environmental studies on Additive Manufacturing I

There is evident need for more research on the environmental impacts of fabbing and making, especially regarding such new processes and materials. Studies are beginning to trickle out, however; each study adds a new piece to an admittedly large and complex puzzle.

I’ll list some of these studies here and hopefully you can access them. (Many of them are in journals or conference proceedings hidden behind a paywall.) These address various aspects in Additive Manufacturing technologies.

It’s important to note that these are usually technologies used in Rapid Prototyping and NOT the kind of stuff we see in a Fab Lab: i.e. desktop, open source Fused Deposit Modelling equipment. And not necessarily finished products or B2C markets, hence the name Rapid Prototyping.

Still, what happens in AM will be relevant to both the development of the personal desktop versions as well as to mass customizers, who are increasingly interested in AM/RM technologies – not to mention the platforms like Shapeways and i.materialise, who already use these systems to make finished products for end-customers and not prototypes.

Drizo, A. & Pegna, J., 2006. Environmental impacts of rapid prototyping: an overview of research to date. Rapid Prototyping Journal, 12(2), pp.64–71.

According to Drizo and Pegna rapid prototyping (RP) or rapid manufacturing (RM) is one of the fastest growing manufacturing technologies but represents less than ten per cent of the total manufacturing sector. Despite a pressing need for more knowledge on the sustainability issues in this area and more study on how to perform environmental impact assessment (EIA), “it is not surprising that the demand for conducting EIA has not received much attention to date”. They present an overview on impact assessment methods, urgent unresolved issues, and the limited coverage of the topic in the literature. One EIA method for instance divided processes into life stages similar to the approach in LCA (life cycle analysis) but geared to RP processes. Assessments using the method allowed identification of key parameters that influenced the technology’s environmental performance.

Three areas that Drizo and Pegna especially emphasize are health and safety, waste, and energy. The authors specifically highlight and write at length about the health and environmental risks due to the toxicity of RP materials and solvents that have not yet been identified. Connected to this are the impacts associated with disposal, as the producers of some of these materials only recommend “incineration and landfilling”. Claims are also made that additive manufacturing can minimize waste, but at the time of writing few studies had been conducted. (Even in 2013 this knowledge gap still exists.) Moreover, there may be issues with the materials after processing that negatively affect their inclusion in closed loop processes. The authors close by offering areas for future research: materials (their toxicity, development of more sustainable materials, etc.), energy consumption, and more comparisons between conventional manufacturing and rapid manufacturing (also in mass customization) to validate claims of less waste and fewer environmental impacts.

Franco, A., Lanzetta, M. & Romoli, L., 2010. Experimental analysis of selective laser sintering of polyamide powders: an energy perspective. Journal of Cleaner Production, 18, pp.1722–1730.

Franco, Lanzetta and Romoli examine energy consumption in Selective Laser Sintering (SLS) through experiment and theoretical modelling, building also on findings and models in other studies not listed in this post (see their references section). They also lament the lack of research done on environmental performance of “laser assisted manufacturing processes relative to the traditional manufacturing methods”.

Since SLS offers the advantage of fabricating very complex parts without needing moulds (in fact, some of the things we’re seeing produced today simply *cannot* be produced with conventional means), these authors emphasize that part accuracy must be an important parameter – and that very few studies have considered the connection between part accuracy and energy-related data. They propose an energy density range (of 0.06-0.08 J/mm2 for the polymer tested) at which “the SLS parts can connect the objective of good dimensional accuracy, good volumetric productivity and reduced energy intensity”. Too high an energy density affects the process’s productivity as it doesn’t actually contribute to the sintering process. One implication of this, according to the authors, is the potential to eliminate the pre-heating phase: important, since most of the energy is consumed by the chamber heaters.

Mognol, P., Lepicart, D. & Perry, N., 2006. Rapid prototyping: energy and environment in the spotlight. Rapid Prototyping Journal, 12(1), pp.26–34.

Mognol, Lepicart and Perry compare three rapid prototyping systems (two 3-D Printing – 3DP – and one SLS system) with the aim to identify the parameters that affect electrical energy consumption, concluding that manufacturing time is the most important. They tested printing of the same part in the three pieces of equipment and in a variety of positions. In two of the machines, reducing the manufacturing time (and thereby energy consumption) was best effected by minimizing the height of the part. The third machine actually has an optimization feature built into it: “the computer calculates the longest diagonal of the part and begins the manufacture at this straight line”, a design that allows faster fabrication. Increased electricity consumption comes instead with a build that requires support, and the authors therefore recommend minimizing the volume of support with this system as the parameter that matters.

By optimizing these parameters as described in the article, electricity consumption was reduced from 43 to 61 per cent depending on the system. What is interesting is the implicit finding that the authors don’t point out as a future area of research and/or development: this idea that environmental impacts can be designed in or OUT of a technology. Besides the optimization of fabrication time mentioned above (i.e. the computer finding the longest diagonal), the authors also point out how a different system ‘waits’ between layers (as it prepares the following layer with the ‘scraper’) so that the manufacturing time is actually made up of “a long waiting time and a short laser sintering time”. The long printing times in AM are quite notorious, so it can’t be too long before more AM equipment manufacturers begin to try to optimize this element: longer printing times usually = money as well as = electricity consumption. This is an area I don’t know well – what the commercial systems do and don’t do at the moment. But one can imagine that there are all kinds of other benefits for producers that have environmental benefits embedded in them, such as material saving and re-use (since RP materials are also more expensive than conventional manufacturing raw materials). See the next article on that subject.

Dotchev, K. & Yusoff, W., 2009. Recycling of polyamide 12 based powders in the laser sintering process. Rapid Prototyping Journal, 15(3), pp.192–203.

Dotchev and Yusoff study the deterioration of polyamide 12 (PA12) powder properties in the laser sintering process (LS) in relation to the temperature and the time during which the material is exposed. They propose a methodology for more efficient powder recycling that also allows more control over fabrication quality and avoids the known ‘orange peel’ texture that occurs if the powder has been recycled many times and/or doesn’t contain a certain amount of new material. What happens currently, according to the authors, is that producers tend to use a high proportion of new material just to avoid the poor ‘orange peel’ surface – without taking into account the used powder’s “thermal history” and therefore scrapping a lot of used powder – in accordance with suppliers’ recommendations. This is both a cost for the producer, since the material costs are high, especially in comparison with conventional injection moulding, and a potential environmental issue “in case of future RM [rapid manufacturing] expansion”. This methodology therefore aims to clarify the relationship between powder degradation and its thermal history by “powder management” – based on measuring the melt flow rate (MFR) of molten PA12 powder.

Telenko, C. & Seepersad, C.C., 2012. A comparison of the energy efficiency of selective laser sintering and injection molding of nylon parts. Rapid Prototyping Journal, 18(1), pp.472–481.

The aim of Telenko and Seepersad’s study is to determine the ‘crossover’ production volume at which it makes sense to produce a part using selective laser sintering (SLS) rather than conventional injection moulding (IM). As the authors point out, there have been studies on this issue regarding financial cost, but not studies according to energy consumption or life cycle inventory (LCI) estimates. In other words there are still hypotheses bouncing around that additive manufacturing (AM) is more ‘sustainable’ than conventional manufacturing but empirical evidence is lacking – and the authors cite Drizo and Pegna (described above) who said the same thing way back in 2006. This study attempts to fill this gap regarding the energy efficiency of SLS compared to IM – and, as the title suggests – using a nylon part as point of comparison.

On the one hand SLS does consume a significant amount of energy. On the other, IM needs a fabricated mould and the accompanying material and energy investments, which means that the producers want to produce [and then sell] a huge amount of the parts/goods to be able to off-set the cost of the mould. SLS allows small production batches at the same cost per piece and in fact – the benefit of AM – allows customization of each piece or each batch to an extent that IM can never reach.

The authors conclude that production with SLS is more energy efficient than IM only with very small production volumes. In this particular study, the crossover production volume – where SLS and IM consumed the same amount of energy – was 150 to 300 parts. The authors also give three recommendations for reducing the energy consumption of SLS (related to optimizing the build height, managing the powder use and recycling, and reducing the time needed for scan and preparation).

Marchelli, G. et al., 2011. The guide to glass 3D printing: developments, methods, diagnostics and results. Rapid Prototyping Journal, 17(3), pp.187–194.

Marchelli, Prabhakar, Storti and Ganter report on their experiments with Three-Dimensional Printing (3DP) and glass. Both SLS and 3DP can use metal and ceramic as materials, but glass has not been utilized in AM. Moreover the authors wish to enable the use of recycled glass as a more sustainable route and include testing of recycled glass powder. In fact, the authors report that as a direct result of their research (at the University of Washington) “industry professionals have exploited recycled glass as a printing medium with Shapeways” and another company (EnVitrum) uses 3DP recycled glass prototypes for green building solutions.

The article summarizes the data and findings for 3DP recycled glass (“shrinkage, apparent porosity, bulk density as functions of peak firing temperature”) and presents information on “how to create the powder binder, which particle sizes are best suited for 3DP, and how to diagnose and optimize printing saturation”. Saturation settings can also be found at Open3D Printing open3dp.me.washington.edu/. (This URL was given in the article and is the lab’s blog. It’s an interesting blog to follow but I haven’t yet found the data they mentioned. Here is one entry that might help: open3dp.me.washington.edu/2009/09/3dp-glass-recipe/.)


Leave a Reply

Your email address will not be published. Required fields are marked *