Fast single piece identification with a 3D scanning LIBS for aluminium cast and wrought alloys recycling

Abstract

In industrial recycling processes secondary metals need to be separated by material grades before they can be further processed. The identification and separation into different material classes lead to higher value-added industrial feedstock and prevent downgrading processes. Secondary raw materials can be used more efficiently resulting in an increased use of waste products in resource-intensive production processes. Laser-induced breakdown spectroscopy (LIBS) offers a contact free, multi-elemental and fast method for the inline quantitative analysis of single objects in moving particle streams. The LIBS method presented in this paper is based on a 3D object detection combined with a scanning LIBS setup. The optical system consists of a pulsed Nd:YAG laser running at 40 Hz, delivering a 200 mJ double pulse for plasma generation. A high performance three axis galvo-scanner guides the laser beam onto single pieces moving at 3 m s⁻¹ through a measuring volume of 600 × 600 × 100 mm³ with a precision of ±1.5 mm. Twenty channels of a high resolution Paschen–Runge spectrometer are simultaneously processed within a few microseconds enabling multi-elemental analysis of different aluminium (Al) alloys. A dimensionless figure of merit is introduced for the evaluation of the analytical performance. Sorting measurements of Al post-consumer scrap charges, consisting of wrought and cast alloys were carried out. After discarding 20% of the data as outliers low and high silicon alloyed Al pieces were identified with a correctness of >96%. In a second sorting scenario the analytical discrimination of 8 different Al alloys of production scrap, requiring high analytical precision, is investigated. A mean identification correctness of wrought Al alloys >95% is demonstrated successfully for the first time.