Since AM constitutes a parallel manufacturing technology allowing the contemporaneous processing of multiple components, any fixed costs incurred can be amortised across multiple components. This implies that there is potentially a strong efficiency-related incentive to maximise the utilisation of the available build volume capacity. However, when considering cost models of the kind developed in this chapter, it is important to realise that capacity utilisation enters in a second way, through the degree of machine utilisation over time, as part of the indirect cost rate, which relies on an estimation of the share of operating hours of overall time. Metrics such as overall equipment effectiveness are useful for the measurement of such aspects of capacity utilisation (see, for example, Bicheno and Holweg 2016).
Further complexity is introduced to the costing framework if the inspection process exhibits capacity or throughput bottlenecks, which is likely to be the case in precision AM if the inspection process is overly time consuming (see Chapter 13 for state-of-the-art inspection times). However, it is assumed that assessing capacity utilisation for inspection processes will carry less complexity than for AM processes due to their serial nature, which obviates the build volume utilisation problem.
Integration with Other Operational Processes
An initial overview of the scope of the cost modelling project can be obtained through a process mapping exercise, as done for example by Baumers and Holweg (2019). Importantly, such an exercise will help define the appropriate boundaries of the cost investigation. As described in the generic AM process shown in Figure 7.1, a number of pre- and post-pro- cessing steps are typically included in a useful AM cost model. In the model presented in Section 7.4, these costs are reflected purely as labour costs (PLC, FLC and VLC). Depending on the nature and characteristics of these secondary processes, significant additional expenses may be incurred which, in turn, warrant inclusion of specific extra items in the cost model.
If the AM process and the inspection process, or other ancillary processes for that matter, are discrete, workflow organisation and scheduling techniques should be used to determine efficient patterns of operation. Well-known scheduling environments include single machine, identical machines in parallel, machines in parallel with different speeds, unrelated machines in parallel, flow shop, flexible flow shop, job shop, flexible job shop and open shop. An exhaustive theoretical treatment of scheduling theories is provided by Pinedo (2012).
A further aspect that has received considerable attention over the past decade is the requirement to adapt designs to the characteristics of the AM process. In particular, the emergence of AM has led to a wave of research on computational design tools capable of exploiting the available design space, thereby leading to products with optimised geometries (for an overview, see Liu et al. 2018 and Chapter 2). In this context, the general reasoning is that optimised product configurations need to take into account all the elements of the product life to avoid island solutions that may sacrifice efficiency. This is likely to be of high relevance for precision AM, where stringent requirements are placed on product features and significant additional expenses may be incurred for inspection.
Relationship between Failure Parameters and Costs of Inspection
The total unit cost model TUC presented in Section 7.4 proposes relationships between overall unit cost and a number of cost parameters, including the instantaneous constant failure rate X, the probability of product rejection PReject and the cost of inspection CInspect. In the case of a closed-loop system, it would be expected that the occurrence of process variation - detected during the AM process itself or during inspection - leads to an automatic correction mechanism. However, this is not currently the case in most applications of AM. Nevertheless, since a process implementation in the commercial manufacturing context demands an acceptable level of product quality and conformance to design parameters (which are of course highly application-specific), it is reasonable to expect a negative relationship between Cjnspe<;t and the failure and rejection parameters in the model, X and PRe)ect.
As illustrated in Figure 7.4, it is possible to posit a mechanism in which a higher expenditure CInspect leads to better or increased feedback of failure and rejection information to the design and production planning activities, for example in the form of an information loop, perhaps as a product of a failure modes and effect analysis (see, for example, Stamatis 2003). The resulting reduction of X and PReject is expected, in turn, to lead to a lower TUC. This suggests a mechanism in which the cost of additional inspection activities is offset over time by decreases in manufacturing cost.
Total cost effect of increases in inspection effort.