Experimental verification of the homogeneity of the compressive mechanical properties of PEEK and Carbon PEEK in the ARGO 500 print volume
Polyetheretherketone (PEEK) is a semicrystalline aromatic thermoplastic polymer belonging to the polyaryletheretherketone (PAEK) family. It is able to combine excellent mechanical characteristics with exceptional chemical resistance in an extremely wide range of temperatures: from cryogenic (<150 ° C) up to a continuous use temperature of 250 ° C. The peculiar properties of this polymer (self-lubrication, self-extinguishing, resistance to hydrolysis, low degassing, and dielectric properties) make it the ideal material for many industrial sectors, from energy to aerospace, from medical to the electrical industry.
Given its chemical-physical characteristics, PEEK is an extremely complex polymer to be processed using Fused Filament Fabrication (FFF) technology due to its viscosity, high melting temperature and semi-crystalline nature. Semi-crystalline polymers have crystalline domains within an amorphous matrix. In the case of PEEK, the crystalline phase can reach values in the order of 40%. [1]
The components made using additive manufacturing technology have a structure characterized by layers consisting of rasters and air gaps.
Interfacial adhesion between adjacent single filaments occurs through heat-induced molecular diffusion. (Figure 1). The quality of the bond depends on the temperature of the surrounding environment and on the variations of the convective conditions inside the build chamber.
In fact, when the semi-fused filament is extruded through the nozzle, during the cooling phase, these ordered regions show a greater volumetric contraction than the amorphous component and a different shrinkage rate depending on the size and shape of the individual domains. These phenomena lead to the formation of residual stresses during cooling which can cause distortions, deformations, detachment of the piece from the printing surface, weak bonding force between the individual layers, high porosity and reduction of the section on which the load rests.
To exercise control over the process and increase its repeatability, the Roboze Automate technological ecosystem provides an insulated stainless steel chamber capable of reaching a homogeneous temperature throughout the print volume (500x500x500 mm). The use of a thermostat working environment allows to reduce the temperature difference between the extruded material (up to 450 ° C) and the external environment (up to 180 ° C), slowing down the cooling process and reducing the forces of retraction and residual stresses within the printed component.
Compression stresses are intrinsically present in many engineering systems both due to the application of a compression load that insists directly on the component, and due to the application of impact or bending loads. Another phenomenon directly related to compression loads is buckling, which severely limits the efficiency of the systems leading to an underutilization of the real properties of the material.
Hence, the present work aims to examine the effect of position variation in the print volume of the Roboze ARGO 500 on the compressive strength of PEEK and Carbon PEEK samples.
Experimental Methods
The study was performed on a Roboze ARGO 500 (dimensions: (X) 500 x (Y) 500 x (Z) 500 mm) fed with a filament having a diameter of 1.75 ± 0.05 mm. This thermoplastic filament was subsequently extruded through a 0.6 mm diameter nozzle. In order to minimize the concentration of water molecules absorbed by the filament due to exposure to the atmospheric environment, before starting the printing process, both spools of polymeric material were subjected to a drying cycle at a temperature of 100 ° C for 12 hours in the HT Dryer (Roboze SpA, Italy).
- Step 1: Design of the CAD model
The FFF process consists in the extrusion and subsequent deposition of a fused filament of polymeric material. The starting point of the process is the CAD model of the object designed directly through a solid modeling software. The design specifications of the CAD model are summarized in Figure 3.
- Step 2: Generation of the mesh and preparation of the print job
Following the design of the solid of interest, to proceed with printing it is necessary to discretize its surface by generating a stereolithographic file (* .stl). This format uses a series of interconnected triangles to recreate the geometric surface of the solid; the result of the discretization process is called "mesh".
Once the * .stl file has been generated, the slicing software (Prometheus) allows you to set up the print job by selecting the quantity and positioning of the parts, arranging the support structures and defining the process parameters.
Depending on their position in the XY plane and along the Z axis, each portion of the tower has been identified with a unique code. The nomenclature used is represented schematically in Figure 5.
- Step 3: Slicing and loading of the machine file
Once the print job has been defined, the slicing software applies the selected parameters, generating a series of instructions (* .gcode) compatible with the machine language provided by the printers.
Once the device has received the file, it is possible to proceed with printing the component.
- Step 4: Layer-by-layer construction of the piece
The printer software reads the * .gcode file containing information about the layers into which the model has been divided and guides the layer-by-layer construction of the physical model. The filament of thermoplastic material is fused with the help of heat, is extruded through a small diameter nozzle and then is deposited layer by layer on the printing platform. Downstream of this process, five towers in PEEK and five in Carbon PEEK were obtained.
- Step 5: Post processing
Once extracted from the build chamber, the printed part was removed from the build plate and the individual towers underwent a milling process in order to obtain the necessary specimens for mechanical characterization. For each tower, test pieces of compressive strength and modulus were obtained at different heights. The various dimensions are represented in Figure 5 a.
In this way it was possible to map five different positions in the XY plane at three different heights along the Z coordinate.
The characteristic dimensions of the specimens are shown below:
- Compressive strength: (b) 12.7mm x (h) 12.7mm x (ℓ) 25.4mm;
- Compression module: (b) 12.7mm x (h) 12.7mm x (ℓ) 50.8mm.
All samples were tested using a universal testing machine (Zwick / Roell Z020, Germany) according to ASTM D695-15.
Before proceeding with the test, the samples were conditioned at 23 ± 2 ° C and 50 ± 5% relative humidity for 40 hours. Subsequently, the width (h) and thickness (b) of the sample at different points along its length were measured by means of a centesimal digital caliper (Mitutoyo Corp, Japan), calculating the minimum value of the cross-section (A). The average values of the individual characteristic dimensions measured on the samples made of Carbon PEEK and PEEK are summarized in the following table.
| Specimen type | h [mm] | b [mm] | A [mm²] |
|---|---|---|---|
| CPEEK Compression modulus | 12,71 ± 0,03 | 12,64 ± 0,27 | 160,87 ± 3,48 |
| CPEEK Ultimate compression strength | 12,72 ± 0,20 | 12,72± 0,31 | 161,87± 4,26 |
| PEEK Compression modulus | 12,73± 0,07 | 12,72 ± 0,05 | 161,80 ± 1,42 |
| PEEK Ultimate compression strength | 12,82 ± 0,02 | 12,80 ± 0,05 | 164,13 ± 0,81 |
Then we proceeded to position the specimen between the surfaces of the compression tool so that the axis of symmetry of the sample coincides with the axis of the piston, making sure that the ends of the specimen are parallel to the surface of the compression tool. The test was therefore started with the test speed set at 1.3 mm / min.
The typical trend of a stress-strain curve for Carbon PEEK is shown below:
Except for the segment AC (typical measurement artifact), Carbon PEEK shows Hookeian (linear) behavior. The point of intersection (C) between the abscissa axis and the extension of the linear section (BD) is the correct point of origin on the abscissa axis from which all the deformations were measured.
The summary of the experimental results for the Carbon PEEK specimens is shown in the table:
| Material | Ec [GPa] | σM [MPa] |
|---|---|---|
| Carbon PEEK | 4,8 ± 0,2 | 207 ± 8 |
The graph below illustrates an overview of the experimental results. The maximum compressive stress values expressed in MPa can be read on the ordinates, while the Z coordinate at which the individual samples have been milled is indicated on the abscissa (note: Z = 1 is the position closest to the printing plane). By means of an appropriate color code, on the other hand, the position in the XY plane of the single specimens has been indicated, as can be deduced from the legend of Figure 10.
The data analysis highlights that most of the tested specimens show compressive strength values on average equal to 207 ± 8 MPa, with peaks of 215 MPa. The statistical dispersion of the data is extremely small along the Z coordinate and in the XY plane. A similar trend is observed for the elastic modulus (Figure 11).
It is important to note that, taking TECAPEEK ® CF30 as a reference, having 3.4 GPa of modulus [2], the response of the compression modulus of Carbon PEEK processed with Roboze technological ecosystem is 29% higher.
As regards to the PEEK specimens, a characteristic example of a stress-strain curve is shown below.
Also in this case the Hookeian behavior of the material is evident, however, the angular coefficient is lower due to the lower intrinsic rigidity of the virgin polymer compared to the composite. The same considerations made previously for Carbon PEEK apply to the neat material’s curve.
| Materia | Ec [GPa] | σM [MPa] |
|---|---|---|
| Carbon PEEK | 4,8 ± 0,2 | 207 ± 8 |
As can be seen from the following graphs, the elastic modulus of PEEK compression remains constant, except for minimal oscillations, as the spatial position inside the build chamber varies, settling at an average value of 4.0 ± 0.2 GPa.
It is important to underline that this value is perfectly in line with the data of a sample of Solvay KT-820 NT made through Injection Molding (4.1 GPa) [3].
Compressive strength results are also excellent. Unlike what was observed for Carbon PEEK, given their ductile nature, the PEEK specimens did not break during the test. At 20% deformation, the average compressive strength value settles at 134 ± 2 MPa. The low average square deviation value testifies to the exceptional stability of the mechanical properties throughout the entire print volume.
The study aimed to evaluate the compressive mechanical properties of PEEK and Carbon PEEK specimens within the print volume of an ARGO 500 in order to assess the independence of these mechanical properties from the position on the XY plane and the height along the Z-axis of construction of the piece. Downstream of the results obtained, it is possible to conclude that this objective has been achieved, as the printed samples show properties of resistance and modulus unchanged with the position, except for slight oscillations. In fact, in all cases, the statistical dispersion of the data never differs by more than 5% from the average measured value. In particular, it should also be emphasized that the modulus values recorded are greater (in the case of Carbon PEEK) or comparable (in the case of PEEK) to the values found in the literature of equivalent samples made using Injection Molding technology.