Simple Determination of the Melt Flow Index of Composite Polymer Filaments Used in Material Extrusion Additive Manufacturing

Abstract

The mechanical properties improving filler introduction into polymer filaments often lead to the formation of defects in 3D-printed products. Studying the bulk modification of polyethylene terephthalate glycol (PETG) filaments with 0.5–1.5 wt% of natural shungite and molybdenum disulfide, we found the melt flow index (MFI) value reduced by 18%–30%. We investigate the effect of bulk modification on the filaments’ rheological properties, develop a technique that eliminates typical additive prototyping defects by regulation of the extrusion parameters in Cura slicing software, and propose an effective method for the MFI-measurement of the 3D-printed filaments that does not require special laboratory equipment (plastometers).

1. Introduction

Currently, there is a high demand for polymer materials with improved mechanical and electrical properties. Such materials are used in the production of structural elements and functional components [1] for robotic devices, unmanned transport systems [2], and high-tech medical and printing equipment [3]. Due to the significant progress achieved over the past decade as a result of fundamental scientific research in the field of polymer materials science [4], the development of composite materials suitable for the implementation of additive prototyping technologies [5] has acquired the status of a scientific and technical task relevant for manufacturing enterprises in various sectors of the economy [6,7] (shipbuilding, aviation and automotive industries, agriculture, medicine, etc.). In addition to regular maintenance and scheduled repairs, for the effective organization of production processes, it is necessary to ensure the quickest replacement of machine parts, machine tools, and other equipment in case of a sudden breakdown. Current logistical difficulties, in general, and the significant remoteness of conveyors from repair shops can cause long downtimes that negatively affect profitability and lead to additional costs. One of the possible solutions to this problem is the use of additive prototyping technologies [8], which allow for the creation of parts of various sizes and shapes quickly and directly on production sites. Material extrusion (MEX) 3D printing [9,10] is one of the most promising forms of additive manufacturing as it has a number of well-known advantages, which, in particular, include the absence of material loss during product molding [11] and the low cost of the filament [12]. The latter is due to the widespread use of polymer materials, such as the copolymer of acrylonitrile, butadiene, and styrene (ABS [13]), polylactide (PLA [14]), polyamide (PA [15]), and polyethylene terephthalate glycol (PETG [16]).

The important aspects of MEX 3D-printed technology applications are the efficient processing and recovery of thermoplastic polymer materials. For example, the possibility of a composite filament creation from the recycled industrial waste of polypropylene and fiberglass is demonstrated in [17]. And the effective innovative technique of graded polyether ether ketone-based functional ingredients synthesis is described in [18].

At the same time, polymers suitable for the practical implementation of 3D printing in the vast majority of cases are significantly inferior in mechanical, electrical, and other properties to the original materials of the failed parts. In some cases, a significant improvement, for example, in the strength properties of 3D-printed products could be achieved by increasing the cohesive, adhesive, and autohesive interaction of the filament layers [19]. The possibility of directed regulation of the bulk and surface properties of polymeric materials is due to the presence of a relationship system, connecting morphological and functional characteristics [20].

Bulk modification methods are usually used to improve the mechanical properties of polymer materials [6]. The results of the bulk modification of a recycled polypropylene-based filament with basalt fiber were presented in [21]. A significant change in the mechanical properties of the material’s extrusion polymer composite as a result of its bulk modification is demonstrated in [22]: it was shown that the waste industrial glass fiber-reinforced polypropylene material can be effectively enhanced and converted to short fiber-reinforced material extrusion filament feedstock for the components printing.

This paper presents an optimized material extrusion (MEX) 3D printing method with feedstock produced from glass fiber-reinforced polypropylene (GFRPP) composite waste and investigates the effect of glass fiber weight fractions (0%, 15%, 30%, and 40%) on the printability and mechanical properties of filaments and printed specimens.

GFRPP feedstock filaments were fabricated by optimizing the filament extrusion process parameters. The highest ultimate tensile strength (112 MPa) was obtained in a filament with 40% fiber weight content. In addition, repeated recycling of filaments was shown to reduce the fiber aspect ratio and enhance the filament quality. It was shown that the increase in glass fiber content from 0% to 40% improved the ultimate tensile strength of MEX 3D-printed specimens from 14.7 MPa to 23.4 MPa, as well as the flexural strength from 11.68 MPa to 66.55 MPa. The increase in fiber weight content from 30% to 40% had a minimal effect on the ultimate tensile strength of the printed samples. In addition, the 40% filaments exhibited a more brittle behavior, rendering them less suitable for MEX 3D printing than 30% GFRPP. So, the waste industrial GFRPP material can be effectively enhanced and converted to short fiber-reinforced material extrusion filament feedstock for the printing of components as a circular economy demonstrator.

The use of filled filaments for 3D printing requires adjustment of the technological parameters of extrusion [23], since the morphology and, as a result, some physicochemical and operational properties of composite materials may differ significantly from the corresponding characteristics of the starting materials. The key technological parameter affecting the quality, speed, and the very possibility of 3D printing is the filament melt flow index [24]—a value (g/10 min) showing the flow rate of the polymer melt at a fixed pressure through a standard capillary at a certain temperature.

According to ASTM D 1238 [25], MFI is the only standardized rheological characteristic of a polymer melt [26]. As is known [25], the processing method (extrusion, pressing, and injection molding) for products made from thermoplastics and composites based on them is selected according to the MFI value. The MFI values are extremely important for regulating the technological parameters of additive prototyping, since an increase in the viscosity of the filament (due to the implementation of fillers of various chemical nature into the polymer) leads to defects such as “non-printing”, low interlayer adhesion, or a malfunction of the extrusion block in 3D printers. The measurement of real flow index values is also necessary for the correct choice of the extrusion parameters during 3D model slicing [27].

According to GOST 11645-2021, “Plastics. Methods for determining the melt flow index of thermoplastics” (Russia), and ASTMD 1238 standards [25], measurements of the flow index of thermoplastics should be carried out in accordance with one of the methods using industrial or laboratory plastometers [28], using a capillary with a diameter (2095 or 1180 mm) significantly larger than the 3D printer’s nozzle (~0.3 mm).

It was demonstrated in [29] that the MFI values for ABS-based composite materials filled with aluminum and copper powders vary significantly depending on the percentage of the metal by weight.

It was shown in [30] that at higher than 40 N forces of extrusion, the melting rate is over-predicted (probably due to the approximation of the shear thinning within the melt film and the capillary).

Our paper shows that the values of the melt flow index measured using a plastometer (in laboratory conditions) and using a 3D printer (in operating conditions) differ significantly; therefore, in order to improve extrusion during additive prototyping, it is necessary to take into account the results of measuring the filament extrusion speed through the nozzle of the 3D printer’s extrusion unit.

2. Materials and Methods

The bulk modification of filaments (Figure 1) was carried out in a modernized twin-screw extruder (Moscow Polytech, Mashplast, Russia), according to the method by [31], at a temperature of extrusion zones from 180 to 240 °C (Figure 2), by mixing polyethylene terephthalate glycol granules (PETG, Shijiazhuang, China) with fillers: natural shungite (average particles size from 2 to 5 microns, LLC NPK “Carbon-Shungite”, Petrozavodsk, Russia) or commercial synthetic molybdenum disulfide (average particles size from 2 to 7 microns, LLC “MetProd”, Krasnogorsk, Russia).

The formation of 100 g of solid-phase product (Figure 1A) for subsequent filament extrusion (Figure 1B) was carried out by the mechanical mixing of PETG granules with shungite or molybdenum disulfide powders at various mass fractions (0.5, 1.0, and 1.5 wt%).

For uniform distribution of the fillers on the granule surfaces, we added from 0.25 to 1.0 mL (depending on the mass content of the ingredients) of polyphenylmethylsiloxane [-Si(CH₃)(C₆H₅)O-]n-silicone oil (Sigma-Aldrich, St. Louis, MO, USA).

For an empirical study of the structural and morphological characteristics of the modified filaments, we used scanning electron microscopy (SEM) with a high-resolution (~1 nm) autoemission scanning electron microscope, JSM-7500 FA (JEOL, Tokyo, Japan), in the mode of detecting secondary electrons at an accelerating voltage of 10 kV (with preliminary deposition of a ~8 nm thick platinum layer in a magnetron-type setup (Jeol Auto Fine Coater 1600) using an Oxford X-max 80 detector with a SATW window at an accelerating voltage of 10 and 20 kV and an electron current of ~1 nA.

The depth of analysis calculated in the Win Casino v2.48 software using the Monte Carlo method for the 10 kV mode was 0.3–0.5 µm.

After optimization for silicon, the sensitivity of the device reached 0.2–1.0 at.%. The average error of the quantitative analysis was ~2%.

To determine the filament’s melt flow index, we used a plastometer, IIRT-5 (Tochpribor, Saint-Petersburg, Russia). The measurements were carried out in accordance with Russian (GOST 11645-2021) and international (ASTMD 1238 [26]) metrological standards.

The proposed technique for determining MFI is based on the weight measurement of the filament (extrudate) that flows within 60 s through the calibrated nozzle (500 µm) of the 3D printer (Figure 3).

To study the effect of fillers on the properties of PETG filaments and PETG-based composites under conditions as close as possible to production, the MFI was determined by filament extrusion through a 500 μm nozzle of a material extrusion 3D printer (Anycubic Mega S, Shenzhen, China) at a temperature of 240 °C. To determine the MFI, the mass of the extrudate was measured on an analytical balance, VibraHT 224 RCE.

The MFI of the filament was determined as follows: on the touch control panel of the 3D printer, the Home >> Tools >> Filament >> Filament In parameters were selected to preheat the extrusion zone to 240 °C. Then, the filament was extruded for a minute through the nozzle of the 3D printer at 40 mm/s into a polymer container, and the extrudate’s mass was measured on an analytical balance with an accuracy of 0.0001 g.

For a study of the effect of volumetric modification modes on the adhesion of cured filament layers, a series of experimental samples in the form of 40 mm diameter and 2 mm thick disks were made (Figure 4).

The quality of the 3D-printed disks was monitored using a stereomicroscope of the Stemi 2000 series (CarlZeiss Microscopy GmbH, Oberkochen, Germany).

The analytical characterization of the surface texture of the 3D-printed disks (Figure 4b) was carried out according to the original technique [33], based on the decomposition of digital images of the SEM images of the sample surfaces into a two-dimensional Fourier series.

The digital image of the analyzed surface is associated with a superposition of biperiodic (with two spatial periods, L_x and L_y) spatial lattices, the wave vectors of which are co-directed to the axes of the own coordinate system of the experimental samples:

M(x,y) = M₀ + ∑_k,l M_k,l,

(1)

where k, l—biharmonic order indices; M_k,l = A_k,l + B_k,l + C_k,l + D_k,l—models of biharmonics; A_k,l = a_k,l∙cos(2π∙k∙x/L_x)∙cos(2π∙l∙y/L_y), B_k,l = b_k,l∙cos(2π∙k∙x/L_x)∙sin(2π∙l∙y/L_y), C_k,l = c_k,l∙sin(2π∙k∙x/L_x)∙cos(2π∙l∙y/L_y), D_k,l = d_k,l∙sin(2π∙k∙x/L_x)∙sin(2π∙l∙y/L_y)—model components; a_k,l, b_k,l, c_k,l, d_k,l—partial amplitudes; 2π∙k∙x/L_x, 2π∙l∙y/L_y—partial phases of biharmonics. The quantitative parameters of the structural models formed in this way are the elements of the morphological spectrum matrix:

μ_k,l = (a _k,l² + b _k,l² + c _k,l² + d _k,l²)^1/2,

(2)

describing the mode structure of the surface roughness of the experimental samples in dimensionless form.

The set of partial morphological spectra, {μ_k,l}_N×N, provides the possibility of forming a generalized statistically significant model of the morphological spectrum, which quantitatively characterizes the structure of a series of partial implementations of 3D-printed disks made under the same conditions.

3. Results and Discussion

The SEM image (Figure 5) shows that the extruded unfilled filament has a homogeneous structure and chemical composition corresponding to PETG (

Table 1. Results of element analysis of PETG filament modified with shungite and MoS₂.

Filler, wt%	Additive SEM Image (Color Corresponds to the Chemical Element) and Mapping of Chemical Elements on the Surface of the Transverse Cleavage of the Filament	Elemental Composition, wt%
		C	O	S	Mo	Al	Si
No filler		77	23	-	-	-	-
MoS2 0.5		75.9	23.5	0.3	0.3	-	-
MoS2 1.5		75.0	20.9	0.8	1.3	-	-
Shungite 0.5		76.7	22.9	-	-	0.2	0.2
Shungite 1.5		75.4	23.7	-	-	0.5	0.4

), and its diameter is 1.65 mm, which is optimal for the manufacture of 3D-printed products by most material extrusion 3D printers since the probability of various defect formation increases when the filament diameter is less than 1.55 or more than 1.75 mm [34].

When shungite and molybdenum disulfide (MoS₂) are added into PETG, a structural transformation of the polymer matrix occurs, as is seen from the analysis results of the transverse cleavage of the filament (Table 1), and sulfur (S), molybdenum (Mo), aluminum (Al), and silicon (Si) appear.

Obviously, structural changes in filled PETG affect the rheological properties of the filaments. When 0.5 wt% shungite is introduced into the PETG filament, the MFI decreases by 18% from 14.6 to 11.9 g/10 min (

Table 2. The melt flow index of bulk-modified filaments from a material extrusion 3D printer measured with the IIRT-5 plastometer.

Filament (Polymer + Filler)		Filler wt%	MFI IIRT-5		MFI-MEX 3-D, g/min	MFIflow, %
			g/10 min	g/min
PETG	No filler	0	14.6	1.46	1.0513	0
	Shungite	0.5	11.9	1.19	0.9235	12.2
		1.0	11.5	1.15	0.6951	33.9
		1.5	11.1	1.11	0.2036	80.6
	Molybdenum disulfide MoS2	0.5	11.1	1.11	0.8509	19.1
		1.0	10.7	1.07	0.9068	13.7
		1.5	10.3	1.03	0.8623	18.0

).

In general, the melt flow index decreases with an increase in the proportion of additives in the composition of the polymer matrix, which correlates with the known literature data [35].

$${\mathrm{M}\mathrm{F}\mathrm{I}}_{\mathrm{f}\mathrm{l}\mathrm{o}\mathrm{w}}=(1-\frac{{\mathrm{M}\mathrm{F}\mathrm{I}}_{\mathrm{c}\mathrm{o}\mathrm{m}\mathrm{p}\mathrm{o}\mathrm{s}\mathrm{i}\mathrm{t}\mathrm{e}}}{{\mathrm{M}\mathrm{F}\mathrm{I}}_{\mathrm{i}\mathrm{n}\mathrm{i}\mathrm{t}\mathrm{i}\mathrm{a}\mathrm{l}}})\times 100\%$$

(3)

where MFI_composite—MFI of bulk modified filaments, g/min; MFI_initial—MFI of filament with no filler, g/min; all measured with material extrusion 3D printer.

The MFI values of bulk-modified filaments obtained by extrusion through a 3D printer nozzle with a diameter of 500 μm decrease proportionally with an increase in the amount of filler (except for the composite 1.5 wt% MoS₂) in comparison with unfilled PETG, which is associated with the formation of agglomerates that impede the flow of polymer melt through a small diameter nozzle (Table 2).

Figure 6 shows the dependences of the melt flow index on the mass fraction of the modifier (molybdenum disulfide or shungite), approximating the measurement results obtained using the IIRT-5 plastometer and the 3D printer with a 500 μm nozzle).

We used the traditional multifactor correlation analysis technique [36] to calculate the set of Pearson correlation coefficients, P_m,n (the pair-factor covariances, cov(f_m; f_n), normalized on the standard deviations, σ_m and σ_n), for the filler content (1), the MFI_IIRT-5 (2), and the MFI_MEX-3D (3) variables:

P_m,n = cov(f_m; f_n)/(σ_m∙σ_n), m,n = {1;2;3}

(4)

cov(f_m; f_n) = (1/N)∙∑_k (f_m,k − mean(f_m))∙(f_n,k − mean(f_n)), k = 1…N

(4a)

σ_z = (1/(N−1))∙∑_k (f_z,k − mean(f_z))², k = 1…N, z = {m;n}

(4b)

where N is the number of the experimental samples (the scope of the statistical errors in Figure 6 is σ-values determined).

The most probable values of the Pearson correlation coefficients between the MFI measured by different methods and the content of the filler (molybdenum disulfide and schungite) in the polymer matrix are presented in

Table 3. The values of the coefficients of paired correlations between the molybdenum disulfide mass fraction and the values of MFI measured under laboratory and production conditions.

Coefficients of Paired Correlations	Filler, wt%	MFI IIRT-5, g/min	MFI MEX 3D, g/min
Filler, wt%	1	−0.87 ± 0.08	−0.72 ± 0.07
MFI IIRT-5, g/min	−0.87 ± 0.08	1	0.94 ± 0.09
MFI MEX 3D, g/min	−0.72 ± 0.07	0.94 ± 0.09	1

and

Table 4. The values of the coefficients of paired correlations between the shungite mass fraction and the values of MFI measured under laboratory and production conditions.

Coefficients of Paired Correlations	Filler, wt%	MFI IIRT-5, g/min	MFI MEX 3D, g/min
Filler, wt%	1	−0.89 ± 0.09	−0.96 ± 0.09
MFI IIRT-5, g/min	−0.89 ± 0.09	1	0.74 ± 0.08
MFI MEX 3D, g/min	−0.96 ± 0.09	0.74 ± 0.08	1

, respectively.

The correlation coefficient between the filler content and MFI for molybdenum disulfide is higher when measured under laboratory conditions (Table 3), while, for shungite, it is higher when measured under production conditions (Table 4).

The 0.94 (±0.09) in Table 3 and 0.74 (±0.08) in Table 4 values also demonstrate that the MFI-printed and MFI-laboratory results are highly correlated in the MoS₂-modified and only middle-correlated in the shungite-modified filament cases.

Apparently, this is due to significant differences in the nature of the structural transformations of the polymer matrix when modified with the considered ingredients: molybdenum disulfide forms single agglomerates with sizes of ~1.5 μm, and shungite (at 1.5 wt%) forms extended thread-like structures of ~20 µm length.

With such ratios of the modifier particle size and the nozzle diameter, the scale limitations for molybdenum disulfide do not become substantial enough for a statistically significant reduction in extrusion rate.

With shungite bulk modification, thread-like structures form a reinforcing mesh, the influence of which is so significant that even with a relatively small modifier content, the melt flow through the nozzle of the 3D printer almost completely stops.

Thus, the source of difficulties in the operation of production equipment in comparison with laboratory equipment is the amount of the filler, and not the insufficient stability of the parameters of the extrusion mode (temperature, pressure).

When calculating the entire set of measurements (both shungite- and molybdenum disulfide-modified filaments), the absolute mean of the “Cura flow parameter” correlation coefficient with the MFI-laboratory values was ≈0.3797855, whereas, with the MFI-printed, the values ≈ 0.9999993. Therefore, when managing the 3D printing process, it is necessary to adjust the extrusion parameter “flow” in the Cura slicer in accordance with the results of production, but not laboratory tests.

Figure 7 shows that the amount of filler in the polymer volume affects not only the changes in MFI, but also the printing and technical properties of the filament and, as a result, the properties of 3D printed products.

The photo (Figure 7A) shows that 3D printed disks made of PETG with 0.5 wt% of shungite, with standard extrusion parameters contain printing defects: voids, non-printing, low interlayer adhesion of the filament, etc. (shown by blue arrows).

Correction of the “flow” parameter to a value of 112.2% (in accordance with Table 2) completely eliminates the above defects (Figure 7B), which is additionally verified by analytical structural modeling methods (

Table 5. Analytical structural modeling of SEM images for initial and bulk-modified filaments.

Filament (Polymer + Filler)	Digital Image	Model	Morphological Spectrum	Maximum Amplitude
PETG (no filler)				10.5 ± 0.7
PETG + shungite 1.5 wt%				19 ± 2

).

Bulk modification of the PETG filament with shungite is accompanied by an increase in the maximum amplitude of the morphological spectrum from 10.5 ± 0.7 to 19 ± 2 standard units (Table 5), which is due to an increase in the viscosity of the polymer melt, and, as a result, a decrease in the adhesive interaction of layers.

4. Conclusions

The bulk modification of the polyethylene terephthalate glycol (PETG) with the fillers (shungite and the molybdenum disulfide) changes the rheological properties of the resulting filaments.

• The melt flow index value of the modified filament decreases with an increase in filler content proportion. The physical reason for the polymer material structure changes is, probably, the enlargement of the filler aggregates with an increase in their content in the polymer matrix (from 0 to 1.5 wt%). This contributes to the increased viscosity of local zone formation due to the elevated friction forces between the filler particles’ agglomerates. A further increase in the amount of filler can lead to the termination of bulk-modified filament extruding through the 3D printer nozzle.
• The MFI decreased by 18%–30% depending on the type and the mass fraction of the filler in the filament under consideration. The mass fraction of the filler is a technological parameter that limits the possibility of high-quality formation of 3D-printed products using material extrusion technology. The bulk modification of the PETG with the shungite (1.5 wt%) has such a significant effect on the viscosity and fluidity of the polymer composite melt that (with standard 3D-printing mode settings) various defects of products and printing difficulties occur (up to the malfunction of the extrusion unit). We’ve quantified the significance of the “filler mass fraction” for the “melt flow index” using the correlation analysis calculating technique.
• The «printing» defects (voids, non-prints, low interlayer adhesion of the filament, etc.) as well as pore networks may appear when using the shungite-modified (1.5 wt%) PETG (as well as other filled filaments) in 3D-printed product manufacturing. It leads to product model distortions and unexpected changes in the surface and other properties.

When studying the rheological properties of the filaments, we found that the MFI values measured during the production tests on a 3D printer and in laboratory conditions (with the industrial plastometer (IIRT-5)) differ significantly from each other.

The traditional MFI laboratory measurements are actually indirect since they require the pre-preparation of raw materials (the production of granulate), and the extrusion equipment (plastometers) operation is significantly different from the 3D printing process conditions. The proposed technique allows direct measurements of the MFI for a filament in the form of a thread directly during the 3D printer extrusion. The technique is more sensitive to the determination of MFI for composites with low concentrations of filler in the polymer matrix (up to 1.5 wt%) since the 3D-printers nozzle diameters (~0.3 mm and more) are much smaller than the plastometer’s capillary diameters (2095 or 1180 mm). Moreover, the speed of establishing the MFI value increases by an order of magnitude since the 3D printer-based measurements do not require: the raw materials (granules) pre-preparation, experience using the scientific equipment (plastometer), and special knowledge of the high-molecular compounds.

We additionally carried out a complete procedure for determining the MFI value for the shungite-modified (0.5 wt%) PETG filament using the plastometer (granules were formed from the filament and were extruded through a nozzle with a diameter of 1180 mm). The differences between the MFIs (calculated on the basis of the MEX 3D-printed (~+12%) and the plastometric (~+18%) measurements amount to ~6% (which is not large according to absolute values); however, using an 18% correction led to an uneven thickness of the product, forming additive layers and some (~5%) overspending of the extrudate. At the same time, an adjustment of the Cura flow parameter by ~12% is sufficient to eliminate the observed (without correction) defects.

The control of the extrusion speed (the “flow” parameter in the Cura slicer corresponding with the filament melt flow index (MFI), which can be measured using the developed original metrological technique) is necessary to ensure the quality of the 3D prototyping process.

We’ve considered only the technological capabilities of the Cura software slicer. We assume that it is possible to adjust the filament flow in these and other slicers, too. Comparing the capabilities of different slicers and their impact on the 3D-printed product’s quality is the goal of our future investigations.