The differences in behavior can be explained by defects that appear from raw material to manufacturing. A “snow-ball effect” is generated because the raw material defects are transferred to the final part as manufacturing defects, affecting the part’s functionality.
3.2.1. PET CF15 Raw Material Defects
PET CF15 raw material was studied under the microscope in order to check dimensional deviations, fiber orientation and arrangement in the matrix before utilization.
Figure 10 below presents a section through the PET CF15 filament used in the research. The section shows deviations from the circularity, with several grooves at the edges. The darker areas visible in the cross-section indicate randomly scattered material voids and gaps. it can be noticed further that the fibers are oriented perpendicularly on the section, with some exceptions that are marked.
Figure 11 shows the same issues enhanced 100 times.
Figure 12 presents a longitudinal section of the filament, enhanced 100 times on the microscope. Material voids scattered randomly are also visible. This image offers a ”map” of the studied section with fiber conglomerates and voids. This explains the unequal material distribution in the fiber and also in the studied specimens. Fibers are orientated mainly longitudinally in the filament. There are also fibers oriented randomly in the matrix, as confirmed by the findings reported in [
11]. Visually fiber length varies between 50 and 200 μm, as concluded by analyzing the plots.
Figure 13 shows the external surface of the filament which is irregular, some randomly oriented carbon fibers can be noticed on the external surface. Based on the orientation of the fibers in the rupture area it can be asserted that the main orientation of fibers is longitudinal. In the detailed view in
Figure 14 areas without any carbon fiber are visible, circled red, as well as areas with fiber conglomerates. Both
Figure 13 and
Figure 14 reveal areas around the fibers with micropores, accounting for the lack of adhesion of the fibers to the matrix.
Figure 14 shows also the areas of fiber breakage., In the section through the filament voids are visible that bear the print of the pulled-out fiber.
Figure 14.
Ultrafuse PET CF15—Longitudinal view of the filament—Scale 600× (Scale length 50 μm)—Hitachi S3400N Type II.
Figure 14.
Ultrafuse PET CF15—Longitudinal view of the filament—Scale 600× (Scale length 50 μm)—Hitachi S3400N Type II.
Figure 15 shows a transversal view of the PET CF15 filament after SEM analysis. The 320× scaled image shows the increased number of voids of different dimensions. Only some of these are marked in the figure given the large amount of this defect. The same figure reveals several areas where fiber orientation deviates from the required one. In
Figure 14 lower fiber density areas are marked with red circles, indicative of a random distribution of fibers.
Figure 16 shows the detail marked by the red box in
Figure 15. The defects highlighted in
Figure 15 are more visible, resulting from the voids that mark the poor adhesion of the fibers to the PET matrix. Also differently oriented fibers are more visible. In this detail image a complete fiber is visible, as well as an increased area without fibers marked with a red ellipse. Carbon fiber conglomerates are also visible confirming an irregular distribution of fibers in the raw material.
Figure 15.
Ultrafuse PET CF15—Transversal view of the filament—Scale 320× (Scale length 100 μm)—Hitachi S3400N Type II.
Figure 15.
Ultrafuse PET CF15—Transversal view of the filament—Scale 320× (Scale length 100 μm)—Hitachi S3400N Type II.
Figure 16.
Ultrafuse PET CF15—Transversal view of the filament—Scale 800× (Scale length 50 μm)—Hitachi S3400N Type II.
Figure 16.
Ultrafuse PET CF15—Transversal view of the filament—Scale 800× (Scale length 50 μm)—Hitachi S3400N Type II.
Raw material issues are presented also in [
32], where the studied material is PAHT-CF15 (Polyamide matrix with 15% Carbon Fibers from Innofil/Ultrafuse (BASF, Ludwigshafen, Germany). In this case the porosity is described during the manufacturing phase because the matrix is not adhering completely to fibers, resulting in material voids.
3.2.2. PET CF15 Tensile Specimen Manufacturing Defects
The defects mentioned in the previous subsection influence the characteristics of the final parts that depend also on the machine functionality, the manufacturing parameters and the state of the environment. All the tested specimens were visually inspected by means of a microscope, in order to identify the manufacturing issues caused by the previously mentioned factors.
Figure 17,
Figure 18,
Figure 19,
Figure 20,
Figure 21,
Figure 22,
Figure 23,
Figure 24,
Figure 25,
Figure 26,
Figure 27,
Figure 28,
Figure 29,
Figure 30,
Figure 31,
Figure 32,
Figure 33,
Figure 34 and
Figure 35 present the identified manufacturing issues of the tested specimens. Material gaps were identified in all specimens, as presented in [
24]. FEM validation of the tests is difficult due to unpredictable manufacturing issues. The volume of manufacturing defects is higher in specimens with a lower infill density of 25%, confirming the geometric instability mentioned in
Section 1. It also offers an explanation for the results obtained for tensile testing, namely the increased number of defects in the considered parts. Also, because of material gaps, the differences in failure of parts with 100% infill are explained by their distribution and volume.
In
Figure 17 a region is highlighted where the deposited material is not bonded, resulting in an area with material inconsistency. Also, due to the gap, the material deposited in its vicinity tends to be affected by slippage resulting in a cross-section of variable area, highlighted by the red polyline in
Figure 17.
Figure 17.
Specimens—Manufacturing defects—Voids between the layers (Scale 100×)—Scale length 200 μm—Nikon T1-SM.
Figure 17.
Specimens—Manufacturing defects—Voids between the layers (Scale 100×)—Scale length 200 μm—Nikon T1-SM.
Figure 18 shows another case of not bonded material in a specimen obtained by material deposition in the thickness direction. The deposited material from the infill area is not adhering to the shell causing gaps.
Figure 18.
Specimens—Manufacturing defects—Voids between the layers (Scale 50×—Scale length 1 mm)—Emspira 3, Leica.
Figure 18.
Specimens—Manufacturing defects—Voids between the layers (Scale 50×—Scale length 1 mm)—Emspira 3, Leica.
In
Figure 19 an area is highlighted where the material is not deposited equally. The layers in the vicinity of the variable width raster have defined edges. Also, the resulted surface is not smooth due to some carbon fibers. The matrix is visibly darker and the fibers whiter, oriented along the direction of material deposition.
Figure 19.
Specimens–Manufacturing defects–Unequal layer thicknesses (Scale 100×)–Scale length 200 μm–Nikon T1-SM.
Figure 19.
Specimens–Manufacturing defects–Unequal layer thicknesses (Scale 100×)–Scale length 200 μm–Nikon T1-SM.
In
Figure 20 the areas are highlighted where the layers lack adhesion., The rupture during tensile tests is initiated in this area. Also, layer 1 is narrower than layer 2.
Figure 20.
Specimens–Manufacturing defects–Lack of adhesion between the layers (Scale 100×)–Scale length 200 μm–Nikon T1-SM.
Figure 20.
Specimens–Manufacturing defects–Lack of adhesion between the layers (Scale 100×)–Scale length 200 μm–Nikon T1-SM.
In
Figure 21 an area of variable layer thickness is highlighted. This type of defect can be due to a variable flow rate that causes a material conglomerate followed by an area of lesser material deposition.
Figure 21.
Specimens–Manufacturing defects–Non-constant thickness (Scale 100×)–Scale length 200 μm–Nikon T1-SM.
Figure 21.
Specimens–Manufacturing defects–Non-constant thickness (Scale 100×)–Scale length 200 μm–Nikon T1-SM.
In
Figure 22 an area is highlighted where a material conglomerate is deposited between the layers, causing a smaller width of the upper layer. Here the width of the specimen is different due to the layer on the left side of the highlighted region.
Figure 22.
Specimens–Manufacturing defects–Material conglomerate (Scale 100×)–Scale length 200 μm–Nikon T1-SM.
Figure 22.
Specimens–Manufacturing defects–Material conglomerate (Scale 100×)–Scale length 200 μm–Nikon T1-SM.
In
Figure 23 the highlighted region displays a faulty material deposition that fails to follow the normal direction of manufacturing. Thus the load path is affected leading to a maximum tensile load in the direction of material deposition.
Figure 23.
Specimens–Manufacturing defects–Inadequate material deposition (Scale 100×)–Scale length 200 μm–Nikon T1-SM.
Figure 23.
Specimens–Manufacturing defects–Inadequate material deposition (Scale 100×)–Scale length 200 μm–Nikon T1-SM.
Figure 24 shows lack of material in the specimen, causing material inconsistency. This defect occurred near the tensile failure area, leading to breakage in this region.
Figure 24.
Specimens–Manufacturing defects–Material voids (Scale 100×)–Scale length 200 μm–Nikon T1-SM.
Figure 24.
Specimens–Manufacturing defects–Material voids (Scale 100×)–Scale length 200 μm–Nikon T1-SM.
Figure 25 and
Figure 26 show cross-sections of the specimens with material void that cause delamination as a failure mode during the tensile testing.
Figure 25.
Specimens–Manufacturing defects–Material voids (Scale 50×)–Scale length 200 μm–Nikon T1-SM.
Figure 25.
Specimens–Manufacturing defects–Material voids (Scale 50×)–Scale length 200 μm–Nikon T1-SM.
Figure 26.
Specimens–Manufacturing defects–Material voids (Scale 100×–Scale length 1 mm)—Emspira 3, Leica.
Figure 26.
Specimens–Manufacturing defects–Material voids (Scale 100×–Scale length 1 mm)—Emspira 3, Leica.
Figure 27 shows a specimen edge of inadequate geometry.– After manufacturing, the deposited material fails to follow the ideal geometry. The higher material flow rate in this area causes the depositing of a larger amount of material.
Figure 27.
Specimens–Manufacturing defects–Inconsistent geometry (Scale 100×)–Scale length 200 μm–Nikon T1-SM.
Figure 27.
Specimens–Manufacturing defects–Inconsistent geometry (Scale 100×)–Scale length 200 μm–Nikon T1-SM.
Figure 28 shows a specimen with a lower infill density. The weak bonding of the infill layer is highlighted. This defect causes an inadequate load path, and failure is initiated in this area.
Figure 28.
Specimens–Manufacturing defects–Weak bonding at the junction of the infill layer (Scale 100×)–Scale length 200 μm–Nikon T1-SM.
Figure 28.
Specimens–Manufacturing defects–Weak bonding at the junction of the infill layer (Scale 100×)–Scale length 200 μm–Nikon T1-SM.
Figure 29 highlights the “stair-effect” that occurs at dimensional variation.. The defect appears in the area of variable width, being limited to the region held by the grips and the region used for tensile analysis. This is more visible in thicker layers.
Figure 29.
Specimens–Manufacturing defects–”Stair effect” (Scale 50×–Scale length 1 mm)–Emspira 3, Leica.
Figure 29.
Specimens–Manufacturing defects–”Stair effect” (Scale 50×–Scale length 1 mm)–Emspira 3, Leica.
Figure 30 refers to the same cross-section. The layers do not overlap 100%. The profile of the offset layers follows the red polyline instead of a desired straight line. This defect is more visible in specimens manufactured in the thickness direction.
Figure 30.
Specimens–Manufacturing defects–Offset layers (Scale 50×–Scale length 1 mm)–Emspira 3, Leica.
Figure 30.
Specimens–Manufacturing defects–Offset layers (Scale 50×–Scale length 1 mm)–Emspira 3, Leica.
For specimens with an infill density of 75% or 25% in
Figure 31 it can be observed that the unsupported layers tend to flow until the material sets, affecting the internal structure of the specimen. This effect is more visible for specimens with a lower infill density. The straight line in
Figure 31 represents the ideal geometry of the specimen. The red rectangle highlights the set material after flowing.
Figure 31.
Specimens–Manufacturing defects–Material flowing (Scale 75×–Scale length 1 mm)–Emspira 3, Leica.
Figure 31.
Specimens–Manufacturing defects–Material flowing (Scale 75×–Scale length 1 mm)–Emspira 3, Leica.
Figure 32 shows a SEM image of various manufacturing defects, such as deficient inter-layer adhesion and areas with visible carbon fiber dislocation from the PET matrix. The same plot shows the failure area after tensile testing, where the main failure mode is delamination. The red box in
Figure 32 marks a detail that is presented in
Figure 33. This is an area of inadequate material deposition that does not follow the direction of the nozzle movement.
Figure 32.
Specimens–Manufacturing defects–Various defects (Scale 42×–Scale length 1 mm)–Hitachi S3400N Type II.
Figure 32.
Specimens–Manufacturing defects–Various defects (Scale 42×–Scale length 1 mm)–Hitachi S3400N Type II.
In
Figure 33 shows the area of deficient material deposition. An increased distance between the adjacent layers can be observed. Also on each layer material voids and fibers dislocated from the PET matrix are visible.
Figure 33.
Specimens–Manufacturing defects–Various defects (Scale 320×–Scale length 100 μm)–Hitachi S3400N Type II.
Figure 33.
Specimens–Manufacturing defects–Various defects (Scale 320×–Scale length 100 μm)–Hitachi S3400N Type II.
Figure 34 shows a random inter-layer area which is irregular and also has material voids. The randomly oriented carbon fibers are also visible. The red box marks the area detailed in
Figure 35.
Figure 34.
Specimens–Manufacturing defects–Various defects (Scale 180×–Scale length 300 μm)–Hitachi S3400N Type II.
Figure 34.
Specimens–Manufacturing defects–Various defects (Scale 180×–Scale length 300 μm)–Hitachi S3400N Type II.
Figure 35 shows an area with voids of different dimensions, with fibers of different orientations dislocated from the matrix. The variable inter-layer distance it is also visible.
Figure 35.
Specimens–Manufacturing defects–Various defects (Scale 400×–Scale length 100 μm)–Hitachi S3400N Type II.
Figure 35.
Specimens–Manufacturing defects–Various defects (Scale 400×–Scale length 100 μm)–Hitachi S3400N Type II.
As a conclusion, all manufacturing defects affect the mechanical properties because of material inconsistency that also affects the load path on the parts and, in the end, its performance. All specimens displayed material voids of different dimensions. Other possible defects cannot be predicted: varying fiber orientation, inconsistent inter-layer gaps or fiber dislocation from the matrix. It is recommended to test for certain failure types in order to define the material behavior, to determine a pattern and the optimal values for the main manufacturing parameters. The rest of the manufacturing parameters must be consistent, and the same machine should be used for printing.
3.2.3. PET CF15 Tensile Specimen Failure Modes
During the tensile tests different failure modes were identified that are caused also by the manufacturing defects. The most common failure issues identified during the visual inspection are presented in
Figure 36,
Figure 37,
Figure 38 and
Figure 39.
Figure 36 shows a specimen after tensile testing, with fibers dislocated from the matrix. This type of failure can appear due to a weaker bonding between the matrix and the fibers, which can also cause fiber breakage.
Figure 36.
Specimens–Failure Modes–Fibers dislocated from the matrix and fiber breakage (Scale 100×)–Scale length 200 μm–Nikon T1-SM.
Figure 36.
Specimens–Failure Modes–Fibers dislocated from the matrix and fiber breakage (Scale 100×)–Scale length 200 μm–Nikon T1-SM.
Figure 37 shows a specimen after tensile testing where the failure is a delamination of the layers. This can be caused by manufacturing defects, such as material gaps or material conglomerates or weak bonding between the layers. Further it can be noticed that layer 2 is wider than its adjacent layers.
Figure 37.
Specimens–Failure Modes–Delamination of the layers (Scale 100×)–Scale length 200 μm–Nikon T1-SM.
Figure 37.
Specimens–Failure Modes–Delamination of the layers (Scale 100×)–Scale length 200 μm–Nikon T1-SM.
Figure 38 shows a specimen after tensile testing and its rupture area; further cracks near the failure area are visible.
Figure 38.
Specimens–Failure Modes–Cracks (Scale 100×)–Scale length 200 μm–Nikon T1-SM.
Figure 38.
Specimens–Failure Modes–Cracks (Scale 100×)–Scale length 200 μm–Nikon T1-SM.
Figure 39 shows a specimen after tensile testing and its rupture area; delamination is present and also cracks starting from the failure region and affecting almost all the layers in the vicinity of the rupture.
Figure 39.
Specimens–Failure Modes–Delamination and cracks (Scale 100×)–Scale length 200 μm–Nikon T1-SM.
Figure 39.
Specimens–Failure Modes–Delamination and cracks (Scale 100×)–Scale length 200 μm–Nikon T1-SM.
The main identified failure modes are delamination and material cracks, with fiber brakeage. Also in many cases the fibers were dislocated from the matrix, confirming the studies that assert that the matrix does not adhere completely to the fibers.