Time-based disassembly method: how to assess the best disassembly sequence and time of target components in complex products

Circular economy (CE) is a new business model that is pressing manufacturing companies to think about closed loop scenarios for materials and products. Design for End-of-Life (DfEoL) and Design for Disassembly (DfD) are key enabling methods for the effective application of this model. The paper presents a time-based method for the calculation of disassembly sequences, adopting basic theories and techniques in this topic and integrating new concepts for the assessment of the disassembly time. The method consists of five steps and starts from the documentations (e.g., CAD model) generally available early in the product development process. The first three steps encompass the product analysis by including (i) the definition of target components from the general assembly, (ii) the analyses of the virtual model, and (iii) the assessment of the so-called level matrix, which is based on the concept of disassembly levels and liaisons characterization among components. The last two steps allow for the assessment of the time-based disassembly sequence by including (iv) the analysis of feasible sequences and (v) the generation of the best disassembly sequence for target components. The method mainly overcomes two issues highlighted in the literature regarding the reliability of the disassemblability analysis using a time-based approach and the quality of results accounting for the real condition of the product at the time of disassembly. The calculation of the effective disassembly time is grounded on a specific repository developed to gather knowledge about the disassembly tasks and related disassembly time. This is the main contribution and novelty of the proposed approach. By using the proposed method, different target components of a washing machine are analysed with the aim of demonstrating the robustness of the method and its consistency in the estimation of disassembly time. A maximum deviation of 10% between the estimated and measured disassembly times is noticed.