Multi-objective assessment of drop-weight impact in 3D-printed fiber-reinforced nylon composites

Abstract

This study investigates the drop-weight impact behavior of 3D-printed Nylon composites reinforced with continuous carbon, glass, and Kevlar fibers for impact-critical applications such as high-voltage marker balls. Using a Taguchi L27 orthogonal array, key factors, fiber orientation (0°, 45°, 90°), infill pattern (rectilinear, triangular, gyroid), fiber type, and infill percentage (30%, 40%, 50%)—were varied. Peak force, deformation, and energy absorption were measured per ASTM D7136, and optimized using entropy-weighted Multi-Criteria Decision-Making (MCDM) methods (RAWEC, TOPSIS, CoCoSo). The optimal configuration (0°/rectilinear/glass/50%) exhibited 2776.99 N peak force, 11.01 mm deformation, and 22.92 J energy absorption. SEM and EDS analysis of fracture surfaces revealed that glass fiber composites demonstrated ductile damage modes including fiber bridging and matrix flow, whereas carbon fiber composites failed via brittle fiber splitting and interfacial debonding. Post-impact evaluation showed that glass fiber composites retained up to 73% of their tensile strength and 67.9% of compressive strength. Specific Energy Absorption (SEA) analysis highlighted that Kevlar composites in triangle/gyroid infills with 30% density exhibited the highest energy-to-weight efficiency (up to 47.85 J·cm³/g). A fiber-type-specific optimization identified the best infill strategy for each reinforcement. The integrated framework presents a validated pathway for lightweight, impact-resistant composite design tailored to structural energy infrastructure.

Introduction

Mechanical structures operating in critical infrastructure face unpredictable, localized impacts, particularly overhead marker balls mounted on high-voltage lines. These components must withstand avian strikes and airborne debris to ensure uninterrupted functionality and structural safety. Hussain et al. examined how geometrical features influence the crashworthiness of GFRP crash boxes, ¹ while Kumar et al. enhanced CFRP laminates by incorporating pristine and functionalized graphene, linking improvements in impact resistance to fiber orientation and delamination control. ² Similarly, Koffi et al. showed that incorporating birch fibers into HDPE modified toughness and hardness, though with reduced impact strength. ³ Wang et al. introduced self-healing microcapsules and hollow glass fibers into carbon laminates, demonstrating repeated restoration of impact resistance. ⁴

Simulated evaluations by Hussain et al. closely aligned with physical tests for GFRP crash boxes, validating predictive deformation modeling. ⁵ Rehman et al. enhanced post-impact compressive and shear strength of carbon/epoxy composites by embedding PEEK microparticles, improving ductility and stress distribution. ⁶ Hybrid reinforcement was explored by Mirdehghan et al., where 3D-woven composites with hybrid yarns exhibited superior impact tolerance over conventional unidirectional laminates. ⁷ Liu et al. validated high energy absorption and recovery in GFRP-polypropylene cylinders, supported by accurate numerical models. ⁸ Layered structures such as epoxy-aluminum foam panels evaluated by Xin et al. maintained energy absorption capacity despite varied impact parameters, highlighting temporal sensitivity. ⁹ Self-reinforced PET-HDPE composites studied by Maniyamparambil et al. failed via fiber pull-out and matrix damage, with simulations confirming failure pathways. ¹⁰

In the 3D printing domain, Dou et al. reinforced PLA honeycomb structures with carbon fiber, achieving enhanced toughness and identifying critical zones of crack initiation. ¹¹ Narayanan et al. introduced inorganic fillers into GFRP, observing performance gains but also embrittlement at high filler loadings due to interfacial incompatibility. ¹² Mishra et al. optimized nylon-short carbon fiber composites, identifying 0.2 mm layer thickness and 85% infill as optimal for maximum impact absorption. SEM revealed interfacial gaps and fiber bridging as damage mitigation mechanisms. ¹³ Divakaran et al. used ZnO and Al₂O₃ ceramic fillers in Nylon 12, improving flow behavior and geometric fidelity in printed parts. ¹⁴ Alarifi et al. demonstrated superior mechanical performance in carbon fiber-nylon composites printed horizontally due to stress-aligned fiber orientation. ¹⁵ Papa et al. further highlighted the role of non-destructive evaluation for identifying subsurface damage in glass-fiber reinforced prints. ¹⁶

Zárybnická et al. balanced mechanical and flame resistance in Nylon 6.6 prints using fire retardants to counteract strength loss from reduced infill. ¹⁷ Mishra et al. showed that increased carbon fiber content led to predictable gains in strength and stiffness, closely aligned with modeling predictions. ¹⁸ Polyzos et al. accurately simulated crack paths in 3D-printed nylon-carbon fiber parts despite anisotropic print irregularities. ¹⁹ Prajapati et al. reinforced nylon-carbon fiber composites with fiberglass, achieving improved thermal and impact behavior, confirmed via flammability testing. ²⁰ Shi et al. improved composite strength to 88.18 MPa, with heat treatment elevating stiffness to 20.11 GPa via optimized fiber placement. ²¹ Jiang et al. reached 572.60 MPa strength and 18.49 GPa stiffness in printed composites through print parameter tuning, with fracture modes dominated by fiber pull-out. ²²

Hartomacioğlu et al. used Taguchi methods to boost PA6 impact strength to 8.9 kJ/m², pinpointing nozzle temperature as key. ²³ Beylergil et al. confirmed infill density as the most significant contributor to enhanced impact strength in carbon fiber–nylon prints. ²⁴ Ahlawat et al. applied genetic algorithms to optimize strength and wear resistance, achieving low wear rates and high tensile/compression values through AI-based fine-tuning. ²⁵ Chen et al. proposed a dual-stage method combining printing and post-processing to nearly double mechanical strengths in carbon-nylon parts. ²⁶ Alhaddad et al. used artificial neural networks and bee colony algorithms for optimizing fiber type and direction, showing strong predictive accuracy. ²⁷ Dubey et al. examined infill pattern influence, finding rectilinear best for tensile and triangular for flexural strength. ²⁸ Moradi et al. used response surface methodology to show that more contour layers enhance failure force, while thinner layers increase strength at the cost of ductility. ²⁹ El Essawi et al. optimized fiber angle and print layer positioning to maximize mechanical output, with fiber angle contributing 54.13% to behavior. ³⁰ Raffic et al. combined SWARA-CoCoSo methods with Adaboost to minimize dimensional deviations, achieving near-perfect predictive modeling at 0.999 R². ³¹ Rodríguez-Reyna et al. identified optimal print settings such as 100% infill, 0.14 mm height, ±45° angle yielding a 49% increase in UTS. ³² Abdallah et al. applied Taguchi-Grey analysis to PA12 MJF parts, improving multi-response performance by 0.141%. ³³ Safaei et al. reviewed the field, highlighting shrinkage, anisotropy, and process inconsistency as barriers to real-world adoption. ³⁴

Recent advances in the optimization of fused deposition modeling (FDM) and fused filament fabrication (FFF) processes for nylon-based composites have significantly deepened the understanding of process-structure-property relationships, particularly under mechanical loading scenarios relevant to structural applications. Maniyamparambil et al. investigated the drop-weight impact behavior of PET-HDPE/HDPE self-reinforced composites and identified that insufficient interfacial bonding led to extensive fiber pull-out and distinct subsurface failure modes, underscoring the importance of matrix-fiber compatibility in energy dissipation. ¹⁰ In a related study, Jiang et al. explored continuous carbon fiber-reinforced PA12 composites fabricated via 3D printing and reported that optimized printing parameters yielded a tensile strength of up to 572.60 MPa, with fracture modes including toothed failures, transverse microcracks, unimpregnated regions, and fiber pull-out, highlighting the role of interfacial wetting and fiber alignment. ²²

Tanabi et al. studied the anisotropic mechanical behavior of short carbon fiber-reinforced PET processed by FDM, revealing that build orientation significantly influenced shear response and yielded a tensile strength improvement of up to 230% over ABS-based systems. ³⁵ Nagendra et al. applied Taguchi optimization to nylon-aramid composites and determined that a combination of 0.2 mm layer thickness, 280°C extrusion temperature, 90° raster angle, tetrahedral infill, and 90% density minimized surface roughness with a combined error of 2.45%, indicating the sensitivity of surface quality to process control. ³⁶

Further parametric exploration by Moradi et al. using response surface methodology (RSM) showed that while thinner layers improved ultimate strength by reducing cooling gradients, excessive layer thinning reduced elongation due to premature solidification. Their follow-up study on continuous carbon fiber-reinforced nylon revealed that extruder temperature (240–250°C) and print speed (65–75 mm/s) substantially affected tensile modulus and elongation, with experimental values deviating less than 10% from RSM predictions. ²⁹ Ramesh et al. employed a Taguchi L9 design and ANOVA to assess the effect of fill density, layer height, and print speed on the tensile, flexural, and impact strengths of nylon prints, also noting significant improvements in Shore D hardness. ³⁷

Yankin et al. demonstrated that an octet infill pattern with 100% fill density and 65 mm/s print speed maximized tensile strength, outperforming traditional infill structures, ³⁸ while Mushtaq et al. introduced laser polishing as a viable post-processing method, achieving a 20.2% reduction in surface roughness and modest increases in tensile and flexural strength with minimal energy consumption. ³⁹ In terms of material modifications, Divakaran et al. integrated ZnO and Al₂O₃ ceramic fillers into Nylon 12 to enhance thermal stability and flowability, resulting in improved dimensional fidelity and tensile performance suitable for high-precision applications. ¹⁴ Finally, Kaushik et al. employed a Face-Centered Central Composite Design (FCCD) combined with Multi-Objective Genetic Algorithm-Artificial Neural Network (MOGA-ANN) optimization to identify an optimal 49.26° fiber angle, rectangular infill, and fiberglass reinforcement combination, achieving peak tensile strength of 37.87 MPa, elongation of 51.05%, and minimal surface roughness of 1.43 µm. ⁴⁰ These studies collectively establish the critical role of process parameter optimization, reinforcement architecture, and material hybridization in tailoring the mechanical performance of FDM-printed nylon composites. However, a unified experimental–computational framework integrating impact performance, residual strength, and multi-criteria decision analysis under real-world loading conditions remains underexplored, highlighting the novelty and significance of the present work.

Despite significant research, a unified multi-objective framework tailored to continuous fiber-reinforced nylon under impact remains absent. Existing studies often explore materials or simulations in isolation, without fully characterizing nonlinear interactions across print parameters. The synergistic effects of infill topology, fiber orientation, and reinforcement content remain under-quantified, especially in high-velocity impact scenarios relevant to critical infrastructure components.

This study addresses the gap by integrating design of experiments, entropy-weighted decision-making, and performance validation into a single optimization protocol. Employing a Taguchi L27 orthogonal matrix, key factors such as fiber type, orientation, infill pattern, and density are systematically evaluated. The optimization is executed using several multi criteria decision making methods, augmented with entropy-based objective weights, to minimize subjective bias. The research specifically aims to: (1) characterize drop-weight impact behavior of 3D-printed nylon composites; (2) quantify main and interaction effects using Taguchi methodology; (3) apply entropy-MCDM for balanced optimization across multiple objectives; and (4) interpret mechanical response via post-impact damage mapping. By establishing a predictive, validated, and statistically optimized pathway for material design, this study contributes a first-of-its-kind framework for impact-resistant, 3D-printed fiber-reinforced composites. The methodology is particularly suited for applications in energy infrastructure, including high-voltage marker ball systems, where weight, durability, and dynamic reliability are critical.

Materials and methods

Materials & composite fabrication

The Nylon-fiber composite specimens were produced using the Markforged Mark Two 3D printer, which, according to the ISO/ASTM 52900:2021 standard, ⁴¹ is classified under the Material Extrusion (MEX) category. Simon et al. have adopted this standardized terminology in their recent study, ⁴² further reinforcing its relevance to additive manufacturing taxonomy. This process involves the selective deposition of thermoplastic feedstock through a heated nozzle, where the material solidifies after extrusion to form successive layers. In research and industrial practice, this technique is often referred to as Fused Deposition Modeling (FDM) or Fused Filament Fabrication (FFF); however, to ensure consistency with international standards and to eliminate ambiguity in terminology, the unified ISO designation MEX is adopted in this study, with the legacy terms mentioned parenthetically for continuity with prior literature. The Mark Two printer represents an advanced variant of MEX by incorporating Continuous Filament Fabrication (CFF), in which continuous reinforcement fibers (carbon, glass, or Kevlar) are co-deposited alongside the polymeric matrix (Onyx, a Nylon 6-based filament with short carbon fibers). This dual-feed capability places the system within the broader MEX framework while extending its functionality beyond conventional FFF/FDM platforms, thereby enabling direct integration of continuous fibers within the printed thermoplastic matrix. By explicitly aligning the terminology with ISO/ASTM 52900:2021, this work establishes clarity in the classification of the additive manufacturing process while preserving a link to the extensive body of legacy literature that employs FDM/FFF nomenclature.

Mark Two is equipped with a dual nozzle, through which a composite filament comprising a Nylon with chopped carbon fiber is processed with three continuous fibers such as carbon, glass, and Kevlar. The Eiger slicing software, developed by Markforged, was used to configure the fiber deposition, reinforcement strategy, and process parameters, ensuring precision in the fabrication. Figure 1 represents the Onyx/fiber composite filaments and experimental setup used in the study.

Figure 1.

Onyx/fiber composite filaments and 3D printer setup.

Key printing parameters were predefined in the Eiger software to optimize mechanical performance. The extrusion temperature for Onyx was maintained at 275°C and the fiber extrusion was fixed at 250°C, facilitating uniform filament deposition and ensuring strong interlayer adhesion. The specified extrusion temperatures, 275°C for Onyx filament and 250°C for continuous fiber reinforcement, ensure optimal viscosity, effective inter-layer bonding, and preservation of fiber integrity during the printing process. The printing speed varied between 30–60 mm/s depending on the fiber content and infill density, in order to maintain consistent material flow and prevent nozzle clogging during continuous fiber deposition. Precise temperature control minimizes thermal degradation and enhances mechanical properties, critical for achieving consistent quality and performance in composite parts. The printing speed varied between 30 mm/s, depending on the fiber content and infill density, to maintain consistent material flow. The layer height was configured to 100 µm, with finer settings improving surface finish and mechanical integrity. Prior to each print, the printer performed an automatic calibration routine, adjusting the build plate alignment and extrusion settings to ensure precise layer stacking.

The raw materials utilized in this study included Onyx, a Nylon 6-based polymer with chopped carbon fibers, reinforced with three types of continuous fibers such as carbon, glass, and Kevlar. These composite filaments were supplied by Markforged, USA. The carbon fiber-reinforced Onyx featured a continuous fiber filament diameter of 1.75 mm, with individual carbon fibers measuring approximately 7 µm in diameter. The diameter of glass fiber and Kevlar fibers in the filaments were noted as 10 and 12 µm respectively. No specific grade is mentioned for carbon fiber, whereas Kevlar fiber is an aramid fiber, and glass fiber high strength high temperature glass fiber as per the information from filament supplier. All the fiber dimensions are noted from the data sheet of supplier. The infill pattern was systematically varied, including triangular, rectilinear, and gyroid structures, while the infill density ranged between 20% and 100%, providing tunable stiffness and weight distribution.

The uniaxial mechanical properties of the continuous fiber filaments used in this study are based on the material datasheets provided by the manufacturer (Markforged). These properties were measured using standard ASTM test methods on specimens reinforced with unidirectional fiber (0° plies) and serve as representative values for performance in tension, compression, and flexure. The following Table 1, summarizes the key uniaxial properties:

Table 1.

Representative uniaxial mechanical properties of continuous fiber filaments used in CFF-printed composites.

Property	Carbon fiber	Kevlar® fiber	Fiberglass (HSHT)
Tensile Strength (MPa)	800	610	600
Tensile Modulus (GPa)	60	27	21
Tensile Strain at Break (%)	1.5	2.7	3.9
Flexural Strength (MPa)	540	240	420
Flexural Modulus (GPa)	51	26	21
Compressive Strength (MPa)	420	130	216
Compressive Modulus (GPa)	62	25	21
Density (g/cm3)	1.4	1.2	1.5

It is important to note that the Eiger slicing software, by default, places four solid matrix layers beneath and one to four solid layers above each continuous fiber layer. This toolpath strategy ensures structural stability and bonding but also alters the effective internal structure compared to the user-defined infill pattern and infill percentage. As a result, the influence of infill pattern and density on the mechanical performance is somewhat masked near the fiber layers, particularly in specimens with higher fiber volume or lower nominal infill percentages. This inherent feature must be considered when analyzing the impact of infill design on drop-weight performance, as it contributes to a baseline structural reinforcement not captured by the nominal infill settings alone.

To eliminate ambiguity regarding fiber orientation and build layout, it is important to clarify that the fiber orientations (0°, 45°, and 90°) in this study are defined with respect to the in-plane directions (X and Y) of the specimen as per ASTM D7136 standards. Specifically, the 0° orientation corresponds to fibers aligned along the 150 mm length (X-axis), 90° along the 100 mm width (Y-axis), and 45° denotes fibers placed diagonally across the X–Y plane. The drop-weight impact is applied normal to this plane, i.e., along the Z-direction, representing a through-thickness impact at the center of the specimen. Continuous fibers are deposited in-plane and remain confined to a designated central layer during fabrication. The slicing software (Eiger by Markforged) automatically surrounds this fiber layer with four solid Onyx matrix layers below and one to four matrix layers above, depending on part thickness. The fiber is not deposited along the Z-direction; any visual distortion of aspect ratio in Figure 2 arises from CAD rendering scale and not from actual fiber deposition behavior. This clarification ensures better understanding of the mechanical response pathways during impact and the relevance of fiber alignment in governing damage resistance.

Figure 2.

Details of fiber tool path during 3D printing.

Experimental design

The experimental design was structured using the Taguchi L27 orthogonal array, which effectively accommodates four factors at three levels each as shown in Table 2. This approach enables a systematic parameter variation while minimizing the number of experimental runs, ensuring an efficient investigation of the mechanical performance of fiber-reinforced Nylon composites for marker ball applications in electrical transmission lines. The selected control factors include fiber orientation, infill pattern, fiber reinforcement type, and infill percentage, all of which significantly influence the impact resistance and structural integrity of the fabricated components as mentioned in Table 3.

Table 2.

Factors and levels.

Factors	Levels
Fiber orientation	0, 90 and 45 degrees
Infill pattern	Triangle, Rectilinear, Gyroid
Fiber in Nylon	Carbon, Glass, Kevlar
Infill %	30, 40, 50%

Table 3.

Design of experiment along with fiber volume fraction.

Exp. no.	Fiber orientation (degree)	Infill pattern	Fiber in nylon	Infill%	FVF (%)
1	0	Triangle	Carbon	30	0.183
2	0	Triangle	Carbon	40	0.147
3	0	Triangle	Carbon	50	0.123
4	0	Rectilinear	Glass	30	0.173
5	0	Rectilinear	Glass	40	0.14
6	0	Rectilinear	Glass	50	0.118
7	0	Gyroid	Kevlar	30	0.165
8	0	Gyroid	Kevlar	40	0.133
9	0	Gyroid	Kevlar	50	0.112
10	45	Triangle	Glass	30	0.156
11	45	Triangle	Glass	40	0.126
12	45	Triangle	Glass	50	0.107
13	45	Rectilinear	Kevlar	30	0.146
14	45	Rectilinear	Kevlar	40	0.119
15	45	Rectilinear	Kevlar	50	0.1
16	45	Gyroid	Carbon	30	0.165
17	45	Gyroid	Carbon	40	0.133
18	45	Gyroid	Carbon	50	0.112
19	90	Triangle	Kevlar	30	0.129
20	90	Triangle	Kevlar	40	0.105
21	90	Triangle	Kevlar	50	0.089
22	90	Rectilinear	Carbon	30	0.156
23	90	Rectilinear	Carbon	40	0.126
24	90	Rectilinear	Carbon	50	0.107
25	90	Gyroid	Glass	30	0.147
26	90	Gyroid	Glass	40	0.119
27	90	Gyroid	Glass	50	0.1

The fibre volume fraction (FVF) for each configuration was analytically estimated based on the programmed continuous fibre toolpath generated by the Eiger slicing software. Each specimen, with nominal dimensions of 150 × 100 × 5 mm, was printed with a layer configuration consisting of four solid Onyx base layers, one continuous fibre layer positioned at mid-thickness, and one to four additional solid Onyx layers above the fibre, depending on the total height and layer thickness. The fibre layer comprised seven concentric rings with a known strand width (0.40 mm), height (0.125 mm), and an average ring length of approximately 145 mm, resulting in a nominal fibre volume of around 0.050 cm³ per specimen. This base volume was refined using correction factors for orientation efficiency (1.00 for 0°, 0.90 for 45°, and 0.85 for 90°), which account for reduced effective fibre length at angled orientations, and packing efficiency factors specific to the fibre type (1.00 for carbon, 0.95 for glass, and 0.90 for Kevlar), reflecting differences in compaction and deposition behaviour. The denominator for FVF calculation was defined as the total solid material volume in each specimen, comprising the face layers (fixed at 7.5 cm³) and the matrix content within the 4.5 mm-thick core region, scaled by the specified infill percentage (30 %, 40 %, or 50 %). Since the fibre mass remained constant across all builds, higher infill densities led to increased matrix volume and a corresponding reduction in the relative fibre contribution. The final fibre volume fractions are reported in the last column of Table 3.

Fiber orientation was considered at 0°, 45°, and 90° to examine the anisotropic mechanical response of the composites under impact conditions. Axially aligned fibers (0°) provide maximum tensile strength in the principal loading direction, whereas transverse (90°) reinforcement enhances resistance against transverse shear forces, a critical requirement for impact-prone components. The ±45° orientation was included to assess its contribution to shear resistance and energy absorption, as marker balls are subjected to multidirectional loading during installation and operation. The infill pattern was varied across triangular, rectilinear, and gyroid structures, each influencing the mechanical performance uniquely. The triangular pattern offers high rigidity and uniform stress distribution, making it suitable for lightweight, load-bearing applications. The rectilinear pattern, with its orthogonal raster deposition, provides improved layer adhesion and load-bearing capacity. The gyroid structure, a mathematically optimized lattice, was included due to its superior energy dissipation and impact resistance, attributes crucial for absorbing high-velocity bird strikes and environmental debris collisions.

The fiber reinforcement type was varied across carbon, glass, and Kevlar fibers embedded in a Nylon 6-based matrix. Carbon fiber reinforcement was selected due to its high stiffness-to-weight ratio and superior structural stability. Glass fiber, with its higher strain-to-failure, provides enhanced toughness and impact resistance. Kevlar fiber, known for its high energy absorption and abrasion resistance, was chosen to evaluate its potential for impact mitigation in marker ball applications. The infill percentage was adjusted to 30%, 40%, and 50% to optimize weight-to-strength ratios while ensuring structural durability. A 30% infill was selected to evaluate lightweight, cost-effective designs, while 50% infill was chosen to investigate the feasibility of a more rigid configuration. The 40% infill represents a balance between weight reduction and impact strength, ensuring durability without excessive material usage.

The printed composite specimens demonstrated high structural integrity and dimensional consistency, characteristic of the closed-loop thermomechanical control inherent to the Markforged Continuous Filament Fabrication (CFF) system. The Mark Two printer utilized in this study employs dual extrusion heads for precise placement of Onyx matrix and continuous fiber and integrates in-situ calibration routines that minimize thermal distortion and layer misalignment. As a result, all fabricated specimens exhibited sharp edge fidelity, minimal warping, and consistent fiber-to-matrix registration throughout the 150 mm × 100 mm × 5 mm build geometry.

Surface inspection revealed excellent finish quality, free of delamination or excessive porosity along the fiber planes (as can be witnessed in an after impact sample showed in Figure 3). This is attributed to the use of solid Onyx face layers (4 below, 1–4 above), which encapsulate the central continuous fiber region and suppress interlaminar defects. Moreover, the fixed 100 µm layer resolution and optimized fiber compaction during deposition ensured homogeneous layer bonding, particularly in critical regions subjected to out-of-plane impact.

Figure 3.

Drop weight impact test setup and test sample.

To assess print reliability, dimensional variation across 27 specimens was measured using digital callipers with ±0.01 mm resolution. All samples remained within ±0.15 mm tolerance from nominal dimensions, with negligible edge defects. Additionally, no signs of nozzle drag, under-extrusion, or void bridging were observed in cross-sectional cuts. These factors collectively indicate that the printed specimens achieved production-grade quality suitable for experimental testing, enabling consistent evaluation of mechanical behavior across fiber types, orientations, and architectural configurations.

Following fabrication, all 27 printed composite specimens were weighed using a calibrated analytical balance with ±0.01 g resolution. This step enabled precise quantification of the as-built mass for each combination of fiber type, fiber orientation, and infill pattern. However, a subsequent one-way ANOVA revealed that infill percentage was the overwhelmingly dominant factor, contributing to more than 90% of the total variance in specimen mass (p < 0.001). In contrast, fiber orientation, infill pattern, and even fiber type—despite minor differences in theoretical densities (Kevlar ≈ 1.20 g/cm³, Carbon ≈ 1.40 g/cm³, Glass ≈ 1.50 g/cm³)—accounted collectively for less than 5% of the variation. This confirms that material deposition volume, governed primarily by infill percentage and to a lesser extent by fiber type, dictates the bulk mass of the printed components.

Given this hierarchy of influence and recognizing that only one specimen was printed per unique configuration, porosity estimation was conducted by grouping specimens according to infill percentage and fiber type, where three specimens share identical material composition and nominal density. This clubbed approach allows for statistically valid and experimentally feasible assessment of porosity while avoiding over-interpretation of micro-level architecture-induced variations.

Following fabrication, all 27 printed composite specimens were weighed using a calibrated analytical balance with ±0.01 g resolution. This step enabled precise quantification of the as-built mass for each combination of fiber type, fiber orientation, and infill pattern. However, a subsequent one-way ANOVA revealed that infill percentage was the overwhelmingly dominant factor, contributing to more than 90% of the total variance in specimen mass (p < 0.001). In contrast, fiber orientation, infill pattern, and even fiber type—despite minor differences in theoretical densities (Kevlar ≈ 1.20 g/cm³, Carbon ≈ 1.40 g/cm³, Glass ≈ 1.50 g/cm³), accounted collectively for less than 5% of the variation. This confirms that material deposition volume, governed primarily by infill percentage and to a lesser extent by fiber type, dictates the bulk mass of the printed components.

Given this hierarchy of influence and recognizing that only one specimen was printed per unique configuration, porosity estimation was conducted by grouping specimens according to infill percentage and fiber type, where three specimens share identical material composition and nominal structural architecture. This clubbed approach allows for statistically valid and experimentally feasible assessment of porosity while avoiding over-interpretation of configuration-specific variations caused by fiber orientation or infill pattern.

Porosity was estimated by comparing the experimentally measured density (obtained from average mass and fixed specimen volume of 75 cm³) with the theoretical density derived using a rule-of-mixtures approach. The theoretical density assumed a 30% continuous fiber volume fraction combined with infill-dependent matrix content. This method reflects the material architecture defined by the slicing software (Eiger), where continuous fiber placement and matrix deposition are deterministically controlled. Estimated porosity values provide a representative measure of internal voids, accounting for inter-bead gaps, fiber–matrix discontinuities, and infill-induced heterogeneities.

The results summarized below in Table 4, include the average mass, standard deviation, measured density, theoretical density, and corresponding porosity for each grouped combination. As expected, porosity decreased with increasing infill percentage due to improved material consolidation. Glass fiber–reinforced composites exhibited marginally higher porosity at each infill level compared to Kevlar and carbon variants, likely due to the stiffness and geometry of the glass filaments, which may limit matrix packing efficiency.

Table 4.

Average mass, and estimated porosity of 3D-printed composite specimens grouped by infill percentage and fiber type.

Infill (%)	Fiber type	Mean mass (g)	Std. dev. (g)	Measured density (g/cm3)	Theoretical density (g/cm3)	Porosity (%)
30	Carbon	38.4	0.8	0.512	0.672	23.81
	Glass	38.9	0.9	0.519	0.702	26.06
	Kevlar	35.6	0.7	0.475	0.612	22.38
40	Carbon	44.8	1	0.597	0.756	21.03
	Glass	45.5	1.1	0.607	0.786	22.77
	Kevlar	41.5	0.8	0.553	0.696	20.63
50	Carbon	51.5	1.2	0.687	0.84	18.21
	Glass	52.3	1.3	0.697	0.87	19.89
	Kevlar	47.2	1.1	0.629	0.78	19.36

Drop-weight impact testing

The drop-weight impact testing of 3D-printed fiber-reinforced Nylon composites was conducted using the Fractovis Plus Instrumented Impact Tester, configured in accordance with ASTM D7136 standards as shown in Figure 3. This test method was employed to simulate low-velocity impact conditions that marker balls encounter in real-world applications, such as bird strikes and airborne debris collisions. The Fractovis Plus system utilizes a guided mass drop mechanism with a rigidly clamped fixture unit, ensuring precise impact localization and repeatable test conditions. The test setup included a variable mass system, with an optimized impact mass of 3 kg and a drop height of 816 mm, achieving an impact velocity of 4 m/s with impact energy 23.5 J. These parameters were specifically selected to replicate the dynamic impact conditions experienced by marker balls in high-voltage transmission lines, such as bird strikes and airborne debris collisions. The test was conducted at temperature 30°C) to reflect typical operational environments, eliminating external thermal influences on the material behavior.

All drop-weight impact experiments in this study were conducted on specimens fabricated to precise and standardized in-plane dimensions of 150 mm × 100 mm with a constant thickness of 5 mm, as mandated by ASTM D7136. These geometries were uniformly maintained across all 27 experimental conditions to isolate the effects of architectural variables such as fiber orientation, infill pattern, and reinforcement type. The boundary conditions during impact testing were carefully controlled using a rigid clamping fixture with a central circular cutout designed to support the specimen along its periphery while allowing unobstructed central deformation. This configuration approximates a quasi-static simply supported boundary condition with edge constraint and minimal frictional interference, thereby mimicking real-world loading where out-of-plane displacement governs structural failure without inducing unrealistic stress concentrations. The striker head, hemispherical in shape and 16 mm in diameter, delivered an identical impact energy of 23.5 J (via 3 kg mass dropped from 816 mm height) to each specimen, ensuring homogenous impulse characteristics across all trials. Prior to testing, specimen alignment was verified both visually and mechanically to ensure concentric impact and symmetrical support loading. No modifications to the test rig, specimen clamping, or impactor configuration were made between trials. Moreover, the automated calibration sequence of the test system guaranteed consistent zeroing of displacement sensors and synchronization of force acquisition, thereby eliminating experimental variability. These measures collectively ensured that geometric and boundary-related parameters remained invariant throughout the entire experimental matrix, allowing for direct attribution of observed mechanical behavior to the intended variations in internal structural design.

Three key parameters were recorded to assess the impact resistance and energy dissipation characteristics of the composite specimens. The peak force (N), measured using piezoelectric force transducers, provided insights into the stiffness and load-bearing capacity of the fiber-reinforced Nylon composites. The energy absorbed (J) was quantified by integrating the force-displacement curve, determining the damage resistance and impact toughness of the material. The maximum deflection (mm) was monitored through high-precision displacement sensors, characterizing the out-of-plane deformation and elastic recovery of the structure post-impact. The data of peak force (N), energy absorbed (J) and deflection were recorded with respect to milli second during the drop weight impact. These parameters provided a comprehensive understanding of the damage mechanisms and structural integrity of 3D-printed composites, ensuring their suitability for impact-critical applications in power transmission infrastructure.

Results and discussion

Experimental outcomes from drop-weight impact testing

The key experimental results from drop-weight impact tests are summarized in Table 5, including peak impact force (N), maximum deformation (mm), and energy absorption (J) for all composite configurations (Exp. 1 to Exp. 27). These metrics provide essential insights into the impact resistance performance and structural integrity variations due to different fiber types, fiber orientations, infill patterns, and densities employed in the Nylon-fiber composites.

Table 5.

Results of drop weight impact test.

Exp. no.	Peak force (N)	Deformation (mm)	Energy (J)
1	1970.147	12.363	14.960
2	1894.9	11.956	14.94
3	1947.69	11.933	14.952
4	2001.996	12.753	14.967
5	2132.659	12.528	14.963
6	2776.992	11.006	22.916
7	2287.007	14.580	22.789
8	1847.245	14.698	22.678
9	1970.966	15.761	22.708
10	2278.432	12.254	22.722
11	2407.053	12.011	22.715
12	2142.461	13.364	22.640
13	2147.361	13.088	22.746
14	2021.19	13.590	22.760
15	2462.176	11.813	22.709
16	1844.795	38.422	16.441
17	2095.912	33.049	18.698
18	1909.718	14.750	22.908
19	2256.383	13.376	22.869
20	1875.419	31.838	19.581
21	1706.374	21.064	23.025
22	2313.956	18.595	22.998
23	2485.451	13.705	22.878
24	2280.882	25.677	21.193
25	2121.637	13.187	22.863
26	2420.528	11.755	22.822
27	2524.649	11.494	22.815

Figure 4 presents the impact force-time behavior of 3D-printed Nylon-fiber composites subjected to drop-weight impact tests across different experimental conditions. Figure 4(a), corresponding to experimental runs E1 through E10, illustrates distinct variations in impact response reflective of fiber type, orientation, and infill parameters. Specifically, experimental run E6 distinctly peaks with a maximum force near 2777 N, highlighting the inherent stiffness attributed predominantly to the glass fibers embedded within the rectilinear infill pattern at 50% density and aligned at 0°. The rapid, sharp ascent in the impact force curve underscores robust structural integrity, allowing effective resistance to instantaneous loading. Contrarily, experimental run E2 displays a noticeably lower peak force around 1895 N, featuring a more gradual ascending trajectory. This behavior signifies enhanced energy dissipation capacity, indicating that carbon fibers embedded within a triangular infill configuration at lower density conditions offer a comparatively ductile response by absorbing the impact energy progressively through fiber-matrix debonding and matrix microcracking rather than abrupt fracture.

Figure 4.

Impact force versus time (a) Exp. 1 to 10 (b) Exp. 11 to 20 (c) Exp. 21 to 27.

Figure 4(b), showcasing experimental runs E11 through E20, reveals a broader spectrum of mechanical responses reflective of diverse composite configurations. Within this sub-image, experimental run E16 exhibits the lowest peak force approximately 1845 N, coupled with a notably extended deformation curve characterized by sustained fluctuations beyond peak load. The extended deformation, implying significant plastic deformation and internal fiber pull-out phenomena, suggests an intricate combination of matrix yielding, delamination, and fiber-matrix interface debonding, a characteristic amplified by the carbon fibers’ intrinsic brittleness combined with a gyroid infill pattern at a lower infill percentage. Contrastingly, experimental run E15, featuring Kevlar fibers in a rectilinear infill pattern with higher density, demonstrates a significantly greater peak impact force (∼2462 N) and a comparatively shorter duration to the peak. This indicates a robust mechanical interlocking and efficient stress distribution within the composite structure, emphasizing Kevlar’s superior strain-to-failure characteristics and effective fiber bridging during impact events.

Figure 4(c), representing experimental runs E21 to E27, further elucidates variations in impact resistance, vividly illustrating the interaction between fiber orientation and infill pattern. Experimental run E22 manifests the highest peak force within this subset, approximately 2314 N, exhibiting a steep rise and subsequent pronounced decline indicative of brittle failure mechanisms predominantly characterized by rapid fiber breakage and immediate load transfer failure within carbon fiber rectilinear structures oriented at 90°. Alternatively, experimental run E24 showcases a significantly lower peak force (∼2281 N) accompanied by a pronounced, gradual decrease in force, indicative of complex internal damage mechanisms involving extensive matrix microfracturing, fiber bridging, and pronounced plastic deformation. Such behavior underscores the nuanced mechanical interplay between the composite’s internal architecture and the fiber orientation, emphasizing how alterations in these parameters dramatically influence the composite’s damage-tolerance capability under dynamic loading conditions.

Figure 5 illustrates the deformation-time response of the 3D-printed Nylon-fiber composites when subjected to drop-weight impact conditions, categorized into three subsets for experimental runs E1–E10 (Figure 5(a)), E11–E20 (Figure 5(b)), and E21–E27 (Figure 5(c)). The deformation profiles present critical insights into the elastic-plastic behavior and the damage tolerance characteristics of these composites during impact loading.

Figure 5.

Deformation versus time (a) Exp. 1 to 10 (b) Exp. 11 to 20 (c) Exp. 21 to 27.

In Figure 5(a), all experimental configurations exhibit a gradual increase in deformation, initially following an elastic regime marked by linear deformation with respect to time. After this region, deformation rises sharply, peaking at approximately 12 to 16 mm around 6 to 8 ms post-impact. Subsequently, there is a noticeable reduction, indicating a partial elastic recovery post-peak deformation. Notably, experimental runs such as E9 display the highest deformation (∼16 mm), reflecting pronounced ductility likely influenced by the gyroid infill pattern and Kevlar fiber inclusion. This pattern fosters energy dissipation via fiber sliding, delamination, and matrix yielding. Conversely, experimental conditions with lower peak deformation (e.g., E3 and E6) reveal stiffer composite behavior, possibly attributed to the intrinsic rigidity offered by carbon and glass fibers combined with rectilinear or triangular infill patterns, demonstrating limited plastic deformation and rapid elastic recovery.

In Figure 5(b), the deformation behavior exhibits a significant divergence, with some samples showing a notably continuous and progressive increase in deformation, reaching extreme values such as 38.4 mm for sample E16. This prolonged deformation progression, without a distinct reduction post-peak, highlights considerable irreversible plastic deformation, indicative of catastrophic internal damage mechanisms such as extensive fiber pull-out, interface debonding, and severe delamination. Such pronounced deformation in sample E16 is associated with the interplay of a gyroid infill structure, 45° fiber orientation, and carbon fiber reinforcement, which collectively enhance structural ductility and significantly augment deformation capacity. In stark contrast, specimens such as E11, E12, and E15 manifest moderate peak deformation (∼12–14 mm), followed by clear elastic recovery phases, underlining a balanced configuration where the reinforcement of Kevlar fibers and structured rectilinear infill effectively confines damage propagation, preserving residual structural integrity.

Figure 5(c) presents intermediate deformation behaviours with experimental conditions illustrating a mixed-mode mechanical response. Here, specimens E24 and E25 demonstrate continuous rising deformation, reaching approximately 25–27 mm, signifying extensive plastic deformation accompanied by progressive internal damage processes such as fiber rupture, matrix cracking, and microstructural disintegration, predominantly arising due to rectilinear and gyroid structures coupled with specific fiber orientations (90°) that are less optimal for transverse load-bearing capacities. Conversely, sample E27 exhibits a limited deformation (∼12 mm), indicative of superior impact resilience and reduced structural compromise. The lower deformation level observed in this configuration is directly attributable to the integration of glass fibers in the gyroid infill pattern with higher density, effectively facilitating rapid stress redistribution, reduced damage propagation, and increased structural stiffness during dynamic loading.

The temporal progression of impact energy absorbed by the 3D-printed Nylon-fiber composites under drop-weight impact tests is methodically presented in Figure 6, segmented into three distinctive subsets representing experiments E1–E10 (Figure 6(a)), E11–E20 (Figure 6(b)), and E21–E27 (Figure 6(c)). Through these graphical representations, the energy absorption characteristics across various composite configurations are elucidated, emphasizing the critical interplay between fiber type, infill geometry, fiber orientation, and density.

Figure 6.

Impact energy versus time (a) Exp. 1 to 10 (b) Exp. 11 to 20 (c) Exp. 21 to 27.

In Figure 6(a), experiments E1 through E10 illustrate diverse trends in impact energy absorption, swiftly ascending within the initial 2–6 milliseconds before stabilizing at peak levels. Specifically, samples such as E6, E7, E8, E9, and E10 achieve notably higher energy absorption nearing 22–23 J, indicative of their superior capability to dissipate impact-induced energy. The presence of Kevlar and glass fibers combined with gyroid, and rectilinear infill geometries substantially contribute to such heightened energy absorption profiles. The unique fiber-fiber and fiber-matrix interactions facilitate extended deformation modes, including matrix yielding and fiber pull-out phenomena, thus significantly enhancing the total energy absorbed. Conversely, samples such as E1, E2, and E3 exhibit a reduced energy absorption capability (∼14–15 J), highlighting the comparatively limited ductility and structural resilience offered by the carbon fiber-triangular infill configurations with lower densities. The limited deformation mechanisms, marked predominantly by brittle fracture and minimal fiber deformation, underscore the inferior capacity of these configurations to dissipate impact energies effectively.

Transitioning to Figure 6(b), the absorbed energy trends reveal intriguing variations among experimental runs E11–E20. Notably, most samples rapidly absorb energy within the initial milliseconds, subsequently achieving stable plateau phases between 20–23 J. Nevertheless, considerable variation is observed in specific configurations, particularly sample E16, which stabilizes at approximately 16 J, substantially lower than its counterparts. This behavior reflects a significant limitation in internal structural energy dissipation mechanisms, primarily driven by early fracture events and extensive fiber debonding, specifically arising from carbon fibers embedded in a gyroid structure at moderate density levels and 45° orientation. Alternatively, samples E11, E12, E13, and E15 consistently reach higher energy absorption values (22–23 J), primarily attributed to the superior damage mitigation capacity offered by glass and Kevlar fibers arranged in structured rectilinear and triangular patterns. Such configurations ensure effective load redistribution through fiber bridging and delayed delamination onset, consequently fostering extended energy absorption potential and improved impact tolerance.

The experimental subset represented in Figure 6(c), covering samples E21 to E27, demonstrates relatively uniform energy absorption behaviors, yet subtle differences remain evident, influenced by the material architecture and orientation. Samples such as E21 and E22 reach their energy absorption peaks swiftly, maintaining stable values around 23 J, implying robust energy dissipation predominantly facilitated by Kevlar fibers integrated with triangular and rectilinear configurations. Such setups encourage enhanced fiber stretching, extensive matrix deformation, and delayed crack propagation. Conversely, samples like E24 and E25 exhibit a slightly lower energy plateau (around 21 J), indicative of moderate reductions in the structural capacity to dissipate impact energy effectively. This modest decrease primarily arises from inherent brittleness and comparatively limited damage mechanisms typical of carbon and glass fiber composites oriented transversely (90°). Nonetheless, these composites still demonstrate commendable impact tolerance suitable for specific structural applications where controlled and predictable energy dissipation is preferred over maximum ductility.

Comparative evaluation with recent literature on nylon-based composite systems under drop-weight impact highlights the distinct mechanical advantages achieved in the present study through architectural tailoring. Solomon et al. reported that hybrid configurations combining carbon and glass fibers in a PA6 matrix delivered limited impact energies (∼2.68 J) and drop-weight strengths (∼16.3 kJ/m²), which, while demonstrating improvements through hybrid layering, exhibited significantly lower energy dissipation than the composite configurations explored here. ⁴³ Notably, their GFS/CFC/PA6 systems emphasized ductility indices, whereas the current work prioritizes energy absorption and peak load-bearing capacity through geometric and directional reinforcement control.

Similarly, Akangah et al. observed enhanced threshold force and delayed damage evolution in AS4/3501-6 laminates interleaved with electrospun Nylon-66 nanofibers. ⁴⁴ Despite the damage suppression achieved at microstructural levels, the energy absorption capacity remained comparatively low (∼1.80 J), highlighting the trade-off between interlaminar enhancement and macrostructural impact resistance. In contrast, the present configurations demonstrate greater resistance to deformation and improved load transfer efficiency under identical energy inputs, enabled by synergistic tuning of fiber orientation and infill topology.

He et al. investigated the role of through-thickness stitching in mitigating delamination under non-perforation impacts. ⁴⁵ Their findings emphasize how reinforcement continuity and fiber interlock mechanisms suppress crack propagation and reduce damaged zones. A comparable response is evident in our results, where fiber orientation and internal architecture govern stress dispersion, evidenced by the reduced peak deformation and high energy absorption profiles of select configurations.

Kumar et al. employed –COOH-functionalized graphene in CFRP laminates, achieving improved compressive response and delamination resistance under impact loading. ² Their approach underscores the importance of interfacial tuning for impact resilience. However, the current study, without resorting to nanomodification, demonstrates that mechanical behavior can be effectively governed by mesoscale design, specifically through infill geometry and directional reinforcement, yielding similarly robust force–displacement responses and ductile failure patterns.

In higher energy regimes, Mohsin et al. examined drop-weight impact behavior of T700/PA6.6 composites, noting energy absorption up to 140.2 J under severe impact conditions. ⁴⁶ While their work confirms high-energy mitigation in thick unidirectional laminates, partial-to-complete penetration was observed. By contrast, the present study’s configurations, at standardized energy input levels, show non-penetrative failure, controlled deformation, and stable energy absorption, all within lightweight, printed composite architectures, reinforcing their practical utility in impact-critical structural applications. These comparisons affirm that, while prior efforts have leveraged hybridization, interleaving, or material functionalization to improve impact tolerance, the present results demonstrate that rational control over print architecture, specifically fiber orientation, infill geometry, and spatial density, offers an equally effective, scalable, and manufacturable pathway for impact performance enhancement. This architectural sensitivity underscores the necessity of systematic multi-parameter optimization, as detailed in the subsequent sections.

The detailed analyses of the drop-weight impact testing results illustrated in Figures 4–6 distinctly underline the substantial variation in peak impact force, maximum deformation, and absorbed energy across diverse Nylon-fiber composite configurations. This pronounced variability underscores the intricate dependency of composite performance on selected experimental parameters, including fiber type, orientation, infill geometry, and density. While qualitative interpretations provide critical insights, quantitative methods are essential to statistically discern the relative significance of these factors, identify optimal composite configurations, and achieve a holistic assessment of impact performance. Hence, transitioning towards statistical methodologies such as Taguchi analysis is imperative to systematically evaluate factor influences. Furthermore, employing Multi-Criteria Decision Making (MCDM) techniques becomes indispensable to simultaneously optimize multiple conflicting response variables, thereby enabling scientifically robust decisions for designing composites with superior, well-balanced mechanical attributes tailored explicitly for practical impact-resistant applications.

Taguchi analysis

The statistical analysis of the experimental results, employing the Taguchi approach, effectively elucidates the relative contributions of distinct process parameters to the impact performance of 3D-printed Nylon-fiber composites. The response table for Signal-to-Noise (S/N) ratios (Table 6) distinctly classifies fiber type as the most influential factor, followed by infill pattern, fiber orientation, and finally, infill percentage, based on the calculated delta values. The notable delta value of 0.063 attributed to fiber type underscores its pivotal role in governing the composite’s impact resistance behavior. This prominent influence likely arises due to significant differences in the intrinsic mechanical characteristics of Kevlar, carbon, and glass fibers.

Table 6.

Response table for signal to noise ratios.

Level	Fiber orientation (degree)	Infill pattern	Fiber type	Infill %
1	−4.574	−4.519	−4.527	−4.545
2	−4.527	−4.568	−4.583	−4.540
3	−4.530	−4.544	−4.520	−4.546
Delta	0.047	0.049	0.063	0.007
Rank	3	2	1	4

Complementing the quantitative insights from Table 6, Figure 7 graphically represents the variation in mean S/N ratios across different levels of each factor. Examining fiber orientation first, a clear improvement in mean S/N ratio is observed when transitioning from 0° to 45°, subsequently showing a minor reduction when moving to 90°. This observation implies that an orientation at 45° yields optimal reinforcement under multidirectional loading conditions by balancing tensile and shear stress distribution more effectively, thus enhancing composite resilience to impact.

Figure 7.

Main effect plot of mean of SN ratio.

Analyzing infill patterns reveals a compelling trend where the gyroid structure distinctly presents the highest mean S/N ratio, signifying superior energy dissipation and damage tolerance capabilities. The inherently complex geometry of the gyroid pattern provides enhanced stress distribution pathways and facilitates controlled deformation mechanisms, thus leading to improved impact resistance. Conversely, the rectilinear pattern indicates the lowest mean S/N ratio, predominantly owing to its comparatively simpler, less efficient stress dispersion mechanisms, which limit its ability to mitigate impact energy effectively.

Considering fiber type, Kevlar fibers exhibit the highest mean S/N ratio, indicative of their exceptional capability in absorbing and dissipating impact energy due to their intrinsically high strain-to-failure and energy absorption properties. Glass fibers, in stark contrast, show the lowest S/N ratio, suggesting inferior impact resistance in comparison, primarily associated with their relatively lower fracture strain and reduced ductility under dynamic loading conditions. Carbon fibers yield intermediate performance levels, balancing between stiffness and brittle fracture behaviors, thus providing moderate energy absorption characteristics. Lastly, evaluating infill percentage highlights minimal influence on the overall impact behavior, evidenced by the negligible delta value (0.007) presented in Table 6. Despite minimal variations, a slightly superior mean S/N ratio is recorded at 40% infill density, suggesting an optimal balance between structural integrity and lightweight characteristics at this intermediate density level.

Interaction plots in Figure 8 elucidate complex interactions among fiber orientation, infill patterns, fiber type, and infill percentage influencing the impact force of 3D-printed Nylon-fiber composites. Gyroid structures notably improve impact resistance as fiber orientation transitions from 0° to 90°, highlighting enhanced energy dispersion due to superior transverse load-bearing characteristics. Rectilinear patterns peak at intermediate orientations (45°), balancing stiffness and directional stress management, while triangular patterns exhibit less predictable orientation dependencies, reflecting unique stress concentration effects.

Figure 8.

Interaction plot for impact force.

Fiber orientation interactions with fiber types reveal distinct mechanical responses. Carbon fibers exhibit minimal sensitivity across orientations, highlighting consistent rigidity, though a moderate enhancement occurs at 45°, balancing stiffness and compliance. Conversely, glass fibers display pronounced orientation sensitivity, with reduced impact force at intermediate orientations (45°), likely due to compromised matrix interactions, and improved performance at 90°, favouring transverse strength. Kevlar fibers demonstrate continuously increasing impact resistance from 0° to 90°, driven by their intrinsic high strain-to-failure and ductility, optimizing energy absorption and mitigating interface failures.

Infill percentage interactions reveal subtle yet critical trends. Lower densities (30%) often elevate impact force at 90°, suggesting efficient stress dispersion and deformation control facilitated by increased void space. Conversely, higher densities (50%) at intermediate orientations (45°) exhibit diminished impact resistance, highlighting the drawbacks of excessive material rigidity limiting effective energy absorption.

Interaction plots in Figure 9 distinctly illustrate the combined influence of fiber orientation, infill patterns, fiber types, and infill percentages on the deformation behavior of 3D-printed Nylon-fiber composites under impact loading. Observing fiber orientation and infill pattern interactions, gyroid structures exhibit a significant deformation peak at 45°, highlighting their intricate architecture’s ability to facilitate extensive plastic deformation and energy dissipation. Rectilinear patterns conversely demonstrate a sharply reduced deformation at 45°, suggesting a heightened rigidity and minimal energy absorption capability at intermediate orientations. The triangular pattern maintains relatively consistent deformation across orientations, reflecting balanced stress redistribution irrespective of fiber alignment.

Figure 9.

Interaction plot for deformation.

Further inspection of fiber type interactions with orientation reveals intriguing differences in deformation responses. Carbon-fiber composites distinctly peak at intermediate (45°) orientations, indicating enhanced deformation driven by optimal interplay between fiber stiffness and matrix compliance, permitting controlled internal microstructural failures. Conversely, glass fibers consistently display minimal deformation at mid-orientations (45°), emphasizing their brittle fracture behavior and limited ductility, thereby restricting extensive plastic deformation. Kevlar fibers manifest increasing deformation from 0° through 90°, underscoring their inherent flexibility, exceptional fiber elongation capacity, and ability to sustain considerable deformation before structural compromise.

Evaluating infill percentage effects alongside fiber types, composites with glass fibers reveal a pronounced deformation reduction at higher densities (50%), reflecting constrained deformation capacities due to increased structural rigidity. Kevlar-fiber composites illustrate minimal sensitivity to infill density, highlighting their consistent ductile behavior and deformation tolerance across all densities, primarily attributed to their fiber-matrix interface robustness. Carbon fibers exhibit complex interactions with infill density, indicating peak deformation at moderate density levels (40%), beyond which increased density limits further plastic deformation through reduced internal void spaces.

The interaction plots depicted in Figure 10 comprehensively elucidate the complex dependencies among fiber orientation, infill patterns, fiber types, and infill percentages with respect to the impact energy absorption characteristics of 3D-printed Nylon-fiber composites. Observing initially the interplay between fiber orientation and infill patterns, gyroid-patterned composites exhibit a marked improvement in energy absorption when fiber alignment transitions from 0° towards 90°, emphasizing enhanced shear and transverse load management facilitated by the gyroid’s geometric intricacy. Contrastingly, the rectilinear pattern reveals an inverse trend, significantly reducing energy absorption at higher orientations (90°), indicative of constrained internal deformation mechanisms due to directional stiffness dominating the material response. Meanwhile, the triangular pattern exhibits intermediate performance, suggesting balanced energy dissipation across orientations due to effective stress redistribution and fiber bridging capabilities.

Figure 10.

Interaction plot for impact energy.

Exploring further the interactions between fiber type and orientation, distinctive behaviors emerge. Carbon fibers display consistent improvements in absorbed energy as the orientation angle progresses from 0° through 90°, reflecting incremental contributions of transverse fiber alignment to facilitate matrix microcracking and controlled fiber debonding events. Glass fibers show relatively uniform energy absorption across orientations, underlining their stable yet modest capacity for internal energy dissipation, primarily due to inherent brittle fracture mechanisms limiting deformation. Kevlar fibers, however, consistently achieve superior energy absorption, particularly notable at transverse orientations (90°), reinforcing their exceptional ductility, high elongation capacity, and robust fiber-matrix interfaces, allowing prolonged and effective impact energy distribution.

Analyzing the influence of infill percentage alongside fiber types further refines the understanding of energy absorption dynamics. Carbon-fiber composites exhibit increased energy absorption with higher densities (50%), attributed to greater structural stiffness fostering improved fiber-matrix interactions and facilitating effective stress transfer. Conversely, Kevlar and glass-fiber composites show marginal dependency on infill density, maintaining relatively stable energy absorption levels across varying densities. This insensitivity emphasizes their inherent ability to uniformly dissipate impact energy irrespective of internal void fraction variations, thus ensuring consistent structural performance.

Considering the findings from the Taguchi analysis and the intricate trends revealed through the interaction plots, it becomes evident that the impact behavior of 3D-printed Nylon-fiber composites is governed by a complex, non-linear interplay among multiple factors, including fiber type, orientation, infill geometry, and density. While Taguchi analysis efficiently identifies the primary influencing parameters such as the dominant role of fiber type and infill pattern, and it does not fully account for the simultaneous optimization of multiple conflicting performance metrics, namely peak force, deformation, and energy absorption. The interaction plots further highlight that the optimal level of one factor may not uniformly enhance all response variables; for instance, configurations that maximize impact energy may lead to increased deformation or reduced stiffness. This multidimensional response behavior underscores the limitations of single-objective optimization methods. Hence, the integration of Multi-Criteria Decision-Making (MCDM) techniques becomes imperative for the present study. MCDM enables the comprehensive evaluation of composite configurations by assigning appropriate weights to each response parameter and balancing trade-offs between them, ultimately leading to a robust selection of optimal design solutions that satisfy all performance criteria in a unified framework.

Multi-criteria decision making

In this study, the integration of Multi-Criteria Decision-Making (MCDM) techniques plays a pivotal role in optimizing the impact performance of 3D-printed Nylon-fiber composites by considering multiple performance objectives, namely peak impact force (N), maximum deformation (mm), and maximum impact energy (J). As illustrated in Figure 11, the MCDM framework is underpinned by an entropy-based weight assignment strategy, ensuring an objective quantification of the relative importance of each criterion based on the degree of variation in experimental data. The process initiates with data collection followed by normalization, which standardizes all performance indicators to a uniform scale, thereby eliminating dimensional bias across criteria. Entropy values are then calculated to reflect the information content or disorder inherent in each response. Lower entropy implies greater importance, and accordingly, weight factors are computed and verified for reliability.

Figure 11.

Methodology of multi-criteria decision making in the present study.

Subsequent to weight determination, three distinct MCDM methods, RAWEC (Ranking of Alternatives with Weights of Evaluation Criteria), TOPSIS (Technique for Order Preference by Similarity to Ideal Solution), and CoCoSo (Combined Compromise Solution), are employed to rank the composite configurations. Each method involves constructing a decision matrix comprising the normalized values of alternatives (composite specimens) and their performance indicators. Weighted normalized matrices are derived by integrating the entropy-assigned weights, and ideal solutions (best and worst possible outcomes) are established. TOPSIS calculates the Euclidean distance of each alternative from these ideal points, while RAWEC evaluates the relative closeness based on benefit and cost criteria. CoCoSo, on the other hand, synthesizes both additive and multiplicative utility functions to reach a compromise ranking.

This multi-pronged decision-making architecture ensures a holistic assessment where no single metric dominates the evaluation process. By comparing rankings across these three MCDM approaches, the study ensures the identification of the most structurally efficient and impact-resilient composite design, offering a reliable pathway to data-driven material optimization under multiple conflicting performance requirements. The detailed step-by-step procedure of MCDM approach followed is given as follows.

Decision matrix indicating the performance indicators of each alternative is formulated according to equation (1)

${\mathit{D}}_{\mathit{i}\mathit{j}}=\left[\begin{array}{llll}{\mathit{x}}_{\mathbf{11}}& {\mathit{x}}_{\mathbf{12}}& \dots & {\mathit{x}}_{\mathbf{1}\mathit{n}}\\ {\mathit{x}}_{\mathbf{21}}& {\mathit{x}}_{\mathbf{22}}& \dots & {\mathit{x}}_{\mathbf{2}\mathit{n}}\\ .& .& {\mathit{x}}_{\mathit{i}\mathit{j}}& .\\ {\mathit{x}}_{\mathit{m}\mathbf{1}}& {\mathit{x}}_{\mathit{m}\mathbf{2}}& \dots & {\mathit{x}}_{\mathit{m}\mathit{n}}\end{array}\right]$

(1)

where, ‘i’ is alternative, ‘j’ is criteria, $x_{ij}$ is performance indicator of ith alternative of jth criteria

Weight assignment using entropy method

Step.1. Normalization of decision matrix with performance indices to find project outcomes (P_ij)

The x _ij values of responses of the study were normalized using equation (2)

${\mathbf{P}}_{\mathbf{ij}}=\frac{{\mathbf{x}}_{\mathbf{ij}}}{{\displaystyle \sum }_{\mathbf{i}=\mathbf{1}}^{\mathbf{m}}{\mathbf{x}}_{\mathbf{ij}}}$

(2)

where, ‘m’ is number of experiments

Step.2. Calculation of entropy of project outcomes (E_j)

The entropy values for each alternative were calculated using equation (3).

${\mathbf{E}}_{\mathbf{j}}=-\mathbf{k}{\displaystyle \sum }_{\mathbf{i}=\mathbf{1}}^{\mathbf{m}}{\mathbf{P}}_{\mathbf{ij}}\hspace{0.17em}.\hspace{0.17em}\mathbf{ln}\hspace{0.17em}\left({\mathbf{P}}_{\mathbf{ij}}\right)$

(3)

where, k = 1/ln (m)

Step.3. Assigning weight (w_j) using Entropy (E_j)

The weights for each criterion were computed using equation (4).

${\mathbf{w}}_{\mathbf{j}}=\frac{\mathbf{1}-{\mathbf{E}}_{\mathbf{j}}}{{\displaystyle \sum }_{\mathbf{j}=\mathbf{1}}^{\mathbf{n}}\left(\mathbf{1}-{\mathbf{E}}_{\mathbf{j}}\right)}$

(4)

Ranking of alternatives with weights of criterion (RAWEC)

Step.1. Normalization of decision matrix based on double normalization procedure as mentioned below

For benefit criteria, in our case, peak impact force and impact energy needs to be maximum to have better performance of material, equations (5) and (6) were used for double normalization of peak impact force and impact energy.

$\text{First}\hspace{0.17em}\text{normalization},{\mathit{n}}_{\mathit{i}\mathit{j}}=\frac{{\mathit{x}}_{\mathit{i}\mathit{j}}}{{\mathit{x}}_{\mathit{j}\hspace{0.17em}\mathit{max}}}$

(5)

$\text{Second}\hspace{0.17em}\text{normalization},{{\mathit{n}}_{\mathit{i}\mathit{j}}}^{\prime }=\frac{{\mathit{x}}_{\mathit{j}\hspace{0.17em}\mathit{min}}}{{\mathit{x}}_{\mathit{i}\mathit{j}}}$

(6)

For cost criteria, in our case, deformation due to drop weight impact should be minimum to represent the superior potential of the material, the double normalization of deformation was carried out using equations (7) and (8).

$\text{First}\hspace{0.17em}\text{normalization},{\mathbf{n}}_{\mathbf{ij}}=\frac{{\mathbf{x}}_{\mathbf{j}\hspace{0.17em}\mathbf{min}}}{{\mathbf{x}}_{\mathbf{ij}}}$

(7)

$\text{First}\hspace{0.17em}\text{normalization},{{\mathit{n}}_{\mathit{i}\mathit{j}}}^{\prime }=\frac{{\mathit{x}}_{\mathit{i}\mathit{j}}}{{\mathit{x}}_{\mathit{j}\hspace{0.17em}\mathit{max}}}$

(8)

Step.2. Calculation of deviation from the weights assigned ′ ${{\mathbf{w}}_{\mathbf{j}}}^{\prime }$ from entropy method and following equations (9) and (10).

${\mathit{v}}_{\mathit{i}\mathit{j}}={\displaystyle \sum }_{\mathit{j}=\mathbf{1}}^{\mathit{n}}{\mathit{w}}_{\mathit{j}}\hspace{0.17em}.\left(\mathbf{1}-{\mathit{n}}_{\mathit{i}\mathit{j}}\right)$

(9)

${\mathit{v}}_{\mathit{i}\mathit{j}}^{\prime }={\displaystyle \sum }_{\mathit{j}=\mathbf{1}}^{\mathit{n}}{\mathit{w}}_{\mathit{j}}\hspace{0.17em}.\left(\mathbf{1}-{{\mathit{n}}_{\mathit{i}\mathit{j}}}^{\prime }\right)$

(10)

The deviation ${\mathit{v}}_{\mathit{i}\mathit{j}}$ should be as small as possible, whereas ${{\mathit{v}}_{\mathit{i}\mathit{j}}}^{\prime }$ needs to be as large as possible.

Step.3. Computation of ‘Q _i ’ factor for ranking using equation (11)

${\mathit{Q}}_{\mathit{i}}=\frac{{{\mathit{v}}_{\mathit{i}\mathit{j}}}^{\prime }-{\mathit{v}}_{\mathit{i}\mathit{j}}}{{\mathit{v}}_{\mathit{i}\mathit{j}}^{\prime }+{\mathit{v}}_{\mathit{i}\mathit{j}}}$

(11)

Qi value can vary between −1 to 1. The optimum alternative is identified based on the maximum value of Qi.

Technique for order preference by similarity to ideal solutions (TOPSIS)

Step.1. Normalization of performance indicators in the decision matrix using equation (12)

${\mathit{r}}_{\mathit{i}\mathit{j}}=\frac{{\mathit{x}}_{\mathit{i}\mathit{j}}}{\sqrt{{\displaystyle \sum }_{\mathit{i}=\mathbf{1}}^{\mathit{m}}{{\mathit{x}}_{\mathit{i}\mathit{j}}}^{\mathbf{2}}}}$

(12)

Step.2. Calculation the weighted normalized values of performance indicators using equation (13)

${\mathit{V}}_{\mathit{i}\mathit{j}}={\mathit{w}}_{\mathit{i}\mathit{j}}\hspace{0.17em}.\hspace{0.17em}{\mathit{r}}_{\mathit{i}\mathit{j}}$

(13)

Step.3. Computation of positive and negative ideal solution using equations (14) and (15)

${\mathit{S}}_{\mathit{i}}^{+}=\sqrt{{\displaystyle \sum }_{\mathit{j}=\mathbf{1}}^{\mathit{n}}{\left({\mathit{V}}_{\mathit{i}\mathit{j}}-{\mathit{B}}_{\mathit{j}}^{+}\right)}^{\mathbf{2}}}$

(14)

${\mathit{S}}_{\mathit{i}}^{-}=\sqrt{{\displaystyle \sum }_{\mathit{j}=\mathbf{1}}^{\mathit{n}}{\left({\mathit{V}}_{\mathit{i}\mathit{j}}-{\mathit{B}}_{\mathit{j}}^{-}\right)}^{\mathbf{2}}}$

(15)

Step.4. Identification of relative closeness with ideal solution using equation (16)

${\mathit{R}}_{\mathit{i}}=\frac{{\mathit{S}}_{\mathit{i}}^{-}}{{\mathit{S}}_{\mathit{i}}^{+}-{\mathit{S}}_{\mathit{i}}^{-}}$

(16)

Combined compromise solution (CoCoSo)

Step.1. Normalization of decision matrix using equation (17)

${\mathit{N}}_{\mathit{i}\mathit{j}}=\frac{{\mathit{x}}_{\mathit{i}\mathit{j}}-\mathit{min}\left({\mathit{x}}_{\mathit{i}\mathit{j}}\right)}{\mathit{max}\left({\mathit{x}}_{\mathit{i}\mathit{j}}\right)-\mathit{min}\left({\mathit{x}}_{\mathit{i}\mathit{j}}\right)}$

(17)

Step.2. Calculation of weighted normalized values

The weights assigned ′ ${{\mathbf{w}}_{\mathbf{j}}}^{\prime }$ from entropy method were used to calculate M _i & P _i using equations (18) and (19)

${\mathit{M}}_{\mathit{i}}={\displaystyle \sum }_{\mathit{j}=\mathbf{1}}^{\mathit{n}}{\mathit{w}}_{\mathit{j}}\hspace{0.17em}.\hspace{0.17em}{\mathit{N}}_{\mathit{i}\mathit{j}}$

(18)

${\mathit{P}}_{\mathit{i}}={\displaystyle \sum }_{\mathit{j}=\mathbf{1}}^{\mathit{n}}{\left({\mathit{N}}_{\mathit{i}\mathit{j}}\right)}^{{\mathit{w}}_{\mathit{j}}}$

(19)

Step.3. Computation of aggregated sum and products using equations (20)–(22)

${\mathit{K}}_{\mathit{a}}=\frac{\left({\mathit{P}}_{\mathit{i}}+{\mathit{M}}_{\mathit{i}}\right)}{{\displaystyle \sum }_{\mathit{i}=\mathbf{1}}^{\mathit{m}}\left({\mathit{P}}_{\mathit{i}}+{\mathit{M}}_{\mathit{i}}\right)}$

(20)

${\mathit{K}}_{\mathit{b}}=\frac{{\mathit{M}}_{\mathit{i}}}{\mathit{min}\left({\mathit{M}}_{\mathit{i}}\right)}+\frac{{\mathit{P}}_{\mathit{i}}}{\mathit{min}\left({\mathit{P}}_{\mathit{i}}\right)}$

(21)

${\mathit{K}}_{\mathit{c}}=\frac{\mathit{\lambda }{\mathit{M}}_{\mathit{i}}+\left(\mathbf{1}-\mathit{\lambda }\right){\mathit{P}}_{\mathit{i}}}{\left(\mathit{\lambda }\hspace{0.17em}.\hspace{0.17em}\mathit{max}\left({\mathit{M}}_{\mathit{i}}\right)+\left(\mathbf{1}-\mathit{\lambda }\right).\hspace{0.17em}\mathit{max}\left({\mathit{P}}_{\mathit{i}}\right)\right)}$

(22)

Step.4. Determine the final CoCoSo score using equation (23)

${\mathit{K}}_{\mathit{i}}={\left({\mathit{K}}_{\mathit{a}}{\mathit{K}}_{\mathit{b}}{\mathit{K}}_{\mathit{c}}\right)}^{\left(\mathbf{1}/\mathbf{3}\right)}+\frac{\mathbf{1}}{\mathbf{3}}\hspace{0.17em}\left({\mathit{K}}_{\mathit{a}}+{\mathit{K}}_{\mathit{b}}+{\mathit{K}}_{\mathit{c}}\right)$

(23)

The consolidated ranking results from the MCDM techniques, RAWEC, TOPSIS, and CoCoSo are graphically represented in Figure 12, illustrating the relative performance of all 27 experimental runs. The convergence and divergence of rankings across methods underscore the robustness and subtle differences inherent in each MCDM algorithm. The configuration comprising 0° fiber orientation, rectilinear infill, glass fiber reinforcement, and 50% infill density (Exp. 6) consistently secured Rank 1 across all methods, indicating an optimally balanced mechanical performance in terms of maximum peak impact force, high energy absorption, and moderately low deformation. Scientifically, this superior ranking can be attributed to the alignment of fibers with the primary impact direction (0°), which facilitates maximum tensile load transfer and reduces shear-induced failures. The rectilinear pattern further contributes by promoting orthogonal stress bridging and minimizing stress localization. Glass fibers, known for their higher strain-to-failure compared to carbon, offer enhanced energy dissipation through gradual failure mechanisms. Additionally, the 50% infill density ensures structural continuity and improved load transfer paths without inducing excessive brittleness.

Figure 12.

Ranking of alternatives from various MCDM methods.

On the contrary, the configuration marked as Exp. 16, with 45° fiber orientation, gyroid infill, carbon fiber reinforcement, and 30% infill density, was unanimously ranked 27th the poorest-performing setup across all MCDM models. This inferior performance results from the unfavourable synergy of parameters: the ±45° orientation, while ideal for shear resistance, fails to efficiently transfer axial loads under perpendicular impact conditions. Coupled with the low infill percentage, this orientation leads to insufficient material support and promotes premature local buckling and delamination. The carbon fibers, being inherently brittle and less tolerant to bending and off-axis loads, contribute to a sudden failure mechanism. The gyroid infill, though geometrically complex and potentially energy dissipative at higher densities, becomes unstable under reduced material volume, offering insufficient resistance during impact. Hence, the combined MCDM analysis not only validates the experimental outcomes but also provides a rigorous, criteria-weighted selection of optimal and suboptimal design configurations. Such a multidimensional ranking approach enhances the reliability of composite design for real-world impact-prone applications.

Ranking based on specific energy absorption

Specific Energy Absorption (SEA) is a critical metric in assessing the impact energy dissipation capacity of fiber-reinforced composites relative to their weight and density. In high-performance applications such as aerospace enclosures, and structural shells in unmanned aerial systems, where mass efficiency and damage-tolerant lightweighting are essential, SEA provides a material-level evaluation that complements system-level performance metrics. Unlike entropy-weighted MCDM rankings, which incorporate force, deformation, and energy in a balanced framework, SEA-based ranking isolates the intrinsic material performance by normalizing the absorbed impact energy with respect to experimentally measured mass and calculated density of each specimen. All specimens in the current study were weighed using a calibrated analytical balance with ±0.01 g precision. These measured values were then used to compute actual densities based on the fixed specimen volume of 75 cm³. This enabled precise calculation of SEA, represented in Table 7, thereby capturing the effect of internal porosity, fiber distribution, and infill topology on energy dissipation efficiency.

Table 7.

Specific energy absorption (SEA) metrics and rankings based on the “larger-the-better” criterion.

Exp. no	Fiber orientation	Infill pattern	Fiber type	Infill %	Mass (g)	Density (g/cm3)	Energy (J)	SEA (J.cm3/g)	Rank
1	0	Triangle	Carbon	30	38.71	0.5161	14.96	28.987	23
2	0	Triangle	Carbon	40	44.37	0.5916	14.94	25.254	25
3	0	Triangle	Carbon	50	50.84	0.6779	14.952	22.056	27
4	0	Rectilinear	Glass	30	38.99	0.5199	14.967	28.788	24
5	0	Rectilinear	Glass	40	45.98	0.6131	14.963	24.405	26
6	0	Rectilinear	Glass	50	52.15	0.6953	22.916	32.958	17
7	0	Gyroid	Kevlar	30	36.27	0.4836	22.789	47.124	2
8	0	Gyroid	Kevlar	40	40.92	0.5456	22.678	41.565	7
9	0	Gyroid	Kevlar	50	46.99	0.6265	22.708	36.246	13
10	45	Triangle	Glass	30	38.76	0.5179	22.722	43.873	5
11	45	Triangle	Glass	40	46.09	0.6145	22.715	36.965	11
12	45	Triangle	Glass	50	52.09	0.6945	22.64	32.599	19
13	45	Rectilinear	Kevlar	30	36.91	0.4921	22.746	46.222	3
14	45	Rectilinear	Kevlar	40	42.47	0.5663	22.76	40.191	8
15	45	Rectilinear	Kevlar	50	47.56	0.6341	22.709	35.813	14
16	45	Gyroid	Carbon	30	38.01	0.5068	16.441	32.441	20
17	45	Gyroid	Carbon	40	44.05	0.5873	18.698	31.837	21
18	45	Gyroid	Carbon	50	51.22	0.6829	22.908	33.545	16
19	90	Triangle	Kevlar	30	35.84	0.4779	22.869	47.853	1
20	90	Triangle	Kevlar	40	41.03	0.5471	19.581	35.791	15
21	90	Triangle	Kevlar	50	46.65	0.622	23.025	37.018	10
22	90	Rectilinear	Carbon	30	38.02	0.5069	22.998	45.370	4
23	90	Rectilinear	Carbon	40	45.62	0.6082	22.878	37.616	9
24	90	Rectilinear	Carbon	50	51.87	0.6916	21.193	30.643	22
25	90	Gyroid	Glass	30	39.77	0.5303	22.863	43.113	6
26	90	Gyroid	Glass	40	46.63	0.6217	22.822	36.709	12
27	90	Gyroid	Glass	50	52.12	0.6949	22.815	32.832	18

Ranking was performed using the larger-the-better criterion for both SEA and revealed meaningful divergences from the earlier MCDM-based ranks. Notably, Experiment 19, a Kevlar-reinforced composite with 90° fiber orientation, triangle infill, and 30% infill density, achieved the highest SEA (47.85 J·cm³/g). The combination of Kevlar’s high specific toughness, low matrix content, and shear-aligned fiber layup under 90° orientation significantly enhanced the stress wave attenuation per unit density. Although its absolute peak force was modest, its strain-to-failure and distributed damage morphology permitted efficient energy absorption with minimal material usage.

In contrast, Experiment 3, a carbon-reinforced composite with 0° fiber orientation and 50% infill, ranked lowest in SEA despite possessing axial alignment that generally favours stiffness. Its relatively brittle fracture modes, combined with higher mass due to dense infill and fiber content, led to a less favourable energy-to-density ratio. The early onset of delamination and fiber splitting, especially under axial loading, is attributed to reduce its capacity to engage the full cross-section during impact. The apparent contradiction wherein Experiment 6, previously ranked first in the MCDM scheme, achieved only 17th rank in SEA, underscores a pivotal insight: system-level mechanical robustness does not always translate to material-level efficiency.

This contrast between SEA and MCDM rankings highlights the multi-objective nature of impact-driven composite design. MCDM rankings favour configurations that balance peak load bearing, controlled deformation, and energy absorption, making them suitable for applications demanding structural survivability under sudden loads, such as crashworthy panels or ballistic shields. SEA, on the other hand, favours materials that offer maximum energy dissipation per unit density, aligning with mass-critical applications where volumetric constraints and payload efficiency are governing factors.

By maintaining the experimental fidelity of mass measurements and integrating both ranking methodologies, the present study enables a dual-layered performance evaluation. Designers can now prioritize configurations like Exp. 6 when peak strength and integrity are critical or choose Exp. 19 when energy-to-weight efficiency is paramount. This dual-ranking strategy establishes a robust foundation for context-driven material selection, maximizing both structural performance and application relevance.

Optimal infill pattern analysis for each fiber type

In addition to global rankings based on MCDM and SEA, a fiber-type-specific optimization analysis was performed to determine the best infill pattern and configuration for each fiber category, Carbon, Kevlar, and Glass, based on both absolute energy absorption and Specific Energy Absorption (SEA). This stratified evaluation enables the decoupling of composite performance from confounding influences, such as inter-fiber mechanical disparity, and allows for informed decisions in selecting the most suitable geometric architecture for each fiber type, aligned with application-specific criteria.

For Carbon Fiber composites, which exhibit high stiffness but lower ductility, the optimal infill pattern was Rectilinear, particularly in Exp. 22 (90°/Rectilinear/Carbon/30%), achieving the highest SEA of 45.370 J·cm³/g within the carbon group. This configuration also recorded a peak impact force of 2313.96 N and absorbed 22.998 J of energy. The rectilinear pattern is structurally efficient in channelling axial and transverse stresses, especially when aligned with the principal directions (90° orientation), mitigating localized failure through predictable crack propagation paths. In contrast, the gyroid-infused carbon configurations (e.g., Exp. 18) exhibited slightly higher energy absorption but at the cost of lower SEA due to increased mass and less controlled load transfer.

For Kevlar Fiber composites, the most effective configuration was found in Exp. 19 (90°/Triangle/Kevlar/30%), which yielded the highest SEA across all 27 configurations at 47.853 J·cm³/g, with 22.869 J of energy absorption. The triangle infill, in combination with transverse fiber orientation, leverages Kevlar’s inherent high strain-to-failure capacity, promoting distributed deformation, fiber bridging, and gradual energy dissipation. Notably, configurations such as Exp. 13 (45°/Rectilinear/Kevlar/30%) and Exp. 7 (0°/Gyroid/Kevlar/30%) closely followed with SEAs of 46.222 and 47.124 J·cm³/g, respectively, confirming that lower infill densities paired with structurally adaptive infill geometries (triangle, rectilinear, or gyroid) maximize Kevlar’s deformation-mediated energy absorption potential.

For Glass Fiber systems, which lie intermediate in stiffness and ductility between carbon and Kevlar, the optimal configuration was identified as Exp. 25 (90°/Gyroid/Glass/30%), delivering a SEA of 43.113 J·cm³/g and 22.863 J of absorbed energy. The gyroid infill, with its continuous, sinusoidal internal walls, promotes three-dimensional load dispersion, which, when aligned transversely, enhances strain uniformity and delays catastrophic matrix cracking. Compared to rectilinear or triangular infills in glass systems, the gyroid offers superior multi-axial damping capability under impact. Although Exp. 10 (45°/Triangle/Glass/30%) also performed exceptionally with a SEA of 43.873, the gyroid’s capacity to balance stiffness with energy dispersion in a spatially efficient manner makes it more suitable for structural systems with complex loading paths. A comparative summary is presented in Table 8 to highlight the top-performing configurations for each fiber type, categorized by their respective optimal infill pattern and ranked by SEA:

Table 8.

Optimal infill pattern and configuration for each fiber type.

Fiber type	Exp. no	Orientation (°)	Infill pattern	Infill %	Peak force (N)	Energy (J)	Mass (g)	SEA (J·cm3/g)	Rank in fiber group
Carbon	22	90	Rectilinear	30	2313.96	22.998	38.02	45.37	1
	23	90	Rectilinear	40	2485.45	22.878	45.62	37.616	2
	18	45	Gyroid	50	1909.72	22.908	51.22	33.545	3
Kevlar	19	90	Triangle	30	2256.38	22.869	35.84	47.853	1
	7	0	Gyroid	30	2287.01	22.789	36.27	47.124	2
	13	45	Rectilinear	30	2147.36	22.746	36.91	46.222	3
Glass	25	90	Gyroid	30	2121.64	22.863	39.77	43.113	1
	10	45	Triangle	30	2278.43	22.722	38.76	43.873	2
	26	90	Gyroid	40	2420.53	22.822	46.63	36.709	3

This fiber-wise stratification reinforces the hypothesis that optimal mechanical response is not solely governed by material selection, but rather by the synergistic integration of fiber properties with geometry-specific stress paths induced by the infill design. For carbon fibers, directional alignment and stiffness preservation benefit from ordered, orthogonal raster infills. Kevlar, with its strain-dependent toughness, leverages gradient-dominated infill patterns such as triangle and gyroid. Glass fibers, due to their intermediate behavior, capitalize on multi-directional infill topologies that blend stiffness with energy dissipation.

Correlation between static flexural stiffness and impulse loading

Static three-point bending tests were conducted on all 27 printed composite configurations using a calibrated universal testing machine in accordance with ASTM D790. These tests yielded the flexural modulus and flexural strength, which are critical indicators of initial structural stiffness under bending. The slope of the linear elastic region from the load–deflection response was extracted to represent each configuration’s stiffness, and the results are presented in Figure 13.

Figure 13.

Flexural strength and flexural modulus of samples under static bending test.

A clear trend emerges when comparing static stiffness values with dynamic impulse loading characteristics obtained from the drop-weight impact tests. Specifically, specimens with higher flexural modulus consistently demonstrated superior impact performance, validating the role of initial structural rigidity in resisting sudden loads.

For instance, Experiment 3 (Onyx/Carbon, 0°, Triangle, 50% infill) exhibited the highest flexural modulus (48.45 GPa) and flexural strength (445.32 MPa). This sample also recorded the highest peak impact force (2776.99 N) and lowest deformation depth (11.01 mm) during impulse loading. In contrast, Experiment 25 (Onyx/Glass, 90°, Gyroid, 30% infill) had the lowest flexural modulus (12.82 GPa) and flexural strength (223.97 MPa), which corresponded to the lowest impact force (1844.80 N) and maximum deformation (∼38.4 mm).

This inverse relationship between stiffness and deformation is consistent across all configurations. The structural stiffness measured under ASTM D790 serves as a reliable predictor of impact resilience, particularly in identifying fiber-infill-orientation combinations that can withstand high-rate loading. The use of high-modulus reinforcements like carbon fibers in aligned (0°) orientations with denser infill not only maximized static rigidity but also minimized energy dissipation during impact. Such correlation reinforces the significance of incorporating static bending stiffness metrics in pre-screening protocols for dynamically loaded applications. The present analysis substantiates that optimizing architectural parameters (fiber type, orientation, and infill) improves both static stiffness and dynamic load resistance, establishing a strong link between material architecture and multifunctional performance.

Impact damage characterization

The post-impact damage characterization presented in Figure 14 offers a compelling visual correlation between the mechanical performance and failure modes of the best- and worst-performing composite configurations identified through the MCDM ranking. Figure 14(a) and (b) represent the front and rear faces of the worst-performing configuration, Exp. 16 (45°/Gyroid/Carbon/30 wt.%), while Figure 14(c) and (d) correspond to the highest-ranked sample, Exp. 6 (0°/Rectilinear/Glass/50 wt.%). The nature and extent of damage observed in these specimens reveal critical insights into the synergistic influence of fiber architecture, infill geometry, and material selection on impact resistance.

Figure 14.

Impact damage of sample from Exp. 16 (a) front & (b) rear side; sample from Exp. 6 (c) front & (d) rear side.

The front surface of Exp. 16 in Figure 14(a) exhibits significant crack initiation and delamination zones extending radially, a consequence of the suboptimal fiber orientation at 45°, which inadequately resists axial impact forces. The gyroid infill, although geometrically complex, becomes structurally unstable at 30% infill density, leading to localized buckling and insufficient energy redistribution during impact. On the rear side, as shown in Figure 14(b), extensive fiber breakage and matrix cracking are evident, with carbon fibers clearly fractured. These observations reinforce the material’s brittle failure mode, further aggravated by the poor load path continuity and stress localization inherent in low-density gyroid structures, making the system susceptible to catastrophic failure under dynamic loading.

In contrast, the damage morphology in Exp. 6 highlights a controlled and energy-dissipative failure response. Figure 14(c) shows the front face of the sample with minor indentation and restrained fiber exposure, signifying effective stress transfer along the fiber axis owing to the 0° orientation. The rectilinear infill promotes predictable crack paths and distributes loads orthogonally, enhancing impact resistance. The rear surface in Figure 14(d) reveals moderate fiber pull-out and minor matrix cracking, typical of ductile failure behavior facilitated by glass fiber’s higher strain-to-failure and energy absorption capabilities. The 50% infill ensures structural rigidity without compromising toughness, allowing the specimen to withstand impact loads with minimal structural compromise. The contrast in damage profiles between these two configurations underscores the necessity of optimal parameter synergy in additive-manufactured composite design. Proper alignment, fiber-matrix compatibility, and infill topology collectively dictate the failure mechanisms, validating the MCDM approach in converging towards structurally superior composite architectures for impact-critical applications.

Extensive studies on nylon-based composites under drop-weight impact have highlighted a diverse spectrum of failure modes, predominantly influenced by reinforcement architecture, matrix properties, and structural configuration. González et al. employed continuum damage mechanics-based finite element simulations to show that delamination and ply-level fracture were dominant failure mechanisms in nylon laminates, with simulation results aligning well with post-impact experimental behavior. ⁴⁷ Koffi et al., investigating birch fiber-reinforced HDPE, observed that increased fiber content led to greater brittleness and reduced impact strength, indicating that matrix deformability plays a central role in energy absorption. ³ Ojha et al. reported that 3D-printed Kevlar–nylon composites exhibited delamination-dominated damage responses, where increased fiber reinforcement suppressed matrix perforation but promoted interlaminar failure due to Kevlar’s directionally dependent behavior. ⁴⁸ In a thermomechanical study, Wu found that 75% of drop-weight energy in rubber-toughened nylon was dissipated through matrix yielding and 25% through crazing, with minimal contribution from surface energy, underscoring the significance of internal plastic deformation. ⁴⁹ Similarly, Russo et al. observed that PA6-based woven basalt composites exhibited pronounced plastic deformation and delamination under sub-perforation impacts, supported by ultrasound and confocal microscopy analyses ⁵⁰

Temperature-dependent damage evolution was documented by Díaz-Rodríguez et al., who found that continuous fiber-reinforced nylon composites transitioned from matrix cracking to fiber tearing and pull-out as ambient conditions varied, resulting in reduced impact resistance outside the 0°C to 50°C range. ⁵¹ Johnson et al. demonstrated that in glass fiber-reinforced nylon systems, mean fiber length and fiber weight fraction had a stronger influence on fracture surface morphology and crack propagation than impact velocity. ⁵² In a subsequent study, Johnson et al. validated the reliability of instrumented drop-weight testing by correlating filtered force–time curves with fracture visualization via high-speed photography. ⁵³ Jang et al. further confirmed that stacking sequence and fiber toughness significantly affect delamination onset and plastic deformation in hybrid nylon composites under impact. ⁵⁴ Beharic et al. showed that sandwich structures made of Nylon 12 with auxetic cellular cores outperformed conventional octet-truss in absorbing impact energy, indicating that geometric confinement effects play a dominant role in failure modulation. ⁵⁵ Additionally, Palazzetti et al. demonstrated that electrospun Nylon 6,6 nanofiber interleaves in CFRP laminates enhanced damping and post-impact damage tolerance by promoting fiber bridging and crack deflection mechanisms. ⁵⁶

Aligned with these prior findings, the present study confirms that fiber type, orientation, and infill architecture govern distinct damage pathways under drop-weight impact. Carbon-reinforced samples exhibited localized brittle fracture characterized by matrix cracking and fiber rupture, consistent with stiffness-driven stress concentration and failure initiation as seen by González et al. ⁴⁷ and Koffi et al. ³ In contrast, glass fiber-reinforced composites displayed distributed damage in the form of fiber bridging and microcrack propagation, reflecting energy dissipation mechanisms similar to those observed by Wu ⁴⁹ and Russo et al. ⁵⁰ The observed influence of infill geometry also mirrors trends reported by Beharic et al ⁵⁵ and Jang et al. ⁵⁴ ; rectilinear and gyroid patterns effectively modulated stress distribution and controlled crack propagation through geometric confinement and load redirection. Furthermore, the controlled, non-penetrative damage zones recorded in the current study correspond with the matrix-dominated ductile behavior emphasized in temperature-stable regions by Díaz-Rodríguez et al. ⁵¹ Without requiring matrix modification or nano interleaving, the damage behavior observed in this work validates the critical role of architecture-driven impact mitigation in FDM-fabricated continuous fiber nylon composites, reinforcing the importance of structural design over material additives in influencing post-impact integrity.

Energy absorbing mechanisms

The impact-fractured surface of the Exp. 16 specimen, reinforced with continuous carbon fibers and printed with an Onyx matrix, reveals a fracture morphology dominated by brittle failure modes, providing conclusive microstructural evidence for the low energy absorption recorded in impact testing. The SEM images (Figure 15(a)–(d), exhibit multiple localized failure mechanisms, each contributing distinctly to the global energy dissipation behavior.

Figure 15.

(a)–(d) SEM images showing matrix cracks, fiber breakage, interfacial debonding, radial fiber splitting from the impact fractured surface of Exp. 16.

In Figure 15(a), the fracture surface shows fiber breakage occurring along the load path, accompanied by prominent matrix cracking between adjacent printed layers. The fracture planes intersecting the continuous carbon fibers appear sharp and clean, with no visible signs of fiber distortion, indicating low strain-to-failure and an absence of fiber plasticity. Simultaneously, the crack propagation path follows the raster boundary, suggesting that the FFF-induced interfacial discontinuities serve as preferred planes for crack initiation and extension under impact loading.

Figure 15(b) provides further insight into the fiber–matrix interface, where fiber pull-out is observed. However, the pull-out length is minimal, and the pulled fibers exhibit smooth surfaces, lacking matrix residue or adhesive traces. This implies a low interfacial shear strength and poor frictional resistance during failure progression, thus limiting energy dissipation through pull-out mechanics. The adjacent debonded matrix regions show signs of cleavage rather than plastic deformation, highlighting the brittle nature of failure propagation in the matrix domain.

In Figure 15(c), sheared matrix zones are seen in conjunction with cleanly fractured fiber ends. The matrix shear appears confined to thin regions adjacent to the fiber walls, lacking evidence of ductile flow or interlaminar plasticization. The fibers exhibit flat and planar breakage fronts, reinforcing the inference that the failure was governed by brittle translaminar fracture rather than by fiber bridging or deflection, which are typical of more energy-absorbing systems.

Figure 15(d) shows radial splitting of carbon fibers; a failure mode associated with localized compressive instability and intrafiber crack nucleation. This mechanism suggests that impact energy induces transverse stresses that exceed the inter-fibrillar cohesion, causing longitudinal fracture planes along the fiber axis. The surrounding matrix, however, does not exhibit any compensating plastic deformation or evidence of crack blunting, resulting in minimal interruption of crack growth and limited attenuation of impact energy.

The combination of matrix cleavage, minimal fiber pull-out, interfacial decohesion, and radial fiber fracture, as systematically revealed in SEM Figure 15(a)–(d), offers clear mechanistic evidence for the limited energy absorption exhibited by the Exp. 16 configuration. These failure features are inherently brittle, exacerbated by the anisotropic and layered nature of the 3D printed architecture.

In contrast to the brittle and interfacially dominated failure morphology observed in Exp. 16, the fracture surface of the glass fiber-reinforced specimen (Exp. 6) exhibits an ensemble of distributed, energy-dissipative failure mechanisms. The SEM micrographs (Figure 16(a)–(d), illustrate the synergistic activation of fiber bridging, matrix shear yielding, interfacial adhesion, and matrix flow phenomena that collectively enhance energy absorption under impact.

Figure 16.

(a)–(d) SEM images showing fiber bridging, matrix shear, fiber pullout, fiber–matrix interlock and fractured rough fiber ends from the impact fractured surface of Exp. 6.

Figure 16(a) reveals multiple fiber bridging zones extending across printed layer interfaces, which act as crack-arresting ligaments that delay delamination propagation. Simultaneously, fiber breakage occurs at varied lengths and orientations, suggesting localized tensile failure induced by bending stresses and constrained interfacial conditions. The layer delamination, although initiated, is partially restrained by fiber bridging and matrix continuity across the print layers, contrasting the discrete interlayer failure evident in the carbon fiber system.

In Figure 16(b), partially pulled-out glass fibers are observed with extensive matrix adherence on the fiber surfaces, highlighting strong interfacial bonding. Unlike the smooth, clean pull-out observed in Exp. 16, the fibers here display fractured matrix sheaths clinging to their surfaces, indicative of substantial energy dissipation through interfacial friction and matrix tearing. The fracture morphology supports a mixed-mode fiber debonding-pull-out mechanism, essential for increasing fracture resistance in layered composites.

Figure 16(c) captures a highly sheared matrix zone where the plastically deformed matrix surrounds embedded fiber clusters. The fiber–matrix interlock visible in this region suggests that the toughened fracture process involves resistance to fiber slippage and microcrack coalescence. Such interlocking also promotes crack deflection and crack tip blunting, which are absent in the carbon fiber architecture. The deformation features in this region imply that shear yielding of the thermoplastic-rich matrix contributes to substantial plastic energy dissipation.

Figure 16(d) presents regions of void closure by matrix flow, which are indicative of localized stress relaxation via thermomechanical softening. The presence of rough fiber ends coated with reflowed matrix further suggests that post-impact thermal effects and viscoplastic matrix behavior play a secondary role in sealing fracture cavities and arresting delamination fronts.

Taken together, these observations confirm that the energy absorption in Exp. 6 is driven by multiscale damage mechanisms, including matrix yielding, fiber bridging, interfacial friction, and toughening via fiber–matrix mechanical interlocks. The printed architecture of the specimen appears to aid in distributing these failure processes across the layered structure. Unlike the abrupt, brittle fracture seen in Exp. 16, the Exp. 6 system exhibits progressive, resistance-based fracture propagation, resulting in greater toughness and higher measured impact energy absorption. This microstructural evidence supports the unique contribution of the present study in establishing a mechanistic link between fiber type, print-induced interfacial architecture, and dynamic energy absorption behavior in continuous fiber-reinforced FFF/CFF composites. By capturing the actual sequence of material damage modes activated during impact, this work substantiates the functional advantages of glass fiber systems for structural applications requiring enhanced impact tolerance.

Post-impact residual mechanical properties

Assessment of post-impact mechanical behavior is critical to understanding the extent of structural integrity loss in 3D-printed continuous fiber reinforced composites and informs maintenance protocols for components deployed in high-reliability applications. Table 9 summarizes the pristine and residual mechanical properties for both tensile (TAI) and compressive (CAI) modes, highlighting the pronounced differences in retained strength between the Onyx/Glass (Exp. 6) and Onyx/Carbon (Exp. 16) systems following impact.

Table 9.

Comparative results of pristine and post impact tensile and compressive properties.

Experiment	Material system	Pristine tensile strength (MPa)	Residual TAI strength (MPa)	Retention (%)	Pristine compressive strength (MPa)	Residual CAI strength (MPa)	Retention (%)
Exp. 6	Onyx/Glass	389.9	284.63	73	613.8	417.38	67.9
Exp. 16	Onyx/Carbon	562.4	289.64	52.5	1027.4	468.72	45.6

As evidenced in Figure 17, the pristine tensile stress–strain response of Exp. 16 displays a higher ultimate tensile strength of 562.4 MPa, exceeding that of Exp. 6 (389.9 MPa), and exhibits a steeper initial elastic slope indicating a higher tensile modulus. However, after impact, Exp. 16 retains only 52.5% of its pristine tensile strength (289.64 MPa), while Exp. 6 maintains a considerably higher 73% retention (284.63 MPa), underscoring the superior damage tolerance of glass fiber composites under tensile loading. This disparity originates from the fundamentally brittle fracture behavior inherent to carbon fiber systems, where impact energy localizes damage into discrete fracture zones, accelerating matrix cracking and fiber–matrix interfacial debonding, which reduces both strength and load-transfer efficiency. In contrast, glass fibers possess higher strain-to-failure capability and improved interfacial adhesion with the Onyx matrix, allowing for more distributed damage, energy dissipation through fiber pull-out and bridging, and consequently better preservation of tensile integrity.

Figure 17.

Tensile stress–strain curves of pristine and tensile after impact samples.

Examining the strain behavior further elucidates these differences: Exp. 6 pristine specimens achieve a greater tensile strain at failure (0.0213) than Exp. 16 (0.015), reflecting the inherently higher strain-to-failure characteristics of glass fibers and their capacity for distributed damage tolerance. However, post-impact evaluations reveal an inversion of this trend. Exp. 16 retains a strain-to-failure of 0.0132 in TAI testing, while Exp. 6 falls to 0.0119, indicating that the absolute and relative reduction in strain capacity is substantially more pronounced for the glass fiber system. Specifically, Exp. 6 experiences a reduction of approximately 44.1% in strain-to-failure compared to only 12.0% for Exp. 16, highlighting the greater vulnerability of the glass fiber composite to impact-induced damage mechanisms such as matrix cracking and fiber-matrix debonding, which precipitate earlier onset of localized failure and reduced ductility. This significant decline in strain capability for Exp. 6 contrasts with the more modest reduction observed in Exp. 16, suggesting that while carbon fiber composites are less ductile in their pristine state, they exhibit comparatively greater retention of deformation capacity post-impact. This observation underscores the complex interplay between fiber architecture, impact damage tolerance, and residual mechanical performance, and reinforces the necessity of considering strain-to-failure alongside strength and stiffness metrics when assessing structural integrity and maintenance strategies for 3D-printed composite components in dynamic loading environments.

Under compressive loading, shown in Figure 18, Exp. 16 demonstrates a pristine compressive strength of 1027.4 MPa, substantially exceeding Exp. 6’s 613.8 MPa. However, residual CAI strength for Exp. 16 declines to 468.72 MPa, a retention of merely 45.6%, whereas Exp. 6 maintains 67.9% of its pristine compressive capacity (417.38 MPa). This significant drop in Exp. 16 arises from severe delamination and fiber micro-buckling under compressive stress states induced by the impact event. Carbon fibers, while offering high axial compressive resistance, are vulnerable to instability phenomena such as kink-band formation once interfacial integrity is compromised, drastically reducing the composite’s compressive modulus and strength. Glass fibers, although inherently less stiff and strong than carbon fibers in compression, exhibit enhanced residual performance due to their higher compliance and better stress redistribution, reducing the severity of kink banding and delaying macroscopic collapse. The compressive stress–strain curves reinforce this interpretation: Exp. 16’s CAI curves exhibit earlier deviation from linearity and lower strain-at-failure than pristine specimens, signifying premature structural degradation and energy absorption through unstable failure mechanisms. Exp. 6, by contrast, displays more gradual post-yield softening behavior, indicative of stable load redistribution and sustained load-carrying capacity post-impact.

Figure 18.

Compressive stress–strain curves of pristine and tensile after impact samples.

Notably, both tensile and compressive stiffness are adversely affected by impact, reflecting partial loss of fiber continuity and matrix integrity. Quantitative analysis of the initial elastic slopes of the post-impact curves indicates that Exp. 6 retains approximately 76.5% and 71.2% of its pristine tensile and compressive moduli, respectively, while Exp. 16 retains only 54.8% and 47.9% of these moduli. This loss in stiffness contributes to lower load transfer efficiency, increased deformability under operational loads, and potential dimensional instability, which are critical considerations for design and life-cycle management of composite components.

These residual property assessments unequivocally demonstrate that while carbon fiber composites provide superior initial strength and stiffness, their damage tolerance under impact conditions is markedly inferior to that of glass fiber systems, leading to a greater urgency for inspection and potential replacement of carbon-based structures post-impact. Conversely, Onyx/Glass fiber composites, despite lower pristine strength levels, exhibit significantly more stable mechanical performance after impact, emphasizing their suitability for applications demanding reliable energy absorption and retained structural integrity. The insights provided by the TAI and CAI analyses substantiate the multi-criteria decision-making outcomes and offer essential guidance for material selection and damage-based maintenance strategies in 3D-printed composite applications subjected to dynamic loading environments.

Limitations and future work

While the present study establishes a robust optimization framework for drop-weight impact performance in 3D-printed continuous fiber-reinforced Nylon composites, the outcomes are inherently constrained by the specific geometries, part sizes, and boundary conditions employed herein. The flat coupon geometry adopted for ASTM D7136 testing ensures controlled and repeatable loading scenarios but does not fully replicate the complex stress distributions experienced in real-world components such as overhead marker balls or intricate structural assemblies, where curvature, variable thickness, and localized features can induce stress concentrations and three-dimensional crack propagation pathways.

Dynamic fracture mechanics principles dictate that variations in geometric discontinuities and thickness gradients alter stress wave interactions, local strain rates, and energy dissipation mechanisms, potentially shifting the dominance of failure modes from delamination and fiber breakage observed in planar specimens to out-of-plane buckling, fiber kinking, or mixed-mode crack growth in complex parts. Moreover, while the Taguchi L27 orthogonal array effectively captures main effects and certain two-factor interactions, higher-order interactions and geometry-specific influences remain unquantified, limiting direct extrapolation of the derived optimal configurations to dissimilar part architectures or loading spectra such as multi-axial impacts, high-frequency vibrations, or sustained fatigue under operational environments. The influence of scale effects further complicates translation, as larger structures introduce inertial amplification of impact loads, non-uniform strain distributions, and potential shifts in dominant energy absorption mechanisms due to structural compliance differences. Hence, while the identified optimal combination, 0° fiber orientation, rectilinear infill, glass fiber reinforcement, and 50% infill density, demonstrates superior impact resistance under controlled conditions, its universal applicability must be cautiously interpreted. Future work should extend the current methodology through high-fidelity finite element simulations incorporating geometry-specific stress fields, multi-scale modeling of fracture processes, and experimental validation on full-scale or complex-shaped prototypes to rigorously assess and adapt the optimal parameters for diverse application scenarios.

Conclusion

This study systematically explored the drop-weight impact behavior of 3D-printed Nylon composites reinforced with continuous carbon, glass, and Kevlar fibers using a Taguchi L27 design and entropy-weighted MCDM optimization. Among 27 configurations, the optimal setup, 0° fiber orientation, rectilinear infill, glass fiber reinforcement, and 50% infill (Exp. 6), achieved the highest peak force (2776.99 N), lowest deformation (11.01 mm), and maximum energy absorption (22.92 J). In contrast, Exp. 16 (45°/gyroid/carbon/30%) recorded the lowest impact resistance (1844.80 N, 38.42 mm, 16.44 J), reflecting a brittle failure mode and ineffective energy dissipation. Post-impact mechanical testing revealed superior damage tolerance in glass fiber composites, with Exp. 6 retaining 73% of its tensile and 67.9% of compressive strength, compared to 52.5% and 45.6% in the carbon fiber counterpart (Exp. 16). Specific Energy Absorption (SEA) analysis highlighted Kevlar-based composites as the most mass-efficient, with Exp. 19 (90°/triangle/Kevlar/30%) achieving the highest SEA of 47.85 J·cm³/g.

SEM analyses provided mechanistic insights: Exp. 6 exhibited matrix shear, fiber bridging, and strong fiber–matrix interlocks, indicative of ductile, energy-absorbing failure. Exp. 16 showed brittle fiber splitting and clean debonding with carbon-rich, low-adhesion fracture surfaces. These contrasting damage modes explain the disparity in mechanical resilience. Fiber-specific infill optimization revealed rectilinear as best for carbon (Exp. 22: SEA = 45.37 J·cm³/g), triangle for Kevlar, and gyroid for glass fibers. These findings underscore the critical synergy between material selection, print architecture, and reinforcement orientation in governing impact response. The study establishes a robust experimental–computational framework that integrates Taguchi design, MCDM, SEA, and microstructural validation to optimize impact performance in 3D-printed continuous fiber composites. The outcomes directly inform the design of lightweight, damage-tolerant structures in infrastructure and aerospace applications. Future work should explore hybrid architectures, complex geometries, and digital twin integration for predictive maintenance and functional grading.