Cast Shop Technology: Molten Metal Cleanliness and Analysis
Sponsored by: TMS Light Metals Division, TMS: Aluminum Committee
Program Organizers: Halldor Gudmundsson, Century - Nordural; Stephan Broek, Kensington Technology Inc
Monday 2:00 PM
March 20, 2023
Room: 31C
Location: SDCC
Session Chair: Volker Ohm, HOESCH Metallurgie GmbH
2:00 PM Introductory Comments Mr. Halldor Gudmundsson
2:05 PM
Electromagnetic Priming of Filtration Systems: Pyrotek EM-DF: Robert Fritzsch1; Joseph Whitworth1; Paul Bosworth1; Jason Midgley1; 1Pyrotek
The reliable and complete priming of ceramic foam and other filter types is a known challenge to casthouses. Process mistakes can lead to significant loss of filtration performance and generate major losses for a casthouse/foundry by reducing the final melt quality. Various technologies have been implemented to support the priming stage for optimized filtration. Using the patented technology to electromagnetic (EM) prime CFFs has shown significant potential over the last decade. The technology has now been developed into a full-scale industrial filtration unit utilizing multiple CFFs of any grade/ppi, delivering performance for highest purity by increased depth filtration and full volume utilization. The EM fields also showed substantial priming effect when retrofitted to existing CFF systems. These EM priming systems allow enhanced priming and regular, undisturbed gravity filtration during the casting process. This paper presents industrial trials of the pilot units, related to metal cleanliness, ease of operation and throughput.
2:30 PM
Automated Metal Cleanliness Analyzer (AMCA): Digital Image Analysis Phase Differentiation and Benchmarking Against PoDFA-derived Cleanliness Data: Hannes Zedel1; Robert Fritzsch1; Ragnhild Aune2; Shahid Akhtar3; 1Metallurgical Insight and Quality; 2Norwegian University of Science and Technology; 3Norsk Hydro
Assessing metal cleanliness of aluminum melts is critical for product quality control, as well as for process optimization. PoDFA is the current standard method for assessing aluminum cleanliness but has limitations in speed and costs due to its manual image processing. The Automated Metal Cleanliness Analyzer (AMCA) method was previously demonstrated to produce cleanliness indicators highly correlating to the main cleanliness indicator of industrial PoDFA analyses on the same samples. In the present work, the features of the AMCA method were expanded, introducing quantitative inclusion characterization and enhanced detection features. The results were systematically benchmarked against industrial PoDFA-derived cleanliness data. The results confirm the equivalence of the total particle area and provide moderate differentiation of inclusion types. Thereby, AMCA shows potential to be used as an alternative to PoDFA, deriving cleanliness data of aluminum samples for generating extensive process data at superior cost-scaling and minimized human biases.
2:55 PM
Automated Image Analysis of Metallurgical Grade Samples Reinforced with Machine Learning: Anish Nayak1; Hannes Zedel2; Shahid Akhtar3; Robert Fritzsch2; Ragnhild Aune2; 1Norwegian University of Science and Technology (NTNU); Institute of Chemical Technology Mumbai (IndianOil Odisha Campus); 2Norwegian University of Science and Technology (NTNU); 3Norsk Hydro, Karmøy Primary Production
Controlling metal cleanliness in primary and secondary aluminium production is critical for ensuring quality of end product and process optimisation. Solidified aluminium melt samples are typically analysed using established techniques such as Porous Disc Filtration Apparatus (PoDFA). The primary bottleneck of PoDFA analyses, the current standard approach of assessing aluminium quality, is the manual analysis of filter micrographs by metallographers. In the present study, an efficient image analysis platform based on a machine learning algorithm capable of quantifying inclusions in PoDFA filter micrographs is developed and benchmarked. Machine learning models, compared to common image analysis techniques using minimal computational resources, allow for improved performance given versatile datasets. The method is intended to enable superior cost-scaling in aluminium metal cleanliness assessments. Future implementation of these procedures will expand on the quantitative differentiation of relevant inclusion types.
3:20 PM Break
3:35 PM
Characterization of Low- and High Mg-containing Aluminum White Dross Using Deterministic Image Analysis of EPMA Scans: Cathrine Solem1; Hannes Zedel1; Ragnhild Aune1; 1Norwegian University of Science and Technology (NTNU)
White dross is a hazardous waste generated during the primary production of aluminum (Al) and consists of a heterogeneous mixture of different oxides and metallic Al in the form of large flakes, lumps, particles, and dust. Due to the heterogeneity of the dross, sampling is challenging. A sampling tool with step-by-step procedures for its use has been published, with recommendations for how to pulverize and analyze different fractions of the dross using X-Ray Diffraction (XRD). In the present study, Electron Probe MicroAnalysis (EPMA) images of white dross samples were collected from the holding furnace in the Al casthouse during the production of Al alloy 5182 (AlMg4.5Mn0.4) and 6016 (AlSi1.2Mg0.4) have been quantified and cross-referenced to XRD analysis. The obtained results suggest that the image analysis method can quantitatively assess the influence of various process parameters on dross characteristics and contribute to optimizing industrial furnace operations aiming to reduce dross formation.
4:00 PM
Assessment of Separation and Agglomeration Tendency of Non-metallic Inclusions in an Electromagnetically Stirred Aluminum Melt: Cong Li1; Thien Dang2; Mertol Gokelma3; Sebastian Zimmermann4; Jonas Mitterecker4; Bernd Friedrich1; 1IME - Process Metallurgy and Metal Recycling Institute, RWTH Aachen University; 2TRIMET Aluminium SE; 3Izmir Institute of Technology; 4Former Student of IME Process Metallurgy and Metal Recycling, RWTH Aachen University
Presence of non-metallic inclusions (NMIs) reduces surface quality and mechanical properties of aluminum products. The development of good NMIs removal practices relies on the understanding of inclusions behaviors with respect to separation and agglomeration particularly in the turbulent flow. In the scenario of electromagnetically induced recirculated turbulent flow, the concerned behaviors of inclusions with different sizes have rarely been investigated experimentally.In the present study, reference materials were prepared with uniformly distributed NMIs (SiC and MgAl2O4) via an ultrasound-involved casting route. Reference materials were charged into an aluminum melt where turbulent flow was promoted via electromagnetic force. Microscopical analysis shows non-significant agglomeration tendency of SiC, MgAl2O4, and TiB2 inclusion. Time-weight filtration curve, PoDFA, and Spark Spectrometer results suggest a strong dependence of separation rate on particle size. Analytical models were established to estimate collision rate of particles and to compare separation rate difference of different sized particles.
4:25 PM
Microalloying of Liquid Al-Mg Alloy Studied In-situ by Laser-induced Breakdown Spectroscopy: Kristjan Leosson1; Sveinn Gudmundsson1; Arne Ratvik2; Anne Kvithyld2; 1DTE ehf.; 2SINTEF
Laser-induced breakdown spectroscopy (LIBS) provides a way to study aluminum melt dynamics in real time. For the aluminum industry, understanding the time-dependent behavior of alkali and alkaline-earth metals is of major importance. In the present work, we use liquid-phase LIBS analysis to study the behavior of such elements in the melt and their interactions, including the effects of Sr and Ca microalloying in Al-Mg alloys, as previous research has suggested that these elements have an inhibiting effect on Mg oxidation.