Table 2.
Geometric mean (geometric standard deviation) average background-corrected airborne particle and total volatile organic compound (TVOC) concentrations and particle sizes by process and task (n = 5 replicates)1.
| Particles | Gases | ||||||
|---|---|---|---|---|---|---|---|
| CNC | FMPS | APS | PID | ||||
| Task | Process | Conc. (#/cm3) * | Conc. (#/cm3) * | Size, nm (GSD)‡ | Conc. (#/cm3) * | Size, nm (GSD)‡ | TVOC (μg/m3) * |
| Pouring (resin) | SLA | 44.0 (2.0) A,† | 14.1 (2.2) A,† | 51.4 (1.3) | 0.1 (2.9) B | 678.2 (1.3) | – |
| DLP | – | – | 55.4 (1.3) | – | 696.6 (1.4) | 1.1 (1.8) D | |
| Printing (resin) | SLA | – | – | 40.6 (1.5) | – | 658.2 (1.2) | 61.0 (1.7) D |
| DLP | – | – | 33.3 (1.5) | – | 654.4 (1.3) | – | |
| Recovery (resin) | SLA | – | – | 41.0 (1.4) | 0.1 (1.9) B,
|
801.0 (1.6) | – |
| DLP | 7.9 (1.6) A | – | 29.8 (1.5) | 0.2 (1.8) B,
|
876.2 (1.6) | 0.4 (4.3) C,D,† | |
| Rinsing (IPA) | SLA | – | 10.5 (3.5) A,& | 44.2 (1.5) | 0.1 (1.4) B,† | 779.8 (1.5) | 1042.5 (4.4) C |
| DLP | 8.8 (2.7) A | 31.7 (3.1) A,
|
29.6 (1.6) | 0.3 (1.1) B,
|
831.8 (1.6) | 0.4 (2.6) C,† | |
| Soaking (IPA) | SLA | 20.7 (1.7) A | – | 48.6 (1.4) | 0.1 (1.8) B | 769.4 (1.5) | 8080.7 (1.9) A |
| DLP | 27.0 (1.9) A | – | 33.1 (1.5) | 0.5 (3.0) B,
|
764.4 (1.5) | 15,176.8 (1.6) A | |
| Drying (IPA) | SLA | – | – | 49.1 (1.5) | – | 695.2 (1.3) | 8322.3 (1.5) A |
| DLP | – | – | 39.4 (1.6) | – | 726.2 (1.4) | 13,384.9 (1.3) A | |
| Curing | SLA | n/a | – | 36.2 (1.3) | n/a | n/a | 1867.2 (1.0) A,B |
| DLP | n/a | – | 44.3 (1.4) | n/a | n/a | 1654.2 (1.1) A,B | |
| Sanding | SLA | – | n/a | n/a | 4.6 (1.9) A,
|
1334.8 (1.9) A | n/a |
| DLP | – | n/a | n/a | 0.3 (4.7) A,
|
1024.5 (1.7) B | n/a | |
CNC = condensation nuclei counter, FMPS = fast mobility particle sizer, GSD = geometric standard deviation of diameter, APS = aerodynamic particle sizer, PID = photoionization detector, TVOC = total volatile organic compounds.
Adjusted regression models (difference among tasks are the same for both printers and there is no difference between printers for each task). Within each column, capital letters for (GM values that are not connected by the same letter are significantly different (p < 0.05). For example using the PID data, the “Drying A” and “Soaking A” tasks are both significantly different than the “Rinsing C” task.
Mixed regression models (significant interaction between printer and task). FMPS size data (all p < 0.05): DLP process—pouring > printing or recovery or rinsing or soaking or drying tasks. APS size data (all p < 0.05): SLA process—sanding > all other tasks; DLP process—sanding > pouring” printing = drying; between processes—SLA sanding > DLP sanding. − = three or more measurements during task were below background (GM not calculated).
One of five raplicate values imputed using; β-substitution method (see Method section).
Two of five; replicate values imputed using β-substitution method (see Method section).
One of four replicate values (data for fifth replicate loss because of instrument error) imputed using β-substitution method (see Method section) n/a = task not monitored with this instrument.