INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)  
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue XII December 2025  
Optimization of 3D‑Printed Patterns Parameters and Two‑Stage  
Burnout Process for Defect Reduction in Propeller Blades Investment  
Casting Shell Mold  
Zolkarnain Marjom*., Ahmad Syazani Ahmad Moktar., Mohamad Ridzuan Mohamad Kamal  
Facuty of Industrial & Manufacturing Technology & Engineering, Universiti Teknikal Malaysia Melaka  
DOI: https://doi.org/10.47772/IJRISS.2025.91200055  
Received: 10 December 2025; Accepted: 17 December 2025; Published: 31 December 2025  
ABSTRACT  
This study investigates the optimization of 3D-printed investment casting patterns and two-stage burnout  
parameters to minimize defects in propeller blade manufacturing. A full factorial design of experiments (2⁴) was  
implemented to analyze the effects of four fused deposition modeling (FDM) parameters—shell thickness, infill  
density, layer height, and internal pattern structure—on burnout performance. Thirty-two PLA patterns were  
fabricated and evaluated through a two-stage burnout process: Stage 1 (200–350ꢀ°C) assessed air permeability,  
while Stage 2 (up to 650ꢀ°C) examined surface integrity using dye penetrant testing and visual crack inspection.  
Statistical analysis using GLM ANOVA revealed that air permeability exhibited no significant main effects but  
was influenced by higher-order interactions, notably Infill× Shell× Pattern (F = 5.067, p = 0.03879) and Layer×  
Shell× Pattern (F = 6.975, p = 0.01779). Dye penetrant indications were dominated by shell thickness (F =  
2135.9, p ≈ 1.84e⁻18), with layer height and multiple interactions also significant. Visual cracking was strongly  
associated with shell thickness (Fisher exact p = 0.00245), with 1ꢀmm shells reducing defects compared to 2ꢀmm.  
The findings underscore that shell thickness is the primary factor for Stage 2 defect mitigation, while Stage 1  
optimization requires joint tuning of shell, infill, and pattern parameters. The proposed two-stage burnout  
workflow enables early identification of critical factor combinations, offering a robust approach for improving  
dimensional integrity and surface quality in additively manufactured investment casting applications.  
Keywords: Additive Manufacturing; Investment Casting; Two‑Stage Burnout; DOE; ANOVA; Air  
Permeability; Dye Penetrant; Cracking.  
INTRODUCTION  
Investment casting (IC) is widely employed for manufacturing components requiring high dimensional accuracy  
and intricate geometries, particularly in aerospace and marine applications [1, 20]. Conventional IC processes  
rely on wax patterns, which require tooling and limit design flexibility [3]. The integration of additive  
manufacturing (AM) technologies has enabled the production of expendable patterns directly from digital  
models, significantly reducing lead time and cost [4, 7].  
3D printing with Fused deposition modeling (FDM) technology is among the most accessible AM techniques  
for producing polymer-based investment casting patterns due to its low cost and material availability [5,13].  
Polylactic acid (PLA) is commonly used; however, its thermal degradation behaviour differs from wax, resulting  
in complex gas evolution during burnout [11]. Insufficient gas evacuation may lead to shell cracking and surface  
defects, especially in enclosed geometries such as propeller blades [15].  
Previous studies often evaluate burnout-related defects only after complete pattern removal [6, 8]. Such  
approaches overlook the importance of early-stage gas evacuation behaviour, which significantly influences  
shell integrity during subsequent high-temperature exposure. Furthermore, the combined effects of infill density,  
internal pattern structure, and shell thickness on permeability and cracking behaviour have not been  
systematically quantified using statistical design methods [9].  
Page 633  
www.rsisinternational.org  
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)  
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue XII December 2025  
To address these gaps, this study adopts a two-stage burnout strategy that separates air permeability evaluation  
from final crack assessment. A full factorial DOE is employed to statistically analyse the influence of key FDM  
parameters on both responses, enabling improved understanding and optimisation of additively manufactured  
investment casting patterns.  
METHODOLOGY  
The creation of the propeller blade begins with design using CATIA software. The model is exported as an STL  
file, imported into Ultimaker Cura software, and printed using PLA filament.  
A full factorial design of experiment (2⁴) was adopted to evaluate the effects of four FDM process parameters:  
infill density (10% and 20%), layer height (0.10 mm and 0.15 mm), shell thickness (1 mm and 2 mm), and  
internal pattern structure (gyroid and concentric). Sixteen experimental runs were conducted using PLA patterns  
fabricated via FDM with two replicates (n = 32).  
Ceramic shells were prepared using a conventional slurry dipping process as described in the investment casting  
literature [1,3]. Burnout was performed in two stages. Stage 1 involved partial burnout between 200 and 350 °C  
to initiate polymer degradation while maintaining shell integrity. Air permeability tests were conducted. Stage 2  
extended heating to 650 °C to ensure complete pattern removal, followed by visual crack inspection and dye  
penetrant testing per ASTM E1417. Figure 1 shows Air Permeability test & Dye Penetration Test (NDT).  
Figure 1. Air Permeability Test & Dye Penetration Test (NDT) Sample  
RESULTS AND DISCUSSION  
A balanced 2⁴ full factorial design with two replicates (n = 32) was analysed for four factors: Infill (10%, 20%),  
Layer height (0.10, 0.15 mm), Shell thickness (1, 2 mm), and Pattern (Gyroid, Concentric). Responses were: Air  
Permeability (after Stage‑1 burnout 200–350ꢀ°C), Dye Penetrant indication (after Stage‑2 burnout up to 650ꢀ°C),  
and Visual Crack (binary).  
Analyses employed GLM ANOVA (Type II) consistent with Minitab practice for balanced factorials. Residual  
diagnostics were performed; exact tests were used for the binary cracking response due to separation at 2 mm  
shell.  
DOE Factors and Levels  
Table 1 shows factors and levels: Infill (10%, 20%), Layer height (0.10 mm, 0.15 mm), Shell thickness (1 mm,  
2 mm), Pattern (Gyroid, Concentric).  
Page 634  
www.rsisinternational.org  
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)  
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue XII December 2025  
Table 1. Full factorial DOE matrix with run order and factor settings.  
Parameters  
Responds  
Run  
Infill  
density  
Layer  
height  
Shell  
Thickness  
Pattern  
Type  
Visual  
Crack  
AirPerm  
Dye  
1
10  
10  
10  
10  
10  
10  
20  
20  
20  
20  
10  
10  
10  
10  
10  
10  
20  
20  
20  
20  
20  
20  
20  
20  
0.15  
0.15  
0.1  
1
1
2
2
1
1
2
2
1
1
2
2
2
2
1
1
2
2
1
1
1
1
2
2
Gyroid  
0
1
1
1
0
1
1
1
1
1
1
1
1
1
0
0
1
1
1
1
0
1
1
1
0.0184  
0.0196  
0.019  
6.45  
6.95  
23.12  
24.7  
8.94  
9.6  
2
Gyroid  
3
Gyroid  
4
0.1  
Gyroid  
0.0204  
0.0191  
0.0205  
0.0193  
0.0208  
0.0198  
0.0212  
0.0207  
0.0222  
0.0189  
0.0203  
0.0178  
0.0189  
0.0196  
0.021  
5
0.1  
Gyroid  
6
0.1  
Gyroid  
7
0.15  
0.15  
0.15  
0.15  
0.15  
0.15  
0.15  
0.15  
0.1  
Gyroid  
15.37  
16.6  
12.55  
13.5  
27.33  
29  
8
Gyroid  
9
Gyroid  
10  
11  
12  
13  
14  
15  
16  
17  
18  
19  
20  
21  
22  
23  
24  
Gyroid  
Gyroid  
Gyroid  
Concentric  
Concentric  
Concentric  
Concentric  
Concentric  
Concentric  
Concentric  
Concentric  
Gyroid  
29.74  
31.6  
0
0.1  
0.4  
0.15  
0.15  
0.15  
0.15  
0.1  
11.62  
12.5  
17.84  
19  
0.0193  
0.0208  
0.0201  
0.0216  
0.0189  
0.0203  
8.44  
9.2  
0.1  
Gyroid  
0.1  
Gyroid  
11.55  
12.4  
0.1  
Gyroid  
Page 635  
www.rsisinternational.org  
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)  
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue XII December 2025  
25  
26  
27  
28  
29  
30  
31  
32  
20  
20  
10  
10  
10  
10  
20  
20  
0.1  
0.1  
0.15  
0.15  
0.1  
0.1  
0.1  
0.1  
1
1
1
1
2
2
2
2
Concentric  
Concentric  
Concentric  
Concentric  
Concentric  
Concentric  
Concentric  
Concentric  
0
1
0
0
1
1
1
1
0.0189  
0.0203  
0.02  
6.65  
7.2  
0
0.0215  
0.0188  
0.0202  
0.0209  
0.0224  
0.5  
20.14  
21.5  
28.54  
30.7  
Air Permeability (Stage‑1, 200–350 °C)  
Mean air permeability was slightly higher at the “upper” levels of each factor: 20% infill (0.02033 vs 0.01977),  
0.15ꢀmm layer height (0.02021 vs 0.01988), 2ꢀmm shell (0.02023 vs 0.01986), and Gyroid pattern (0.02012 vs  
0.01998).  
Main‑effects  
means  
showed  
small  
differences  
(Figure  
2);  
significant  
three‑way  
interactions  
(Infill×Shell×Pattern; Layer×Shell×Pattern) indicate multi‑factor control of permeability (Figure 3).  
Figure 2. Main‑effects plots for Air Permeability (Stage‑1 Burnout).  
Page 636  
www.rsisinternational.org  
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)  
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue XII December 2025  
Figure 3. Interaction plots for Air Permeability (Stage‑1 Burnout).  
Dye Penetrant Indication (Stage‑2, up to 650 °C)  
Dye readings increased sharply with shell thickness and moderately with layer height (Figure 4); several  
interactions were highly significant, confirming combined factor effects (Figure 5).  
Figure 4. Main‑effects plots for Dye Penetrant Indication (Stage‑2 Burnout).  
Page 637  
www.rsisinternational.org  
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)  
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue XII December 2025  
Figure 5. Interaction plots for Dye Penetrant Indication (Stage‑2 Burnout).  
Visual Cracking (binary)  
Crack proportion was 100% at 2 mm shell versus 50% at 1 mm (Fisher exact p = 0.00245; Chi‑square p =  
0.00427). Other factors showed non‑significant differences. Quasi‑complete separation suggests exact tests over  
GLM.  
Comprehensive ANOVA: Air Permeability (Type II)  
No main effects reached significance (p > 0.05). Two three‑way interactions were significant:  
Infill×Shell×Pattern (F = 5.067, p = 0.03879) and Layer×Shell×Pattern (F = 6.975, p = 0.01779) (Table 2).  
Practically, early‑stage gas evacuation depends on combined settings of shell, pattern, and either infill or layer  
height; optimisation should evaluate these factors jointly. Residual diagnostics indicated non‑normality  
(Shapiro–Wilk pꢀ≈ꢀ5.8×10⁻⁷); a log transformation preserved the same interaction significance but did not restore  
normality (Shapiro–Wilk pꢀ≈ꢀ4.0×10⁻⁷)  
Table 2. Full ANOVA for Air Permeability (sum of squares, F, and p-values).  
Term  
Sum Sq  
F
p-value  
0.13424  
0.36148  
0.31121  
0.68892  
C(Infill)  
C(Layer)  
C(Shell)  
C(Pattern)  
0.000002  
0.000001  
0.000001  
0.000000  
2.489  
0.883  
1.094  
0.166  
Page 638  
www.rsisinternational.org  
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)  
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue XII December 2025  
C(Infill):C(Layer)  
0.000002  
0.000000  
0.000000  
0.000001  
0.000000  
0.000000  
0.000000  
0.000001  
0.000005  
0.000007  
2.270  
0.385  
0.071  
0.694  
0.038  
0.342  
0.196  
0.817  
5.067  
6.975  
0.636  
—
0.15140  
0.54375  
0.79373  
0.41707  
0.84787  
0.56676  
0.66361  
0.37942  
0.03879  
0.01779  
0.43677  
—
C(Infill):C(Shell)  
C(Layer):C(Shell)  
C(Infill):C(Pattern)  
C(Layer):C(Pattern)  
C(Shell):C(Pattern)  
C(Infill):C(Layer):C(Shell)  
C(Infill):C(Layer):C(Pattern)  
C(Infill):C(Shell):C(Pattern)  
C(Layer):C(Shell):C(Pattern)  
C(Infill):C(Layer):C(Shell):C(Pattern) 0.000001  
Residual 0.000016  
Comprehensive ANOVA: Dye Penetrant (Type II)  
Shell thickness is the dominant main effect (F = 2135.9, p ≈ 1.84e-18), with Layer height also significant (F =  
33.55, p ≈ 2.75e-05). Several interactions (Infill×Shell, Infill×Pattern, Shell×Pattern, Infill×Layer×Shell) are  
highly significant, confirming that thicker shells amplify defect severity under certain infill‑pattern and layer  
combinations (Table 3). Operationally, lowering shell thickness and controlling the interacting settings reduces  
Stage‑2 surface indications.  
Table 3. Full ANOVA for Dye Penetrant Indication (sum of squares, F, and p-values).  
Term  
Sum Sq  
1.244  
F
p-value  
C(Infill)  
1.770  
2.020e-01  
2.752e-05  
1.840e-18  
6.446e-01  
1.079e-03  
7.627e-15  
6.851e-05  
2.023e-10  
2.108e-02  
1.201e-08  
C(Layer)  
23.581  
1501.383  
0.155  
33.547  
2135.898  
0.221  
C(Shell)  
C(Pattern)  
C(Infill):C(Layer)  
C(Infill):C(Shell)  
C(Layer):C(Shell)  
C(Infill):C(Pattern)  
C(Layer):C(Pattern)  
C(Shell):C(Pattern)  
11.127  
523.180  
19.924  
138.819  
4.598  
15.830  
744.287  
28.344  
197.487  
6.541  
79.097  
112.525  
Page 639  
www.rsisinternational.org  
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)  
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue XII December 2025  
C(Infill):C(Layer):C(Shell)  
C(Infill):C(Layer):C(Pattern)  
C(Infill):C(Shell):C(Pattern)  
C(Layer):C(Shell):C(Pattern)  
263.064  
63.253  
2.779  
374.241  
89.985  
3.953  
1.598e-12  
5.694e-08  
6.417e-02  
8.011e-09  
3.985e-10  
—
83.754  
119.150  
180.222  
—
C(Infill):C(Layer):C(Shell):C(Pattern) 126.683  
Residual  
11.247  
CONCLUSION  
Shell thickness is the primary lever for Stage‑2 defects; 1 mm shells reduce dye and cracking versus 2 mm.  
Stage‑1 permeability is controlled by higher‑order interactions—avoid single‑factor tuning. Pattern should be  
tuned jointly with shell, infill and layer height due to strong interactions. The two‑stage burnout workflow  
enables early identification of sensitive combinations and robust Stage‑2 inference via exact tests.  
ACKNOWLEDGEMENT  
The authors would like to thank the Faculty of Industrial and Manufacturing Technology and Engineering at  
Universiti Teknikal Malaysia Melaka (UTeM) for financial, educational, and technical support throughout this  
research.  
REFERENCES  
1. Campbell, J. (2015). Complete casting handbook: Metal casting processes, metallurgy, techniques and  
design (2nd ed.). Elsevier. https://doi.org/10.1016/C2013-0-16361-1  
2. Abisuga, O. A., Doran, K., & de Beer, D. (2022). Study of investment casting process for 3D printed  
jewellery  
design.  
MATEC  
Web  
of  
Conferences,  
370,  
04002.  
https://doi.org/10.1051/matecconf/202237004002  
3. Jones, S., & Yuan, C. (2003). Advances in shell moulding for investment casting. Journal of Materials  
Processing Technology, 135(2–3), 258–265. https://doi.org/10.1016/S0924-0136(02)00828-1  
4. Hague, R., Campbell, I., & Dickens, P. (2003). Implications on design of rapid manufacturing. Proc.  
IMechE Part C: J. Mech. Eng. Sci., 217(1), 25–30. https://doi.org/10.1243/095440603762554587  
5. Chhabra, M., & Singh, R. (2011). Rapid casting solutions: A review. Rapid Prototyping Journal, 17(5),  
328–350. https://doi.org/10.1108/13552541111156469  
6. Snelling, D., Li, Q., Meisel, N., Williams, C., & Batra, R. (2015). The effects of 3D printed molds on  
metal  
casting.  
Journal  
of  
Manufacturing  
Processes,  
19,  
1–7.  
https://doi.org/10.1016/j.jmapro.2015.05.002  
7. Ngo, T. D., Kashani, A., Imbalzano, G., Nguyen, K. T. Q., & Hui, D. (2018). Additive manufacturing  
(3D printing): A review of materials, methods, applications and challenges. Composites Part B:  
Engineering, 143, 172–196. https://doi.org/10.1016/j.compositesb.2018.02.012  
8. Bassoli, E., Gatto, A., Iuliano, L., & Violante, M. G. (2007). 3D printing technique applied to rapid  
casting. Rapid Prototyping Journal, 13(3), 148–155. https://doi.org/10.1108/13552540710750898  
9. Singh, R., & Sachdeva, A. (2011). Optimization of investment casting process parameters using DOE.  
Journal  
of  
Manufacturing  
Technology  
Management,  
22(4),  
468–484.  
https://doi.org/10.1108/17410381111126401  
10. Kumar, S., & Kruth, J. P. (2010). Composites by rapid prototyping technology. Materials & Design,  
31(2), 850–856. https://doi.org/10.1016/j.matdes.2009.06.045  
11. Yarlagadda, P. K. D. V., & Cheng, C. K. (2001). Thermal stresses in investment casting shells. Journal  
of Materials Processing Technology, 110(1), 1–8. https://doi.org/10.1016/S0924-0136(00)00804-6  
Page 640  
www.rsisinternational.org  
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)  
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue XII December 2025  
12. Ding, D., Pan, Z., Cuiuri, D., & Li, H. (2015). Wire-feed additive manufacturing of metal components.  
International  
Journal  
of  
Advanced  
Manufacturing  
Technology,  
81,  
465–481.  
https://doi.org/10.1007/s00170-015-7077-3  
13. Kumar, P., Ahuja, I. P. S., & Singh, R. (2012). Application of fusion deposition modelling for rapid  
investment casting. International Journal of Materials Engineering Innovation, 3(3), 204–221.  
https://doi.org/10.1504/IJMATEI.2012.050613  
14. Ferreira, J. C., Mateus, A., & Alves, N. (2007). Rapid tooling aided by reverse engineering. Journal of  
Materials Processing Technology, 183(2–3), 412–419. https://doi.org/10.1016/j.jmatprotec.2006.10.026  
15. Zhang, Y., Chou, K., & He, Y. (2016). Numerical investigation of gas escape in investment casting shells.  
Journal of Manufacturing Science and Engineering, 138(11), 111005. https://doi.org/10.1115/1.4033865  
16. Bassoli, E., Atzeni, E., Salmi, A., & Calignano, F. (2018). Additive manufacturing for investment casting  
applications. Procedia CIRP, 70, 3–8. https://doi.org/10.1016/j.procir.2018.02.014  
17. Singh, S., Ramakrishna, S., & Berto, F. (2020). 3D printing of polymer composites. Composites Part B:  
Engineering, 191, 107938. https://doi.org/10.1016/j.compositesb.2020.107938  
18. ASTM E1417/E1417M-16 (2016). Standard practice for liquid penetrant testing. ASTM International.  
https://doi.org/10.1520/E1417_E1417M-16  
19. Gibson, I., Rosen, D., & Stucker, B. (2015). Additive manufacturing technologies (2nd ed.). Springer.  
https://doi.org/10.1007/978-1-4939-2113-3  
20. Kalpakjian, S., & Schmid, S. (2014). Manufacturing engineering and technology (7th ed.). Pearson  
Education.  
Page 641  
www.rsisinternational.org