Particle emissions and exposure to hazardous chemical agents during binder jetting utilising a silica sand

Abstract

Background: There are research gaps in additive manufacturing (AM) concerning particle number concentrations, particle emission rates, area concentrations, and personal respiratory exposure in real-world workplaces. Binder jetting (BJ) machines, have the potential to release particles, including ultrafine particles, into the air. AM operators might be exposed to particle emissions and hazardous chemical agents (HCAs) namely, respirable crystalline silica, and respirable particulates not otherwise specified (PNOS) during the pre-processing phase, processing phase, and post-processing phase when using silica sand. More research is needed regarding particle number concentrations, particle emissions, area concentrations and respiratory exposure of AM operators to respirable crystalline silica and respirable PNOS during BJ utilising a silica sand. Aims and objectives: The research aim of this dissertation was to determine the powder characteristics and chemical composition of uncoated, coated and used silica sand and to quantify particle number concentrations, particle emission rates, area concentrations, and personal respiratory exposure of the AM operator to respirable crystalline silica and respirable PNOS during the three AM phases at a South African AM research facility. The research objectives were: (i) To determine the powder characteristics and chemical composition of uncoated, coated, and used silica sands during BJ; (ii) To quantify particle number concentrations and to determine emission rates (ERs) of particles released during BJ, and (iii) to assess area concentrations and personal respiratory exposure of the AM operator to respirable crystalline silica and respirable PNOS during BJ. Methodology: Powder characterisation of uncoated, coated, and used silica sand was determined through particle size distribution (PSD) and shape analysis using a Malvern Morphologi particle analyser and scanning electron microscopy (SEM), while the and wavelength dispersive X-ray fluorescence (WD-XRF) was used to determine chemical composition. Direct-reading instruments including, a Grimm-Portable Laser Aerosol Spectrometer model 11-A and a Nanozen DustCount® 9000 Z1 optical particle counter, were used to quantify particle number concentrations and emission rates of particles released during the three AM phases. Both area monitoring and personal exposure monitoring were conducted during the AM operator’s full shift to determine area concentrations and personal respiratory exposure to respirable crystalline silica and respirable PNOS in the AM research facility. This was conducted by means of time-integrated sampling (GilAir Plus pump with a 2-piece cassette and aluminium cyclone) and real-time monitoring (DustCount® used to obtain real-time data and filter analysis results). This study was conducted over six days: four days were allocated for printing parts, while two days were allocated for the carbon dioxide (CO2) decay method (as part of the air exchange rate (AER) and ER calculations). Results: The PSD and SEM results could not be compared to the safety data sheet (SDS) since the particle size distribution and shape of silica sand were not stated in the SDS. PSD results indicated that uncoated, coated and used silica sand fell into the inhalable size fraction (< 100 μm). PSD analysis indicated that 10% of particles [d(0.1)] in uncoated (0.58 ± 0.03 μm), coated0.61 ± 0.08 μm), and used silica sand (0.58 ± 0.04 μm) were respirable sized particles. PSD analysis indicated that 50% [d(0.5)] of uncoated, coated, and used silica sand particles were smaller than 1.29 ± 0.36 μm, 4.27 ± 0.08 μm and 2.42 ± 0.04 μm respectively. Statistically significant differences was found for particles [d(0.5)]. PSD analysis indicated that 90% [d(0.9)] of uncoated, coated, and used silica sand particles were smaller than 42.21 ± 29.96, 104.7 ± 23.33 μm and 54.9 ± 49.93 μm respectively.SEM images and PSD results supported these findings since particles < 100 μm in size were detected. PSD analysis and SEM images confirmed that silica sand particles were smooth and non-spherical in shape. All three silica sand samples contained high crystalline silica content, ranging from 96.83 to 98.20%, which corresponded with the >90% stated in the SDS. The highest peak particle number concentration and particle ER were 680.51 p/cm3 and 3.06 × 105 p/min respectively for particles < 1 μm in size. Real-time monitoring using the DustCount® indicated that particles sized 0.375 μm in size were the most prevalent during the AM process with detectable respirable crystalline silica and respirable PNOS during the AM phases. Personal exposure 8-hour Time Weighted Average (TWA) concentrations were measured at 0.01 ± 0.00 mg/m3 for respirable crystalline silica and 0.04 ± 0.03 mg/m3 for respirable PNOS. It was found that respirable crystalline silica measured during printing was equal to 10% of the Time-weighted Average-Occupational Exposure Limit-Maximum limit (TWA-OEL-ML) of 0.1 mg/m3. Conclusion: PSD analysis and SEM images of uncoated, coated and used silica sand indicated particles in the inhalable fraction (< 100 μm). Particles < 1 μm in size were emitted during the three AM phases, indicating that BJ using silica sand are a high emitter of submicron particles. A statistically significant difference was found between the uncoated, coated, and used silica sand particle according to the d(0.5) PSD results. There was a notable difference in the particle number concentrations and particle ERs measured by the direct-reading instruments, with the DustCount® yielding much lower particle number concentrations and particle ERs. When comparing the particle number concentrations measured by the DustCount® in area one (in front of AM machine) and area two (back of AM machine), higher particle number concentrations were measured in area two. Time-integrated sampling and real-time monitoring of respirable crystalline silica and respirable PNOS indicated that all 8-hour TWA personal exposures complied with their respective TWA-OELs. However. It’s important to note that the respirable crystalline silica concentrations during personal monitoring was equal 10% of the TWA-OEL-ML.