The final project report from the feasibility study led by Dr Martin Baumers at the University of Nottingham is now online. “The economics of 3D Printing: A total cost perspective” describes results of the work undertaken between the University of Nottingham, the University of Oxford, and Digits2Widgets.
Martin Baumers, Matthias Holweg and Jonathan Rowley
AM processes are generally associated with two advantages over conventional manufacturing techniques. Firstly, they avoid many of the tooling-related constraints on the geometries that can be achieved through conventional manufacturing processes. Secondly, AM allows the efficient creation of products in very low volumes, down to a single unit, enabling the manufacture of customised or highly differentiated products.
The technological opportunities that AM presents are not in question. We do however still lack a fundamental understanding of the economics that underpin the application of this technology, which is a fundamental precursor to developing a business case for its application. In this report we present the findings of a project aiming to develop our understanding of the underlying economics
It is frequently claimed that the generic advantages associated with AM will lead to flourishing supply chain innovation challenging the existing paradigm of centralised mass manufacturing. However, the successful and meaningful adoption of AM will depend on a favourable business case of which, at present, key aspects are not
fully measured and understood. This underlying research addresses the identified gap.
As a central element for making the business case towards AM adoption, existing costing approaches have largely focused on capital investments and consumables, with an emphasis on build materials. Analyses of such “well-structured costs” have observed that utilising the available machine capacity forms a prerequisite for efficient operations. This is also a core principle of traditional manufacturing, which is directed at achieving economies of scale and, as a result, has led to the formation of global supply chains in many industries.
This stands in contrast to AM, where the underlying reason for the different requirements towards full utilisation is that the technology is inherently parallel, allowing the contemporaneous deposition of multiple geometries. Moreover, existing analyses of AM resource consumption have largely ignored hidden or so-called “ill-structured” costs relating to build failure, part rejection and ancillary manual processes, such as support removal and surface finishing. This omission has come at the expense of industrial applicability, also leading to a lack of realistic decision tools for the support of AM technology adoption which are an essential prerequisite for successful diffusion.
Over the duration of this project, we set out to develop new methodologies and conducted a series of experiments to build up a body of data supporting a realistic and comprehensive costing model. Overall, 20 build experiments were carried out on state-of-the-art polymeric Laser Sintering (LS) and metallic Selective Laser Melting (SLM) platforms.
As polymeric LS constitutes one of the most commonly adopted technologies for the additive manufacture of end-use components and is capable of delivering useful material properties, the project has concentrated on LS in its experimental work. The key methodologies employed in the analysis of LS, together with reached results, are presented in this report.
We have identified three aspects that have proven to be of special significance for the formulation of the total cost perspective for AM:
- it is known that the unit cost achievable with LS is dependent on the degree of build volume utilisation. This relationship underlies the approach taken in this project;
- AM processes do not operate in isolation. They are embedded in a sequence of ancillary process steps that can, as the project has identified, be adequately captured through process mapping;
- at the current state of technology, AM processes are prone to build failure events of various sorts, which all have a detrimental effect on cost and thus need to be incorporated in any costing model.
We further demonstrate that there is a relationship between the quantity of parts included within a build volume and the resulting unit cost. We show that sub-normal machine utilisation leads to higher unit cost, as one would expect. We also show that once the process operates at technical efficiency (optimal build space utilisation) there are no cost benefits from repeating the build process.
Based on the experimental results we develop a total cost model that accommodates both manual process inputs and interventions as well as the risk of build failure. The methodology developed within this project thus provides the basis upon which any economic case for AM, associated network effects, and potentially redistributed manufacturing can be built.