Characteristics, formation mechanisms, and control methods ​of emissions from 3D printing​

• Main Author: Ding, Shirun
• Other Authors: Ng Bing Feng
• Format: Thesis-Doctor of Philosophy
• Language: English
• Published: Nanyang Technological University 2022
• Subjects: Engineering::Environmental engineering::Hazardous substances
• Online Access: https://hdl.handle.net/10356/159265
• Institution: Nanyang Technological University

Additive manufacturing (AM) has been an active area of research due to its extensive application in many fields such as mechanical engineering, marine engineering, construction, bioengineering, and electronic engineering to manufacture products with complicated geometries and advanced material properties. However, particle emissions from desktop fused deposition modeling (FDM) 3D printers were first reported in 2013 [1]. Since then, most of the studies have focused on the characterizations of the particulate matter (PM) and volatile organic compounds (VOCs) emitted from FDM 3D printers [2], which have adverse health effects on users exposed for prolonged durations. To predict the characteristics of emissions and provide a scientific basis to control emissions, it is necessary to understand the formation of emissions. This work has investigated the characteristics, formation mechanisms, and control methods of hazardous emissions from 3D printers. Since most attention has been focused on desktop FDM printers, particle emissions from other kinds of 3D printing techniques are also of concern to occupational health but have since been less explored. Here, on-site particle concentration levels were examined for polymer filament-based FDM, metal- and polymer-based powder bed fusion (PBF), metal powder-based directed energy deposition (DED), and ink-based material jetting (MJ). Particle concentrations in the operating environments of users were measured using a combination of particle sizers including scanning mobility particle sizer (10-420 nm) and optical particle sizer (0.3-10 µm). The number and mass concentrations of submicron particles emitted from a desktop open-type FDM printer for acrylonitrile-butadiene-styrene (ABS) and polyvinyl alcohol (PVA), approached and significantly exceeded the nanoparticle reference limits (4 × 104 #/cm3), respectively. On the other hand, caution should be taken in the pre- and post-processing of metal and polymer powder. Specifically, one to ten micrometers of particles were observed in the air during the sieving, loading, and cleaning of powder, with transient mass concentrations ranging from 150 to 9000 µg/m3 that significantly exceeded the threshold level (150 µg/m3) suggested for indoor air quality. Automatic systems that enable ‘closed powder cycle’ or ‘powder-free handling’ should be adopted to prevent users from unnecessary particle exposure. While the standard chamber method has been widely adopted to measure particle emissions from an FDM printer, there was obvious inconsistency and uncertainty in terms of particle emission rates (PER, #/min) being measured, owing to different measurement conditions and calculation models used. Here, a dynamic analysis of the size-resolved PER was conducted through a comparative study of the chamber and flow tunnel measurements. Two models to resolve PER from the chamber and a model for flow tunnel measurements were examined. It was found that chamber measurements for different materials underestimated PER by up to an order of magnitude and overestimated particle diameters by up to 2.3 times, while the flow tunnel provided more accurate results. Field measurements of the time-resolved particle size distribution (PSD) in a typical room environment could be predicted well by the flow tunnel, while the chamber measurements could not represent the main PSD characteristics (e.g., particle diameter mode). Secondary aerosols (>30 nm) formed in chambers were not observed in field measurements. Flow tunnel was adopted as a possible alternative for the study of 3D printer emissions to overcome the disadvantages in standard chamber methods and as means to predict exposure levels. For the characterization of emissions from FDM printers, the percentage of PM in total emissions (i.e., the nucleation ratio of evaporated substances) was still unknown. Here, we directly measured particle emission yields from the extrusion process of filaments using an FDM 3D printer in a chamber and at the same time, indirectly measured total evaporated substances yields using a proposed weight-loss method (called TVOCWL). The nucleation ratio of evaporated substances was estimated by comparing the particle and TVOCWL results. It was found that TVOCWL mass yields were 0.03%, 0.21%, and 2.14% for polylactic acid, ABS, and PVA, respectively, at 220℃. Among TVOCWL, particle mass accounted for 1% to 5% of TVOCWL mass depending on the type of filaments used. Important research gaps in the mechanisms leading to the formation of both UFP and VOC from FDM 3D printers remain. Here, we further characterize the formation mechanisms of emissions from polymer filaments commonly used in FDM 3D printing. The temporal relationships between the amount and species of VOCs at different operating thermal conditions were obtained through a combination of evolved gas analysis (EGA) and thermogravimetric analysis (TGA). This is to capture physicochemical reactions, in which the furnace of EGA or TGA closely resembled the heating process of the nozzle in the FDM 3D printer. It was generally observed that emissions initiated at the start of the glass transition process and peaked during liquefaction for filaments. Initial increment in emissions during liquefaction and the relatively constant decomposition of products in the liquid phase were two main VOC formation mechanisms. Also, fumes from an FDM printer were directly captured using a laser imaging method. It was observed that fumes originate from the printer nozzle and newly deposited layers during printing, where control measures should be targeted. Having amassed in-depth knowledge in the formation mechanisms of emissions from FDM printers, control methods were investigated to decrease the amount or change the chemical composition of the emissions. It was found that low heating rates had the potential to restrain the formation of carcinogenic monomer, styrene, from ABS while reusing filaments or pre-conditioning for filaments before use has the potential to decrease the level of emissions. On the other hand, a sucking ring design around a printer nozzle was proposed, that can prevent the diffusion of fumes produced. The removal efficiency of the sucking ring for both particle and VOC emissions was higher than 90%. In summary, this thesis has systematically and comprehensively investigated the characteristics, formation mechanisms, and control methods of particle and VOC emissions from desktop FDM 3D printers. Investigations into on-site particle emissions from a wide spectrum of AM techniques were also conducted, which provided a preliminary understanding of possible particle emissions from a diverse range of 3D printers. For FDM printers, the proposed flow tunnel method provided an alternative to the standard chamber method, to more accurately measure the characteristics of particle emissions. In addition, the proposed TVOCWL measurement method based on the weight-loss analysis provided new insights into the characteristics of the emissions. Separately, investigations in the formation mechanisms contributed to the understanding of the formation of particle and VOC emissions from FDM printers. To minimize the exposure of operators to potentially hazardous emissions, the proposed sucking ring design is a simple and effective method to control emissions from FDM printers. We also recommend the use of closed-type 3D printers with control measures (e.g., use of an enclosure with internal air filtration) and “green materials” with fewer additives. For other industrial-scale 3D printing techniques, additional automatic systems need to be incorporated for pre- and post-processing to achieve ‘powder-free handling’.