- Email: [email protected]
Effect of Fiber Orientation in Fatigue Properties of FRAM Components
ScienceDirect ScienceDirect Procedia Manufacturing 00 (2018) 000–000
Available online at www.sciencedirect.com
Available Availableonline onlineatatwww.sciencedirect.com www.sciencedirect.com
www.elsevier.com/locate/procedia
ScienceDirect ScienceDirect
www.elsevier.com/locate/procedia
Procedia Manufacturing 00 (2018) 000–000
Procedia Manufacturing 26 (2018) 892–899 Procedia Manufacturing 00 (2017) 000–000
www.elsevier.com/locate/procedia 46th SME North American Manufacturing Research Conference, NAMRC 46, Texas, USA
46th SME North American Manufacturing Research Conference, NAMRC 46, Texas, USA
Effect of Fiber Orientation in Fatigue Properties of FRAM Components Effect of Fiber Orientation in Fatigue Properties of FRAM Manufacturing Engineering Society International Conference 2017, MESIC 2017, 28-30 June Components b, Astrit Imeria,d, Ismail Fidan(Pontevedra), *, Michael Spain Allenc, Garrett Perryd 2017, Vigo c Department of a,d Mechanical Engineering, Tech University, 38505, USA d Astrit Imeri , Ismail Fidanb,Tennessee *, Michael AllenCookeville, , Garrett Perry a
of Manufacturing and Engineering Technology, Tennessee Tech Cookeville, 38505,Trade-off USA CostingDepartment models for capacity optimization inUniversity, Industry 4.0: Department of Mathematics, Tennessee Tech University, Cookeville, 38505,38505, USA USA Department of Mechanical Engineering, Tennessee Tech University, Cookeville, Center for used Manufacturing Research,Technology, Tennessee University, 38505, USA38505, USA Department of Manufacturing andcapacity Engineering Tennessee Tech Cookeville, University, Cookeville, between andTech operational efficiency b
c
a
b
d
c Department of Mathematics, Tennessee Tech University, Cookeville, 38505, USA Center for Manufacturing a Research, Tennessee a,* Tech University, b Cookeville, 38505, b USA * Corresponding author. Tel.: +1-931-372-6298; fax: +1-931-372-3813. E-mail address: [email protected] a University of Minho, 4800-058 Guimarães, Portugal * Corresponding author. Tel.: +1-931-372-6298; fax: +1-931-372-3813. b Unochapecó, 89809-000 Chapecó, SC, Brazil E-mail address: [email protected] d
A. Santana , P. Afonso , A. Zanin , R. Wernke
Abstract
Abstract The purpose of this research study is to evaluate the effect of fiber pattern and infill type on the tension-tension fatigue properties Abstract of Fiber Reinforced Additive Manufacturing (FRAM) specimens with a load ratio of 0.1. As for the design of the study the The purpose this research is to fiber evaluate the effectand of fiber patterninto and “concentric” infill type onand the “isotropic” tension-tension fatigue depending properties specimens areofgenerated withstudy different orientations are divided specimens, Under the concept oflaid "Industry 4.0", production processes will be resistant pushed to be of Fiber Reinforced Manufacturing (FRAM) fatigue specimens with load ratio of specimens 0.1. As increasingly fortothe design of study the on the pattern the fiberAdditive is down. From experimental testing, thea most failure turninterconnected, outthe to be carbon information onwith awith real timeisotropic basis necessarily, much more Ininfill. this context, capacity optimization specimens arebased generated fiberand, orientations divided intoefficient. “concentric” andAs “isotropic” specimens, depending fiber reinforced nylon 0 different rings infill and 1 and ringare isotropic with concentric a conclusion, it is noticed that goes thefiber traditional aim ofFrom capacity maximization, contributing for organization’s profitability and on thebeyond pattern is rings laid down. experimental fatigue testing, the fatigue mostalso resistant specimens tofor failure turn out to be value. carbon increasing the the number of for concentric infill specimens increases the performance, while isotropic infill increasing fibernumber reinforced nylon with 0decreases rings infillperformance. and 1 ring isotropic with concentric As a conclusion, it is instead noticed that Indeed, lean management and isotropic continuous improvement approaches suggestinfill. capacity optimization of the of rings actually the fatigue increasing the number of ringsoffor concentric infill specimens fatigueisperformance, while for isotropic maximization. The study capacity optimization andincreases costing the models an important research topic infill that increasing deserves the number of rings decreases theand fatigue performance. contributions fromactually both the practical theoretical perspectives. This paper presents and discusses a mathematical © 2018for Thecapacity Authors. management Published by Elsevier B.V. model based on different costing models (ABC and TDABC). A generic model has been © 2018 The Authors. Published by Elsevier B.V. Peer-reviewand under responsibility of the scientific committee of design NAMRI/SME. developed it was used to analyze idle capacity and to strategies maximization of organization’s Peer-review under responsibility of the scientific committee of the 46th SME Northtowards Americanthe Manufacturing Research Conference. © 2018 The Authors. Published by Elsevier B.V. value. The trade-off capacity maximization vs operational efficiency is highlighted and it is shown that capacity Peer-review under Manufacturing; responsibility Fiber of theReinforced scientificAdditive committee of NAMRI/SME. Keywords: Additive Manufacturing; Composite Materials; Fatigue
optimization might hide operational inefficiency.
© 2017 The Authors. Published by Elsevier B.V. Keywords: Additive Manufacturing; Fiber Reinforced Additive Manufacturing; Composite Materials; Fatigue shape partsEngineering and allows Society usage of over 90% of the material Peer-review under responsibility of the scientific committee of the Manufacturing International Conference 1. Introduction [1]. The cost constraint to produce a part is reduced is not 2017.
shape parts and allowsthe usage over 90% ofjigs, the fixtures, material as significant because needoffor complex 1. Introduction Amongst many production technologies, AM is one of the latest industry manufacturing processes. as significant because the need for complex fixtures, customization of the parts are also no longerjigs, a constraint. Amongst technologies, AM is that one Contrary tomany the production conventional manufacturing or moldings is eliminated. Complexity and AM is being utilized at an of the waste latestandindustry processes. creates leftovers,manufacturing AM produces near-net 1. Introduction customization of the parts are also no longer a constraint. Contrary to the conventional manufacturing that AM is being utilized at an creates waste and leftovers, AM produces near-net The cost of idle is a fundamental for companies and their management of extreme importance 2351-9789 © 2018 Thecapacity Authors. Published by Elsevier information B.V. Peer-review responsibility of theIn scientific committee of NAMRI/SME. in modern under production systems. general, it is defined as unused capacity or production potential and can be measured 2351-9789 2018 The Authors. Published by Elsevier B.V.hours of manufacturing, etc. The management of the idle capacity in several©ways: tons of production, available
The cost constraint to produce a part is reduced is and not Keywords: Cost Models; ABC; TDABC; Capacity Management; Idle Capacity;[1]. Efficiency orOperational moldings is eliminated. Complexity
Peer-review underTel.: responsibility the761; scientific committee NAMRI/SME. * Paulo Afonso. +351 253of 510 fax: +351 253 604of741 E-mail address: [email protected]
2351-9789 © 2017 The Authors. Published by Elsevier B.V. Peer-review under responsibility of the scientific committee of the Manufacturing Engineering Society International Conference 2017. 2351-9789 © 2018 The Authors. Published by Elsevier B.V. Peer-review under responsibility of the scientific committee of the 46th SME North American Manufacturing Research Conference. 10.1016/j.promfg.2018.07.115
2
Astrit Imeri et al. / Procedia Manufacturing 26 (2018) 892–899 Author name / Procedia Manufacturing 00 (2018) 000–000
increasing rate in sectors such as aerospace and motorsports due to offering high geometrical complexity and short manufacturing lead times [2]. Another main driver of the adoption of AM is the lightweight production of parts [3]. Traditionally, AM has focused on metals and polymers. However, recent techniques like FRAM have widened the scope of material used in AM. Yet, the field of FRAM has been narrowly studied. Generally, the fibers used in FRAM have been discontinuous [4]. In recent years, a printer the Markforged Mark Two [5] seen in Figure 1, has appeared on the market and is able to print with continuous fibers.
893
properties of Acrylonitrile Butadiene Styrene (ABS) [9]. Since continuous FRAM is a new technology not many of the properties have been tested and analyzed. Although the individual composite materials’ mechanical properties have been investigated in detail [10], there have been no studies on the details of FRAM depending on the infill and orientation configurations. In this study, the effect of geometrical orientation of fiber lay down patterns on the fatigue properties of FRAM was investigated. Nomenclature FRAM Fiber Reinforced Additive Manufacturing AM Additive Manufacturing MKF Markforged Mark Two printer ASTM American Society for Testing and Materials CAD Computer-Aided Design STL Stereo Lithography C Concentric I Isotropic CF Carbon Fiber FG Fiberglass KV Kevlar R Load Ratio
Figure 1. Markforged Mark Two 3D Printer. This printer functions by first constructing a matrix of Nylon or Onyx and then overlaying that matrix with fiber filament layer by layer. The way the fibers are laid down is important, because it changes the structure of the component. Microstructure and macrostructure are crucial for mechanical properties and performance of the parts. Hence, testing the materials for mechanical properties is critical for the utilization of the parts in industry. A very important property is fatigue. Fatigue is officially defined and stated by the American Society for Testing and Materials ( ASTM) [6] as the process of progressive localized permanent structural change occurring in a material subjected to conditions that produce fluctuating stresses and strains at some point or points that may culminate in cracks or complete fracture after a sufficient number of fluctuations. According to Stephens et al. [7] 50-90% of all mechanical failures are fatigue failures. Hence, a thorough investigation in fatigue properties of materials is paramount for designing parts and systems. Generally, the fatigue studies conducted in AM were on the conventional polymer AM materials. Fischer et al studied the behavior of FDM parts printed with Ultem 9085 [8]. Zhang et al worked on the fatigue
2. Test models The ASTM standard used for fatigue testing was E606M [11]. The dimensions of the specimen with a according to the standard are shown in Figure 2. The ASTM specimens were modelled in SolidWork 2016. All specimens, carbon fiber, Kevlar, and fiberglass were printed with MKF.
Figure 2. Fatigue specimen dimensions. T= 5mm; R=10mm; H=10mm; W=20mm. The MKF printer has dual head extrusion nozzles, which print the base and reinforcement materials. One of the nozzles prints only nylon or onyx while the other one prints the reinforcement materials. For the nylon material, the filament comes in a form of 1.75 mm diameter. Kevlar and fiberglass come in diameters of 0.3 mm, while the carbon fiber comes in a diameter of 0.35 mm.
894
Astrit Imeri et al. / Procedia Manufacturing 26 (2018) 892–899 Author name / Procedia Manufacturing 00 (2018) 000–000
To produce a part the digital file is needed. The CAD file is sliced into layers and is converted to STL file. The MKF slicer is Eiger, which is a web-based slicer. From the STL file in the slicer the numerical control code is generated, which commands the printer head what moves to make and when to extrude materials so the part is produced. A schematic of the printer is shown in Figure 3. It should be noted that in Figure 3, the two extrusion heads are in line and there is not only one as it might be percepted from the figure. Both base materials and reinforcement fibers are kept in rolled spools of filament. Due to the sensitivity of the materials to moisture they are kept in a water tight Pelican 1430 dry box for protection. The filaments are pulled into the extrusion heads with the help of the stepper motors.
Figure 3. Schematic of MKF printer extrusion process. The temperature of the extrusion heads are between 265-270oC. At this temperature, while passing through the extrusion head, the base material becomes molten, and begins to solidify once it leaves the nozzle. The first layers of the part are nylon. The fiber reinforcement materials which are not in a molten state are laid down horizontally layer by layer into the nylon matrix. The walls or outer borders of the part are also made up of nylon. Therefore, in every layer, the walls are printed first followed by the reinforcement layers if necessary. This process is concluded with nylon being printing on the opposite surface. Each part produced using FRAM AM follows this procedure layer by layer. In the specimens made out of carbon fiber and nylon there are 40 layers. Of those 40 layers, 32 are fiber reinforced, which is the maximum allowable number of layers to be reinforced. However, for the
3
nylon specimens reinforced with Kevlar or fiberglass there are 50 layers in total. Of those 50 layers, 42 are fiber reinforced, which again is the maximum allowable number of layers to be reinforced. Reason for this difference in the number of layers and maximum allowable number of layers to be reinforced is the different filament diameters for carbon fiber and Kevlar and fiberglass. This difference in the number of layers between carbon fiber and other fibers filled specimens results in a slight fiber volumetric fraction difference (up to 0.1 cm3). Due to this small variation in the fiber volume, the effect of fiber volumetric fraction was not analyzed in this research study. For nylon the used infill in this study was rectilinear, because it shows maximum strength in uniaxial loadings according to Fernandez-Vicente et al [12]. The infill density used is 75%, due to having similar strength values with 100% infill, but time-wise the 75% was more efficient to produce. The slicing software controls the design parameters and based on these parameters different specimens were generated. Changing these controllable settings, specimens with different number of rings, different infill type were generated. Specimens can be grouped into two categories, based on the type of the fiber reinforcement fill type. The two categories are: concentric and isotropic. The horizontal cross-section view of the different fill types are shown respectively in Figure 4a. The yellow lines represent the fiber reinforcement material, while the black portion represents the base material, Nylon.
Figure 4a. Left: Concentric fill. Right: Isotropic fill. The number of different types of generated specimens was six, for each material. To better visualize the horizontal cross-section view for a layer of these specimens is presented in Figure 4b.
4
Astrit Imeri et al. / Procedia Manufacturing 26 (2018) 892–899 Author name / Procedia Manufacturing 00 (2018) 000–000
895
3. Testing machine
Positive stop
Flat Wedge Grippers
Table 1: Abbreviation, naming and description Abbreviation # of rings and Description fiber fill type CF/FG/KV2 Rings Two concentric 2RC concentric rings of fiber in each fiber layer on the Nylon matrix CF/FG/KV3 Rings Three concentric 3RC concentric rings of fiber in each fiber layer on the Nylon matrix CF/FG/KV0 Rings isotropic Isotropic fill with 0RI 0 degree angle direction in each fiber layer on the Nylon matrix CF/FG/KV1 Rings isotropic Isotropic fill with 1RI one concentric ring on the outer border of the fiber fill CF/FG/KV2 Rings isotropic Isotropic fill with 2RI two concentric ring on the outer border of the fiber fill CF/FG/KV3 Rings isotropic Isotropic fill with 3RI three concentric rings on the outer border of the fiber fill
Extensometer
Having many specimens, naming them was a challenge too. Hence, the abbreviation, number of rings and infill type, and description are presented in Table 1.
Specimen
Figure 4b. All infill types and orientations generated.
The specimens were tested using a MTS 810 as seen in Figure 5. This model is a closed loop servohydraulic machine. The testing variables, such as, frequency, amplitude, and average load were all controlled and input through MTS’s Multipurpose Testware interface. This machine has two gripping heads. The lower head is actuated in order to load the specimen while the upper head is used only for gripping purposes. The gripping pressure applied to the larger portions of the test specimens, Figure 2, was 4 MPa. This value was decided upon because higher gripping pressure caused an immediate crack in the specimens while lower gripping pressure caused the specimen to slip.
Figure 5. Specimen clamped in jaws of the MTS810. Specimen alignment is important when testing for mechanical properties. Poor specimen alignment can be a significant contributor to premature failure. It can also affect repeatability and reproducibility. Therefore, wedge grips which incorporated positive stops were used to make sure that the specimen was aligned as shown in Figure 6.
Positive stop
Figure 6. Positive stops on the flat wedge grip.
Astrit Imeri et al. / Procedia Manufacturing 26 (2018) 892–899 Author name / Procedia Manufacturing 00 (2018) 000–000
896
The machines software interface provides a live data stream to the user for the force, displacement, and iteration which were the testing variables of interest. This assures the user that the machine is operating correctly. 4. Determination of failure The point of comparison between the different fiber reinforcement type and orientation was the number of cycles until total separation. The total separation was defined two different ways. The first was complete failure of the specimen into two different parts as seen in Figure 7. The second was when a visible crack had formed in either the Nylon or the inlaid fibers. For this study, the first way of total separation was used. In addition, since the goal of this study is to determine if a correlation exists between geometric orientations and fiber reinforcement material in relation to fatigue properties some of the specimens that did not break were truncated after 10000 cycles. 5. Testing Three specimens of each type presented in Table 1 were produced in order to have a better statistical representation. The type of load for the testing specimens was tensile-tensile load, with a load ratio R
Table 2: Results from the first load (3.33-0.33 kN) Load (kN) Material Rings
5
of 0.1. Different load ratios were not tested in this study but may be later. The Intermediate Value Property Theorem (IVP) was used to determine the magnitude of the loads used for testing. First, a high load of 17.4 kN and low load of 3.33 kN were chosen. The logic behind this selection is to eliminate the specimens that will fail immediately. Further, if one specimen failed during the first cycle, meaning that the load is excessive, the second and the third tests were eliminated. Then intermediate values were also selected and testing was repeated. The testing magnitudes for the maximum loads were:17.4 kN, 13.8 kN, 10.3 kN, 3.33 kN. Tables 2-5 provide a comprehensive summary of all the specimens and loads results. They are arranged from the lowest to the highest load. The results of testing at the first load (3.33-0.333 kN) are presented in Table 2. At this load the specimens with concentric infill were tabulated only. Isotropic infill specimens did not break at all at this load. From the results, the three rings concentric infill for any of the materials gave higher number of cycles. Sequently, the three rings isotropic specimens were then tested in the second load. The second load (10.3-1.03 kN) results are shown in Table 3, the concentric specimens with 3 rings, that showed higher number of cycles in the first load, failed in the first cycle.
Type
Test #1
Test #2
Test #3
3.33-0.33 3.33-0.33
Carbon Fiber Carbon Fiber
2 3
Concentric Concentric
510 10000+
222 -
417 -
3.33-0.33
Kevlar
2
Concentric
45
190
9
3.33-0.33
Kevlar
3
Concentric
383
72
181
3.33-0.33
Fiberglass
2
Concentric
168
132
137
3.33-0.33
Fiberglass
3
Concentric
644
561
644
Except the three rings concentric specimens failing in the first cycles, in the second load too, the isotropic with three concentric rings for any material failed in the first cycle. The Kevlar specimens with two rings isotropic infill, contrary to fiberglass and carbon fiber specimens, failed in the first cycle. Kevlar and fiberglass isotropic with concentric one ring failed within a range of cycles under 10000 (the truncating number of cycles). However, carbon fiber specimen did not break in that range.
Author name / Procedia Manufacturing 00 (2018) 000–000
6
Astrit Imeri et al. / Procedia Manufacturing 26 (2018) 892–899
Table 3: Results from the second load (10.34-1.03 kN) Load (kN) Material Rings
897
Type
Test #1
Test #2
Test #3
10.34-1.03
Fiberglass
0
Isotropic
750
1161
989
10.34-1.03
Fiberglass
1
Iso+Conc
699
996
744
10.34-1.03
Fiberglass
2
Iso+Conc
19
26
31
10.34-1.03
Fiberglass
3
Iso+Conc
1
-
-
10.34-1.03
Carbon Fiber
0
Isotropic
10000+
-
-
10.34-1.03
Carbon Fiber
1
Iso+Conc
10000+
-
-
10.34-1.03
Carbon Fiber
2
Iso+Conc
3753
3995
1747
10.34-1.03
Carbon Fiber
3
Iso+Conc
1
1
10.34-1.03
Kevlar
0
Isotropic
10000+
-
-
10.34-1.03
Kevlar
1
Iso+Conc
54
285
196
10.34-1.03
Kevlar
2
Iso+Conc
1
-
-
10.34-1.03
Kevlar
3
Iso+Conc
1
-
-
10.34-1.03
Carbon Fiber
3
Concentric
1
-
-
10.34-1.03
Fiberglass
3
Concentric
1
-
-
After truncating the specimens from the previous load, in the 13.8-1.38 kN, the isotropic with two concentric rings specimens of carbon fiber and fiberglass failed in the first cycle. Kevlar specimens with one ring isotropic infill failed in the first cycle, while isotropic with zero rings had just a small number of cycles. The only specimens that had more than 10000 cycles were the carbon fiber with zero rings. Table 4: Results from the third load (13.8-1.38 kN) Load (kN) Material Rings
Results are presented in Table 4. For the highest load case, the 17.4-1.74 kN, testing all other specimens failed in the first cycle other than carbon fiber specimens with no rings. However, even in these specimens a cracking sound was heard in the first cycle. Results from the final load are presented in Table 5
Type
Test #1
Test #2
Test #3
13.8-1.38
Fiberglass
0
Isotropic
5
66
20
13.8-1.38
Fiberglass
1
Iso+Conc
15
24
38
13.8-1.38
Fiberglass
2
Iso+Conc
1
-
-
13.8-1.38
Carbon Fiber
0
Isotropic
10000+
-
-
13.8-1.38
Carbon Fiber
1
Iso+Conc
324
275
209
13.8-1.38
Carbon Fiber
2
Iso+Conc
1
-
-
13.8-1.38
Kevlar
0
Isotropic
3
7
6
13.8-1.38
Kevlar
1
Iso+Conc
1
-
-
.
Astrit Imeri et al. / Procedia Manufacturing 26 (2018) 892–899
898
Author name / Procedia Manufacturing 00 (2018) 000–000
7
Table 5: Results from the fourth load (17.4-1.74 kN)
Load (kN)
Material
Rings
Type
Test #1
Test #2
Test #3
17.4-1.74
Fiberglass
0
Isotropic
1
-
-
17.4-1.74
Fiberglass
1
Iso+Conc
1
-
-
17.4-1.74
Carbon Fiber
0
Isotropic
19
4
12
17.4-1.74
Carbon Fiber
1
Iso+Conc
1
-
-
17.4-1.74
Carbon Fiber
2
Iso+Conc
1
-
-
17.4-1.74
Kevlar
0
Isotropic
1
-
-
6. Discussion Specimens of nylon reinforced with carbon fiber, fiberglass or Kevlar were printed with different infill types. These specimens were tested at four different loads with a load ratio (R) of 0.1. The first type of the infill was concentric infill with two and three rings. For specimens with concentric infill increasing the rings from two to three proved to have an impact on the number of cycles to failure. This result is independent of the type of fiber material used for infill. However, the concentric specimens failed in the first cycle for the higher loads. Hence, it can be concluded that for low loads, such as the 3.33-0.33, concentric type infill works well. Additionally, to make a part more resistant to failure due to fatigue the number of rings can be increased up to the geometric limit for both isotropic and isotropic with concentric rings. It should be noted that for Kevlar filled specimens with two concentric rings in Table 2, there seems to be unusual variation of the number of cycles up to failure. Since for each type of the specimen, there were three experiments done and from the other experiments there was not such variations, this issue could have happened due to printer anomalies while printing the specimens. Parts are sensitive to any irregularity while printing. Similar situation is observed with one specimen of carbon fiber filled isotropic with zero rings. The three ringed isotropic with concentric infill failed for the second load (10.34-1.03 kN) for all fiber materials. However, at this load, the two ring infill of carbon fiber and fiberglass had finite number of cycles up to failure, which give an intuition that decreasing the number of cycles increased the resistance to breaking. For the second load case the type of material used for the infill was also important. The Kevlar specimen, contrary to carbon fiber and fiberglass, failed immediately unlike the first load case which was material independent.
In the third load case (13.8-1.38 kN) the specimens with two concentric rings with isotropic infill were truncated because they failed in the first cycle. The one ring specimens did not fail during the first cycle; therefore showing better fatigue resistance than the two ring isotropic with concentric filled specimens. This indicates that decreasing the number of rings increases the fatigue life of the specimens. Based on the results of the experiments the nylon parts failed first followed by the the fiber reinforced specimens. However, for different fibers different patterns of failing modes were observed. Fiberglass and Kevlar with nylon specimens would deflect more than carbon fiber with nylon ones. However, for the same loads the shape of the broken specimens was similar. Failed specimens of different materials at 10.34-1.03 kN are shown in Figure 7.
Kevlar filled with two Fiberglass filled with rings two rings Carbon Fiber filled with two rings
Figure 7. Broken specimens at 10.34-1.03 kN
8
Astrit Imeri et al. / Procedia Manufacturing 26 (2018) 892–899 Author name / Procedia Manufacturing 00 (2018) 000–000
7. Conclusion Therefore, in conclusion the experiments showed that there is a correlation between filler material type and the pattern the reinforcing material is extruded. Among the materials, tested carbon fiber was shown to have better fatigue resistance than the other reinforcing materials. Based on the trend of number of cycles, for concentric infill, adding more rings improves the fatigue life of the specimen. While for isotropic infill, increasing the number of rings, it weakens the fatigue performance. The more rings the less the number of cycles up to failure. In addition, this study revealed that the isotropic with zero rings and one ring isotropic with concentric infill types showed better results than other specimens, with zero rings slightly topping the one ring isotropic with concentric infill specimens. Acknowledgements This work is part of a larger project funded by the Advanced Technological Education Program of the National Science Foundation, DUE #1601587. The funding provided by the National Science Foundation is greatly appreciated. References [1] Swolfs, Y. and Pinho, S.T., 2016, September. Designing and 3D printing continuous fibrereinforced composites with a high fracture toughness. In Proceedings of the 31st Technical Conference of the American Society for Composites (pp. 1-8). DESTech Publications. [2] S. Rawal, J. Brantley, N. Karabudak, Additive manufacturing of Ti–6Al–4V alloy components for spacecraft applications. In Recent Advances in Space Technologies (RAST), 2013 6th International Conference on, 2013. [3] R. Huang, et al., Energy and emissions saving potential of additive manufacturing: the case of lightweight aircraft components, J. Clean. Prod. (0) (2015). [4] Quan, Z., A. Wu, M. Keefe, X. Qin, J. Yu, J. Suhr, et al. 2015. "Additive manufacturing of multidirectional preforms for composites: opportunities and challenges," Mater. Today, 18(9):503-512. [5] Markforged [Online] https://markforged.com/about/ [Accessed on November 2, 2017].
899
[6] ASTM E1823-13, Standard Terminology Relating to Fatigue and Fracture Testing, ASTM International, West Conshohocken, PA, 2013, p.1034 www.astm.org” [7] Stephens et al., 2001. Metal fatigue in engineering 2nd ed., New York: Wiley. [8] Fischer, M. & Schöppner, V. JOM (2017) 69: 563. https://doi.org/10.1007/s11837-016-2197-2 [9] Zhang, H., Cai, L., Golub, M., Zhang, Y., Yang, X., Schlarman, K. and Zhang, J., 2018. Tensile, creep, and fatigue behaviors of 3D-printed acrylonitrile butadiene styrene. Journal of Materials Engineering and Performance, 27(1), pp.57-62. [10] Markforged [Online] https://static.markforged.com/markforged_composite s_datasheet.pdf [Accessed on January 31, 2018] [11] ASTM E606 / E606M-12, Standard Test Method for Strain-Controlled Fatigue Testing, ASTM International, West Conshohocken, PA, 2012, www.astm.org [12] Fernandez-Vicente, M. et al., (2016). Effect of Infill Parameters on Tensile Mechanical Behavior in Desktop 3D Printing. 3D Printing and Additive Manufacturing, 3(3), pp.183–192.
Available online at www.sciencedirect.com
Available Availableonline onlineatatwww.sciencedirect.com www.sciencedirect.com
www.elsevier.com/locate/procedia
ScienceDirect ScienceDirect
www.elsevier.com/locate/procedia
Procedia Manufacturing 00 (2018) 000–000
Procedia Manufacturing 26 (2018) 892–899 Procedia Manufacturing 00 (2017) 000–000
www.elsevier.com/locate/procedia 46th SME North American Manufacturing Research Conference, NAMRC 46, Texas, USA
46th SME North American Manufacturing Research Conference, NAMRC 46, Texas, USA
Effect of Fiber Orientation in Fatigue Properties of FRAM Components Effect of Fiber Orientation in Fatigue Properties of FRAM Manufacturing Engineering Society International Conference 2017, MESIC 2017, 28-30 June Components b, Astrit Imeria,d, Ismail Fidan(Pontevedra), *, Michael Spain Allenc, Garrett Perryd 2017, Vigo c Department of a,d Mechanical Engineering, Tech University, 38505, USA d Astrit Imeri , Ismail Fidanb,Tennessee *, Michael AllenCookeville, , Garrett Perry a
of Manufacturing and Engineering Technology, Tennessee Tech Cookeville, 38505,Trade-off USA CostingDepartment models for capacity optimization inUniversity, Industry 4.0: Department of Mathematics, Tennessee Tech University, Cookeville, 38505,38505, USA USA Department of Mechanical Engineering, Tennessee Tech University, Cookeville, Center for used Manufacturing Research,Technology, Tennessee University, 38505, USA38505, USA Department of Manufacturing andcapacity Engineering Tennessee Tech Cookeville, University, Cookeville, between andTech operational efficiency b
c
a
b
d
c Department of Mathematics, Tennessee Tech University, Cookeville, 38505, USA Center for Manufacturing a Research, Tennessee a,* Tech University, b Cookeville, 38505, b USA * Corresponding author. Tel.: +1-931-372-6298; fax: +1-931-372-3813. E-mail address: [email protected] a University of Minho, 4800-058 Guimarães, Portugal * Corresponding author. Tel.: +1-931-372-6298; fax: +1-931-372-3813. b Unochapecó, 89809-000 Chapecó, SC, Brazil E-mail address: [email protected] d
A. Santana , P. Afonso , A. Zanin , R. Wernke
Abstract
Abstract The purpose of this research study is to evaluate the effect of fiber pattern and infill type on the tension-tension fatigue properties Abstract of Fiber Reinforced Additive Manufacturing (FRAM) specimens with a load ratio of 0.1. As for the design of the study the The purpose this research is to fiber evaluate the effectand of fiber patterninto and “concentric” infill type onand the “isotropic” tension-tension fatigue depending properties specimens areofgenerated withstudy different orientations are divided specimens, Under the concept oflaid "Industry 4.0", production processes will be resistant pushed to be of Fiber Reinforced Manufacturing (FRAM) fatigue specimens with load ratio of specimens 0.1. As increasingly fortothe design of study the on the pattern the fiberAdditive is down. From experimental testing, thea most failure turninterconnected, outthe to be carbon information onwith awith real timeisotropic basis necessarily, much more Ininfill. this context, capacity optimization specimens arebased generated fiberand, orientations divided intoefficient. “concentric” andAs “isotropic” specimens, depending fiber reinforced nylon 0 different rings infill and 1 and ringare isotropic with concentric a conclusion, it is noticed that goes thefiber traditional aim ofFrom capacity maximization, contributing for organization’s profitability and on thebeyond pattern is rings laid down. experimental fatigue testing, the fatigue mostalso resistant specimens tofor failure turn out to be value. carbon increasing the the number of for concentric infill specimens increases the performance, while isotropic infill increasing fibernumber reinforced nylon with 0decreases rings infillperformance. and 1 ring isotropic with concentric As a conclusion, it is instead noticed that Indeed, lean management and isotropic continuous improvement approaches suggestinfill. capacity optimization of the of rings actually the fatigue increasing the number of ringsoffor concentric infill specimens fatigueisperformance, while for isotropic maximization. The study capacity optimization andincreases costing the models an important research topic infill that increasing deserves the number of rings decreases theand fatigue performance. contributions fromactually both the practical theoretical perspectives. This paper presents and discusses a mathematical © 2018for Thecapacity Authors. management Published by Elsevier B.V. model based on different costing models (ABC and TDABC). A generic model has been © 2018 The Authors. Published by Elsevier B.V. Peer-reviewand under responsibility of the scientific committee of design NAMRI/SME. developed it was used to analyze idle capacity and to strategies maximization of organization’s Peer-review under responsibility of the scientific committee of the 46th SME Northtowards Americanthe Manufacturing Research Conference. © 2018 The Authors. Published by Elsevier B.V. value. The trade-off capacity maximization vs operational efficiency is highlighted and it is shown that capacity Peer-review under Manufacturing; responsibility Fiber of theReinforced scientificAdditive committee of NAMRI/SME. Keywords: Additive Manufacturing; Composite Materials; Fatigue
optimization might hide operational inefficiency.
© 2017 The Authors. Published by Elsevier B.V. Keywords: Additive Manufacturing; Fiber Reinforced Additive Manufacturing; Composite Materials; Fatigue shape partsEngineering and allows Society usage of over 90% of the material Peer-review under responsibility of the scientific committee of the Manufacturing International Conference 1. Introduction [1]. The cost constraint to produce a part is reduced is not 2017.
shape parts and allowsthe usage over 90% ofjigs, the fixtures, material as significant because needoffor complex 1. Introduction Amongst many production technologies, AM is one of the latest industry manufacturing processes. as significant because the need for complex fixtures, customization of the parts are also no longerjigs, a constraint. Amongst technologies, AM is that one Contrary tomany the production conventional manufacturing or moldings is eliminated. Complexity and AM is being utilized at an of the waste latestandindustry processes. creates leftovers,manufacturing AM produces near-net 1. Introduction customization of the parts are also no longer a constraint. Contrary to the conventional manufacturing that AM is being utilized at an creates waste and leftovers, AM produces near-net The cost of idle is a fundamental for companies and their management of extreme importance 2351-9789 © 2018 Thecapacity Authors. Published by Elsevier information B.V. Peer-review responsibility of theIn scientific committee of NAMRI/SME. in modern under production systems. general, it is defined as unused capacity or production potential and can be measured 2351-9789 2018 The Authors. Published by Elsevier B.V.hours of manufacturing, etc. The management of the idle capacity in several©ways: tons of production, available
The cost constraint to produce a part is reduced is and not Keywords: Cost Models; ABC; TDABC; Capacity Management; Idle Capacity;[1]. Efficiency orOperational moldings is eliminated. Complexity
Peer-review underTel.: responsibility the761; scientific committee NAMRI/SME. * Paulo Afonso. +351 253of 510 fax: +351 253 604of741 E-mail address: [email protected]
2351-9789 © 2017 The Authors. Published by Elsevier B.V. Peer-review under responsibility of the scientific committee of the Manufacturing Engineering Society International Conference 2017. 2351-9789 © 2018 The Authors. Published by Elsevier B.V. Peer-review under responsibility of the scientific committee of the 46th SME North American Manufacturing Research Conference. 10.1016/j.promfg.2018.07.115
2
Astrit Imeri et al. / Procedia Manufacturing 26 (2018) 892–899 Author name / Procedia Manufacturing 00 (2018) 000–000
increasing rate in sectors such as aerospace and motorsports due to offering high geometrical complexity and short manufacturing lead times [2]. Another main driver of the adoption of AM is the lightweight production of parts [3]. Traditionally, AM has focused on metals and polymers. However, recent techniques like FRAM have widened the scope of material used in AM. Yet, the field of FRAM has been narrowly studied. Generally, the fibers used in FRAM have been discontinuous [4]. In recent years, a printer the Markforged Mark Two [5] seen in Figure 1, has appeared on the market and is able to print with continuous fibers.
893
properties of Acrylonitrile Butadiene Styrene (ABS) [9]. Since continuous FRAM is a new technology not many of the properties have been tested and analyzed. Although the individual composite materials’ mechanical properties have been investigated in detail [10], there have been no studies on the details of FRAM depending on the infill and orientation configurations. In this study, the effect of geometrical orientation of fiber lay down patterns on the fatigue properties of FRAM was investigated. Nomenclature FRAM Fiber Reinforced Additive Manufacturing AM Additive Manufacturing MKF Markforged Mark Two printer ASTM American Society for Testing and Materials CAD Computer-Aided Design STL Stereo Lithography C Concentric I Isotropic CF Carbon Fiber FG Fiberglass KV Kevlar R Load Ratio
Figure 1. Markforged Mark Two 3D Printer. This printer functions by first constructing a matrix of Nylon or Onyx and then overlaying that matrix with fiber filament layer by layer. The way the fibers are laid down is important, because it changes the structure of the component. Microstructure and macrostructure are crucial for mechanical properties and performance of the parts. Hence, testing the materials for mechanical properties is critical for the utilization of the parts in industry. A very important property is fatigue. Fatigue is officially defined and stated by the American Society for Testing and Materials ( ASTM) [6] as the process of progressive localized permanent structural change occurring in a material subjected to conditions that produce fluctuating stresses and strains at some point or points that may culminate in cracks or complete fracture after a sufficient number of fluctuations. According to Stephens et al. [7] 50-90% of all mechanical failures are fatigue failures. Hence, a thorough investigation in fatigue properties of materials is paramount for designing parts and systems. Generally, the fatigue studies conducted in AM were on the conventional polymer AM materials. Fischer et al studied the behavior of FDM parts printed with Ultem 9085 [8]. Zhang et al worked on the fatigue
2. Test models The ASTM standard used for fatigue testing was E606M [11]. The dimensions of the specimen with a according to the standard are shown in Figure 2. The ASTM specimens were modelled in SolidWork 2016. All specimens, carbon fiber, Kevlar, and fiberglass were printed with MKF.
Figure 2. Fatigue specimen dimensions. T= 5mm; R=10mm; H=10mm; W=20mm. The MKF printer has dual head extrusion nozzles, which print the base and reinforcement materials. One of the nozzles prints only nylon or onyx while the other one prints the reinforcement materials. For the nylon material, the filament comes in a form of 1.75 mm diameter. Kevlar and fiberglass come in diameters of 0.3 mm, while the carbon fiber comes in a diameter of 0.35 mm.
894
Astrit Imeri et al. / Procedia Manufacturing 26 (2018) 892–899 Author name / Procedia Manufacturing 00 (2018) 000–000
To produce a part the digital file is needed. The CAD file is sliced into layers and is converted to STL file. The MKF slicer is Eiger, which is a web-based slicer. From the STL file in the slicer the numerical control code is generated, which commands the printer head what moves to make and when to extrude materials so the part is produced. A schematic of the printer is shown in Figure 3. It should be noted that in Figure 3, the two extrusion heads are in line and there is not only one as it might be percepted from the figure. Both base materials and reinforcement fibers are kept in rolled spools of filament. Due to the sensitivity of the materials to moisture they are kept in a water tight Pelican 1430 dry box for protection. The filaments are pulled into the extrusion heads with the help of the stepper motors.
Figure 3. Schematic of MKF printer extrusion process. The temperature of the extrusion heads are between 265-270oC. At this temperature, while passing through the extrusion head, the base material becomes molten, and begins to solidify once it leaves the nozzle. The first layers of the part are nylon. The fiber reinforcement materials which are not in a molten state are laid down horizontally layer by layer into the nylon matrix. The walls or outer borders of the part are also made up of nylon. Therefore, in every layer, the walls are printed first followed by the reinforcement layers if necessary. This process is concluded with nylon being printing on the opposite surface. Each part produced using FRAM AM follows this procedure layer by layer. In the specimens made out of carbon fiber and nylon there are 40 layers. Of those 40 layers, 32 are fiber reinforced, which is the maximum allowable number of layers to be reinforced. However, for the
3
nylon specimens reinforced with Kevlar or fiberglass there are 50 layers in total. Of those 50 layers, 42 are fiber reinforced, which again is the maximum allowable number of layers to be reinforced. Reason for this difference in the number of layers and maximum allowable number of layers to be reinforced is the different filament diameters for carbon fiber and Kevlar and fiberglass. This difference in the number of layers between carbon fiber and other fibers filled specimens results in a slight fiber volumetric fraction difference (up to 0.1 cm3). Due to this small variation in the fiber volume, the effect of fiber volumetric fraction was not analyzed in this research study. For nylon the used infill in this study was rectilinear, because it shows maximum strength in uniaxial loadings according to Fernandez-Vicente et al [12]. The infill density used is 75%, due to having similar strength values with 100% infill, but time-wise the 75% was more efficient to produce. The slicing software controls the design parameters and based on these parameters different specimens were generated. Changing these controllable settings, specimens with different number of rings, different infill type were generated. Specimens can be grouped into two categories, based on the type of the fiber reinforcement fill type. The two categories are: concentric and isotropic. The horizontal cross-section view of the different fill types are shown respectively in Figure 4a. The yellow lines represent the fiber reinforcement material, while the black portion represents the base material, Nylon.
Figure 4a. Left: Concentric fill. Right: Isotropic fill. The number of different types of generated specimens was six, for each material. To better visualize the horizontal cross-section view for a layer of these specimens is presented in Figure 4b.
4
Astrit Imeri et al. / Procedia Manufacturing 26 (2018) 892–899 Author name / Procedia Manufacturing 00 (2018) 000–000
895
3. Testing machine
Positive stop
Flat Wedge Grippers
Table 1: Abbreviation, naming and description Abbreviation # of rings and Description fiber fill type CF/FG/KV2 Rings Two concentric 2RC concentric rings of fiber in each fiber layer on the Nylon matrix CF/FG/KV3 Rings Three concentric 3RC concentric rings of fiber in each fiber layer on the Nylon matrix CF/FG/KV0 Rings isotropic Isotropic fill with 0RI 0 degree angle direction in each fiber layer on the Nylon matrix CF/FG/KV1 Rings isotropic Isotropic fill with 1RI one concentric ring on the outer border of the fiber fill CF/FG/KV2 Rings isotropic Isotropic fill with 2RI two concentric ring on the outer border of the fiber fill CF/FG/KV3 Rings isotropic Isotropic fill with 3RI three concentric rings on the outer border of the fiber fill
Extensometer
Having many specimens, naming them was a challenge too. Hence, the abbreviation, number of rings and infill type, and description are presented in Table 1.
Specimen
Figure 4b. All infill types and orientations generated.
The specimens were tested using a MTS 810 as seen in Figure 5. This model is a closed loop servohydraulic machine. The testing variables, such as, frequency, amplitude, and average load were all controlled and input through MTS’s Multipurpose Testware interface. This machine has two gripping heads. The lower head is actuated in order to load the specimen while the upper head is used only for gripping purposes. The gripping pressure applied to the larger portions of the test specimens, Figure 2, was 4 MPa. This value was decided upon because higher gripping pressure caused an immediate crack in the specimens while lower gripping pressure caused the specimen to slip.
Figure 5. Specimen clamped in jaws of the MTS810. Specimen alignment is important when testing for mechanical properties. Poor specimen alignment can be a significant contributor to premature failure. It can also affect repeatability and reproducibility. Therefore, wedge grips which incorporated positive stops were used to make sure that the specimen was aligned as shown in Figure 6.
Positive stop
Figure 6. Positive stops on the flat wedge grip.
Astrit Imeri et al. / Procedia Manufacturing 26 (2018) 892–899 Author name / Procedia Manufacturing 00 (2018) 000–000
896
The machines software interface provides a live data stream to the user for the force, displacement, and iteration which were the testing variables of interest. This assures the user that the machine is operating correctly. 4. Determination of failure The point of comparison between the different fiber reinforcement type and orientation was the number of cycles until total separation. The total separation was defined two different ways. The first was complete failure of the specimen into two different parts as seen in Figure 7. The second was when a visible crack had formed in either the Nylon or the inlaid fibers. For this study, the first way of total separation was used. In addition, since the goal of this study is to determine if a correlation exists between geometric orientations and fiber reinforcement material in relation to fatigue properties some of the specimens that did not break were truncated after 10000 cycles. 5. Testing Three specimens of each type presented in Table 1 were produced in order to have a better statistical representation. The type of load for the testing specimens was tensile-tensile load, with a load ratio R
Table 2: Results from the first load (3.33-0.33 kN) Load (kN) Material Rings
5
of 0.1. Different load ratios were not tested in this study but may be later. The Intermediate Value Property Theorem (IVP) was used to determine the magnitude of the loads used for testing. First, a high load of 17.4 kN and low load of 3.33 kN were chosen. The logic behind this selection is to eliminate the specimens that will fail immediately. Further, if one specimen failed during the first cycle, meaning that the load is excessive, the second and the third tests were eliminated. Then intermediate values were also selected and testing was repeated. The testing magnitudes for the maximum loads were:17.4 kN, 13.8 kN, 10.3 kN, 3.33 kN. Tables 2-5 provide a comprehensive summary of all the specimens and loads results. They are arranged from the lowest to the highest load. The results of testing at the first load (3.33-0.333 kN) are presented in Table 2. At this load the specimens with concentric infill were tabulated only. Isotropic infill specimens did not break at all at this load. From the results, the three rings concentric infill for any of the materials gave higher number of cycles. Sequently, the three rings isotropic specimens were then tested in the second load. The second load (10.3-1.03 kN) results are shown in Table 3, the concentric specimens with 3 rings, that showed higher number of cycles in the first load, failed in the first cycle.
Type
Test #1
Test #2
Test #3
3.33-0.33 3.33-0.33
Carbon Fiber Carbon Fiber
2 3
Concentric Concentric
510 10000+
222 -
417 -
3.33-0.33
Kevlar
2
Concentric
45
190
9
3.33-0.33
Kevlar
3
Concentric
383
72
181
3.33-0.33
Fiberglass
2
Concentric
168
132
137
3.33-0.33
Fiberglass
3
Concentric
644
561
644
Except the three rings concentric specimens failing in the first cycles, in the second load too, the isotropic with three concentric rings for any material failed in the first cycle. The Kevlar specimens with two rings isotropic infill, contrary to fiberglass and carbon fiber specimens, failed in the first cycle. Kevlar and fiberglass isotropic with concentric one ring failed within a range of cycles under 10000 (the truncating number of cycles). However, carbon fiber specimen did not break in that range.
Author name / Procedia Manufacturing 00 (2018) 000–000
6
Astrit Imeri et al. / Procedia Manufacturing 26 (2018) 892–899
Table 3: Results from the second load (10.34-1.03 kN) Load (kN) Material Rings
897
Type
Test #1
Test #2
Test #3
10.34-1.03
Fiberglass
0
Isotropic
750
1161
989
10.34-1.03
Fiberglass
1
Iso+Conc
699
996
744
10.34-1.03
Fiberglass
2
Iso+Conc
19
26
31
10.34-1.03
Fiberglass
3
Iso+Conc
1
-
-
10.34-1.03
Carbon Fiber
0
Isotropic
10000+
-
-
10.34-1.03
Carbon Fiber
1
Iso+Conc
10000+
-
-
10.34-1.03
Carbon Fiber
2
Iso+Conc
3753
3995
1747
10.34-1.03
Carbon Fiber
3
Iso+Conc
1
1
10.34-1.03
Kevlar
0
Isotropic
10000+
-
-
10.34-1.03
Kevlar
1
Iso+Conc
54
285
196
10.34-1.03
Kevlar
2
Iso+Conc
1
-
-
10.34-1.03
Kevlar
3
Iso+Conc
1
-
-
10.34-1.03
Carbon Fiber
3
Concentric
1
-
-
10.34-1.03
Fiberglass
3
Concentric
1
-
-
After truncating the specimens from the previous load, in the 13.8-1.38 kN, the isotropic with two concentric rings specimens of carbon fiber and fiberglass failed in the first cycle. Kevlar specimens with one ring isotropic infill failed in the first cycle, while isotropic with zero rings had just a small number of cycles. The only specimens that had more than 10000 cycles were the carbon fiber with zero rings. Table 4: Results from the third load (13.8-1.38 kN) Load (kN) Material Rings
Results are presented in Table 4. For the highest load case, the 17.4-1.74 kN, testing all other specimens failed in the first cycle other than carbon fiber specimens with no rings. However, even in these specimens a cracking sound was heard in the first cycle. Results from the final load are presented in Table 5
Type
Test #1
Test #2
Test #3
13.8-1.38
Fiberglass
0
Isotropic
5
66
20
13.8-1.38
Fiberglass
1
Iso+Conc
15
24
38
13.8-1.38
Fiberglass
2
Iso+Conc
1
-
-
13.8-1.38
Carbon Fiber
0
Isotropic
10000+
-
-
13.8-1.38
Carbon Fiber
1
Iso+Conc
324
275
209
13.8-1.38
Carbon Fiber
2
Iso+Conc
1
-
-
13.8-1.38
Kevlar
0
Isotropic
3
7
6
13.8-1.38
Kevlar
1
Iso+Conc
1
-
-
.
Astrit Imeri et al. / Procedia Manufacturing 26 (2018) 892–899
898
Author name / Procedia Manufacturing 00 (2018) 000–000
7
Table 5: Results from the fourth load (17.4-1.74 kN)
Load (kN)
Material
Rings
Type
Test #1
Test #2
Test #3
17.4-1.74
Fiberglass
0
Isotropic
1
-
-
17.4-1.74
Fiberglass
1
Iso+Conc
1
-
-
17.4-1.74
Carbon Fiber
0
Isotropic
19
4
12
17.4-1.74
Carbon Fiber
1
Iso+Conc
1
-
-
17.4-1.74
Carbon Fiber
2
Iso+Conc
1
-
-
17.4-1.74
Kevlar
0
Isotropic
1
-
-
6. Discussion Specimens of nylon reinforced with carbon fiber, fiberglass or Kevlar were printed with different infill types. These specimens were tested at four different loads with a load ratio (R) of 0.1. The first type of the infill was concentric infill with two and three rings. For specimens with concentric infill increasing the rings from two to three proved to have an impact on the number of cycles to failure. This result is independent of the type of fiber material used for infill. However, the concentric specimens failed in the first cycle for the higher loads. Hence, it can be concluded that for low loads, such as the 3.33-0.33, concentric type infill works well. Additionally, to make a part more resistant to failure due to fatigue the number of rings can be increased up to the geometric limit for both isotropic and isotropic with concentric rings. It should be noted that for Kevlar filled specimens with two concentric rings in Table 2, there seems to be unusual variation of the number of cycles up to failure. Since for each type of the specimen, there were three experiments done and from the other experiments there was not such variations, this issue could have happened due to printer anomalies while printing the specimens. Parts are sensitive to any irregularity while printing. Similar situation is observed with one specimen of carbon fiber filled isotropic with zero rings. The three ringed isotropic with concentric infill failed for the second load (10.34-1.03 kN) for all fiber materials. However, at this load, the two ring infill of carbon fiber and fiberglass had finite number of cycles up to failure, which give an intuition that decreasing the number of cycles increased the resistance to breaking. For the second load case the type of material used for the infill was also important. The Kevlar specimen, contrary to carbon fiber and fiberglass, failed immediately unlike the first load case which was material independent.
In the third load case (13.8-1.38 kN) the specimens with two concentric rings with isotropic infill were truncated because they failed in the first cycle. The one ring specimens did not fail during the first cycle; therefore showing better fatigue resistance than the two ring isotropic with concentric filled specimens. This indicates that decreasing the number of rings increases the fatigue life of the specimens. Based on the results of the experiments the nylon parts failed first followed by the the fiber reinforced specimens. However, for different fibers different patterns of failing modes were observed. Fiberglass and Kevlar with nylon specimens would deflect more than carbon fiber with nylon ones. However, for the same loads the shape of the broken specimens was similar. Failed specimens of different materials at 10.34-1.03 kN are shown in Figure 7.
Kevlar filled with two Fiberglass filled with rings two rings Carbon Fiber filled with two rings
Figure 7. Broken specimens at 10.34-1.03 kN
8
Astrit Imeri et al. / Procedia Manufacturing 26 (2018) 892–899 Author name / Procedia Manufacturing 00 (2018) 000–000
7. Conclusion Therefore, in conclusion the experiments showed that there is a correlation between filler material type and the pattern the reinforcing material is extruded. Among the materials, tested carbon fiber was shown to have better fatigue resistance than the other reinforcing materials. Based on the trend of number of cycles, for concentric infill, adding more rings improves the fatigue life of the specimen. While for isotropic infill, increasing the number of rings, it weakens the fatigue performance. The more rings the less the number of cycles up to failure. In addition, this study revealed that the isotropic with zero rings and one ring isotropic with concentric infill types showed better results than other specimens, with zero rings slightly topping the one ring isotropic with concentric infill specimens. Acknowledgements This work is part of a larger project funded by the Advanced Technological Education Program of the National Science Foundation, DUE #1601587. The funding provided by the National Science Foundation is greatly appreciated. References [1] Swolfs, Y. and Pinho, S.T., 2016, September. Designing and 3D printing continuous fibrereinforced composites with a high fracture toughness. In Proceedings of the 31st Technical Conference of the American Society for Composites (pp. 1-8). DESTech Publications. [2] S. Rawal, J. Brantley, N. Karabudak, Additive manufacturing of Ti–6Al–4V alloy components for spacecraft applications. In Recent Advances in Space Technologies (RAST), 2013 6th International Conference on, 2013. [3] R. Huang, et al., Energy and emissions saving potential of additive manufacturing: the case of lightweight aircraft components, J. Clean. Prod. (0) (2015). [4] Quan, Z., A. Wu, M. Keefe, X. Qin, J. Yu, J. Suhr, et al. 2015. "Additive manufacturing of multidirectional preforms for composites: opportunities and challenges," Mater. Today, 18(9):503-512. [5] Markforged [Online] https://markforged.com/about/ [Accessed on November 2, 2017].
899
[6] ASTM E1823-13, Standard Terminology Relating to Fatigue and Fracture Testing, ASTM International, West Conshohocken, PA, 2013, p.1034 www.astm.org” [7] Stephens et al., 2001. Metal fatigue in engineering 2nd ed., New York: Wiley. [8] Fischer, M. & Schöppner, V. JOM (2017) 69: 563. https://doi.org/10.1007/s11837-016-2197-2 [9] Zhang, H., Cai, L., Golub, M., Zhang, Y., Yang, X., Schlarman, K. and Zhang, J., 2018. Tensile, creep, and fatigue behaviors of 3D-printed acrylonitrile butadiene styrene. Journal of Materials Engineering and Performance, 27(1), pp.57-62. [10] Markforged [Online] https://static.markforged.com/markforged_composite s_datasheet.pdf [Accessed on January 31, 2018] [11] ASTM E606 / E606M-12, Standard Test Method for Strain-Controlled Fatigue Testing, ASTM International, West Conshohocken, PA, 2012, www.astm.org [12] Fernandez-Vicente, M. et al., (2016). Effect of Infill Parameters on Tensile Mechanical Behavior in Desktop 3D Printing. 3D Printing and Additive Manufacturing, 3(3), pp.183–192.