relevant themes in the area of welding processes. Quality management in welded construction involves well-regulated procedures which are, however, so vast that they are of great complexity. In fact, there are hundreds of rules to apply to each specific case. This makes it imperative to have a welding coordinator, at least in cases of greater responsibility. In fact, the application of the rules presents yet another difficulty: along with the breadth of the rules, each country may adopt certain special requirements in their own case. Thus, it is essential to know the necessary path for the proper development of responsible work. First, one must know the various welding processes, their advantages and limitations; then one must identify the regulations applicable to the work concerned; and finally, one must know what tests and qualifications are required, including the expected defects and acceptance criteria. One purpose of this work is to show the necessary path for the implementation of a welding quality management system in metal construction, as well as to present the numerous applicable norms, both for the welder and the welding coordinator, but also for the process and the product. This paper also presents an analysis of how to obtain the CE marking (Conformité Européene) for a welded metal construction and which normative ramifications apply.

show abstract

Section: Introductionmentioning

confidence: 99%

Quality Assessment and Process Management of Welded Joints in Metal Construction—A Review

Péreira

Melo

2020

Metals

View full text Add to dashboard Cite

show abstract

“…Various sensors are exploited for this purpose. [20][21][22] Radiation intensity detectors like photodiodes are a common choice, 23,24 they offer high temporal resolution and with appropriate filters can detect light at different wavelengths, 25 such as visible light, laser reflected radiation, or infrared. 26 Despite several efforts in the literature, the assessment of quality attributes along with the use of a noninvasive monitoring device able to detect such changes remains an open question.…”

Section: Introductionmentioning

confidence: 99%

Process development and coaxial sensing in fiber laser welding of 5754 Al-alloy

Garavaglia

Demir

Zarini³

et al. 2019

Journal of Laser Applications

View full text Add to dashboard Cite

The use of Al-alloys is increasing in the automotive industry due to the pressing necessity to reduce weight and fuel consumption. Several parts concerning the car body are assembled through welding, where a high-quality seam is a key requirement. For this purpose, laser welding stands as an appealing option. On the other hand, laser welding of Al-alloys is a complex process due to the high reflectivity, reactivity, and crack susceptibility of these materials. In many cases, such issues limit the applicability of the autogeneous welding, which is an advantageous feature of laser welding. High-brilliance fiber lasers have been an enabling technology for improving the weldability of Al-alloys. However, laser welding of Al-alloys, especially in a lap-joint configuration, requires robust processing conditions able to maintain seam quality for each weld in high volumes even with part tolerances and tooling variability. Accordingly, this work discusses the process development and monitoring in laser welding of 5754 Al-alloy. In particular, the process was carried out in a double lap-joint configuration with 1 mm sheets, commonly used in automotive applications. A 3 kW fiber laser with in-source integrated monitoring capability was employed as the light source. The process feasibility zone was investigated as a function of laser power and welding speed, while the effect of focal position was investigated for the weld robustness. Weld seam types and defects were identified, as well as the monitoring signals associated light back-reflected from the process.

show abstract

“…and laser parameters (e.g., beam size, power, scan rate, etc.). [39] Machine learning offers a route to convert sensor data into realtime assessments; however, this requires a wealth of labeled sensor data that traditionally is too time consuming and/or expen sive to assemble. [8] Furthermore, many types of LPBF defects arise due to inherent www.advmattechnol.de process monitoring and rectification impossible.…”

Section: Introductionmentioning

confidence: 99%

“…[8] Furthermore, many types of LPBF defects arise due to inherent process monitoring and rectification impossible. [39] Machine learning offers a route to convert sensor data into realtime assessments; however, this requires a wealth of labeled sensor data that traditionally is too time consuming and/or expen sive to assemble. In this manuscript, this critical issue of generating labeled video data for machine learning is solved.…”

Section: Introductionmentioning

confidence: 99%

Machine‐Learning‐Based Monitoring of Laser Powder Bed Fusion

Yuan

Guss

Wilson

et al. 2018

Adv Materials Technologies

View full text Add to dashboard Cite

variability in the powder properties, [9] bed thickness nonuniformity, [10] and laser parameters and scan paths that result in improper power melting. [11] Thus, even after optimizing LPBF operating para meters and identifying suitable processing windows, [12] rapid build qualification, improved quality, and higher production yields require methods of monitoring the melt pool and/or powder bed in situ, i.e., during a build, that enable realtime process feedback and automated quality detection. [13,14] The majority of LPBF process moni toring approaches rely on noncontact sensing [15] from optical, thermal, [16] and/ or acoustic [17,18] sensors. These sen sors provide assessments of spatial and spectral features of the melt pool, [19,20] process plume, [21] degree of spatter, [22][23][24][25] overhang layers, [26] or print bed. High speed image sequences of the melting process, [27] scans of the powder bed, [10,28] beam quality, [29] and/or thermal monitoring [30] are all routinely collected forms of in situ monitoring data. Making use of this data requires methods that can extract rel evant diagnostic information. For instance, before initiating laser melting, automated computer vision algorithms can characterize metal powder feedstocks, [31] and image analysis of newly spread powder can reveal nonuniformities in the powder bed thickness. [32] Aminzadeh et al. demonstrated layerbylayer detection of fusion defects from images using a Bayesian clas sifier. [33] Realtime events such as material ejecta are detectable by applying manually set thresholds to highspeed nearinfrared images of the melt pool. Also, increasing the µm pixel −1 image resolution relative to the standard deviation of measured track width, σ measured , may result in improved predictions of the final track width, δ predicted , and topography. [34] Reducing laser power proportionally to an integrated signal from a photodiode cali brated against a camera results in smoother overhang struc tures. [35] Using images of the print bed taken after laser melting, a level sets method can detect intentionally created defects, [36] machine vision algorithms can identify pore defects, [37] and multifractal image analysis can characterize layers with balling, cracks and pores, and no defects. [38] Visual imaging equipment is appealing to LPBF monitoring systems because it is relatively inexpensive and provides noncontact sensing. [13] As with most additive manufacturing systems, analysis of LPBF sensor data currently occurs postbuild, rendering A two-step machine learning approach to monitoring laser powder bed fusion (LPBF) additive manufacturing is demonstrated that enables on-the-fly assessments of laser track welds. First, in situ video melt pool data acquired during LPBF is labeled according to the (1) average and (2) standard deviation of individual track width and also (3) whether or not the track is continuous, measured postbuild through an ex situ height map analysis algorithm. This procedure generates three ground truth labeled datasets for supervised...

show abstract

Quality assessment in laser welding: a critical review

Cited by 182 publications

References 109 publications

Quality Assessment and Process Management of Welded Joints in Metal Construction—A Review

Quality Assessment and Process Management of Welded Joints in Metal Construction—A Review

Process development and coaxial sensing in fiber laser welding of 5754 Al-alloy

Machine‐Learning‐Based Monitoring of Laser Powder Bed Fusion

Contact Info

Product

Resources

About