Sage Journals: Discover world-class research

Abstract

The present study investigates the influence of Fused Deposition Modelling (FDM) process parameters on the mechanical properties and surface quality of multi-material PLA-ABS Bi-layer Laminate Structures fabricated at elevated printing speeds. A Taguchi L9 orthogonal array was used to evaluate the effects of printing speed (150–250 mm/s), layer height (0.1–0.3 mm), and infill pattern (concentric, octagram spiral, and Hilbert curve) on tensile properties and surface roughness. The results show that a printing speed of 150 mm/s, a layer height of 0.1 mm, and concentric infill provided optimal mechanical performance, yielding a Young’s modulus of 1252.29 MPa, an ultimate tensile strength of 42.17 MPa, and an elongation at break of 5.79%. Surface roughness analysis indicated the minimum roughness (R_a = 3.4 µm) at a printing speed of 250 mm/s and a layer height of 0.2 mm using concentric infill. Statistical evaluation using mean effect plots, signal-to-noise ratios, and ANOVA revealed that layer height predominantly governs surface quality, while infill pattern and printing speed significantly influence tensile behaviour. Optical fractography and scanning electron microscopy showed that fracture initiation occurred at interlayer interfaces and inter-raster voids. Well-bonded specimens exhibited ductile micro-void coalescence, whereas poorly bonded samples failed by interlayer separation and brittle fracture propagation. These findings demonstrate that appropriate selection of layer height and infill pattern enables rapid fabrication of PLA-ABS Bi-layer Laminate components with improved mechanical performance and surface finish.

Keywords

Optimization FDM PLA ABS Bi-layer laminate Taguchi

Introduction

Additive Manufacturing (AM) enables the fabrication of three-dimensional components through layer-by-layer material deposition directly from computer-aided design (CAD) models.^1,2 Compared with conventional subtractive manufacturing, AM offers high material efficiency, enhanced design freedom, and the ability to fabricate complex geometries with minimal waste.^3,4 Consequently, AM has emerged as a key enabling technology for rapid prototyping and functional components within the Fourth Industrial Revolution.^5,6 Among various AM techniques, such as Stereolithography (SLA), Selective Laser Sintering (SLS), and Binder Jetting, Fused Deposition Modelling (FDM) also referred to as Fused Filament Fabrication (FFF) or Material Extrusion (MEX) is the most widely adopted process due to its cost-effectiveness, operational simplicity, and broad material availability.^7,8 In FDM, a semi-molten thermoplastic filament is extruded through a heated nozzle and deposited along predefined toolpaths, where it solidifies and bonds with previously deposited layers.^9,10 However, the mechanical performance of FDM-fabricated parts is highly sensitive to process parameters and thermal history, leading to anisotropic properties that require careful optimization for industrial us.¹¹

FDM is compatible with a wide range of thermoplastics, among which PLA and ABS are the most commonly used. PLA is valued for its ease of processing, dimensional accuracy, and biodegradability.¹² But its low thermal stability and limited ductility restrict its application in load-bearing components.^13,14 In contrast, ABS exhibits superior toughness, impact resistance, and thermal stability, although it is prone to warping and shrinkage during cooling, necessitating controlled processing conditions.^15,16 As a result, ABS is generally preferred for mechanically demanding functional parts, whereas PLA is primarily used in biomedical and prototyping applications.¹⁷ The specific applications of these materials are summarized in Table 1. The mechanical performance of FDM components is significantly influenced by the internal mesostructured determined by the infill pattern. Akhoundi and Behravesh established that the infill pattern markedly affects tensile and flexural properties, with concentric patterns frequently producing enhanced strength due to improved raster alignment with loading vectors.¹⁸ Recent research by Gohar et al. have emphasized the significant influence of cell shape on the performance of sandwich systems.¹⁹ Heidari-Rarani et al. underscored the necessity of controlling layer height and printing speed to enhance the tensile qualities of PLA-based components.²⁰ This research expands upon these principles by exploring the domain of high-speed dual material printing.

Table 1.

Industrial and functional applications of ABS and PLA materials.

Material	Key properties	Specific applications	Citation
ABS	High impact strength, toughness, heat resistance, durability	Automotive components (bumpers, dashboards), aerospace parts, electronic housings, consumer goods (e.g., LEGO bricks), and functional marine spare parts	^8,15,21
PLA	Biodegradable, biocompatible, high stiffness, low melting point	Biomedical scaffolds, tissue engineering, drug delivery systems, rapid prototyping, artistic models, and sustainable packaging	^9,12,22
Composite/multi-material	Customizable mechanical properties, enhanced functionality	Sandwich structures for crashworthiness, soft robotics, multi-material sensors, and vibration-damping components combining stiffness and flexibility	^16,23,24

Parametric optimization of FDM processes has been widely investigated to improve the mechanical performance and surface quality of printed parts.²⁵ The final properties of FDM components depend on a complex interaction of parameters such as layer height, infill density, raster angle, printing speed, and nozzle temperature.^13,25 The mesostructured generated by slicer settings strongly influences tensile behaviour, with raster orientation and infill density identified as dominant factors.²⁶ Maximum tensile strength is generally achieved at 100% infill and raster angles aligned with the loading direction, whereas perpendicular raster orientations often result in reduced strength due to weaker inter-raster bonding. Interlayer adhesion is another critical factor governing mechanical performance, as poor bonding between layers can lead to premature failure under tensile loading.^13,27 SEM-based fracture analyses indicate predominantly ductile failure in PLA, while ABS exhibits brittle or mixed-mode fracture depending on processing conditions.^8,11 Recent studies on multi-material printing, including PLA/ABS Bi-layer Laminate, emphasize the need for precise process control to ensure adequate interfacial bonding between dissimilar materials.^4,21 While extensive research exists on FDM parameters, studies focusing on high-speed multi-material printing (>150 mm/s) remain limited. Most prior work, such as the interface studies by Khan et al. and surface optimization by Kechagias et al., was conducted at standard speeds (30–60 mm/s).^26,28 This study addresses this gap by optimizing the interfacial bonding and mechanical performance of PLA-ABS Bi-layer laminates specifically under high-velocity deposition regimes. Unlike fibre-reinforced composites which rely on matrix-fibre transfer, this work isolates the thermal fusion efficacy of incompatible polymers (amorphous ABS vs semi-crystalline PLA) when deposited at speeds up to 250 mm/s.²⁵

Surface roughness remains a key quality concern in FDM due to the inherent stair-stepping effect of layer-by-layer deposition.²⁸ Layer height has been consistently reported as the most influential parameter, with reduced layer height improving surface finish.^22,29 Printing speed and nozzle temperature also affect surface quality, as improper thermal conditions and excessive speeds can cause defects such as stringing, vibration-induced waviness, and surface irregularities.³⁰ Despite extensive research, systematic optimization of FDM at high printing speeds is still limited. Most studies focus on moderate speeds, whereas industrial applications increasingly demand higher production rates.^8,14 High-speed printing introduces challenges such as reduced interlayer adhesion, under-extrusion, and increased surface roughness.^29,31 Moreover, limited work has examined the combined influence of high printing speed, infill pattern, and layer height on both mechanical and surface properties of PLA and ABS using robust statistical approaches.^8,9

To address these gaps, the present study focuses on optimizing key FDM parameters to enhance the manufacturability of PLA and ABS at elevated printing speeds.³² The effects of layer height, infill pattern, and high printing speed on tensile strength, surface roughness, and fracture behaviour are investigated using a Taguchi L9 DoE approach.^9,33 The outcomes aim to establish practical guidelines for high-speed FDM, balancing part quality with production efficiency.

Materials and methods

Material

This study selected ABS and PLA as the constituent materials for the bi-laminate constructions due to their complementing mechanical properties and widespread industrial use. PLA generally provides elevated tensile strength (∼50–60 MPa) and rigidity, although it is characterized by brittleness and inadequate thermal resistance.^16,34 Conversely, ABS offers enhanced toughness, impact resistance, and thermal stability.²¹ The integration of a rigid PLA base with a resilient ABS top layer seeks to create a composite that alleviates the inherent drawbacks of each polymer. Moreover, these materials demonstrate stable rheological properties conducive to high-speed extrusion (150–250 mm/s), in contrast to flexible elastomers or high-temperature thermoplastics, which frequently encounter printability issues at these velocities.^15,16 When arranged in a Bi-layer Laminate configuration, this combination enables improved functional performance, such as enhanced environmental durability, by shielding the hygroscopic and biodegradable PLA layer with moisture-resistant ABS.³⁵ The mechanical and physical properties of the filaments are summarized in Table 2. To ensure optimal print quality and mechanical performance, the filaments were dried prior to fabrication to minimize defects caused by moisture absorption. ABS filaments were dried in a filament dryer (Bambu X1 Carbon) at 80°C for 4–6 h before extrusion. PLA, due to its organic and hygroscopic nature, exhibits higher moisture absorption than ABS and was therefore dried at 60°C for 6–8 h. Although ABS shows comparatively lower moisture uptake, it can still degrade under humid conditions.^36,37 In experimental setups, both PLA and ABS filaments are commonly vacuum dried prior to filament synthesis to eliminate residual moisture and volatiles, as illustrated in Figure 1. The overall experimental workflow is depicted in Figure 1. The process begins with the preparation of PLA and ABS filaments (densities of 1.24 g/cm³ and 1.04 g/cm³, respectively, with a diameter of 1.75 mm), followed by filament drying, CAD modelling, and computer-aided slicing.²⁸ Dual-material dog-bone specimens were then fabricated using an FDM printer equipped with an extruder, nozzle, and heated bed. Surface roughness was measured using a digital profilometer, tensile testing was performed on a universal testing machine with adjustable crosshead speed, and fracture morphology was examined using SEM to identify interfacial features and infill-related defects at the PLA-ABS interface.³⁸

Table 2.

Mechanical properties of the PLA and ABS filament.

Properties	PLA	ABS
Density g/cm³	1.24	1.04
Diameter (mm)	1.75	1.75
Ultimate tensile strength (σ_UTS)	50 MPa^25,39	30 MPa⁴⁰
Young’s modulus (E)	3.5 GPa^16,25	1.5 GPa^16,25
Elongation at break (Ɛ_f)	3–10%^16,25	10%^16,25

Figure 1.

Schematic workflow of fabrication and characterization of dual-material PLA-ABS tensile specimens by fused filament fabrication.

Printing parameters

The selection of processing parameters is critical in Fused Deposition Modelling (FDM), as these variables directly govern the mechanical integrity, surface morphology, and dimensional accuracy of the fabricated Bi-layer Laminate. Based on the identified literature gap related to high-speed manufacturing, three key processing parameters, Layer Height (LH), Infill Pattern (IP), and Printing Speed (PS), were selected for optimization using the Taguchi method.⁴¹ The specific levels of these parameters are summarized in Table 3. Layer height (0.1–0.3 mm) is one of the most influential parameters affecting surface roughness and dimensional resolution.²⁸ A lower layer height (0.1 mm) is generally associated with improved surface finish and reduced porosity.⁴ In contrast, a higher layer height (0.3 mm) is often employed to decrease printing time and has been reported to enhance tensile strength in certain composite systems due to increased thermal mass effects.^7,42 The infill pattern (Concentric, Octagram spiral, and Hilbert curve) significantly influences the internal mesostructured and, consequently, the mechanical behaviour of printed components. The Concentric pattern is known to promote ductile fracture behaviour and higher tensile strength by aligning raster with the external contours.⁴³ The Hilbert curve and Octagram spiral were selected to examine the influence of complex and continuous tool paths under high-speed printing conditions. However, intricate patterns such as the Hilbert curve may induce vibrations due to frequent directional changes, potentially altering the fracture mechanism from ductile to brittle.⁴³ Printing speed (150–250 mm/s) is a critical factor in achieving high productivity. While most previous studies restrict printing speeds to 20–90 mm/s to ensure part quality,⁸ the present work focuses on the high-speed capability of FDM. Earlier optimization studies have reported printing speeds up to 125 mm/s for ABS.⁴⁴ Investigating higher speeds in the range of 150–250 mm/s addresses industrial demands for rapid manufacturing, despite the associated risks of reduced interlayer adhesion and increased surface roughness at elevated velocities.^4,8

Table 3.

Variable parameters and factor levels.

Variable parameters	Factor representation	Level
Variable parameters	Factor representation	1 (low)	2 (medium)	3 (high)
Layer height (mm)	LH	0.1	0.2	0.3
Infill pattern	IP	Concentric	Octagram spiral	Hilbert curve
Printing speed (mm/s)	PS	150	200	250

All printing parameters other than the selected variable parameters were kept constant in accordance with manufacturer guidelines and optimized values reported in the literature to isolate the effect of the studied variables. An overview of these fixed parameters is provided in Table 4. For PLA, a nozzle temperature of 210°C was selected, as it lies within the recommended processing range of 190–220°C, ensuring adequate melt viscosity without thermal degradation.²⁵ For ABS, a nozzle temperature of 240°C was used to ensure sufficient melting and improved interlayer bonding.⁴⁵ While a 100% infill density was established for all specimens to emulate solid components, the Infill Pattern was altered to examine the dynamic impacts of toolpath trajectory at elevated printing velocities. Intricate designs such as the Hilbert Curve necessitate frequent directional alterations, potentially causing vibrations and under-extrusion at velocities over 150 mm/s, therefore resulting in voids despite a directive for 100% density. Conversely, the Concentric pattern employs extensive, uninterrupted deposition trajectories that preserve uniform nozzle pressure, essential for attaining genuine solidity in high-speed FDM.^16,44 A standard 0.4 mm nozzle was employed, as commonly adopted in FDM studies due to its balanced compromise between dimensional accuracy and deposition rate.²⁸ For PLA, a bed temperature of 60°C was maintained to enhance adhesion and minimize warping.⁷ The infill patterns and layer height variations of the tensile specimens are illustrated in Figure 2. The octagram spiral, concentric, and Hilbert curve infill geometries are shown in Figure 2(a)–(c). The dual-material configuration, with layer dimensions of 2–2 mm (PLA-ABS) and a total height of 4 mm, is highlighted in Figure 2(d)–(f), emphasizing the structural design adopted for optimal mechanical performance. Dog-bone specimens composed of PLA-ABS materials were produced utilizing a Bambu Lab X1 Carbon FDM printer. This printer utilizes a single-nozzle hot-end in conjunction with an Automatic Material System (AMS), in contrast to dual-extruder systems. The AMS enables multi-material production by automatically retracting one filament and introducing another into the same 0.4 mm nozzle, facilitating seamless material transitions within a single layer or across layers.³⁷ The printing sequence was established as PLA (bottom) followed by ABS (top) to ensure thermal stability. Since PLA has a lower glass transition temperature (Tg ≈ 60°C) compared to ABS (Tg ≈ 105°C), printing PLA on the build plate (maintained at 60°C) ensures good bed adhesion without softening.⁴⁶ If ABS were printed first, the required higher bed temperature (90–100°C) would likely cause the subsequent PLA layers to exceed their Tg, leading to severe deformation and poor dimensional accuracy.¹⁶ Parameters such as nozzle temperature (210°C for PLA, 240°C for ABS) and raster angle were fixed to isolate the effects of printing speed and infill pattern. High nozzle temperatures were selected based on literature to ensure sufficient flow and intermolecular diffusion at elevated printing speeds (up to 250 mm/s).^16,31The cooling fan was disabled to prevent thermal shock and warping in the ABS layers.³²

Table 4.

Constant parameters during printing.

S. no	Constant parameters	Values
1.	Top/bottom surface layer	2
2.	Top/bottom IP	Monatomic
3.	Wall loop	2
4.	Build plate	PEI
5.	Nozzle temperature	210°C(PLA) and 240°C (ABS)
6.	Bed temperature	60°C
7.	Nozzle diameter	0.4 mm
8.	ID	100%
9.	Print orientation	Flat
10.	Fan	Off

Figure 2.

Infill pattern and layer height orientation (a) octagram spiral, (b) concentric, (c) Hilbert curve, (d) LH 0.1 mm, (e) LH 0.2 mm and (f) LH 0.3 mm.

Design matrix

A Design of Experiments (DoE) approach was employed to systematically investigate the influence of selected process parameters, printing speed, layer height, and infill pattern, on the mechanical and surface properties of the printed Bi-layer Laminate.⁴⁷ The experimental matrix was designed using a Taguchi L9 orthogonal array, which accommodates three factors at three levels. Although the Taguchi L9 (3⁴) array is a fractional factorial design and does not explicitly assign interaction columns due to saturation, it is highly efficient for evaluating the main effects of the selected parameters. The potential influence of interactions is indirectly addressed through Signal-to-Noise (S/N) ratio analysis, which identifies robust parameter settings that minimize performance variability arising from unmodeled interactions and process noise.^1,7 Confirmation experiments were performed to verify that the optimal factor levels predicted from the main effects analysis are valid in practice. Table 5 lists the combinations of coded levels (1, 2, and 3) and their corresponding real parameter values for the nine experimental trials. The Taguchi L9 array was chosen for its statistical efficiency, reducing the number of required experiments from 27 in a full factorial design to 9, thereby saving both time and material resources.⁴⁸ By minimizing the influence of process variability and isolating noise effects, this approach ensures statistical robustness and enables reliable Analysis of Variance (ANOVA) for key quality characteristics such as surface roughness and circularity. Furthermore, the balanced orthogonality of the L9 array allows the individual effects of each process parameter to be evaluated independently, avoiding confounding effects.³³ This design also supports Signal-to-Noise (S/N) ratio analysis, facilitating the identification of optimal parameter combinations that enhance mechanical performance and reduce surface-related defects.

Table 5.

Design matrix for Taguchi L9 orthogonal array.

Run order	Coded values			Actual values
Run order	PS	LH	IP	PS (mm/s)	LH (mm)	IP
1.	1	1	1	150	0.1	Concentric
2.	1	2	2	150	0.2	Octagram spiral
3.	1	3	3	150	0.3	Hilbert curve
4.	2	1	2	200	0.1	Octagram spiral
5.	2	2	3	200	0.2	Hilbert curve
6.	2	3	1	200	0.3	Concentric
7.	3	1	3	250	0.1	Hilbert curve
8.	3	2	1	250	0.2	Concentric
9.	3	3	2	250	0.3	Octagram spiral

Sample preparation

Tensile specimens were fabricated in accordance with the ASTM D638 Type IV standard, which is widely used for evaluating rigid thermoplastic polymers.²⁷ The specimen geometry, shown in Figure 1, has an overall length of 115 mm to ensure adequate gripping during testing and a gauge length (L) of 19 mm, which serves as the primary deformation region.^23,31 Rounded transitions with defined radii of curvature (R6 and R14) were incorporated to minimize stress concentrations and promote fracture within the gauge section. The digital workflow began with three-dimensional modelling in AutoCAD (Computer-Aided Design), followed by exporting the geometry in STL (Stereolithography) format. This standard file format discretizes the CAD model into triangulated surface meshes, ensuring compatibility with additive manufacturing systems.³⁰ The STL files were then imported into Bambu Slicer, where slicing was performed to generate G-code (Geometric code) containing machine-readable instructions for nozzle movement, extrusion rate, and material transitions.⁴⁹ The configuration of the Bambu X1 Carbon printer is shown in Figure 3(a). Key components include a heated PEI (polyetherimide) build plate to enhance adhesion, an enclosed thermal chamber to maintain a stable printing environment, a Micro Lidar depth sensor for real-time surface detection and first-layer calibration, a precision servo-controlled filament feeding system, and a multi-material hot-end assembly with independent nozzles and cooling fans. This configuration enables simultaneous extrusion of PLA and ABS. The specimens were fabricated using a bi-layer laminate strategy rather than a traditional sandwich core configuration.^23,35 To ensure equal volumetric contribution (50% PLA/50% ABS), the slicing software was configured to deposit PLA for the bottom half of the specimen thickness (0-2 mm), followed by ABS for the top half (2-4 mm). The Automatic Material System (AMS) executed the filament change precisely at the mid-plane layer. PLA was used to fabricate the bottom layers of the specimens (Figure 3(b)), while ABS was deposited on the top layers (Figure 3(c)). A total of nine tensile specimens with alternating material configurations (labelled 1–9 in Figure 3(b) and (c)) were printed sequentially. The slicing parameters were optimized to ensure 100% infill density, precise nozzle positioning, and seamless material switching between PLA and ABS. This approach resulted in well-defined interfacial regions suitable for tensile testing and comparative mechanical characterization across the material variation matrix.^16,21

Figure 3.

Fabrication of ASTM type IV tensile samples: (a) Bambu X1 carbon 3D printer, (b) bottom side showing PLA and (c) top side showing ABS material.

Tensile and surface roughness testing

Figure 4 illustrates the tensile testing procedure. Tensile tests were conducted using a Zwick Roell Z010 universal testing machine. Figure 4(a) shows the complete testing setup, including the control panel connected to a dedicated computer interface for real-time data acquisition and load displacement monitoring, upper and lower jaws for specimen gripping, and an emergency stop button for safety. To eliminate atmospheric effects such as thermal expansion or humidity-induced degradation, all nine ASTM D638 Type IV tensile specimens with different print speeds, layer heights, and infill patterns were tested under controlled room-temperature conditions.³⁰

Figure 4.

Tensile testing: (a) universal testing machine hardware, (b) specimen positioning and load direction of tensile testing, (c) operating system of surface profilometer, (d) specimen jaw assembly of tensile testing and (e) surface profilometer.

To ensure uniaxial stress application, specimens were carefully aligned with the loading direction and positioned between the upper and lower jaws, as shown in Figure 4(b). Pneumatic grips with serrated surfaces were used to provide stable clamping, minimize slippage during elongation, and accommodate the tapered specimen geometry, as illustrated in Figure 4(d). A constant crosshead speed of 1 mm/min was applied to maintain quasi-static loading conditions and accurately capture E (MPa), σ_UTS (MPa), and Ɛ_f (%) without dynamic effects. Three replicate specimens were tested for each parameter combination to improve statistical reliability, resulting in a total of 27 tests based on the L9 orthogonal array. The average values of tensile tests are reported in Table 6.⁴³ Surface roughness measurements were performed on the gauge section of the tensile specimens using a contact stylus surface profilometer, as shown in Figure 4(c) and (e). Each cleaned and conditioned specimen was securely placed on the sliding stage with the stylus aligned parallel to the tensile loading direction.⁴ The stylus was brought into contact with the surface under a controlled force and scanned along a predefined path. The recorded vertical displacements were filtered to remove waviness, and the software calculated the line roughness from multiple traces, from which average values were obtained (e.g., Sample no. 1, R_a = 4.14 µm).^3,9

Table 6.

Total result of the full experimentation.

Run order	Input			Output
	Input			Tensile result			Surface roughness
	PS (mm/s)	LH (mm)	IP	E (MPa)	σ_UTS (MPa)	Ɛ_f (%)	R_a (µm)
1.	150	0.1	Concentric	1252.29	42.17	5.79	4.14
2.	150	0.2	Octagram spiral	992.36	35.60	4.98	3.70
3.	150	0.3	Hilbert curve	857.55	34.65	5.07	6.3
4.	200	0.1	Octagram spiral	1041.19	31.60	4.17	3.58
5.	200	0.2	Hilbert curve	1026.48	32.83	4.62	4.6
6.	200	0.3	Concentric	845.88	35.00	5.06	14.67
7.	250	0.1	Hilbert curve	444.6	18.39	4.88	4.48
8.	250	0.2	Concentric	1082.48	41.91	5.39	3.4
9.	250	0.3	Octagram spiral	785.29	33.27	5.56	10.43

Results and discussions

Tensile and surface roughness result

Table 6 summarizes the tensile and surface roughness responses of the ABS-PLA Bi-layer Laminate specimens fabricated using the L9 matrix. The key response parameters reported include E, σ_UTS, Ɛ_f, and R_a. Among all conditions, the combination of printing speed (150 mm/s), layer height (0.1 mm), and concentric infill pattern (Sample 1) yields the best overall performance, with E = 1252.29 MPa, σ_UTS = 42.17 MPa, and a maximum Ɛ_f = 5.79%, while maintaining a relatively low surface roughness (R_a ≈ 4.14 μm).⁵⁰ In contrast, increasing the layer height to 0.3 mm or employing Hilbert and octagram spiral infill patterns generally leads to a reduction in both E and σ_UTS. In extreme cases, such as Sample 6 (200 mm/s, 0.3 mm, concentric), the mechanical performance deteriorates significantly, with a very low modulus (E = 444.6 MPa) and strength (σ_UTS = 18.39 MPa), despite moderate ductility (Ɛ_f ≈ 4.88%) and high surface roughness (R_a ≈ 14.67 μm). This behaviour indicates poor interlayer bonding at coarse layer heights and the presence of internal defects, which also explains the elevated Ra values observed for Samples 3, 6, 7, and 9.⁵¹

At a higher printing speed of 250 mm/s, the Bi-layer Laminate structure retains relatively high strength for selected infill conditions. For example, Sample 8 (concentric infill, layer height = 0.2 mm) exhibits σ_UTS = 41.91 MPa, E = 1082.48 MPa, Ɛ_f = 5.39%, and a low R_a ≈ 3.40 μm. This result demonstrates that concentric infill combined with thin-to-medium layer heights promotes effective PLA-ABS load transfer and improved surface quality. Conversely, Hilbert and octagram spiral patterns at the same speed (Samples 7 and 9) show reduced E and σ_UTS, along with moderate to high R_a, confirming that complex infill trajectories and thick layers promote stress concentrations, premature debonding, and rougher surfaces in this Bi-layer Laminate system.⁵² A critical trade-off was observed between production efficiency and mechanical performance. While the lowest speed (150 mm/s) yielded the maximum tensile strength of 42.17 MPa (sample1), increasing the printing speed to 250 mm/s (Run 8) resulted in a tensile strength of 41.91 MPa. This represents a negligible strength reduction of only ∼0.6%, whereas the print time was reduced by approximately 40%. These finding tests earlier studies that reported important property degradation above 100 mm/s⁸. The results indicate that with the ‘Concentric’ infill pattern, industrial-grade high-speed printing (250 mm/s) is viable for PLA-ABS laminates without compromising structural integrity.⁵³

The tensile stress–strain curves in Figure 5 further support these trends. At a printing speed of 150 mm/s (Figure 5(a), Samples 1–3), all curves exhibit an initial linear elastic region followed by gradual strain hardening up to σ_UTS. However, Sample 1 shows a steeper slope, higher peak stress, and greater strain at fracture, indicating a stiffer and more ductile response when concentric infill and minimum layer height are employed.⁵⁴ Samples 2 and 3 (octagram spiral and Hilbert) display lower slopes and earlier softening, consistent with stress localization within the core infill. At printing speeds of 200 and 250 mm/s (Figure 5(b) and (c)), concentric infill samples (Samples 6 and 8) demonstrate higher σ_UTS and extended plastic deformation prior to failure, whereas Samples 5 and 7 show reduced modulus and premature load drops due to weak interfacial bonding. Overall, the PLA-ABS Bi-layer Laminate structure behaves as a semi-ductile laminate, in which stiff PLA skins dominate elastic loading while the ABS core accommodates strain, with optimized thin layers, concentric infill, and moderate-to-high printing speeds providing a favourable balance between E, σ_UTS, and Ɛ_f.⁵⁵ The bi-layer design (50% PLA/50% ABS) produces a non-uniform stress distribution due to the stiffness disparity between the layers. Given that PLA has a greater Young’s Modulus (E ≈ 3.5 GPa) than ABS (E ≈ 2.0 GPa),¹⁹ the PLA substrate predominantly supports the tensile load in the elastic region. The interface operates as a shear transfer zone; effective load-bearing capacity of up to 42 MPa demonstrates that the interfacial shear strength surpassed the stress differential between the strained PLA and ABS layers, thereby averting premature delamination.¹²

Figure 5.

Tensile stress–strain curves of ABS-PLA Bi-layer Laminate specimens fabricated with different printing parameters: (a) PS = 150 mm/s (Samples 1–3), (b) PS = 200 mm/s (Samples 4–6), and (c) PS = 250 mm/s (Samples 7–9).

Figure 6 presents the surface profile traces measured along the tensile gauge length for all nine samples and confirms that conditions yielding superior mechanical performance also produce smoother surfaces (Figure 6(a)–(i)). Samples 1, 2, 4, and 8 exhibit relatively stable, low-amplitude roughness oscillations around the mean line (R_a ≈ 3–4 μm), corresponding to uniform filament deposition and good layer coalescence in the PLA-ABS Bi-layer Laminate. In contrast, Samples 3, 6, 7, and 9 show pronounced peaks and valleys, consistent with their higher R_a values and degraded tensile behaviour, particularly for Sample 6. These surface profiles indicate the presence of voids, under-extrusion marks, and interlayer gaps associated with thick layers and non-concentric infill paths.⁵⁶

Figure 6.

Surface roughness profiles of ABS-PLA Bi-layer laminate specimens: (a) samples 1, (b) samples 2, (c) samples 3 (d) samples 4, (e) samples 5, (f) samples 6, (g) samples 7, (h) samples 8, and (i) samples 9.

Figure 7 graphically summarizes the result of tensile testing and roughness measurement. Figure 7(a) highlights the superior E of Sample 1, attributed to thin layer height and concentric infill promoting uniform load transfer, whereas the reduced modulus of Sample 5 is associated with Hilbert-induced defects. Figure 7(b) confirms that Samples 1 and 8 achieve peak σ_UTS through effective skin loading, largely independent of printing speed. Figure 7(c) demonstrates that the ABS core sustains high Ɛ_f (>5.5%) when concentric infill is used. Finally, Figure 7(d) identifies the optimal R_a of Sample 8, attributed to enhanced filament fusion at higher printing speed, in contrast to the severe surface waviness of Sample 6 caused by excessive layer height.³⁵ To validate the Bi-laminate approach, the optimized PLA-ABS specimen (sample 1 σ_UTS = 42.17 MPa) was compared to standard mono-material data reported in the literature under similar FDM conditions. The laminate strength exceeds that of pure ABS specimens, which typically range between 30 and 35 MPa,^16,25 but remains lower than pure PLA specimens, which can reach 50–60 MPa.¹⁹ This indicates that the PLA-ABS Bi-laminate successfully synergizes the materials, it offers higher strength than a pure ABS part while providing the superior thermal and impact resistance of the ABS surface, which pure PLA lacks.⁵⁷

Figure 7.

Tensile and surface roughness data: (a) E (MPa), (b) σ_UTS (MPa), (c) Ɛ_f (%), (d) Ra (μm).

Taguchi analysis

Mean affect plots

Figure 8 presents the mean effect plots obtained from the Taguchi L9 analysis, highlighting the individual influence of printing speed, layer height, and infill pattern on the mechanical and surface characteristics of the 3D-printed Bi-layer Laminate structure. These plots show the average response at each factor level and assist in identifying optimal parameter settings in the absence of interaction effects.²⁷ Figure 8(a) shows the mean effects on E (MPa). A lower printing speed of 150 mm/s yields the maximum stiffness, which can be attributed to improved interlayer bonding and reduced void formation under slower deposition conditions. Layer height exhibits an increasing trend, with the highest E observed at 0.3 mm, as thicker layers enhance load transfer despite potential anisotropy.³⁰ The infill pattern significantly affects stiffness, with concentric and octagram configurations outperforming the Hilbert pattern, which introduces stress concentrations due to its complex toolpath.⁴³

Figure 8.

Main effect plots from Taguchi L9 analysis showing individual influences of printing speed, layer height, and infill pattern on (a) Young’s modulus (MPa), (b) ultimate tensile strength (MPa), (c) elongation at break (%), and (d) surface roughness (μm).

Figure 8(b) illustrates the response of σ_UTS (MPa). The maximum strength is achieved at a printing speed of 150 mm/s, resulting from enhanced filament fusion and reduced thermal gradients during extrusion. The optimal layer height is 0.2 mm, balancing interlayer bonding while avoiding over-extrusion effects at higher values. Among infill patterns, the octagram spiral provides the highest σ_UTS, followed by the concentric pattern, whereas the Hilbert pattern shows inferior performance, emphasizing the importance of continuous deposition paths for load transfer.¹² Figure 8(c) shows the variation in Ɛ_f (%), where layer height is the dominant factor. Ductility increases progressively from 0.1 to 0.3 mm due to reduced strain localization in thicker layers. Printing speed displays a U-shaped trend, with higher strain capacity at 150 and 250 mm/s compared to 200 mm/s. The concentric infill pattern produces the highest Ɛ_f, while the Hilbert and octagram patterns exhibit lower ductility due to non-uniform strain distribution.²⁹

Figure 8(d) presents the mean effects on R_a (µm). Higher printing speed (250 mm/s) reduces surface roughness by promoting rapid solidification and layer stabilization, thereby limiting sagging. A layer height of 0.2 mm provides the optimal surface finish by avoiding fine-layer stair-stepping at 0.1 mm and excessive waviness at 0.3 mm. The Hilbert infill pattern increases R_a due to frequent directional changes and overhangs, whereas concentric and octagram patterns yield smoother surfaces. Overall, the trends in Figure 8 indicate that a printing speed of 150 mm/s and a layer height of 0.2–0.3 mm combined with octagram or concentric infill enhance mechanical performance, while a printing speed of 250 mm/s and a layer height of 0.2 mm favour improved surface finish, enabling balanced parameter selection for hybrid optimization.³¹

Single to noice ratio plots

Figure 9 presents the signal-to-noise (S/N) ratio plots obtained from the Taguchi L9 orthogonal array analysis, corresponding to the printing parameters and experimental responses previously shown in the mean effect plots. Unlike mean effect plots, the S/N ratios (higher-the-better criterion) emphasize parameter robustness by maximizing the signal (mean response) relative to noise (variation), thereby guiding the selection of stable processing conditions. The trends observed in Figure 9 are consistent with those in Figure 8 and highlight operating conditions that minimize response variability. Figure 9(a) illustrates the S/N ratios for E (MPa).⁹ A printing speed of 150 mm/s yields the highest S/N ratio, indicating reduced stiffness variability due to improved cooling and inter-layer adhesion. A layer height of 0.3 mm further enhances robustness by promoting a more homogeneous microstructure. Among the infill patterns, concentric and octagram spirals outperform the Hilbert pattern, as the latter introduces higher variability due to its complex toolpath and non-uniform extrusion.³¹

Figure 9.

Signal-to-noise (S/N) ratio plots from Taguchi L9 analysis showing robustness effects of printing speed, layer height, and infill pattern on (a) Young’s modulus (MPa), (b) ultimate tensile strength (MPa), (c) elongation at break (%), and (d) surface roughness (μm).

Figure 9(b) shows the S/N analysis for σ_UTS (MPa). The maximum S/N ratio is achieved at a printing speed of 150 mm/s, which enhances strength stability through improved inter-layer bonding. A layer height of 0.2 mm provides an optimal balance between deposition accuracy and defect formation. The octagram spiral infill exhibits the highest robustness, attributed to its symmetric load-transfer paths, whereas the Hilbert pattern increases variability due to irregular fibre bridging. Figure 9(c) presents the S/N ratios for Ɛ_f (%).³⁸ A layer height of 0.3 mm is optimal for ductility stability, as thicker layers reduce strain inconsistencies through improved matrix coalescence. A printing speed of 150 mm/s is preferred, since higher speeds increase variability due to warping effects. The concentric infill pattern provides the highest S/N ratio, indicating more uniform deformation and reduced variability in crack initiation compared to Hilbert and octagram patterns.⁴⁵ Figure 9(d) depicts the S/N ratios for R_a (μm) using the lower-the-better criterion. A printing speed of 250 mm/s results in the most stable surface finish due to rapid layer solidification, while a layer height of 0.2 mm improves Ra consistency by limiting resolution-related defects. Non-Hilbert infill patterns, particularly concentric and octagram, produce lower noise in surface roughness due to simplified toolpaths. Overall, the S/N analysis confirms that a printing speed of 150 mm/s, a layer height of 0.2–0.3 mm, and octagram or concentric infill patterns yield robust mechanical and surface properties.⁵⁸

ANOVA analysis

Tables 7 and 8 present the ANOVA results obtained from the Taguchi L9 design. Each statistical term serves a specific purpose: DF represents degrees of freedom, Seq SS denotes the sequential sum of squares attributed to each factor, Adj SS and Adj MS represent the adjusted sum of squares and mean square after accounting for other factors, F is the F-statistic used to evaluate the significance of factor effects relative to experimental noise, and P indicates the probability of rejecting the null hypothesis. The coefficients R² and R²_adj describe the overall model fit, representing the explained variance and its adjustment for added factors, respectively.³³ For σ_UTS, Table 7 shows that the infill pattern contributes the highest Seq SS value of 184.71, accounting for approximately 48% of the total variation and yielding the highest F value of 2.25. This dominance results in an R² value of 78.64%, which directly confirms the trend observed in Figure 8(b). Although the corresponding P value (0.308) exceeds 0.05, this is attributed to the limited error degrees of freedom (DF = 2) rather than a lack of influence. The S/N analysis in Table 8 further supports this observation, where the infill pattern shows the highest Seq SS (14.81, 42%) and F value of 1.82, producing an R² of 77.19% and validating the robustness peak observed in Figure 9(b).³³ For E (MPa), Table 7 indicates comparable contributions from printing speed and infill pattern, with Seq SS values of 113,426 and 122,767 and similar F values of 0.92 and 0.99, respectively. This balanced contribution results in an R² of 70.86%, consistent with the comparable slopes observed in Figure 8(a). The relatively high residual MS limits R²_adj to 0.00%, which is typical for modulus measurements. The S/N analysis in Table 8 maintains this parity, with F values close to unity and an R² of 70.43%, supporting the plateau-like response shown in Figure 9(a).⁵⁹

Table 7.

Analysis of variance for means.

	Source	DF	Seq SS	Adj SS	Adj MS	F	P	R²	R² (adj)
σ_UTS (MPa)	PS (mm/s)	2	62.04	62.04	31.02	0.76	0.57	78.64%	14.57%
	LH (mm)	2	55.71	55.71	27.85	0.68	0.596
	IP	2	184.71	184.71	92.36	2.25	0.308
	Residual error	2	82.14	82.14	41.07
	Total	8	384.6
E (MPa)	PS (mm/s)	2	113426	113426	56713	0.92	0.52	70.86%	0.00%
	LH (mm)	2	63267	63267	31633	0.52	0.66
	IP	2	121906	121906	60953	0.99	0.502
	Residual error	2	122767	122767	61383
	Total	8	421366
Ɛ_f (%)	PS (mm/s)	2	0.8756	0.8756	0.43781	2.64	0.275	82.71%	30.84%
	LH (mm)	2	0.1372	0.1372	0.06861	0.41	0.707
	IP	2	0.5722	0.5722	0.28608	1.73	0.367
	Residual error	2	0.3314	0.3314	0.16568
	Total	8	1.9164
R_a (µm)	PS (mm/s)	2	12.624	12.624	6.312	0.81	0.552	87.10%	48.38%
	LH (mm)	2	84.164	84.164	42.082	5.42	0.156
	IP	2	8.034	8.034	4.017	0.52	0.659
	Residual error	2	15.53	15.53	7.765
	Total	8	120.352

Table 8.

Analysis of variance for single to noise ratios.

	Source	DF	Seq SS	Adj SS	Adj MS	F	P	R²	R² (adj)
σ_UTS (MPa)	PS (mm/s)	2	6.292	6.292	3.146	0.77	0.564	77.19%	8.78%
	LH (mm)	2	6.42	6.42	3.21	0.79	0.559
	IP	2	14.805	14.805	7.403	1.82	0.354
	Residual error	2	8.13	8.13	4.065
	Total	8	35.647
E (MPa)	PS (mm/s)	2	15.612	15.612	7.806	0.98	0.505	70.43%	0.00%
	LH (mm)	2	7.125	7.125	3.562	0.45	0.691
	IP	2	15.146	15.146	7.573	0.95	0.512
	Residual error	2	15.906	15.906	7.953
	Total	8	53.788
Ɛ_f (%)	PS (mm/s)	2	2.7556	2.7556	1.3778	2.91	0.256	83.94%	35.76%
	LH (mm)	2	0.4923	0.4923	0.2461	0.52	0.658
	IP	2	1.6981	1.6981	0.849	1.79	0.358
	Residual error	2	0.9464	0.9464	0.4732
	Total	8	5.8924
R_a (µm)	PS (mm/s)	2	10.585	10.585	5.293	0.55	0.644	88.03%	52.12%
	LH (mm)	2	126.789	126.789	63.394	6.63	0.131
	IP	2	3.201	3.201	1.6	0.17	0.857
	Residual error	2	19.114	19.114	9.557
	Total	8	159.689

For Ɛ_f (%), printing speed emerges as the dominant factor. Table 7 shows the highest Seq SS (0.876) and F value (2.64), while Table 8 reports a corresponding S/N F value of 2.91. These results yield R²_adj values of 30.84% and 35.76%, quantitatively validating the U-shaped response observed in Figures 8(c) and 9(c). The comparatively low residual MS values enable clear detection of this effect, while infill pattern and layer height show secondary influence with F values below 2.⁶⁰ Surface roughness, R_a (µm), is strongly governed by layer height. Table 7 shows that layer height contributes 84.16 Seq SS, accounting for approximately 70% of the total variation, with an F value of 5.42. Similarly, Table 8 reports an even higher contribution of 126.79 Seq SS (79%) and an F value of 6.63. These dominant effects produce the highest R² values of 87.10% and 88.03%, with R²_adj values of 48.38% and 52.12%, respectively. These findings are in excellent agreement with the trends observed in Figures 8(d) and 9(d) and are further supported by the low residual MS values, indicating high surface predictability.⁶¹ The ANOVA results demonstrate strong statistical consistency between quantitative analysis and experimental trends. The Seq SS rankings closely follow the visual spreads in Figures 8 and 9, F values ranging from 0.5 to 6.6 confirm that factor effects exceed experimental noise, and high R² values (69–88%) indicate substantial explanatory power. The moderate R²adj values reflect the limited DF while maintaining statistical credibility, confirming that the Taguchi screening successfully identifies reliable process parameter combinations for industrial fabrication.⁶² The ANOVA revealed distinct dominant factors for different responses. Infill Pattern was the most significant factor for σ_UTS, contributing 47.9% to the total variation, confirming that raster alignment is critical for load bearing.³⁶ Conversely, Layer Height was the dominant factor for R_a, contributing 78.8%, which aligns with the geometric relationship between layer thickness and surface profile.²⁸ Printing Speed showed a significant contribution (∼30%) to Ɛ_f, influencing the thermal bonding history of the polymer chains.

Fractography of tensile fracture surfaces

The macroscopic fracture morphology of the tensile-tested specimens was examined using optical imaging, and the results are shown in Figure 10. Clear differences in fracture characteristics are observed between the ABS and PLA regions. The fracture surface of ABS in Figure 10(a) (Sample 1) exhibits a rough and irregular morphology with uneven edges and visible tearing zones. The crack path is non-linear, showing local deviations that indicate noticeable deformation prior to final failure.⁶³ The ABS side view in Figure 10(c) confirms that fracture occurred within the gauge section, demonstrating proper tensile loading. A slight reduction in cross-sectional area near the fracture region suggests localized plastic deformation. Similarly, the ABS side view of Sample 5 in Figure 10(e) shows fracture within the central region of the specimen. The fracture surface of ABS in Sample 5 (Figure 10(g)) reveals a distinct layered structure, where crack propagation follows a step-like path along the printed layers. This behaviour indicates that interlayer interfaces acted as preferential crack paths, highlighting the influence of layer bonding on the failure behaviour of FDM-fabricated parts.⁶⁴

Figure 10.

Optical and fractographic analysis of specimens. Sample 1: (a) ABS fracture surface (Sample 1), (b) PLA fracture surface (Sample 1), (c) sample 1 ABS side view, (d) sample 1 PLA side view, (e) sample 5 ABS side view, (f) sample 5 PLA side view, (g) ABS fracture surface (sample 5), (h) PLA fracture surface (sample 5).

In contrast, the PLA fracture surfaces exhibit comparatively flatter and sharper features. The PLA fracture surface in Figure 10(b) (Sample 1) shows a localized fracture region with a cleaner separation between the fractured halves. The presence of a pull-out region suggests localized detachment, likely associated with weaker inter-raster or interlayer bonding.⁶⁵ The side views in Figure 10(d) (Sample 1) and Figure 10(f) (Sample 5) display relatively straight fracture paths across the cross-section with minimal visible necking. The PLA fracture surface in Figure 10(h) (Sample 5) appears more planar with limited tearing, indicating rapid crack propagation. Overall, the optical observations show that ABS exhibits a rough fracture surface and clear signs of plastic deformation, reflecting a semi-ductile failure response. In contrast, PLA displays sharper and more planar fracture features, characteristic of a brittle tensile failure mode. The layer-oriented fracture morphology, particularly evident in ABS (Figure 10(g)), confirms that failure behaviour is strongly governed by interlayer adhesion inherent to the FDM process.¹

To gain a deeper understanding of the failure mechanisms (Figure 11), the fractured tensile specimens were examined using Scanning Electron Microscopy (SEM). The analysis was performed using a TESCAN MAIA3 XMH Field Emission Scanning Electron Microscope (FE-SEM). Imaging was carried out at an accelerating voltage of 5 kV and a working distance of approximately 9–10 mm in Secondary Electron Imaging (SEI) mode, which provides detailed surface topography and enables clear observation of interlayer bonding and fracture features.⁶⁶ While direct interlaminar shear strength (ILSS) tests were not conducted in this study, the global tensile strength serves as a macroscopic indicator of interfacial bonding quality. In multi-material FDM parts, weak adhesion typically results in delamination failure at stresses significantly lower than the yield strength of the constituent materials.²⁶ The high σ_UTS observed in optimal specimens (sample 1 is > 42 MPa) suggests that the PLA-ABS interface achieved sufficient thermal fusion to transfer load effectively. SEM micrographs (Figure 11) confirm this, showing rough fracture surfaces and material pull-out at the interface, characteristic of cohesive failure, rather than the smooth surfaces associated with adhesive delamination failure.⁵²

Figure 11.

SEM images of tensile-fractured specimens: (a) layered PLA-ABS structure of Sample 1, (b) voids and partially bonded, (c) micro voids and ductile fracture features, (d) fractured PLA layers of Sample 5 showing interlayer debonding and (e) infill gaps and non-uniform layer bonding in Sample 5.

The low-magnification SEM micrograph of Sample 1 (Figure 11(a)) clearly reveals distinct PLA and ABS layers, confirming the layered architecture produced during the FDM process. The fracture surface appears uneven, with visible separation between adjacent layers, indicating that interlayer interfaces significantly influenced the failure process. This morphology highlights the anisotropic nature of the printed material, where interlayer bonding governs the mechanical response. At higher magnification (Figure 11(b)), micro-voids, small cavities, and regions of partially bonded rasters are observed.⁶⁷ These defects suggest that crack initiation occurred at inherent structural discontinuities, such as inter-raster gaps formed during deposition. Incomplete fusion between neighbouring filaments created localized stress concentration sites, promoting crack nucleation under tensile loading. Further magnification (Figure 11(c)) shows rough fracture regions with micro-voids and torn surfaces, which are characteristic of ductile fracture behaviour involving plastic deformation followed by void growth and coalescence. These observations are consistent with optical analysis, which indicated noticeable deformation in Sample 1 prior to fracture and a comparatively improved tensile response.⁶⁷

In contrast, the SEM micrograph of Sample 5 (Figure 11(d)) shows fractured PLA layers with pronounced interlayer debonding and delamination. Several regions of the fracture surface appear relatively planar, indicating rapid crack propagation along weak layer interfaces.⁶⁸ Compared with Sample 1, fewer features associated with plastic deformation are observed, suggesting a more brittle mode of failure. The higher-magnification image in Figure 11(e) further reveals infill gaps, localized layer distortion, and non-uniform bonding between adjacent layers.⁶⁷ These discontinuities act as stress raisers, facilitating early crack initiation and unstable crack growth under tensile loading. The presence of such defects explains the reduced tensile performance of Sample 5, as poor interlayer fusion limits the material’s ability to sustain applied stress. From this SEM examination, it is clear that the fracture behaviour of the printed specimens is governed by interlayer adhesion, raster bonding quality, and void content. Specimens with uniform layer bonding exhibit ductile fracture features, whereas those with significant interlayer defects predominantly fail through interlayer separation and rapid crack propagation.⁶⁶

Comparative analysis

To rigorously validate the efficacy of the proposed high-speed FDM bi-layer laminate approach, the experimental outcomes of the present study were benchmarked against standard mono-material and multi-material data reported in recent literature. Table 9 shows that the current high-speed PLA/ABS bi-layer printing approach performs better than many standard FDM baselines. It achieved a maximum tensile strength of 42.17 MPa, which is higher than typical pure ABS and multi-material values, indicating effective interfacial bonding and good load transfer. The process also operated successfully at much higher print speeds of 150–250 mm/s, demonstrating the feasibility of high-speed fabrication. In terms of surface quality, the best roughness value of 3.40 µm was significantly better than commonly reported standard ABS/FDM ranges. However, the Hilbert curve pattern caused catastrophic failure due to vibration and under-extrusion at very high speed. Whereas, the concentric toolpath emerged as the optimal pattern because it improved strength and reduced shear-related failure.

Table 9.

Quantitative comparison of current high-speed laminate versus literature baselines.

Parameter	Current study (high-speed PLA/ABS)	Literature baseline (standard FDM)	Phenomenological analysis
Max tensile strength (σUTS)	42.17 MPa (Run 1)	30 – 40 MPa (pure ABS) and ∼30 – 35 MPa (multi-material)¹⁶	The laminate’s strength exceeds pure ABS baselines, indicating that the bi-layer interfacial bonding successfully transfers structural loads without premature delamination
Operating print speed	150 – 250 mm/s	15 – 110 mm/s¹³	This study successfully operates at 2x to 5x standard industry speeds, proving high-speed viability for continuous toolpaths
Best surface finish (Ra)	3.40 µm (Run 8)	7.18 – 13.4 µm (standard ABS) and ∼9 – 35 µm (FDM)^11,28	Achieved exceptionally low surface roughness at peak velocity (250 mm/s), outperforming many low-speed optimized prints
Catastrophic failure mode	Hilbert curve (E = 444.6 MPa, σUTS = 18.39 MPa)	Complex/fractal patterns prone to severe voids¹³	Rapid acceleration/deceleration at 250 mm/s induces severe machine vibration and severe under-extrusion
Optimal toolpath	Concentric	0° raster/linear³⁶	Concentric patterns mimic axial (0°) alignment, maximizing load-bearing capacity and suppressing shear failure

Conclusion

This study successfully developed a systematic framework for the parametric optimisation of FDM-printed PLA + ABS Bi-layer Laminate using a Taguchi L9 design of experiments. The main conclusions are summarised as follows:

• The parameter combination of PS = 150 mm/s, LH = 0.1 mm, and concentric infill (Sample 1) produced the best mechanical performance, with E = 1252.29 MPa, σUTS = 42.17 MPa, and Ɛf = 5.79%. This result confirms that thinner layers enhance interlayer fusion, leading to improved stiffness and strength.

• Poor interlayer bonding at higher layer height (LH = 0.3 mm) significantly reduced mechanical performance. Sample 6 exhibited high surface roughness (Ra = 14.67 µm) and lower strength, indicating that coarse layers promote interlayer defects.

• The infill pattern strongly influenced mechanical behaviour. Even at PS = 250 mm/s and LH = 0.2 mm, the concentric infill configuration (Sample 8) achieved high strength (σUTS = 41.91 MPa, E = 1082.48 MPa) with low surface roughness (Ra = 3.4 µm), demonstrating efficient load transfer.

• Optical microscopy and SEM analyses revealed micro-voids, inter-raster gaps, and layer debonding as the dominant crack initiation sites, directly linking reduced tensile properties to poor filament fusion and internal defects.

• The fracture mode transitioned from ductile behaviour (micro-void coalescence in Sample 1) to brittle interlayer separation (Sample 5 and high-LH samples). This confirms that the mechanical reliability of PLA-ABS FDM parts is primarily governed by interlayer adhesion quality.

Footnotes

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Ratnesh Kumar Raj Singh

References

Juneja

Chohan

Kumar

, et al. Impact of process variables of acetone vapor jet drilling on surface roughness and circularity of 3D-Printed ABS parts: fabrication and studies on thermal, morphological, and chemical characterizations. Polymers 2022; 14: 1367. https://doi.org/10.3390/polym14071367, Epub ahead of print 1 April 2022.

Patel

Gupta

Saket

, et al. Effect of infill pattern on the mechanical properties of PLA and ABS specimens prepared by FDM 3D printing. Proceedings of the Institution of Mechanical Engineers, Part E: Journal of Process Mechanical Engineering. 2024 Jun; 11: 09544089241258744. https://doi.org/10.1177/09544089241258744

Golhin

Tonello

Frisvad

, et al. Surface roughness of as-printed polymers: a comprehensive review. Int J Adv Manuf Technol 2023; 127: 987–1043. https://doi.org/10.1007/s00170-023-11566-z

Hiremath

Dsouza

Chiniwar

, et al. Exploring the impact of epoxy coated 3D-printed polymers on surface roughness and mechanical behavior: an experimental and numerical study. Results in Engineering. Sep; 2024123: 102779. https://doi.org/10.1016/j.rineng.2024.102779

Butt

Bhaskar

Mohaghegh

. Investigating the effects of ironing parameters on the dimensional accuracy, surface roughness, and hardness of FFF-printed thermoplastics. Journal of Composites Science. 2022 May; 6(5): 121. https://doi.org/10.3390/jcs6050121

Qayyum

Saqib

Hussain

, et al. Predicting flexural properties of 3D-printed composites: a small dataset analysis using multiple machine learning models. Elsevier. https://www.sciencedirect.com/science/article/pii/S2352492824031179 (accessed 6 March 2026).

Kechagias

Zaoutsos

. An assessment of PLA/wood with PLA core sandwich multilayer component tensile strength under different 3D printing conditions. J Manuf Process 2024; 131: 1240–1249. https://doi.org/10.1016/j.jmapro.2024.09.098

Kamer

Temiz

Yaykaşli

, et al. Comparison of mechanical properties of tensile test specimens produced with ABS and PLA material at different printing speeds in 3D printer. J Facult Eng Archit Gazi Univ 2022; 37: 1197–1211.

Sultana

Sarker

Dhar

. Parametric optimization and sensitivity analysis of the integrated Taguchi-CRITIC-EDAS method to enhance the surface quality and tensile test behavior of 3D printed PLA and ABS parts. Heliyon. 2025 Jan 15;11(1). https://doi.org/10.1016/j.heliyon.2024.e41289

10.

Mishra

Bharat

Kumar

. Multi-objective optimization of 3D printed PLA/carbon fibre composite using a combined approach of gray relational analysis and ANOVA. J Thermoplast Compos Mater 2025; 38: 4208–4227. https://doi.org/10.1177/08927057251339246

11.

Ouazzani

El Jai

Akhrif

, et al. An experimental study of FDM parameter effects on ABS surface quality: roughness analysis. Int J Adv Manuf Technol 2023; 127: 151–178. https://doi.org/10.1007/s00170-023-11435-9

12.

Dairabayeva

Perveen

Talamona

. Tensile properties of the mono-material and multi-material PLA and ABS sandwich structures. In: 2024 15th international conference on mechanical and intelligent manufacturing technologies, ICMIMT 2024, Cape town, South Africa, 17-19 May 2024. Institute of Electrical and Electronics Engineers Inc., 2024, pp. 168–172.

13.

Abeykoon

Sri-Amphorn

Fernando

. Optimization of fused deposition modeling parameters for improved PLA and ABS 3D printed structures. Int J Lightweight Mater Manuf 2020; 3: 284–297. https://doi.org/10.1016/j.ijlmm.2020.03.003

14.

Utz

Ferbert

Geis

, et al. Mechanical properties and predictive analysis of multi-material polypropylene-glass fiber sandwich structures produced by material extrusion. Polym Compos 2025; 46: 10239–10253. https://doi.org/10.1002/pc.29617

15.

Kuzmanić

Vujović

Petković

, et al. Influence of 3D printing properties on relative dielectric constant in PLA and ABS materials. Progr Addit Manuf 2023; 8: 703–710. https://doi.org/10.1007/s40964-023-00411-0

16.

Khodaee

Abedini

Kami

. Effects of fused filament fabrication (FFF) process parameters on tensile and flexural properties of ABS/PLA multi-material. J Braz Soc Mech Sci Eng. 2024 Oct; 46(10): 628. https://doi.org/10.1007/s40430-024-05189-y

17.

Gohar

Hussain

Ali

, et al. Mechanical performance of honeycomb sandwich structures built by FDM printing technique. J Thermoplast Compos Mater 2023; 36: 182–200. https://doi.org/10.1177/0892705721997892

18.

Akhoundi

Mechanics

AB-E

. Effect of filling pattern on the tensile and flexural mechanical properties of FDM 3D printed products. Exp Mech 2019; 59: 883–897. https://doi.org/10.1007/s11340-018-00467-y

19.

Gohar

Hussain

Ali

. Mechanical performance of honeycomb sandwich structures built by FDM printing technique. J Thermoplast Compos Mater 2023; 36: 182–200. https://doi.org/10.1177/0892705721997892

20.

Heidari-Rarani

Ezati

Sadeghi

, et al. Optimization of FDM process parameters for tensile properties of polylactic acid specimens using Taguchi design of experiment method. J Thermoplast Compos Mater 2022; 35: 2435–2452. https://doi.org/10.1177/0892705720964560

21.

Sumalatha

Supraja Reddy

Chennaiah

, et al. Experimental investigation on combined ABS and PLA materials using 3D printing FDM machine. In: Lecture Notes in Mechanical Engineering. Springer Science and Business Media Deutschland GmbH, 2024, pp. 371–382.

22.

Günaydın

Rea

Kazancı

. Energy absorption enhancement of additively manufactured hexagonal and re-entrant (auxetic) lattice structures by using multi-material reinforcements. Addit Manuf. 2022 Nov 1; 59: 103076. https://doi.org/10.1016/j.addma.2022.103076

23.

Arifvianto

Satiti

Salim

, et al. Mechanical properties of the FFF sandwich-structured parts made of PLA/TPU multi-material. Progr Addit Manuf 2022; 7: 1213–1223. https://doi.org/10.1007/s40964-022-00295-6

24.

Tunay

Deniz

Kazel

. Experimental investigation on crashworthiness behaviors of 3D-Printed cellular composite sandwich structures. Polym Compos 2025; 46: 14433–14446. https://doi.org/10.1002/pc.30071

25.

Radadiya

Gandhi

. A study of tensile characteristics for glass and carbon fiber along with sandwiched reinforced ABS composites. J Inst Eng: Series C 2022; 103: 1049–1057. https://doi.org/10.1007/s40032-022-00848-2

26.

Pakhare

Kalia

Nadimpalli

SPV

, et al. Interface fracture characterization of 3D-printed rigid/flexible dissimilar polymers. Progr Addit Manuf 2024; 9: 2235–2247. https://doi.org/10.1007/s40964-024-00575-3

27.

Kumar

Singh

. Multi-material 3D printed PLA/PA6-TiO2 composite matrix: rheological, thermal, tensile, morphological and 4D capabilities. Advances in Materials and Processing Technologies 2022; 8: 2329–2348. https://doi.org/10.1080/2374068x.2021.1912527

28.

Kechagias

. Surface roughness assessment of ABS and PLA filament 3D printing parts: structural parameters experimentation and semi-empirical modelling. Int J Adv Manuf Technol 2024; 134: 1935–1946. https://doi.org/10.1007/s00170-024-14232-0

29.

Fadillah

Suryanto

Suprayitno

. Study on effect of 3D printing parameters on surface roughness and tensile strength using analysis of variance. Journal of Mechanical Engineering Science and Technology (JMEST) 2023; 7: 96. https://doi.org/10.17977/um016v7i22023p096

30.

Amiri

Zolfaghari

Shakeri

. 3D printing of glass fiber reinforced acrylonitrile butadiene styrene and investigation of tensile, flexural, warpage and roughness properties. Polym Compos 2022; 43: 6287–6299. https://doi.org/10.1002/pc.26937

31.

Sudin

Shamsudin

Daud

, et al. Examining the effect of annealing parameters on surface quality and tensile strength of ABS 3D-Printed materials. El-Cezeri Journal of Science and Engineering 2024; 11: 307–316.

32.

Ghaznavi Youvalari

Alizadeh Kaklar

Mohamadi

. Investigation of mechanical properties in PLA, ABS and epoxy resin parts fabricated by 3D printing technology. Sci Rep. 2025 Jul 30; 15(1): 27777. https://doi.org/10.1038/s41598-025-13866-8

33.

Kafshgar

Rostami

Aliha

MRM

, et al. Optimization of properties for 3D printed PLA material using Taguchi, ANOVA and multi-objective methodologies. In: Procedia Structural Integrity. Elsevier B.V., 2021, pp. 71–77.

34.

Ahmed

Hussain

Gohar

, et al. Impact toughness of hybrid carbon fiber-PLA/ABS laminar composite produced through fused filament fabrication. Polymers 2021; 13: 3057. mdpi.com. Epub ahead of print 2021. https://doi.org/10.3390/polym13183057

35.

Dairabayeva

Perveen

Talamona

. Investigation on the mechanical performance of mono-material vs multi-material interface geometries using fused filament fabrication. Rapid Prototyp J 2022; 29: 40–52. https://doi.org/10.1108/rpj-07-2022-0221

36.

Çakan

. Effects of raster angle on tensile and surface roughness properties of various FDM filaments. J Mech Sci Technol 2021; 35: 3347–3353. https://doi.org/10.1007/s12206-021-0708-8

37.

Hassan

Alomayri

Noaman

. 3D printed concrete for sustainable construction: a review of mechanical properties and environmental impact. Arch Comput Methods Eng 2025; 32: 2713–2743. https://doi.org/10.1007/s11831-024-10220-5

38.

Ekrem

Yilmaz

Bilgisi

, et al. Yüksek hızlı 3B Yazıcıda basılmış PLA. PETG ve ABS Numunelerin Mekanik Özellikleri; 7: 2667–7989.

39.

Chouhan

Raj Singh

Mudgal

. Investigation of tensile properties of FDM-printed PLA-CF infill in PLA shell components. J Thermoplast Compos Mater. 2025 Dec 19; 08927057261421156. https://doi.org/10.1177/08927057261421156

40.

Chouhan

Raj Singh

Mudgal

. Optimization of mechanical properties of 3D printed ABS: a comparative study of response surface methodology and TOPSIS. J Polym Res 2026; 33: 36. https://doi.org/10.1007/s10965-025-04754-w

41.

Azeem

Nasir

. Tensile behavior of sandwich structures using various 3D printed core shapes with polymer matrix composites. Int J Adv Manuf Technol 2025; 137: 5665–5682. https://doi.org/10.1007/s00170-025-15478-y

42.

Ayrilmis

Kariz

Šernek

, et al. Effects of sandwich core structure and infill rate on mechanical properties of 3D-printed wood/PLA composites. https://doi.org/10.1007/s00170-021-07382-y/Published

43.

Caliskan

Sevim

Demirbas

. Tensile behavior of functionally graded sandwich PLA-ABS produced via fused filament fabrication process. Mech Adv Mater Struct 2024; 31: 261–270. https://doi.org/10.1080/15376494.2023.2244943

44.

Selvam

Mayilswamy

Whenish

, et al. Multi-objective optimization and prediction of surface roughness and printing time in FFF printed ABS polymer. Sci Rep. 2022 Oct 7; 12(1): 16887. https://doi.org/10.1038/s41598-022-20782-8

45.

Kechagias

Ninikas

Petousis

, et al. Laser cutting of 3D printed acrylonitrile butadiene styrene plates for dimensional and surface roughness optimization. Int J Adv Manuf Technol 2022; 119: 2301–2315. https://doi.org/10.1007/s00170-021-08350-2

46.

Mishra

Negi

Bharat

, et al. Effect of FFF parameters on PLA/WD composite properties: comparative analysis using Taguchi and GRA approach. J Reinforc Plast Compos. 2026 Mar; 45(5–6): 1198–210. https://doi.org/10.1177/07316844251314979

47.

Mani

Karthikeyan

Kalaiselvan

, et al. Optimization of FDM 3-D printer process parameters for surface roughness and mechanical properties using PLA material. In: Materials Today: Proceedings. Elsevier Ltd, 2022, pp. 1926–1931.

48.

Khan

Tariq

Abas

, et al. Parametric investigation and optimisation of mechanical properties of thick tri-material based composite of PLA-PETG-ABS 3D-printed using fused filament fabrication. Composites Part C: Open Access. 2023 Oct 1; 12: 100392. https://doi.org/10.1016/j.jcomc.2023.100392

49.

Gómez Amador

Venturini Avendano

González

, et al. Mechanical characterization and testing of multi-polymer combinations in 3D printing. Heliyon. 2025 Feb 15;11(3). https://doi.org/10.1016/j.heliyon.2025.e42420

50.

Yousefi

Kennedy

. Experimental and numerical analysis of additively manufactured foamed sandwich beams. Compos Struct. 2023 May 15; 312:116866. https://doi.org/10.1016/j.compstruct.2023.116866

51.

Zaoutsos

Kechagias

. Optimizing bonding conditions between multilayer FFF material extrusion. Mater Manuf Process 2025; 40: 499–506. https://doi.org/10.1080/10426914.2024.2437753

52.

Duarte

de Queiroz

HFM

Neto

JSS

, et al. Evaluation of durability of 3D-printed multi-material parts for potential applications in structures exposed to marine environments. Procedia Structural Integrity. Elsevier B.V., 2024, pp. 299–308.

53.

Bharat

Jain

Bose

PSC

. A comprehensive overview on additive manufacturing processes: materials, applications, and challenges. In: International conference on production and industrial engineering, Jalandhar, India, 10-12 March 2023. Springer 2024, pp. 95–105.

54.

Khosravani

Zolfagharian

Jennings

, et al. Structural performance of 3D-printed composites under various loads and environmental conditions. Polym Test. 2020 Nov 1; 91: 106770. https://doi.org/10.1016/j.polymertesting.2020.106770

55.

AL-Khafaji

AbdulAmeer

Husain

, et al. On mechanical behavior and optimization of elliptical-rectangular absorbers made by ABS and PLA in both experimental and numerical methods. J Braz Soc Mech Sci Eng. 2023 Jul; 45(7): 383. https://doi.org/10.1007/s40430-023-04199-6

56.

Ermolai

Nagîţ

Sover

, et al. Design and testing of multi-material shape-changing flexure hinges for fused filament fabrication. Bulletin of the Polytechnic Institute of Iaşi Machine constructions Section 2022; 68: 19–30. https://doi.org/10.2478/bipcm-2022-0022

57.

Khan

Ali

Hussain

, et al. An experimental study on interfacial fracture toughness of 3-D printed ABS/CF-PLA composite under mode I, II, and mixed-mode loading. J Thermoplast Compos Mater 2021; 34: 1599–1622. https://doi.org/10.1177/0892705719874860

58.

La Gala

Ceretti

DVA

Fiorio

, et al. Comparing pellet- and filament-based additive manufacturing with conventional processing for ABS and PLA parts. J Appl Polym Sci. 2022 Nov 20; 139(44): e53089. https://doi.org/10.1002/app.53089

59.

Vălean

Foti

Razavi

SMJ

, et al. Static and fatigue behavior of 3D printed PLA and PLA reinforced with short carbon fibers. J Mech Sci Technol 2023; 37: 5555–5559. https://doi.org/10.1007/s12206-023-2307-3

60.

Bouhsiss

En-naji

Kartouni

, et al. Mechanical characterization and statistical study of experimental tensile test results of ABS specimens. Chaos Solitons Fractals. 2025 Apr 1; 193: 116107. https://doi.org/10.1016/j.chaos.2025.116107

61.

Piramanayagam

Kalimuthu

Nagarajan

, et al. Experimental investigation and statistical analysis of additively manufactured onyx-carbon fiber reinforced composites. J Appl Polym Sci. 2021; 138(18): 50338. (accessed May 10).

62.

Raj

Tyagi

Sharma

, et al. Optimizing the process parameters with statistical and soft computing techniques for enhanced mechanical properties of acrylonitrile butadiene styrene material samples fabricated via fused filament fabrication technique. Progr Addit Manuf 2025; 10: 2559–2583. https://doi.org/10.1007/s40964-024-00767-x

63.

Tyagi

Kumar

, et al. 3D printing of acrylonitrile butadiene styrene–jute fiber-based sandwich composites: mechanical properties and fractography analysis. J Mater Eng Perform 2025; 34: 13672–13685. https://doi.org/10.1007/s11665-024-10122-4

64.

Mudakavi

Sreesha

Kumar

, et al. A comprehensive experimental investigation on mechanical properties and fracture morphology of particulate composites via material extrusion-based 3D printing. Results Mater. 2022 Dec; 116: 100348. https://doi.org/10.1016/j.rinma.2022.100348

65.

Jain

Upadhyay

Sahai

, et al. Comparing the flexural and morphological properties of dissimilar FFF-fabricated polymer composites. J Thermoplast Compos Mater 2024; 37: 167–191. https://doi.org/10.1177/08927057231170790

66.

Moradi

Hashemi

Kasaeian-Naeini

. Experimental investigation of parameters in fused filament fabrication 3D printing process of ABS plus using response surface methodology. Int J Adv Manuf Technol. 2023 May; 1: 1–8. https://doi.org/10.1007/s00170-023-11468-0

67.

Moradi

Beygi

Yusof

, et al. 3D printing of acrylonitrile butadiene styrene by fused deposition modeling: artificial neural network and response surface method analyses. J Mater Eng Perform 2023; 32: 2016–2028.

68.

Morales

Porras

Maranon

, et al. Effect of polypropylene-grafted-maleic anhydride and rice husk particle size on 3D-Printing filament composites from recycled polypropylene. Polym Compos. 2025 Nov 10;46(16):15125-41. https://doi.org/10.1002/pc.30118

Optimization of FDM printing parameters for high-speed fabrication of PLA-ABS Bi-layer laminate structures

Abstract

Keywords

Introduction

Materials and methods

Material

Printing parameters

Design matrix

Sample preparation

Tensile and surface roughness testing

Results and discussions

Tensile and surface roughness result

Taguchi analysis

Mean affect plots

Single to noice ratio plots

ANOVA analysis

Fractography of tensile fracture surfaces

Comparative analysis

Conclusion

Footnotes

Declaration of conflicting interests

Funding

ORCID iD

References