Three-Dimensional Printed Lightweight Composite Foams

Abstract

The goal of this paper is to enable three-dimensional (3D) printed lightweight composite foams by blending hollow glass microballoons (GMBs) with high density polyethylene (HDPE). To that end, lightweight feedstock for printing syntactic foam composites is developed. The blend for this is prepared by varying the GMB content (20, 40, and 60 volume %) in HDPE for filament extrusion, which is subsequently used for 3D printing. The rheological properties and the melt flow index (MFI) of blends are investigated for identifying suitable printing parameters. It is observed that the storage and loss modulus, as well as complex viscosity, increase with increasing GMB content, whereas MFI decreases. Further, the coefficient of thermal expansion of HDPE and foam filaments decreases with increasing GMB content, thereby lowering the thermal stresses in prints, which promotes the reduction in warpage. The mechanical properties of filaments are determined by subjecting them to tensile tests, whereas 3D printed samples are tested under tensile and flexure tests. The tensile modulus of the filament increases with increasing GMB content (8–47%) as compared to HDPE and exhibit comparable filament strength. 3D printed foams show a higher specific tensile and flexural modulus as compared to neat HDPE, making them suitable candidate materials for weight-sensitive applications. HDPE having 60% by volume GMB exhibited the highest modulus and is 48.02% higher than the printed HDPE. Finally, the property map reveals a higher modulus and comparable strength against injection- and compression-molded foams. Printed foam registered 1.8 times higher modulus than the molded samples. Hence, 3D printed foams have the potential for replacing components processed through conventional manufacturing processes that have limitations on geometrically complex designs, lead time, and associated costs.

Introduction

Traditional manufacturing of thermoplastic-based closed cell foams is realized through injection or compression molding processes.^1,2 These methods require tooling for fabricating complex parts, which can be expensive and time consuming. However, research on additive manufacturing (AM) indicates that the fused filament fabrication (FFF) is among the widely utilized techniques to create complex functional parts.³ Further, AM eliminates the standard constraints on the component size along with producing highly complex parts with zero tooling cost, lower energy, and material consumption.⁴⁻⁶ Although most polymers are currently used in FFF-based three-dimensional printing (3DP), the development of lightweight thermoplastic filaments for specialized applications is still in its infancy. Thermoplastic composites are used in semi-structural and many engineering applications as they are environmentally friendly and offer the flexibility of processing using various methods.⁷ Commonly used thermoplastic polymers such as polymethylmethacrylate,⁸ polylactide,⁹ acrylonitrile butadiene styrene,^10,11 polycarbonate,¹² and polyetherimide¹³ filaments produced from their respective blends^14,15 are used in industrial 3D printers as feedstock materials. Polymers such as high density polyethylene (HDPE),¹⁶ polypropylene,¹⁷ polyamide,¹⁸ polycaprolactone,¹⁹ polybutylene terephthalate,²⁰ and so forth, have limited studies because of warpage and delamination associated issues that can be addressed by adding various inorganic/organic fillers using compounding methods.

Inorganic and organic solid fillers have been used extensively in thermoplastic industries.²¹⁻²⁴ Reinforcing filler particles in the matrix has several benefits, including a reduction in resin costs as well as flexibility in tailoring properties.²⁵ Mechanical, surface, electrical, and magnetic properties can be altered using such fillers.^2,26 The most commonly used fillers are Al₂O₃,²⁷ glass,²⁸ iron particles,²⁹ carbon, and glass fibers.³⁰ Hollow spherical particles, such as fly ash and glass micro balloons (GMBs) as fillers in the matrix, were investigated using conventional processing methods wherein higher tooling cost and complex geometrical design restrictions pose many challenges.³¹⁻³⁴ Closed-cell composite foams (hollow microballoons reinforced in the matrix) provide greater versatility in designing underwater vehicle structures, including internal descriptions of instrumentation housings, buoyancy chambers,³⁵ and so forth. By changing the volume % of these hollow fillers in the matrix, tailor-made properties can be achieved for many different applications.^26,36 Achieving these properties depends on particle survival in these lightweight foams and processing methods used for synthesizing them.^2,37 Developing a lightweight filament with minimum to no particle breakage should significantly enhance specific properties in 3DP of the components for weight-sensitive applications such as in nose cones of remotely operated underwater vehicles or even printing the entire body in the tubular design form with all the internal structural details all at once. Automotive and aerospace components without any joints, if realized through printing, can add structural stability with enhanced performance. Adhesive joints are the weakest entities in the structure as the pressurization/depressurization leads to foam fracture in a marine environment. 3DP of foams can eliminate adhesive bonding of multiple blocks making them to work well in deep sea environments. In order to manufacture complex shapes and contours and eliminate the need for adhesive bonding, foam printing and the associated development of specialized lightweight filaments is the need of the hour for marine, automotive, and aerospace components.

The feedstock filament development poses processability challenges because of density differences between the constituents, filler dispersion, and rheological behavior.³⁸ Further, developed composite filaments must be in the desired diameter to be fed into commercially available 3D printers with sufficient flexibility for spooling.³⁹ These properties allow the filaments to be printed through the printer nozzle without any breakage leading to a block-free layered deposition of prints with dimensional stability.²⁸ Similarly, the print quality rendered by the FFF is governed by various parameters such as extrusion temperature, the temperature of the bed and nozzle, the orientation of print, percentage of infill (filling of the space), layer height, and raster width.⁴⁰ In addition, semi-crystalline polymer printing includes issues such as shrinkage/warpage, build plate adhesion, and post-print removal.^28,41 Adding thermally stable inorganic fillers to semi-crystalline polymers minimize shrinkage and make products dimensionally stable.⁴² The composite components, on the other hand, show a considerable variation in thermal properties and experience distinct thermal cycles during the subsequent processing. Hence, the appropriate choice of processing temperatures and cooling rates ensures quality prints,⁴³ and realizing such lightweight foams is crucial. The reinforcing matrix with hollow fillers results in the reduction of the matrix volume % leading to lightweight composite structures known as syntactic foams. These closed cell composite foams may be categorized into two, three, and multi-phase systems depending on the different types of phases present. They have better mechanical properties and can produce complex functional parts that can replace HDPE, thereby lowering carbon footprints.⁴⁴⁻⁴⁶ Naturally available fly ash cenospheres have numerous surface defects⁴⁷ as compared to engineered microballoons, and hence, engineered GMBs are chosen in the current work. Introducing hollow GMB particulate fillers in a matrix can impart significant weight reduction and can be effectively exploited for weight-sensitive structures. Weight (density) and strength are essential material properties and are crucial for aeronautical, naval, and automotive components. In manufacturing low-cost lightweight thermoplastics without compromising the mechanical properties of the material, GMBs are candidate fillers exhibiting promising behavior.² GMB-based foams are shown to have better mechanical properties than fly ash-based composite foams.^7,48 Owing to better mechanical responses and biocompatibility, HDPE finds its application in chemical containers, milk jugs, household utilities, and other structural applications.^49,50 Replacing HDPE with the engineered GMBs and realizing a 3D printed lightweight complex structure having enhanced specific mechanical properties is of paramount interest for weight-sensitive structural applications. Nonetheless, for such applications, hollow particle survival, warpage, and layer adhesion need to be carefully investigated, which is the focus of this paper.

GMB-based 3D printed foam structures can be effectively used in many structural applications⁵¹⁻⁵³ owing to the higher specific mechanical properties, as mentioned earlier. The inclusion of such stiffer GMB particles in the HDPE matrix can make the prints more dimensionally stable.⁵⁴ The load carrying capacity of 3D printed parts depends on the infill %.^3,55 In this paper, GMB/HDPE blends are prepared with varying filler content and tested for the melt flow index (MFI) and rheology. Filaments extruded from the blend are investigated under differential scanning calorimetry (DSC) and tensile tests. Extruded lightweight filaments are then fed into a 3D printer, and the fabricated prints are investigated under DSC, coefficient of thermal expansion (CTE), rheology, flexural, and tensile tests. Finally, a property map is plotted to compare the 3D printed GMB/HDPE composite foam results with other composite results realized using conventional fabrication processes. Such a comparison acts as a guiding tool for material selection based on specific end product requirements.

Experimental Section

Materials

Hollow GMBs (iM30k, 3M Corporation, Singapore) are used as fillers, with an average diameter of 15.3 μm, a density of 0.6 g/cm³, crushing strength of 27,000 psi, and a wall thickness of 1.4 μm. HDPE (HD50MA180) used as the matrix is purchased from IOCL, Mumbai, India (Table 1), with a 3 mm mean granule size. GMBs are varied at 20, 40, and 60% by volume in H (HDPE matrix) and are designated as H20, H40, and H60, respectively. GMB content lower than 20% results in nonuniform dispersion in the HDPE matrix, while higher than 60% results in increased viscosity of the melt, leading to microballoon breakage as observed in the pilot experiments. Figure 1 shows the micrographs of GMB and HDPE used in the current work. A smooth surface without any surface defects is observed for both GMB (Figure 1a) and HDPE (Figure 1b). Further, GMB particles are spherical in shape, aiding uniform resin spread during processing.

Table 1. Typical Characteristics of HDPE Granules^a.

property	typical value
MFI	20 g/10 min
Density	0.950 g/cm3
flexural modulus	750 MPa
vicat softening point	124 °C
tensile strength @ yield	22 MPa
elongation @ yield	12%

^aAs mentioned by the supplier.

Figure 1.

Micrographs of as-received (a) GMB and (b) HDPE.

Preparation of Blend, MFI, and Rheological Properties

A Brabender (16CME SPL) is used for blending HDPE and GMB. Blending speed and temperature are set at 10 rpm and 160 °C, respectively, based on the pilot experiments with the objective of avoiding the GMB breakage to prepare H20, H40, and H60 compositions.^2,26 The representative image of the H60 blend is shown in Figure 2a. Dynisco LMI5000 MFI equipment is used for measuring MFI (ASTM D1238) of H–H60 pellets, which helps in setting an appropriate multiplier in printing by isolating different temperature settings for different compositions. The study of rheological properties is essential to know the effect of the filler on manufacturing conditions. An Anton Paar rotational rheometer, MCR 502, is used to investigate the influence of fillers on the rheology of the developed blends. A 25 mm diameter and 1 mm thick specimen are used for a frequency sweep of 0.1–10 Hz at 150 °C at a 5% loading rate.

Figure 2.

Representative (a) blend of GMB/HDPE and (b) extruded H60 feedstock filament.

Filament Development and 3DP

The extrusion process is carried out using a 25SS/MF/26 single screw extruder supplied by Aasabi Machinery Pvt. Ltd., Bombay with an L/D ratio (flight length of screw to its outside diameter) of 25:1. The composite blends are pre-heated at 80 °C for 24 h to eliminate the moisture, if any, before gravity feeding them into the extruder hopper. Foam pellets (Figure 2a) are fed into an extruder having a barrel temperature profile of 145–150–155–145 °C (feed–die segment). The screw speed is set at 25 rpm. Take-off unit speed is maintained at 11.5 rpm to extrude the filaments of 2.85 ± 0.05 mm in diameter (Figure 2b).

Obtained H–H60 filaments are used as the input material for 3DP. A commercially available FFF-based Star 3D printer supplied by AHA 3D Innovations Pvt. Ltd., Jaipur has two nozzles of 0.5 mm diameter. The suitable values of temperature and flow rates based on the pilot experiments are set to achieve completely rigid parts with 100% infill for comparing with fully dense molded components. Printing at higher temperatures can help to achieve temperature distribution uniformly alongside the annealing effect, yielding in better adhesion of layers and dimensionally stability. Nozzle and bed temperatures below 225 and 80 °C, respectively, resulted in improper material flow through nozzles and nonuniform bonding of the raft with the HDPE plate, which is placed on the glass bed of the printer. Higher material flow through the nozzle and HDPE plate distortion is observed for the temperatures above 245 and 100 °C, respectively, for the nozzle and printer chamber. The experimental strategy followed for identifying suitable printing parameters based on layer deposition, defects, layer adhesion, post printing removal, and warpage are discussed in the later section. Samples are printed on the HDPE plate. After printing, samples are left on the build plate until it reaches room temperature to minimize the warpage. Later, prints are used for characterization.

DSC and CTE Investigations

PerkinElmer DSC-6000, USA, is used to estimate melting and crystallization on filaments and prints of H–H60 compositions. The specimen (10 mg) is heated in a 30 μL Al crucible for a 0–200 °C temperature range with isothermal curing at 200 °C for about 3 min. Later, samples are brought to zero degrees at a rate of 10°/min, eliminating thermal history because of earlier processing steps. Post cooling at 0 °C for 3 min, the samples are heated again from 0 to 200 °C. DSC plots display endothermic and exothermic peaks, representing melting enthalpy at cold crystallization. Crystallinity % (α_Cryst) is assessed as⁵⁶

1

where, ΔH_m = heat of fusion in J/g and ΔH_m^* = heat of fusion/gram for HDPE, 293 J/g⁵⁷ and W_GMB is the weight fraction of GMBs. A dilatometer, CIPET, Chennai, is used to estimate CTE for prints (ASTM D696-13) having a dimension of 75 × 12.7 × 3 mm. CTE values qualitatively exhibit warpage and dimensional stability information.⁵⁸

Void Content and Density Estimations

According to ASTM D792-13, filaments and prints, experimental densities are calculated. Using the rule of mixture, the theoretical density is determined by

2

where, m, f, c, V, and ρ are the matrix, filler, composite, volume fraction, and density, respectively. The difference in theoretical and experimental densities gives % void content and is given by⁵⁹

3

Such matrix porosity (void) in prints implies raster gaps though the infill is 100%. These air gaps developed while printing leads to three-phase foam structures helping in enhancing the energy absorbing capabilities.

Tensile and Flexural Investigations

Filament and 3D printed samples are tensile tested using Zwick Roell Z020, USA, with a 20 kN load cell. The total length of the filament is 176 mm, with a 76 mm distance between the grips. The test is carried out by maintaining a constant 5 mm/min loading rate. An extensometer (gauge length 50 mm) is used to measure the strain. The printed samples are tested according to ASTM D638-14, at similar cross-head displacement using a 25 mm extensometer gauge length. An initial load elongation of 0.1 MPa is recorded using an extensometer. For flexural testing of prints (127 × 12.7 × 3.2 mm), a three-point bending configuration (ASTM D790-17) is adopted with a preload of 0.1 MPa, loading rate of 1.37 mm/min with a span length to depth ratio of 16:1. Flexural modulus is computed using

4

where, d: thickness, b: width, m: slope, and L: span length.

Flexural stress (σ_fm) is calculated using

5

A minimum of five samples are tested for all the experimental investigations, and the average values are reported. Micrographs of as-fabricated freeze-fractured and post-test filaments and prints are taken by gold sputter covering (JFC-1600) using JSM 6380LA JEOL, Japan.

Results and Discussion

MFI and Rheology of GMB/HDPE

Flowability is quantified by MFI. An increase in GMB content reduces MFI because of filler resistance to the flow of the polymer.⁶⁰ HDPE has recorded the highest MFI (17.94 g/10 min) when compared with H20 (13.76), H40 (8.11), and H60 (4.85). MFI decreased by 23.29, 54.79, and 72.97%, with increasing GMB by 20, 40, and 60 volume %, respectively.^60,61 Decreased MFI needs to be carefully looked into either by raising the temperature of printing or increasing the print extrusion multiplier, especially for foams with higher filler loadings. The printing temperature is kept constant for H–H60 to consolidate the warpage, and hence, a multiplier factor is changed for higher GMB %. An increase in filler infusion increases the melt viscosity of the polymer⁶² and is observed in the entire frequency sweep (Figure 3a). At a higher frequency, HDPE shows a shear-thinning region. H20–H60 shows similar behavior with a slight increase in η′ and is due to the restriction of polymer chain movements by GMBs. Among foams, H60 shows the highest η′. At 0.1 and 50 rad/s, complex viscosities for H, H20, H40, and H60 are in the range of 1080.52–636.75, 2045.4–1048, 2729.6–1324.2, and 4331.4–1701.5 Pa·s, respectively. Compared to H (11,808 Pa at 50 rad/s), foams have a higher storage modulus owing to the presence of a greater number of stiffer particles (Figure 3b). The storage modulus increases from 20,019 to 32,163 Pa for H20–H60 foams. HDPE and H20 display standard homopolymer-like terminal behavior at lower frequencies because of the complete relaxation of polymer chains.⁶³ Compared to pure HDPE, H20 has a higher modulus. The plateau region is observed at a lower frequency for H40 and H60, indicating viscoelasticity. G″ increases with increasing frequency and filler content for all the samples (Figure 3c). The loss modulus for H–H60 ranges between 107.56–429.56 Pa, respectively, at 0.1 rad/s, which is ∼4 times for H60 as compared to H. Such a multifold increase in G″ could be because of the restrained matrix flow around stiff intact GMBs. Rheological and MFI properties act as a guideline for selecting appropriate process parameters for the printing of quality components.

Figure 3.

(a) Complex viscosity, (b) storage, and (c) loss modulus vs frequency for blends.

Physical and Microstructural Characterization

Performance and behavior of extruded foam filaments are influenced by the interaction of the filler–matrix, filler %, and matrix porosity. For filaments to be used in a 3D printer, adequate spooling stiffness and strength are needed. Hence, tests to find the density, morphology of the extruded filament, and tensile properties are performed before printing to check the quality, stiffness, and strength necessary for filament feasibility to be used in a commercially available printer. Table 2 presents density estimations, void %, and the weight reduction potential of filaments and prints.

Table 2. Physical Properties of Filament (F) and Prints (Pnt).

			ρexp (kg/m3)		ϕv (%)		weight saving potential (%)
composition	ϕf (vol %)	ρth (kg/m3)	F	Pnt	F	Pnt	F	Pnt
H	0	950	942 ± 8	927 ± 12	0.84	2.42
H20	20	880	858 ± 15	826 ± 13	2.50	6.14	8.92	10.90
H40	40	810	780 ± 11	746 ± 18	3.70	7.90	17.20	19.53
H60	60	740	683 ± 12	668 ± 10	7.70	9.73	27.49	27.94

The experimental and theoretical densities of HDPE filaments are very close, indicating lower void formations because of its hydrophobic nature. Mechanical properties of HDPE and foams are influenced by the void’s presence, as an effective load-bearing area reduces. An increase in GMB content increases void content in filaments (0.84–7.70%) and prints (2.42–9.73%). Higher void content in print, as compared to filaments, indicate that matrix porosity is transferred from the filament to prints. Further, additional porosity of 1.58, 3.64, 4.2, and 2.03% is observed in H, H20, H40, and H60 prints, respectively. Such additional porosity in prints amid 100% infill is because of air gaps between the raster. These additional porosities form three-phase (HDPE, GMB, and raster gap) syntactic foams enhancing the damping capabilities further. Figure 4 presents the micrographs post knife cut GMB/HDPE filaments to check the cross section and dispersion of GMBs in the HDPE matrix.

Figure 4.

Extruded filament micrograph of (a) cross-sectional view for representative H20. H60 at (b) lower and (c) higher magnifications.

The circular cross section in Figure 4a of the representative H20 filament affirms the suitability of the chosen extrusion parameters. Figure 4b shows a low magnification micrograph of H60, showing the uniform distribution of intact GMB particles and few voids in the compliant HDPE matrix. Such pores/voids, if transferred during 3DP, may increase three-phase syntactic foam compliance resulting in higher damping. Poor interfacial bonding between GMB and HDPE is clearly evident from a higher magnification micrograph of H60 (Figure 4c) and is obvious as constituent materials are used without any surface treatment to avoid additional processing time, cost, and difficulty in correlating properties with inconsistently coated layer thickness.

DSC Investigations of Filaments and Prints

Thermal behavior (T_Cryst, T_Melt, and CTE) of H–H60 is presented in Table 3 for looking into warpage-related issues. DSC plots for H–H60 are presented in Figure 5. For pure HDPE, the endothermic peak is observed at 108 °C, which is noted to be in an increasing trend for foams. The decrease in the level of the endotherm and crystallization temperature rise with higher GMB content is also noted in Figure 5. This strongly affirms the fact that, while HDPE cools, the nucleation of the melt occurs on the filler surface at relatively higher temperatures, forming thicker crystal lamellas leading to higher T_Cryst.⁶² Melt inertia is ignored as the crystallization temperature of foams varies in a very narrow range of 2.2% as compared to H (Table 3). An increase in the filler volume % has an insignificant influence on T_Melt of both filaments and prints, as seen from Table 3, indicating (i) additional thermal history imposed by 3DP post-extrusion has not induced higher residual thermal stresses and (ii) printing temperature can be kept constant for all the samples. A decrease in α_Cryst (56.68%) for foam filaments is observed with increasing GMB content relative to HDPE.

Table 3. Thermal Behavior of H–H60.

	TCryst (°C)		αCryst (%)		TMelt (°C)
composition	F	Pnt	F	Pnt	F	Pnt	CTE × 10–6 (°C)	CTE % reduction w.r.t H
H	105.70	110.82	59.54	61.74	131.47	130.88	135 ± 3.29
H20	112.67	113.12	49.12	50.72	132.51	131.24	106 ± 3.85	21.48
H40	112.92	113.23	33.71	37.01	130.45	131.29	88 ± 2.65	34.81
H60	112.59	113.27	25.79	28.59	130.86	130.90	75 ± 1.15	44.44

Figure 5.

DSC for crystallization peaks: cooling cycle in (a) filaments and (c) prints. Melting peaks from the heating cycle (2nd) in (b) filaments and (d) prints.

Printed samples also show similar behavior where α_Cryst dropped from 61.74 (H) to 28.59% (H60). Compared to filaments, the corresponding prints have higher α_Cryst and are anticipated to provide higher dimensional stability and reduce warpage-related issues. Extruded filaments are subjected to a sort of quenching as it passes through the water bath immediately after the extrusion. Thereby, very little time and energy are available for the melt crystallization of filaments^64,65 as compared to prints wherein samples cool slowly within the printer chamber. Because of the resistance offered by GMB to the flow of the polymer chain α_Cryst decreases in foams along with the reduction of the crystal domain of H.^32,66,67 Hence, dimensionally stable foam prints without any warpage can be successfully 3D printed, having a potential weight saving of ∼28% (Table 2).

Investigations on Filament Tensile Properties

The tensile response is governed by the dispersion of reinforcement, filler size, matrix interaction, and inherent properties of the matrix.⁶⁸ In order to use filament as the feedstock material in the 3D printer, it must meet specific requirements like shape retention without excessive bending to absorb frictional forces while going through drive rollers.⁶⁹ Bending can be avoided by keeping the filament rigid enough to withstand the push of the drive roller without damaging the associated printer elements. Figure 6 presents the tensile stress–strain plots of filaments. Stiffer intact GMB particles increase filament modulus by 8.17, 14.40, and 46.81% in H20, H40, and H60, respectively, as compared to H (Table 6 and Figure 6b). The HDPE filament is strained to more than 1000% without any breakage because of its ductility. However, only up to 400% strain is graphed in Figure 6a. H40 and H60 failed within ∼25% strain, as seen from Figure 6b. H20 exhibits more than 40% strain with the highest ultimate tensile strength (UTS) of 12.63 MPa among foams. A higher amount of the matrix in H20 resists the tensile load effectively by the plastic deformation of the entire cross section, as observed from Figure 6c. The marked area, as shown in Figure 6c, shows the formation of a new surface at the bulk scale, enhancing strain. H60 has the highest void content of 7.7% (Table 2) among foams resulting in much earlier filament fracture because of a reduction in the effective area arising from elongated pore coalescence (locations 1–4 in Figure 6d). Nevertheless, H60 exhibits the highest modulus because of a higher number of intact GMB particles (marked area in Figure 6d).

Figure 6.

Representative filament stress–strain plot of (a) H and (b) H20–H60. SEM of (c) H20 and (d) H60 filament post tensile tests.

Table 6. Filament and Prints Tensile Response.

	modulus (MPa)		UTS (MPa)		elongation at UTS (%)		fracture strength (MPa)		fracture strain in %
Composition	F	Pnt	F	Pnt	F	Pnt	F	Pnt	F	Pnt
H	722 ± 16.73	810.25 ± 16.73	16.4 ± 0.22	17.68 ± 0.21	17.90 ± 0.26	15.04 ± 0.23		6.68 ± 0.11		93.00 ± 1.03
H20	781 ± 17.95	865.56 ± 17.79	10.45 ± 0.42	12.8 ± 0.35	12.63 ± 0.33	5.68 ± 0.29	8.93 ± 0.23	10.39 ± 0.29	44.27 ± 0.23	30.48 ± 0.10
H40	826 ± 14.27	1125.68 ± 12.41	9.25 ± 0.39	9.49 ± 0.49	5.27 ± 0.35	3.11 ± 0.31	7.01 ± 0.19	8.24 ± 0.25	23.81 ± 0.22	21.66 ± 0.06
H60	1060 ± 18.53	1199.26 ± 11.53	7.16 ± 0.17	8.45 ± 0.18	2.39 ± 0.21	4.69 ± 0.11	5.90 ± 0.14	7.78 ± 0.19	16.53 ± 0.31	14.49 ± 0.07

Strength decreases with increasing filler content because of weaker bonding between GMB and HDPE, as seen in Figure 4c. Further, with increasing GMB content, HDPE volume decreases, lowering the ductile phase substantially, resulting in lower strength values. Filament strength can be increased by surface treating of GMB particles that lead to enhanced interfacial bonding, which is not within the scope of this work. Such a surface treatment approach needs careful attention as coupling agents increase the brittleness and can hamper spooling flexibility. The focus of the current work is the development of lightweight composite foam filaments for 3DP using as-received constituent materials, so that the processing time and cost are minimum, and enhances industrial adaptability for components where modulus and comparable strength are the design criteria.

3DP of GMB/HDPE

All the samples are printed in a rectilinear pattern having a print orientation in the Y-axis. A layer thickness of 0.32 mm is set to provide adequate clearance between the nozzle and the printed part. A printing speed of 35 mm/s is kept constant for all the samples to improve the surface finish and lower the warpage. The infill is kept at 100% to achieve structural stability in addition to the comparative analysis with fully dense injection and compression-molded samples. A multiplier is set to 1 and 1.2 for H–H40 and H60, respectively, based on the MFI estimations. Up to 60% MFI reduction, layers are deposited without any difficulties with 35 mm/s printing speed. With reducing MFI above 60%, blocking of the nozzle is experienced, and hence a multiplier of 1.2 was set for H60 for a given nozzle temperature setting. The nozzle temperature is set above the vicat softening point (124 °C) of HDPE. The printing and bed temperatures are varied across two temperature settings of 225, 245 and 80, 100 °C, respectively, for the reasons mentioned in the earlier section.

The experiments are carried out based on 3DP of HDPE as it exhibits maximum warpage compared to foams. Nonetheless, with the suitable printing parameters of H, H20–H60 samples are also printed. Table 4 and Figure 7 show the observations pertaining to the experimental tests carried out to identify suitable printing parameters. All H–H60 compositions are printed on the HDPE plate for optimum bonding between the first print layer with the base. Table 4 and Figure 7 show that the printing and bed temperatures of 245 and 80 °C, respectively, are best suited for quality printing of H–H60. Hence, all the samples are printed using the parameters listed in Table 5.

Table 4. Experimental Test of 3DP Parameters.

printing temperature (°C)	print bed temperature (°C)	observation	Figure 7
225	80	improper layer deposition	7a
		interlayer defects	7b
225	100	merging of the bottom layer with plate	7c
245	80	proper layer deposition,	7d
		absence of interlayer defects,
		easier removal of print from the plate,
		no warpage
245	100	maximum warpage, Defective part	7e

Figure 7.

Challenges in 3DP of HDPE (Table 4). (a) Improper layer deposition, (b) interlayer defects, (c) excessive diffusion, (d) defect-free print, and (e) highest warpage.

Table 5. Printing Values Utilized in the Current Work.

parameters	value
temp. of nozzle (°C)	245
printing bed temp (°C)	80
layer height (mm)	0.32
extrusion multiplier	H–H40: 1, H60: 1.2
print speed (mm/s)	35
print pattern	rectilinear
orientation of part	Y-axis
infill (%)	100 (±45° to x-axis)

3D printed samples are cooled within the build chamber till room temperature is reached. The printed samples exhibit consistent bonding between the layers with the least warpage (Figure 8a). The marked area, as shown in Figure 8a, indicates very good seamless diffusion between the layers at higher magnification. This fact reaffirms the suitability of printing parameters utilized in the present work (Table 5). Micrographs of freeze-fractured HDPE print show very few voids (Figure 8b), while the H60 micrograph (Figure 8c) exhibits uniform GMB dispersion along with elongated voids. Such elongated voids at higher filler volume % are the result of lower MFI and reduced melt viscosity. Printed samples have more void content compared to filaments (Table 2) because of air gaps between adjacent raster (Figure 8d). Air gaps are observed to be increasing with GMB content because of the lower matrix phase, higher melt viscosity, and reduced CTE values. Such air gaps might enhance damping and compressive capabilities, as mentioned earlier.

Figure 8.

Micrograph of printed (a) H in thickness direction and (b) freeze fractured across the thickness (c) H60 and (d) associated raster gaps in H60.

CTE of Prints

The addition of GMB in the HDPE matrix lowers CTE, as seen from Table 3.^70,71 At higher printing temperatures, dimensional stability can be achieved by adding GMB into HDPE. This indicates that the warpage can be avoided to a greater extent in printed components with dimensional stability and lower residual thermal stresses.⁷⁰ The entrapped gas inside the hollow GMB offers resistance against heat flow, which results in lower thermal conductivity. Also, there is a large difference in the CTE values of both the constituents, which is GMB and HDPE. Further, CTE also helps in understanding the raster diffusion mechanism and air gap formation in 3DP. Warpage, which is clearly evident from Figure 7e, is a crucial and challenging factor while printing neat HDPE because of higher CTE values. Nonetheless, appropriate printing and bed temperatures can effectively address this issue. Dimensionally stable prints are observed in foams because of lower CTE as a result of the presence of lower thermal conductive gases within hollow GMB that limit the flow of heat.^72,73 Among foams, H60 print showed the lowest CTE leading to minimal raster diffusion, thereby resulting in air gaps (Figure 8d). Such air gaps make syntactic foams lighter (∼2–4%), as seen from Table 2.

Tensile Response of Prints

A similar trend of the stress–strain response is exhibited by 3D printed H–H60 samples as in filaments, and the values are listed in Table 6. Breakage of the pure HDPE filament is not seen even after a strain of up to 1000%, while HDPE print could sustain only up to ∼45% strain, indicating a behavioral change from the ductile to brittle phase post 3DP. HDPE is extruded twice, once during filament formation, and second in a printer nozzle extruder. Such multiple extrusion cycles result in polymer chain alignment, associated cross-linking because of thermal processing leading to the hardening process. Failure strain for 3D printed H40 and H60 foams are 21.66 and 14.49%, respectively, whereas H20 shows up to 30.48% strain. In the case of HDPE, a long necking region is clearly observed (Figure 9a) because of raster fibrillation, resulting in broom-like fibrous ends. Such fibrous ends are a result of new surface formations because of extensive plastic deformation (micrograph of the marked area in Figure 9a). H40 and H60 foam prints show no necking region and fracture in a typical brittle manner, which is also seen from the fractographic area wherein matrix plastic deformation is hardly seen (Figure 9b).

Figure 9.

Fractographic analysis of representative 3D printed (a) H and (b) H60 post tensile test. Reprinted Figure 9a (photograph) with permission from [Patil, B.; Bharath Kumar, B. R.; Bontha, S.; Balla, V. K.; Powar, S.; Hemanth Kumar, V.; Suresha, S. N.; Doddamani, M. Eco-friendly lightweight filament synthesis and mechanical characterization of additively manufactured closed cell foams. Compos. Sci. Technol. 2019,183, 107816]. Copyright [2019] [Elsevier].

All the microballoons are observed to be intact, signifying potential weight saving of ∼28% (Table 2) is successfully achieved post printing. Intact GMB particles at a higher filler % make the matrix responsible for load carrying, which succumbs early owing to induced brittleness post-printing. Comparative analysis between the filament and printed coupons pertaining to modulus and strength show an increase by 12.22, 10.83, 36.28, 13.14, and 7.8, 22.49, 2.59, and 18.02%, respectively. GMB/HDPE prints results are compared with injection-molded cenosphere/HDPE foams. 3D printed HDPE shows appreciable UTS with a higher elastic modulus of 53.17% when compared with injection-molded foams. The 3D printed foam specimen elongation at UTS and fracture strength are 47.45% and ∼3 times higher than that of the injection-molded specimen.³⁷ Modulus of foam increases with GMB % (Table 6). Among foams, H60 displays the highest modulus and is 48.02% higher than the HDPE print. 3D printed H–H60 registered 1.5–1.8 times higher modulus than molded counterparts with zero tooling cost. Foam print fracture strength is 1.16–1.56 times higher when compared with H. For weight-sensitive applications, specific properties of foams are essential because printing allows flexibility in developing integrated (joint less) components with complex designs. Among all foams, H60 and H20 exhibit the highest specific modulus and strength, respectively. Table 7 shows the GMB/HDPE weight-saving potential through estimations of E/ρⁿ (n = 1, 2, and 3). Values in Table 7 clearly indicated that 3D printed GMB/HDPE foams can be used effectively in buoyancy modules and automotive and aerospace components of integrated complex designs.

Table 7. Weight-Saving Quantification Parameters of H and their Foams.

composition	E/ρ (MPa/kg/m3)	E/ρ2 (MPa/(kg/m3)2 × 10–3)	E/ρ3 (MPa/(kg/m3)3 × 10–6)
H	0.87	0.94	1.02
H20	1.05	1.27	1.54
H40	1.51	2.02	2.71
H60	1.80	2.69	4.02

Flexural Behavior of Prints

Foams displayed brittle fracture (Figure 10a) as compared to HDPE, which did not fail until 10% strain (Figure 10b). GMB inclusion induces brittleness in the compliant HDPE matrix. A crack initiated from the tensile side and propagated along the loading direction until it meets the compressive side. This is a typical flexural failure mode. It is interesting to note here that the crack did not propagate along with the deposited layers, confirming again the suitable printing parameters (Table 5) chosen for printing. As seen in Table 8, intact GMB particles (Figure 10c) increases the modulus with higher filler loadings. The H60 modulus is 1.37 times higher than H, while strength is observed to be decreased because of poor interface bonding between constituent elements and raster gaps (Figure 11).

Figure 10.

(a) Fractured foam samples post-flexural test. Representative (b) stress–strain plots for prints and (c) H60 micrograph post flexure test.

Table 8. Flexural Response of H–H60 Prints.

composition	modulus in MPa	strength in MPa	fracture strength in MPa	fracture strain in %	specific modulus in MPa/kg/m3	specific strength in MPa/kg/m3× 10–3
H	990 ± 11.28	25.4 ± 0.12			1.068	27.40
H20	1210 ± 19.56	21.0 ± 0.58	20.34 ± 0.32	6.88 ± 0.09	1.465	25.42
H40	1280 ± 11.87	17.1 ± 0.47	16.89 ± 0.41	6.04 ± 0.11	1.716	22.92
H60	1360 ± 11.23	15.1 ± 0.72	15.00 ± 0.79	3.15 ± 0.07	2.036	22.60

Figure 11.

3D printed representative H60 micrograph showing raster gaps.

GMBs embedded in the HDPE matrix increases the specific modulus by ∼2 times compared to H. The modulus of H–H60 printed foams is higher by 1.39–1.08 times against molded counterparts, whereas strength is observed to be higher and comparable in the case of H and H20 foams, respectively. Drop-in strength by 1.14 and 1.27 is noted for printed H40 and H60, respectively, against fully dense molded samples and is obvious owing to higher matrix porosity resulting from raster gaps.⁷⁴ With increasing filler loadings, these raster gaps volume increases because of lower CTE values. Nevertheless, these gaps can be minimized by the overlapped deposition of layers and will be explored in future investigations. Tensile and flexural strength is observed to be decreasing as constituent materials are used in as-received conditions, as mentioned earlier. Furthermore, filler addition increases amorphous fraction leading to a more restrained matrix flow and polymer chain mobility resulting in weaker interfaces. Enhancing the bonding between the constituents through appropriate coupling agents might increase the strength but at the expense of a substantial reduction in ductility, which may hamper filament extrusion and the 3DP process.

Property Graphs

Figures 12 and 13(2,26,37) show the tensile and flexural response, respectively, as a function of composite density fabricated using different processing routes. Hollow particle-filled composites exhibit promising properties to be exploited in weight-sensitive applications as compared to solid-filled material systems. The density of GMB-based 3D printed foams is in between the injection- and compression-molded foams. The tensile modulus of printed composites outperforms injection- and compression-molded composites (Figure 12a) except for wood-filled composites. The strength of GMB-based printed foams is comparable to compression- and injection-molded samples (Figure 12b). Flexural modulus of GMB-based 3D printed composites is greater than other syntactic foams realized by conventional manufacturing processes (Figure 13a).

Figure 12.

Tensile (a) modulus and (b) strength of the HDPE composite.^2,26,37

Figure 13.

Flexural (a) modulus and (b) strength of the HDPE composite.^2,26,37

Flexural strength is comparable to composites produced from compression and injection molding (Figure 13b). Figure 14 helps materials designers/scientists and industrial practitioners in deciding the composition of 3D printed GMB/HDPE foam based on the properties required as per the envisaged application. The choice of suitable extrusion and printing conditions with no particle breakage results in substantial weight reduction by ∼28%, as shown in the current work. Such a weight reduction for complex-integrated 3D printed components would enhance the performance with reduced carbon footprints.

Figure 14.

Comparative chart of the 3D printed GMB/HDPE properties.

Conclusions

GMB-based lightweight composite foam feedstock is successfully synthesized to be used on a commercial printer for weight-sensitive applications. Filaments and 3D printed samples are tested for mechanical characterization to check their adaptability and feasibility for 3DP applications, and a summary of the results is presented below:

• void contents increase in filaments and prints by 0.84–7.70 and 2.42–9.73%, respectively, with increasing filler content. 3D printed foams exhibit a 3 phase foam structure.

• An increase in GMB content decreases the MFI of HDPE.

• Loss modulus, storage modulus, and complex viscosity increase with increasing GMB content. At a lower frequency, complex viscosity is maximum and decreases as the frequency increases showing shear-thinning behavior in GMB/HDPE blends. Loss and storage modulus showed an increasing trend with an increase in GMB % and frequency.

• Filler content has no significant effect on peak melting temperature (T_Melt) of filaments and prints. The degree of crystallinity (α_Cryst) decreases by 56.68 and 53.69% for foam filaments and prints, respectively, with increasing GMB content as compared to HDPE. Compared to filaments, the corresponding prints have a higher α_Cryst and are a positive sign toward dimensional stability and warpage-related issues.

• Stiffer intact GMB particles increase the filament modulus by 8.17–46.81% in H20–H60, respectively, as compared to H.

• The addition of GMB in HDPE decreases the CTE of prints substantially, making the prints dimensionally more stable.

• Among foams, H60 displays the highest modulus, which is 48.02% higher than the HDPE print. 3D printed H–H60 registered 1.5–1.8 times higher modulus than the molded counterparts. Printed H20–H60 has 1.16–1.56 times higher fracture strength than the printed H.

• GMBs embedded in the HDPE matrix increases the specific modulus by ∼2 times compared to H. The modulus of H–H60 printed foams is higher by 1.39–1.08 times against molded counterparts.

• 3D printed GMB/HDPE foams having substantial weight saving potential (28%) with superior specific mechanical properties and reduced carbon footprints are successfully realized.

The current work successfully demonstrated the development of lightweight feedstock filament with the intention to widen available material choices for commercially available 3D printers. GMB/HDPE-integrated complex geometrical components can be printed without any warpage, as presented in this work. Strength enhancement can be realized by surface modification of the constituent materials along with the strategy of overlapping raster and is the focus of future investigations.

Glossary

Nomenclature

ρ

density

ρ_c

composite density

ρ_f

density of filler

ρ_m

density of matrix

ρ_th

theoretical density

ρ_exp

experimental density

V_f

filler volume %

V_m

matrix volume %

T_Cryst

crystallization temperature

T_Melt

peak melting temperature

η′

complex viscosity

G′

storage modulus

G″

loss modulus

ϕ_f

GMB volume %

ϕ_v

void content

α_Cryst

degree of crystallinity

Author Contributions

All authors contributed equally.

The authors declare no competing financial interest.