ABSTRACT
Fused deposition modeling (FDM) is a popular technique for polymer additive manufacturing. However, the hygroscopic nature of thermoplastic filaments can lead to moisture-related defects during extrusion. When moisture is absorbed and vaporizes inside the nozzle, bubbles form, resulting in voids within the extrudate and deposited roads. This can compromise inter-road bonding and diminish mechanical performance. This study examines how the initial moisture content of ABS filaments affects the tensile behavior of parts fabricated by FDM. ABS filaments were conditioned to seven different moisture levels through water immersion for periods ranging from 0 to 12 hours, with moisture content quantified using the loss-in-weight method (ASTM D6980). ASTM D638 Type I specimens were printed under consistent processing conditions and tested in tension (n = 5 per condition). The results showed that ultimate tensile strength (UTS) decreased as filament moisture content increased, with a maximum reduction of 9.3% observed at 0.69% moisture compared to the dried condition (0.05%). Ductility was assessed by measuring strain at break, and its relationship with moisture content is illustrated in Fig. 4, along with statistical analysis (one-way ANOVA and post-hoc comparisons). These findings offer valuable insights for moisture management and quality control in ABS FDM processes.
-
KEYWORDS: Additive manufacturing, 3D printer, ABS filament, Moisture content, Tensile strength
-
KEYWORDS: 적층 제조, 3D 프린터, ABS 필라멘트, 수분 함량, 인장 강도
NOMENCLATURE
AM = Additive Manufacturing
ME = Material Extrusion
FDM = Fused Deposition Modeling
ABS = Acrylonitrile Butadiene Styrene
STL = StereoLithography
UTM = Universal Testing Machine
UTS = Ultimate Tensile Strength
M (%) = Average Moisture Content
Wi = Current Specimen Mass, g
Wo = Oven-dry Specimen Mass, g
1. Introduction
Additive manufacturing (AM), commonly referred to as 3D printing, enables the direct fabrication of parts with complex geometries and has therefore been adopted in a wide range of industrial applications. According to ISO/ASTM 52900, AM technologies are classified into seven categories based on the material-forming principle [
1]. Among them, Material Extrusion (ME) builds parts by extruding material through a nozzle and depositing it layer-by-layer. Fused Deposition Modeling (FDM), originally developed by Stratasys, is one of the most widely used ME processes because of its material availability, relatively low equipment cost, and ability to produce functional thermoplastic components [
2].
A variety of thermoplastics are processed via FDM, including ABS, PC, and PLA. In particular, acrylonitrile–butadiene–styrene (ABS) remains a popular choice due to its balanced performance in strength, stiffness, impact resistance, thermal stability, and postprocessability, making it suitable for applications such as electronic housings, automotive parts, and sporting goods. However, the mechanical properties of FDM-printed ABS parts are known to be sensitive to process conditions. Previous studies have shown that printing parameters such as extrusion temperature, print speed, layer thickness, and raster strategy can significantly affect tensile performance by altering interlayer bonding and internal defect populations [
3-
5].
In addition to these parameters, filament moisture content is an important practical factor that can degrade print quality and mechanical performance. During storage, filament spools may absorb moisture when they are not sealed with desiccant, particularly under hot and humid seasonal conditions (e.g., summer or monsoon-season environments) where high relative humidity can persist for extended periods. When moisture-containing filament is heated to extrusion temperatures, the absorbed water can vaporize and form bubbles, which may become trapped as voids in the extruded roads and interlayer regions. Such moisture-induced porosity can reduce effective load-bearing area and weaken interlayer adhesion, thereby lowering the measured tensile strength of printed parts [
6]. Earlier studies reported moisture-related degradation in FDM parts: Kim et al. found that moisture and temperature both influence the mechanical properties of FDM components [
7], and Zaldivar et al. showed that elevated initial moisture in ULTEM 9085 filament increased internal porosity and reduced mechanical strength [
8].
In this study, we systematically investigate how the initial moisture content of a commercially available ABS filament influences (i) microstructural features associated with bubble/void formation during extrusion and deposition and (ii) tensile properties of FDM-printed specimens. Filaments were conditioned to multiple moisture levels spanning a practical lower bound achieved by controlled drying and an upper bound representative of severe moisture exposure during unsealed storage. Tensile specimens were fabricated according to ASTM D638, while other printing parameters were controlled to isolate moisture effects. Tensile responses were evaluated using ultimate tensile strength (UTS) and strain at break as the primary indicators of strength and ductility, respectively. The results provide quantitative evidence on moisture-driven defect formation and its mechanical consequences, offering guidance for filament handling and storage practices in ABS FDM.
2. Materials and Methods
2.1 Moisture Conditioning and Moisture Measurement
A commercial ABS filament (Cubicon ABS, 1.75 mm diameter; CUBICON, Korea) was used. To obtain different initial moisture contents, filament segments were immersed in water at 23°C for 0, 2, 4, 6, 8, 10, and 12 h, following the water-immersion procedure for plastics described in ASTM D570 [
7]. After immersion, surface water was removed using a lint-free wiper, and the filament was briefly dried with ambient hot air to eliminate residual surface moisture prior to weighing. Conditioned filaments were stored in sealed bags until printing to minimize moisture change.
The filament moisture content was determined using the loss-inweight method in accordance with ASTM D6980 [
10]. Each filament segment was dried at 130°C for at least 2 h and then further dried until the mass change was less than 0.02% per 2 min. The moisture content was calculated as:
where Wi is the mass before oven drying and Wo is the oven-dry mass.
The lowest moisture content obtained after the applied drying and weighing protocol was 0.05%, which was treated as the practical reference condition (dried filament) in this study. This value represents the minimum moisture level measurable under the present procedure and does not imply an intrinsic lower-limit moisture content of ABS or an absolute zero-moisture state. Accordingly, 0.05% is used as the baseline moisture content (M0) throughout this paper.
Moisture content in
Table 1 is reported as a single measurement for each immersion time due to the limited availability of conditioned filament. Replicate measurements will be performed in a follow-up study to quantify measurement variability.
ASTM D638 Type I tensile specimens were designed and exported as STL files and sliced using Bambu Studio (Bambu Lab). Specimens were printed on a commercial FDM printer (X1 Carbon; Bambu Lab, China) with a 0.4 mm nozzle. To isolate the effect of filament moisture content, all printing parameters were fixed across conditions. The raster/toolpath strategy was concentric with 100% infill, and specimens were printed flat on the build plate.
Table 2 summarizes the fixed parameters used for specimen fabrication.
Tensile tests were performed according to ASTM D638 [
11] using a universal testing machine (UTM) under displacement control at a crosshead speed of 20 mm/min. Tests were conducted at 20°C. For each moisture condition, five specimens were tested (n = 5). Engineering stress-strain curves were calculated from the force and displacement data.
Ultimate tensile strength (UTS) was defined as the maximum engineering stress. Strain at break was defined as the engineering strain at specimen fracture (complete separation) and was selected as the primary ductility indicator because it is widely reported for polymer tensile behavior and directly reflects defect-driven crack initiation and post-yield localization. The physical implications of strain at break in moisture-affected FDM parts are discussed in Section 4.
2.4 Statistical Analysis
All tensile properties are reported as mean ± standard deviation. The effect of filament moisture content on tensile properties was evaluated using one-way analysis of variance (ANOVA) at a significance level of α = 0.05. When ANOVA indicated a significant effect, Tukey’s honestly significant difference (HSD) post-hoc test was applied for pairwise comparisons. For the moisture uptake data, the early-stage relationship between moisture uptake and square root of time was evaluated using linear regression to assess consistency with Fickian diffusion behavior.
3. Results
3.1 Microstructure Observation
Microstructural observations were performed to (i) verify moisture-induced bubble/void formation during extrusion and (ii) quantify how such defects manifest on a deposited layer and potentially reduce effective inter-road bonding areas. A variable-magnification stereo microscope (HD200VP-UM, AmScope, USA; ×15–45) was used for all observations.
Automatic thresholding was applied consistently to all images in ImageJ.
Microstructural observations were conducted to confirm moisture-induced bubble/void formation during extrusion and to evaluate how these defects appear in deposited layers. All images were obtained using a variable-magnification stereo microscope (HD200VPUM, AmScope, USA; ×15–45).
Bubble/void formation during extrusion.
Fig. 2 schematically illustrates that absorbed moisture can vaporize and expand in the heated nozzle, forming bubbles in the extrudate during FDM. To verify this behavior experimentally, filament strands extruded under identical printing conditions were collected immediately after extrusion (before deposition) and compared. As shown in
Fig. 3(a), the dried filament condition produced a smooth, uniform extruded strand, whereas the moisture-conditioned filament exhibited clear bubble/void features and local irregularities in the extrudate (
Fig. 3(b)).
Bubble-related defects and reduced bonding in deposited layers. As summarized in
Fig. 4, bubble-containing extrudate can lead to a porous deposited structure and reduce the effective contact/bonding area between adjacent roads and layers.
Single-layer observations and ImageJ quantification.
Representative magnified images of a single deposited layer on the build plate are shown in
Fig. 5 for different filament moisture contents (0.05–0.72%). With increasing moisture content, more frequent and larger bubble-derived pores were observed, along with a rougher and less uniform road morphology. To quantify the severity of bubble-related surface damage, the images in
Fig. 5 were analyzed using ImageJ by separating smooth/continuous regions from bubble-damaged regions (pore openings and surrounding disrupted areas). The bubble-damaged area fraction, calculated as the damaged area divided by the total analyzed area, increased with filament moisture content (
Table 3), providing quantitative evidence that higher initial moisture content promotes bubble-related defects in the deposited layer.
Table 1 shows that filament moisture content increased monotonically with immersion time, from 0.05% (0 h) to 0.72% (12 h). To assess whether the uptake behavior is consistent with Fickian diffusion, the moisture-content increase from the dried baseline was plotted against the square root of immersion time for the first 6 h. An approximately linear relationship was observed (R
2 = 0.956), indicating that the early-stage uptake is approximately Fickian under the present conditioning conditions. At longer times, the uptake rate decreased, suggesting an approach toward saturation.
3.3 Tensile Properties
Fig. 7 shows representative engineering stress–strain curves of ABS FDM specimens printed using filaments with different initial moisture contents. Across all conditions, the curves exhibit similar overall deformation behavior, while the peak stress decreases as moisture content increases, indicating a moisture-dependent reduction in tensile strength.
The moisture dependence of ultimate tensile strength (UTS), defined as the maximum engineering stress in each curve, is summarized in
Fig. 8. UTS decreased from 39.34 ± 0.42 MPa at 0.05% moisture to 35.68 ± 0.85 MPa at 0.69% moisture (maximum reduction: 9.29%). One-way ANOVA confirmed a significant effect of filament moisture content on UTS (F = 29.50, p = 7.19 × 10
-11). A linear fit of the mean UTS values versus moisture content is also provided in
Fig. 8, showing an overall negative trend (R
2 = 0.835).
Fig. 9 summarizes the strain at break for each moisture condition. The mean strain at break remained in a relatively narrow range from 0.05% to 0.69% moisture, whereas the 0.72% condition exhibited a notably higher value (3.34 ± 0.19%). One-way ANOVA indicated that moisture content had a significant effect on strain at break (F = 18.99, p = 1.08 × 10
-8). The linear regression shown in
Fig. 9 captures the overall increasing trend but with a low coefficient of determination (R
2 = 0.215), implying substantial scatter and/or nonlinearity, with the highest-moisture condition contributing strongly to the trend.
4. Discussion
The microscopy observations support the proposed mechanism that moisture absorbed in the filament vaporizes during high-temperature extrusion, generating bubbles that evolve into voids within the extrudate and deposited roads. These voids can reduce (i) the effective load-bearing cross-sectional area and (ii) the effective inter-road/interlayer bonded area, thereby degrading load transfer across deposited roads and interfaces. Consequently, even if the intrinsic stress carried by the polymer matrix (i.e., the true stress within the solid portion) remains comparable, the measured engineering (nominal) stress—defined using the gross specimen cross-section—decreases because part of the cross-section is occupied by pores and imperfect interfaces. This interpretation is consistent with the statistically significant reduction in UTS with increasing moisture content (one-way ANOVA: p = 7.19 × 10-¹¹), as well as the overall negative trend captured by the linear fit in
Fig. 8 (R
2 = 0.835).
Strain at break is a commonly used indicator of ductility. Because the base polymer and printing parameters were held constant in this study, the deformation capacity of the ABS matrix itself is not expected to change substantially with moisture content. Instead, moisture primarily introduces void- and interface-related defects that reduce strength through effective-area and bonding losses. In the present results, strain at break remained within a relatively narrow range for 0.05–0.69% moisture, whereas the highest-moisture condition (0.72%) exhibited a noticeably higher strain at break. Although the overall one-way ANOVA indicated a significant effect across all groups (p = 1.08 × 10-8), the weak linear fit for strain at break (R2 = 0.215) suggests substantial scatter and/or nonlinearity, implying that ductility is less systematically governed by moisture content than UTS within the investigated range. The remaining variation in strain at break is plausibly dominated by specimen-to-specimen scatter associated with the local distribution of defects, fracture initiation site, and the stochastic nature of void-driven failure.
From a practical perspective, the investigated moisture range (0.05–0.72%) spans conditions from well-dried filament to severe exposure that may occur when spools are stored or handled without sealed packaging or desiccant, particularly under hot and humid summer environments (e.g., monsoon-season storage in Korea or similar climates). In industrial practice, pre-drying is recommended to mitigate moisture-related defects; however, drying requires time and energy and may not always be feasible. Therefore, quantifying strength degradation as a function of filament moisture content can support quality-control decisions: if measured moisture remains within an acceptable range, unnecessary drying can be avoided; otherwise, drying, stricter storage protocols, and/or conservative design margins should be applied.
5. Conclusions
This study quantified the influence of the initial moisture content of ABS filament on moisture uptake behavior, microstructural defect formation, and tensile properties of FDM-printed parts.
(1) Filament moisture content increased from 0.05% to 0.72% with 0–12 h water immersion at 23°C. The early-stage uptake was approximately consistent with Fickian diffusion, as evidenced by an approximately linear relationship between (Mt-M0) and t for t ≤ 6h (R2 ≈ 0.95).
(2)Microscopy confirmed that bubble-derived defects increased with moisture content: moisture-conditioned filaments produced bubble/void features in the extrudate and a higher pore density in deposited single-layer roads.
(3) Ultimate tensile strength (UTS) decreased significantly with increasing moisture content (one-way ANOVA: p = 7.19 × 10-¹¹). The maximum reduction was 9.29% (from 39.34 ± 0.42 MPa at 0.05% to 35.68 ± 0.85 MPa at 0.69%).
(4) Strain at break showed relatively small changes from 0.05% to 0.69% moisture, while the highest-moisture condition (0.72%) exhibited a notably higher strain at break (3.34 ± 0.19%). Overall differences among groups were statistically significant (one-way ANOVA: p = 1.08 × 10-8), suggesting that moisture can influence ductility, although the trend was less systematic than that observed for UTS.
These findings provide quantitative guidance for moisture management and storage/preprocessing of ABS filaments to improve the reliability of mechanical performance in FDM applications.
FOOTNOTES
-
ACKNOWLEDGEMENT
This work was supported by Korea Institute for Advancement of Technology (KIAT) grant funded by the Korea Government (MOTIE) (No. RS-2023-KI002686, HRD Program for Industrial Innovation).
Fig. 1ASTM D638 type-1 specimen dimensions for tensile strength test
Fig. 2Schematic illustrating moisture vaporization in the heated nozzle and bubble formation in the extrudate during FDM
Fig. 3Extruded filament strand collected immediately after extrusion (before deposition): (a) 0.05% and (b) 0.72% moisture content. Scale bar: 500 μm
Fig. 4Conceptual mechanism linking moisture-induced bubbles to a porous deposited structure and reduced effective inter-road/interlayer bonding area
Fig. 5Single-layer images on the build plate at different moisture contents: (a) 0.05%, (b) 0.37%, (c) 0.43%, (d) 0.63%, (e) 0.72%. Scale bar: 0.5 mm
Fig. 6Fickian diffusion check for moisture uptake (early stage, up to 6 h): moisture-content increase from the dried baseline plotted against the square root of immersion time, with a linear fit (R2 = 0.956)
Fig. 7Representative engineering stress–strain curves of ABS FDM specimens printed with different initial filament moisture contents
Fig. 8Ultimate tensile strength (UTS) as a function of filament moisture content (mean ± SD, n = 5), with a linear fit
Fig. 9Strain at break as a function of filament moisture content (mean ± SD, n = 5), with a linear fit
Table 1Moisture content as a function of immersion time (23°C)
Table 1
|
Absorption Time [Hr] |
0 |
2 |
4 |
6 |
8 |
10 |
12 |
|
Moisture content [%] |
0.05 |
0.37 |
0.43 |
0.46 |
0.63 |
0.69 |
0.72 |
Table 2Fixed printing parameters used for specimen fabrication
Table 2
|
Parameter |
Value |
|
Additive angle [°] |
0 |
|
Nozzle temperature [°C] |
260 |
|
Bed temperature [°C] |
90 |
|
Nozzle speed [mm/s] |
270 |
|
Infill density [%] |
100 |
|
Infill pattern |
concentric |
|
Layer height [mm] |
0.2 |
|
Nozzle diameter [mm] |
0.4 |
|
Layer angle [°] |
0 (In fill pattern is concentric) |
Table 3ImageJ-based quantification of the bubble-damaged area fraction from
Fig. 5 as a function of filament moisture content
Table 3
|
Moisture content [%] |
0.05 (a) |
0.37 (b) |
0.43 (c) |
0.63 (d) |
0.72 (e) |
|
Damaged area fraction [%] |
0 |
32 |
38 |
41 |
45 |
REFERENCES
- 1. ISO/ASTM 52900. 2022;Additive manufacturing - General principles - Fundamentals and vocabulary.
- 2. Lee, I. H., Kim, H. C., Ahn, D. G., (2020), ISO/ASTM 52900 Korean terminologies for additive manufacturing according to the ISO/ASTM 52900 standard, Journal of the Korean Society for Precision Engineering, 37(12), 929-936.
- 3. Woo, I. Y., Lyu, M. Y., (2020), Variations in the tensile strength in material extrusion-type 3D printed specimens for various tool paths, Polymer(Korea), 44(6), 769-775.
- 4. Na, D. H., Kim, S. G., (2022), A study on tensile strength according to various output conditions of PLA+ materials using 3D printing, Transactions of Materials Processing, 31(2), 89-95.
- 5. Lee, Y. S., Han, S. R., Kim, K. A., (2022), A study on the tensile strength of different PLA materials under FDM method 3D printing layer conditions, Journal of the Korean Society of Mechanical Technology, 24(4), 591-596.
- 6. Chuang K. C., Grady J. E., Arnold S. M., Draper R. D., Shin E., Patterson C., Santelle T., Lao C., Rhein M., Mehl J.. 2015;A fully nonmetallic gas turbine engine enabled by additive manufacturing, Part II: Additive manufacturing and characterization of polymer composites, (Report NASA/TM—2015-218749). https://ntrs.nasa.gov/api/citations/20150011644/downloads/20150011644.pdf.
- 7. Kim, E., Shin, Y. J., Ahn, S. H., (2016), The effects of moisture and temperature on the mechanical properties of additive manufacturing components: fused deposition modeling, Rapid Prototyping Journal, 22(6), 887-894.
- 8. Zaldivar, R. J., Mclouth, T. D., Ferrelli, G. L., Patel, D. N., Hopkins, A. R., Witkin, D., (2018), Effect of initial filament moisture content on the microstructure and mechanical behavior of ULTEM 9085 3D printed parts, Additive Manufacturing, 24, 457-466.
- 9. ASTM D570. 2022;Standard test method for water absorption of plastics.
- 10. ASTM D6980. 2017;Standard test method for determination of moisture in plastics by loss in weight.
- 11. ASTM D638. 2022;Standard test method for tensile properties of plastics.
Biography
- SeokHwan Jung
M.S. candidate in the Department of Mechanical Engineering, Dankook University. His research interests are in precision engineering, manufacturing systems, and smart materials.
- JiHwan Park
Undergraduate student in the Department of Mechanical Engineering, Dankook University. His research interest is precision engineering.
- Sung Han Rhim
Professor in Department of Mechanical Engineering, Dankook University. His research interests are in precision engineering, manufacturing systems, and digital health.
