Environmental Performance Evaluation of a Fast Mask Image Projection Stereolithography Process Through Time and Energy Modeling

Businesses and consumers are looking for sustainable technologies as they make purchasing decisions today. A sustainable technology can be thought of as one that operates in a manner that is economically viable, environmentally benign, and beneficial to society. In recent years, many advantages of AM (also known as 3D printing, solid freeform fabrication, and rapid manufacturing) over traditional subtractive and formative manufacturing processes have been discussed at length. Subtractive manufacturing processes remove material from billets or stock material, and formative processes require specialized materials, labor, and manufacturing techniques to deform and shape material [1,2]. Additive processes assemble material layer-by-layer from digital inputs of computer-aided design (CAD) models to produce final or net-shape parts [3]. The ability to produce customizable and functional parts on demand, the elimination of tooling, and the expansion of the product design space portend mainstream AM to aid in economic and social development [4]. Progress made through research has enabled the growth of new and innovative techniques and functionally viable products, framing layer-by-layer manufacturing processes as feasible alternatives to subtractive and formative techniques [5].

Given that AM enables the production of geometrically complex parts from a wide range of materials, a tremendous advantage over traditional processes can be found in material utilization, which is nearly one-to-one [3]. In fact, many benefits of sustainability can be realized through AM due to the optimization of part design, which can lead to high-performance functional parts with minimal mass [6,7]. The fabrication of microfluidic heat exchangers and other microdevices using AM can significantly reduce material use, while also increasing the functional productivity of the products [8]. However, in recent years, several studies have been conducted on the environmental impacts of AM, and their findings have been mixed [4,5,9]. While the advantages provided by a reduction in material consumption, tooling, and harmful chemicals used in machining are well known, the benefits have been tempered by findings that additive processes tend to be energy inefficient and contain hidden wastes [8]. In reality, more efforts are required to fully understand the breadth of sustainability factors and improve the efficiency of additive techniques [5,10–12].

The investigation herein reviews the complementary roles of sustainability and AM. The synergy of sustainability and AM, the role of each in design and their benefits for society, various indicators and factors (e.g., energy consumption), and models for sustainability assessment of AM are considered. This review is followed by an environmental assessment of a stereolithography (SL) process, called fast MIP-SL, for production of parts in a more efficient manner than traditional SL processes. The study reviews the motivations and methods, presents an EIA for the production of several products, and discusses the results of the assessment. Furthermore, modeling of process energy consumed based on design parameters (shape complexity and dimensional size), part build time, and washing time within the process is presented. The work reflects the interrelationship between process energy consumed and environmental impact, and the model developed is validated using data from real-time process energy monitoring. Finally, challenges and future work toward sustainable AM are discussed.

Sustainability is a varied conception in today's world and has been largely motivated as a result of a series of environmental incidents and disasters, as well as fears from chemical contamination and resource depletion [1,13,14]. According to the United Nations Brundtland Commission Report [15], sustainable development is “development that meets the needs of the present without compromising the ability of the future generations to meet their needs.” It has been suggested that sustainable development is a function of three major dimensions, namely, economic, social, and environmental [16]. Further, sustainable manufacturing can be interpreted in the engineering context [17], which requires the “design of human and industrial systems to ensure that humankind's use of natural resources and cycles does not lead to diminished quality of life due to either losses in future economic opportunities or to adverse impacts on social conditions, human health, and the environment.” Considering manufacturing systems as a business function, the U.S. Department of Commerce [18] defined sustainable manufacturing as “the creation of manufactured products using processes that are non-polluting, conserve energy and natural resources, and are economically sound and safe for employees, communities, and consumers.” While these and other definitions have been proposed for sustainable manufacturing (e.g., by Haapala et al. [1] and Sutherland and Gunter [14]), they espouse the fundamental tenets of economic, environmental, and social responsibility, as mentioned above for sustainable development.

Sustainability as a systems approach requires a balance between resource consumption and waste generation at a rate at which the environment can assimilate and reproduce nutrients and resources [19]. For a manufacturing system to have continuous development and also constitute sustainability, it should be considered a system where material and energy flows are in a closed loop [1]. Thus, engineering researchers have a duty to provide advancements in manufacturing processes, equipment, and systems, and to reduce material consumption, energy use, waste production, and environmental impacts with a focus on simultaneous product and process design. Herein, this perspective is presented through the case of a new AM approach by reviewing its sustainability performance from cradle to gate (i.e., material extraction, material processing, and part production), in terms of environmental performance and energy consumption.

As defined by ASTM International [20], AM is a process of making objects from three-dimensional solid model data by joining materials, usually in a layer-by-layer fashion. While the most popular applications in AM still involve rapid prototyping for testing the form, fit, and function of a design, the technology is growing as a reliable method to design and manufacture functional products of value [5,21]. A key aspect of AM and its future success is the ability of the technology to quickly produce parts at high volumes and produce components customized to application- or customer-specific needs. The layer-based process allows for the design of almost any geometry, a drastic expansion of the previously constrained design space.

As introduced above, the benefits that make AM advantageous compared to traditional subtractive and formative processes are compatible with the principles of environmental responsibility, economic growth, and social prosperity. These benefits include elimination of tooling, the ability to manufacture complex geometries, optimized product design, increase product functionality, and the selective placement of material only where necessary, which contribute to a reduction in waste and an increase in process efficiency [6]. It has been shown that the ability to update, repair, and remanufacture tooling presents opportunities for significant reductions in energy consumption, emissions, and costs [2]. The optimal design of products can be exploited to increase product performance and add value through embedded functionality. Furthermore, benefits to the supply chain can be realized through the displacement of inefficient and detrimental production processes, improvement of supply chain flexibility, elimination of work-in-process and stock obsolescence, compression of the supply chain, manufacturing closer to the distribution location, and implementation of on-demand (just-in-time) manufacturing [6,22]. AM, therefore, has the potential to directly and indirectly impact the life cycle of products by increasing affordability, longevity, and likability of products and to reduce the burden placed on the environment by manufacturing processes [4,23,24].

One of the technical barriers to the adoption of AM for finished product production is incomplete integration of homogenous design with heterogeneous CAD and closed-loop AM [5,7]. This integration would promote social and environmental responsibility by enhancing the desirability of designed products that are unlimited by the traditional materials selection and geometric definition approach. Thus, sustainable design and manufacturing principles are applied herein to elucidate the environmental performance of the newly developed fast MIP-SL process through a life-cycle impact assessment (LCIA) approach.

Diegel et al. [25] described sustainable design as “design which aims to achieve triple-bottom line ideals by striving to produce products that minimize their detriment to the environment while, at the same time, achieving acceptable economic benefits to the company and, wherever possible, having a positive impact on society.” As highlighted above, AM presents numerous opportunities that have the potential to benefit the environmentally responsible design of products. Many products are not manufactured to their optimal geometry and utilize extraneous materials required by casting/molding, forming, and machining processes [2]. Material use can be reduced while maintaining functional performance through the optimization of product design to lattice or honeycomb structures—which are only possible through AM processes [5,6]. Zhang et al. [26] proposed a design optimization strategy for variable density hexagonal cellular structures. The research related cellular structures with continuous micromechanics models and achieves efficiency through continuous topology optimization strategy. In addition to reduced material use, these lighter weight products can reduce carbon emissions over a product's life cycle by reducing the energy needed to transport and convey the product. Thus, the optimal design of products, which is usually constrained by traditional manufacturing techniques, can be exploited to increase product performance and add value through embedded functionality.

Research in sustainable product design has usually focused on lowering the environmental impacts of material, resource, and energy use, while it often ignores understanding “design quality” as a method to maximize product longevity [24,25,27]. The tradeoffs between optimal design and manufacturability of a product, and the tradeoffs between custom fit and characteristics of large-scale enterprise focused on process and cost efficiency can negatively impact design quality and, consequently, longevity. Diegel et al. argued that AM has the potential to address both of these factors and is, therefore, an effective tool to enable sustainable product design [25]. Since EIA requires broad, comprehensive analysis, the environmental performance of various designs is explored in Sec. 2.3 by examining a set of products with different geometries and shape complexities, including simple 2.5-dimensional, shell, and multiscale structures with fine features.

Despite the demonstrated success of AM, EIA is a challenging but necessary task [28]. In particular, the joint efforts of design and manufacturing engineers with environmental scientists are essential in understanding the fundamental environmental impacts of newer technologies and materials. Evaluation of the extent of the impact of these processes and their resulting products is required to define regulations and the spatial distribution that will enable the control and prevention of the potential harm, along with estimates of the cost required to deal with related issues.

EIA of AM is limited in literature. One study, conducted by Luo et al. [11], proposed a method for evaluating the environmental performance by dividing the AM process into individual elements (i.e., material preparation, energy consumption, material toxicity, and waste disposal). Each element was considered within the various life stages to cope with process complexity [11]. Their method was demonstrated for an SL process with focus on the energy consumption rate (ECR), which is the amount of energy consumed per unit mass of material used. The environmental impact was calculated as the ECR multiplied by an electricity consumption factor (0.57 mPts/kWh), and comparisons were made for three different equipment models used for the SL process [29]. Even though the underpinning method is life-cycle assessment (LCA), analysis fails to account for potential toxicological health and environmental impacts that can occur from handling, use, and disposal of photo resin materials. This is simply because the toxicity and environmental impacts of many AM materials and chemical solvents used for their removal have not been identified to date [28]. The available safety data sheets are limited to older generation, epoxy resin materials. The majority of these data recognize that severe eye and skin irritation and possible allergic skin reactions might occur as a result of handling or inhaling vapor from those materials [28].

In addition to photopolymer resins, information about the chemical solvents used for the removal of excess resin on the products is unknown. Data about environmental mobility (air, water, and soil) are unavailable [30]. Similarly, data on human toxicity are limited [28]. Another important issue that needs to be addressed relates to the impacts and consequences of energy consumption of these processes, as discussed in Sec. 2.3.1.

Energy generation and industrial activity contribute significantly to the overall emission of greenhouse gases, which are thought to be the key driver of global warming [12]. The reduction of energy consumption in manufacturing is vital to limiting the overall emission of greenhouse gases. AM allows the production of multiple components in a parallel manner, without the need for tooling [31,32]. The single-step nature of AM provides transparency to the energy utilized in the process. However, AM has several drawbacks in terms of product quality, processing speed, and cost [28,31]. From the perspective of energy consumption, AM processes are usually not as efficient as conventional manufacturing processes [33]. Equipment often has peripheral devices; thus, basic power consumption and processing time are the two main considerations in energy consumption calculations [28].

AM processes involve the construction of a part that may consist of thousands of layers, and each may take several minutes to complete. Thus, production may require significantly more time than conventional manufacturing processes. AM requires a significant amount of energy; thus, energy consumed per volume of material is high [34]. The Advanced Manufacturing Office of the U.S. Department of Energy, however, has suggested that AM saves energy by eliminating distributed manufacturing processes and material waste [35]. Since energy savings are product-specific and vary extensively, it is not possible to map the energy utilization of the entire AM sector, and this conclusion cannot be generalized [36].

Previous studies suggested that there may be an increase in ECR as productivity increases [11,12,37,38]. In SL processes, this effect is related to the solidification rate of the raw materials [19]. It is seen that AM provides an advantage for large build volumes, as well as faster build rates for products having a small number of parts [39]. However, AM processes are inefficient compared to conventional processes, and therefore, the optimization of energy consumption in AM is essential in order to reduce the environmental impact [8]. Translating process parameter behavior into a process model can reduce variability in the process and help control energy consumption. To achieve that goal, mapping the interrelationships of process energy consumption with environmental performance metrics, e.g., through energy modeling, is required. As described in Section 2.3.2, environmental modeling and impact assessment can provide an indication of the broader ramifications of materials and energy use.

Until recently, little effort has been invested in the development of environmental models representing the life cycle of additively manufactured products. In the 2009 Roadmap for Additive Manufacturing, Bourell et al. [5] stated that achieving important AM sustainability goals will require a total life-cycle analysis and a comprehensive sustainability evaluation of each AM process. This includes analysis of four life-cycle stages: premanufacturing, manufacturing, use, and postuse. It is also imperative to ensure the development of design for sustainable additive manufacturing (DFSAM) [10]. As defined by Rosen [40], DFAM is the “synthesis of shapes, sizes, geometric mesostructures, and material compositions and microstructures to best utilize manufacturing process capabilities to achieve desired performance and other life-cycle objectives.” DFSAM was developed in the context of sustainability [10]. It includes both DFAM and EIA in the development of products and processes. This merging is critical, as innovative product design and manufacturing activities in the coming century will require the integration of life-cycle data and sustainable design principles for the improvement of products and processes [5]. This integration will be enabled through material, process, and system modeling approaches.

In the past decade, models have begun to emerge for assessment, prediction, and optimization of environmental impacts and efficiency of AM processes. For instance, Le Bourhis et al. [10] proposed a new LCA-based methodology to evaluate the environmental impact of a part from its CAD model for a direct metal deposition process. The process model is based on electricity, fluid, and material consumption, unlike previous energy-only assessments [4,11,34,38]. Through predictive modeling of process inputs, the work aims to minimize consumption of all material and energy fluxes during manufacturing by integrating the model into design activities [41]. The process model enables environmental evaluation of different manufacturing strategies for the same part, based on the CAD model.

Verma and Rai [42] proposed another modeling approach based on multistep optimization to enable material and energy efficiency of AM technology. The approach aims to minimize material waste and energy consumption for finished parts as well as on a layer-by-layer basis. A process model was formulated for selective laser sintering, and it was claimed to be easily extendable to other AM processes. The proposed approach is generic and does not appear to require part geometry data, e.g., complexity, curvature, or feature identification. Model development and experimental analysis demonstrated the ability of the proposed optimization techniques to define manufacturing process plans and compete with current layer-by-layer slicing approaches.

Faludi et al. [9] used LCA to conduct a comprehensive comparison of subtractive and AM processes exploring the types of ecological impacts and their sources for the two manufacturing approaches. The study compared fused deposition modeling (FDM), inkjet printing, and computer numerically controlled (CNC) machining of polymeric materials. In order to conduct a fair comparison, part production was modeled on a part-per-year basis, which was then followed by a calculation of ecological impacts. The comprehensive cradle-to-grave study found that it cannot be unconditionally stated that AM technology has an environmental impact advantage over subtractive processes, specifically in terms of material waste or energy consumption. The relative impact of AM processes depends primarily on machine utilization; therefore, the best strategy for enhancing environmental performance is to have the fewest number of machines, each running the most jobs possible [9].

In spite of these efforts, a more comprehensive understanding has not emerged to assist designers from a DFSAM perspective during early phases of design. In particular, design decision making greatly influences the AM product cost, performance, and environmental impact [7,43]. Hence, fundamental relationships that relate design attributes and AM process parameters to product environmental performance would greatly facilitate sustainable design. Sec. 3 explores this direction further by focusing on the environmental performance of a newly developed process, termed fast MIP-SL.

As iterated above, the rapid pace of advancements in AM technologies necessitates EIA to ensure their responsible development. Given the potential of these technologies to enhance environmentally responsible manufacturing and economic development across the world, engineering research should investigate the relative environmental impacts of these technologies. Thus, this study aims to conduct an EIA of the novel fast MIP-SL process presented in Sec. 3.1 using an LCIA approach.

SL was the first commercialized AM technology. Currently, MIP-SL, also known as digital light projection SL, is one of the most commonly used AM technologies. SL uses light to solidify liquid photopolymer resin one layer at a time. In the MIP-SL process, light is patterned using a digital micromirror device (DMD) as a digital mask image to selectively cure the liquid photopolymer resin. A DMD is a micro-electromechanical system device that enables simultaneous control of ∼1 × 10⁶ small mirrors to turn pixels on or off at over 5 kHz. An illustration of the DMD chip and its use in a fast MIP-SL system are shown in Fig. 1.

Similar to other AM technologies, a typical MIP-SL process starts with a CAD model, which is then sliced into two-dimensional layers with a certain layer thickness. Each resulting slice is stored as a bitmap to be displayed on the dynamic mask. The light radiation reflected by the “on” micromirrors projects the sliced bitmap image onto the resin surface to cure a layer. An automated vertical (Z) stage is used to raise the platform in a resin vat. The technology addresses the need to develop AM machines with higher throughput [5]. To improve build speed, a fast MIP-SL process was developed and demonstrated [44], while mask image planning to control deformation in the process has recently been investigated [45].

In fast MIP-SL, a two-way movement separation mechanism was developed and adopted for the first time. In particular, a polydimethylsiloxane (PDMS) film coating was applied to the bottom surface of the resin vat, and additional sliding motion of the build platform is enabled in the horizontal (X) direction. The PDMS coating and the two-way movement design significantly reduced the separation time between the cured layers and the resin tank, and hence, reduced total build time. Stage movement is performed quickly to separate the newly cured layers from the bottom of the tank, followed by recoating the bottom layer to maintain a uniform and thin liquid resin layer. The fast MIP-SL process demonstrated the capability to significantly shorten build time without affecting part quality. Thus, an assessment of the relative performance of the fast MIP-SL on an environmental impact basis is desired. The approach for conducting the assessment is presented below.

First, production of six parts (Fig. 2) with different shape complexities was used in testing the performance of the approach [46]. Two different layer thicknesses commonly used in the MIP-SL process were also tested. A 50 μm layer thickness was used in the production of a gear model (Fig. 2(a)). For all other models (Figs. 2(b)–2(f)), a 100 μm layer thickness was utilized during production. The mask image projection time was 0.35 s for each layer, except for the base. The projection waiting time was set at 0.1 s. Due to larger layer thickness, longer image exposure and projection waiting times (0.45 s and 0.3 s, respectively) were used for the gear. Accordingly, the platform (Z-axis) movement also took more time—0.32 s and 0.42 s, for 50 μm and 100 μm layer thicknesses, respectively.

A relatively short waiting time was adopted after sliding movement in the X-direction. To obtain good surface quality, it is critical that the small gap is filled completely with still liquid resin before curing. Therefore, a shorter waiting time (50 ms) was used for parts with smaller cross-sectional areas, while a longer time (100 ms) was used for parts with larger cross-sectional areas. In addition, two types of resins, SI 500 and Acryl R5 (Envisiontec, Inc., Dearborn, MI), with different curing characteristics were tested. For the same layer thickness, the curing of Acryl R5 took about 0.1 s longer than SI 500 [47,48]. Based on the similarities in chemistry and build time for both resins, only SI 500 was considered for the EIA.

An EIA using an LCIA approach was conducted for the six part designs shown in Fig. 2. The functional unit for the analysis was 1000 units of each manufactured part (to represent production at scale). Environmental impacts were assessed for each part, including materials, energy involved in the process, and generated waste. The life-cycle inventory (LCI) was developed with reference to a study by Luo et al. [11], who analyzed the environmental performance of SL as a rapid prototyping process. The inputs for the process include resin (Perfactory SI500) for part fabrication, ethoxylated alcohol for cleaning, and electrical energy for processing. Outputs include the finished part and generated waste. In order to measure energy consumed in the process, real-time process energy monitoring was performed using a Fluke 435 Series II three-phase power quality and energy analyzer with four 6-kA flex connectors (Fluke Inc., Everett, WA) [49]. Also, Fluke-80i-110S AC/DC (0.5–100 A DC-1 kHz/20 kHz, 0.46 in.) current probes were used to match the AM machine specifications [50]. Relevant LCI data for the analysis, including energy consumption, mass of the fabricated objects, build times, and material waste, are shown in Table 1.

To determine the environmental impacts of the process, the ReCiPe 2008 LCIA method with world ReCiPe H/A weighting was selected because of its categorization of impacts [51]. The impact categories were addressed at the endpoint level, with three indicators: damage to human health, damage to ecosystems diversity, and damage to resource costs. In this research, the hierarchist perspective was applied, which is based on the most common policy principles with regard to timeframe and other issues [51]. LCI data were imported to LCA software (simapro 8.1), which generated the relative environmental impact results presented below (Fig. 3). It is seen that production of the gear has the lowest impact, while the greatest environmental impact can be attributed to the shell. However, upon analyzing the environmental impact of the manufactured parts without taking into account the postprocessing, it is seen that the brush, which has the lowest mass and second shortest build time, has the lowest impact, while the head, which has the greatest mass and longest build time, contributes the highest environmental impact (Fig. 4).

The distribution of environmental impacts for different process stages involved in the production of the shell (with postprocessing) and head (without postprocessing) parts is displayed in Figs. 5 and 6, respectively, in order to identify and understand the influence of individual process stages on the environmental impact of the parts.

The impact distribution was similar for other parts under analysis. From Fig. 5, it can be seen that impacts for the shell are primarily attributed to total energy use (fast MIP-SL process energy use and postprocessing energy use), with negligible impacts attributed to production of resin and wastes. Energy use remains a dominant impact factor for the head (Fig. 6); however, the production of the resin is also responsible for significant environmental impacts, due to proportionately lower MIP-SL machine energy requirements.

The total energy consumed in part production is proportional to the total time consumed per part manufactured using the fast MIP-SL process (Fig. 7). From the figure, it can be seen that machine energy use varies linearly (y = 0.454x, R²= 1) with build time (t_b). It can also be seen that the head, which exhibited the longest build time and environmental impacts, requires the most machine energy, while the teeth, which had the shortest build time, used the least amount of energy. The total energy consumed (E_T) is the sum of machine energy use (E_M) and postprocessing energy (E_P), as shown in the following equation:

From Fig. 8, it is seen that postprocessing energy (E_P) also varies linearly (y = 2.9959x, R²= 1) with postprocessing time (t_p). The shell, which exhibited the longest postprocessing time due to its shape complexity, requires the most energy, while the gear part which exhibited the least postprocessing time requires the least energy.

Hence, the relationship between total energy use and total time consumed per part can be formulated as a linear function.

To calculate machine energy and postprocessing energy, the modeling build time and postprocessing time per part are necessary.

The part build time for the fast MIP-SL process is the sum of projection time (t_p) and axes-translation time (t_a), as shown in Eq. (2). The projector time is the total time taken to project the mask images of all layers of the part.

Postprocessing time is calculated as the time required to wash the fabricated part in a washing station. From experimentation, it is found that postprocessing time varies as a function of shape complexity and dimensional size.

Design data required for modeling postprocessing time are represented in Table 2. Based on various part design conditions, an index number can be calculated to represent the shape complexity (S_C) of each part. The index is given a value based on the sum of four design criteria:

1. (1)

   Configuration: If the part design is 2.5D, S_C = 0, and if the part design is 3D, S_C = 1.

2. (2)

   Aspect ratio: If the height (z)/smallest cross-sectional size is < 10, S_C = 0, else S_C = 1.

3. (3)

   Feature scale ratio: If the triangle number (n)/volume is < 5000, S_C = 0, else S_C = 1.

4. (4)

   Internal features: If the part has long, small channels or holes, S_C = 1, else S_C = 0.

where t_base is the base time required to wash the part (dependent on operator's skill level), A is the part outside surface area (m²), A_b is the base outside surface area (m²), N_i is the number of internal channels, A_i is the internal face surface area (m²), and K₁, K₂, and K₃ denote degree of difficulty in washing, as defined below:

If S_c ≥ 1, K₁ = 1, else K₁ = 0.

If S_c ≥ 2, K₂ = 1, else K₂ = 0.

If S_c ≥ 3, K₃ = 1, else K₃ = 0.

The recorded time values obtained from the AM software and the energy values measured using a power quality analyzer are displayed in Table 1. Figure 9 shows the variation of the proposed model from real-time data obtained from the experiments. Calculated energy values for individual parts manufactured along with the measured energy values for the parts are represented against the total time of the process. A squared error of 0.015 is present between the calculated and measured total energy values.

The low error value between the calculated and measured energy values provides model verification proving that correct parameters are modeled, and the model has been accurately developed and implemented.

The current work evaluated the environmental performance of a novel fast MIP-SL process and identified energy as the dominant impact factor. Energy modeling and time modeling approaches were used to identify the interdependency of AM part design attributes and process energy consumption. It is noted that translating the design and process parameter behavior into a process model can reduce process variations and help control process energy consumption from the part design stage. The environmental performance was evaluated using an LCIA method. The process was modeled from material production through waste disposal to obtain LCI data. The LCIA was performed within simapro 8.1 software for the modeled process using the ReCiPe 2008 method.

The environmental performance study elucidated the effects of material, energy, and wastes in the fast MIP-SL process with respect to damage to human health, ecosystem quality, and resource depletion. Energy, identified as the most impactful factor, was also modeled. Models for calculation of total process time (part build time and postprocessing time) and process energy were developed using design data, from the 3D CAD model, and machine specifications. The models were validated using real-time energy values measured using a power quality analyzer. It was seen that the time values calculated using the models have a negligible variation from the time values obtained from AM software application output, while calculated and measured energy values have low variation. The small error (squared error of 0.015) verifies and validates model accuracy and resemblance to the real-world situation.

These process models can aid in predicting and controlling the process time and energy consumed for different designs. This understanding can promote social, economic, and environmental responsibility through informed design decision making. With design information being the main input to process planning, robust decision making can reduce/eliminate changes in build, planning, and monitoring strategies for the process. Hence, variation in product quality can be controlled with proper planning and efficient process implementation, thereby increasing the product desirability. The developed models can also be used to reduce or eliminate the undesirable processing time and energy costs associated with uninformed design decision making by allowing designers to navigate the interrelationships among 3D CAD model data and energy.

Although this study evaluates the key sources of environmental impacts for the fast MIP-SL process, the LCIA conducted highlights several concerns. The presented approaches cannot be easily extended to the evaluation of other AM processes due to a myriad of technology innovations and the variability in the parameters (process and material) associated with these technologies. The results of future studies of the SL process would be strengthened by examining more products with varying shape complexity to obtain more data and improve the accuracy and functionality of the developed models. Also, a comparison of the SL process with similar AM processes would aid in defining a foundational baseline for product and process sustainability comparisons. This would also help develop generalized energy and build time models for AM processes. Defining this information will enable engineers to suggest improvements to products and processes, enabling more sustainable AM.

The authors acknowledge the U.S. Department of Energy through the Industrial Assessment Center at Oregon State University for its support of this research.

A =

part outside surface area (m²)

A_b =

base outside surface area (m²)

A_i =

internal face surface area (m²)

AM =

additive manufacturing

ASTM =

American Society for Testing and Materials

CAD =

computer-aided design

CNC =

computer numerical control

DFAM =

design for additive manufacturing

DFSAM =

design for sustainable additive manufacturing

DMD =

digital micromirror device

E_M =

machine energy (kWh)

E_P =

postprocessing energy (kWh)

E_T =

total process energy (kWh)

ECR =

energy consumption rate

EIA =

environmental impact assessment

FDM =

fusion deposition modeling

i =

total number of layers

l_t =

layer thickness (mm)

LCA =

life-cycle analysis

LCI =

life-cycle inventory

LCIA =

life-cycle impact assessment

MIP-SL =

mask image projection stereolithography

MSDS =

material safety data sheet

n =

triangle number

N_i =

number of internal channels

P_M =

rated power of MIP-SL machine

P_P =

rated power of postprocessing setup

PDMS =

polydimethylsiloxane

S_C =

shape complexity

SETAC =

Society of Environmental Toxicology and Chemistry

SL =

stereolithography

STL =

surface tessellation file

t_a =

axes translation time (s)

t_b =

part build time (s)

t_base =

base time for washing the part (s)

t_p =

projector time (s)

t_pp =

postprocessing time (s)

z =

height of the part (mm)