Glass Fiber Reinforced Acrylonitrile Butadiene Styrene Composite Gears by FDM 3D Printing

3D printing of gears via fused deposition modeling (FDM) has been recently introduced as a low‐cost efficient manufacturing method. Different materials have been 3D printed and Acrylonitrile Butadiene Styrene (ABS) with excellent mechanical properties has been found to be promising. However, 3D printed ABS gears possess a high level of abrasion rate. This paper introduces a new class of ABS‐based gears reinforced by different amounts of milled E‐glass fibers and 3D printed by FDM with acceptable thermo‐mechanical properties and performance. A set of thermo‐mechanical tests is carried out to provide an insight into the influence of adding glass fibers on the glass transition temperature (Tg), hardness and teeth bending strength, teeth failure force, weight lost, abrasion resistance, mechanical wear, and performance of composite gears. The mechanical behaviors of driving and driven gears are examined in high and room temperatures with or without lubrication. Microstructure and gear profile analysis of 3D printed layers, worn surfaces, and fracture locations are also conducted by SEM images and profile projector. The newly developed glass fiber reinforced ABS gears reveal a high level of thermo‐mechanical performance in terms of hardness, mechanical strength, bending force, abrasion and wear resistance compared to pure 3D printed ABS gears.


Introduction
Gears are mechanical parts, manufactured in various shapes and sizes, for various purposes such as transporting or changing the DOI: 10.1002/admi.202300337direction of force between axles.Throughout years of refinement, gears have drastically changed in material, shape, and size; yet they remain among the most used mechanical parts across various industries such as automotive, aerospace, marine, instruments, and home appliance manufacturing.Various gear shapes include simple, spiral, helical, bevel, and comb and are made of different materials such as metal alloys, steel, or plastics. [1]Despite their high strength, metal alloys suffer from issues like maintenance and repair cost, high wear and corrosion, acoustic pollution, and lack of self-lubrication.These issues have driven the industries toward polymer gears as an alternative to metal gears, [2] Some benefits of polymer gears that have led to their successful adoption by various industries include the ability to absorb and reduce noise, ease of manufacturing, low density, and thus low weight.On the other hand, they exhibit a lower strength compared to steel gears under high torque loads during long work duration that leads to more tooth wear and failure. [3]   3D printing or additive manufacturing (AM) is one of the most modern manufacturing methods, that works by adding a material layer by layer until the design product gets its final shape.3D printing technology produces parts with low weights and complex shapes that are exceptionally difficult to be fabricated by other methods. [4]3D printed electronics is an emerging field of high importance, enabling fabrication of devices with embedded or conformal electronic circuits.Binuclear copper complex can be directly inkjet printed onto 3D plastic objects, resulting in resistivity as low as 2.38 cm, 72% conductivity of bulk copper. [5]Hybrid materials with improved functionalities are used for 3D printing of various devices. [6]Currently, Fused Deposition Modeling (FDM) is one of the cheapest and most common AM methods, in which a string spool of thermoplastic filaments is used as base material.In the FDM method, a filament made of a polymer thermoplastic material is passed through a nozzle and warmed up.3D printing method can also be used for producing polymer-based composites, fiber reinforced composites, and nanocomposites. [7]10] For example, the sustainability of low-cost 3D printed parts requires consideration of energy consumption, quality metrics, time, quality, and flexibility.Careful choices must be made during the pre-production phase to ensure viability and sustainability. [8]Moreover, according to another study, [9] 3D printing has been used in dental implants since 2000.Surface roughness, printing time, shape quality, and surface hardness were reported for the optimized implant.In another investigation, [10] FDM-related parameters are important for 3D-printed components, and researchers investigated the effect of these parameters on the flexural strength of PET-G.GWO (Grey Wolf algorithm) was used to suggest a good combination of parameter settings to maintain good flexural strength with a gain of 15%.GWO was used to suggest a good combination of parameter settings to maintain good flexural strength with an enhancement of 15%.
Acrylonitrile butadiene styrene (ABS) filaments are among the most prevalent industrial thermoplastics in FDM 3D printing method.They are relatively strong, and resistant to wear, heat, and shock. [11]Glass fibers are made up of thin cilium oxide filaments and used directly in various industries as a reinforcing material for other materials like polymers. [12]Petrove et al., [13] produced polymer gears with polyether ether ketone (PEEK) base material covered with various materials like Boron Nitride (BN), molybdenum sulfide (Mo2SO4), Polytetrafluoroethylene, and graphene; and performed corrosion tests.Results showed that a decrease in temperature reduces corrosion rate, and corrosion resistance increases for the sample coated with polytetrafluoroethylene materials.Zhang et al., [14] used materials like Taulman material, carbon fiber alloy nylon such as 618 Nylon, 645 Nylon645, and alloy 910 spec to produce gears using FDM 3D printing in order to analyze the effect of crystallization and layered formation method on corrosion rate.Test results showed that corrosion mostly begins on the pitch line of gears.Gears produced using Nylon618 experience tooth surface corrosion, but no material peel-off, whereas corrosion causes some material to peel off from the surface of gears produced with other materials.Mao et al., [15] analyzed and compared polymer gears made of pure polyoxymethylene and reinforced with 28% weight of glass fiber    produced via injection.Glass fibers caused a 50% increase in fracture force and a 20% reduction in the corrosion rate.Moreover, an increase in rotation speed causes increased corrosion rate in both gears.During testing, due to the applied force, the surface and length of the glass fibers reduced.Singah et al., [16] produced polyethylene gears, reinforced with (0%, 15%, and 30%) glass fibers with the aim of reducing acoustic gear noise.Results showed that any increase in rotation speed, causes an exponential raise in noise, and pure polyethylene creates much more noise compared to the reinforced gears.Chemzou et al., [17] produced different filament gears using 3D printing by changing the procedural parameters and studied their effects on mechanical prop-  erties.Results showed that lower nozzle velocity allows higher plasticity, surface quality, and gear density.Moreover, in similar research, Zhang et al., [18] tweaked four 3D printed production parameters, namely temperature, printing speed, support temperature, and filling ratio, in order to produce a lightweight gear with an acceptable hardness.They used neural network and genetic algorithm methods to produce the gear.Results showed that the two factors of filling ratio and support temperature have the least effect on the gear's hardness, and the best outcome was produced when process parameters, temperature and print speed were 250 °C and 70 mm s −1 , while the support temperature was 25 °C, and dispersion rate was 80%.Jain et al., [19] investigated fracture and fatigue behavior of Nylon and Stall gears and found that most defects in Stall gears are due to fatigue, while most defects in Nylon gears are due to fracture under the pitch line/circle close to the root line.They illustrated that a bending stress up to 15 MPa on a tooth has no effect on the gear's performance.However, a further increase of up to 20 MPa led to a drastic plastic transformation.They also proved that the effect of torque and force is seven to eight times more than the effect of rotation speed on the surface of the tooth.Based on the published results of Harsha et al., [20] gears made of Nylon produced through the FDM method have the least rate of abrasion, and ABS results in the highest gear abrasion.Zhang et al., [18] investigated the effects of gear FDM parameters like nozzle temperature, printing speed, platform temperature, and filling ratio, and optimized their values using genetic and machine learning algorithms to achieve the best possible gear performance inside the gearbox.Heat and abrasion behaviors of the printed PA12 gears were analyzed by Kalani et al. [21] They showed that the abrasion rate in high rotational velocities, is higher than that of low rotational velocities.Using polymeric base gears can reduce the weight of the system and the production cost.Producing polymeric gears by 3D printer  eliminates the need to fabricate molds with various dimension and tooth numbers and modulus promoting sustainability.Although using polymeric gears may lead to less strength and different temperature durability, but adding reinforcement materials such as glass fibers could increase the yield strength and wear resistance of the gear tooth.Table 1 shows a comparative analysis of gears manufactured by FDM method based on the literature review.
ABS copolymer is composed of polybutadiene groups dispersed in styrene and acrylonitrile parts, resulting in toughness, dimensional stability, and chemical resistance. [26,27]Rheological or melt flow parameters of ABS and its composites play a key role in additive manufacturing and affects 3D-printed products performance.Melt flow index (MFI) rates of neat ABS and EG (expandable graphite)-containing composites were evaluated by the MFI test, yielding lower MFI values of EG (expandable graphite)-containing composites. [28]A reduction in MFI was observed as the amount of EG increased.EG addition at the lowest amount (5%) resulted in a slight reduction of melt flow behavior of ABS, suggesting processing conditions for ABS may not be affected during fabrication.EG plates dispersed uniformly into the ABS matrix due to acid treatment, resulting in improved tensile strength and storage modulus. [28]n other investigations, [29] the MFI values of ABS and composites were found to be in a narrow range; hence, CF (carbon fiber) additions had no effect on the processing conditions of ABS.Resized and PU-coated short CF were compounded with ABS matrix.CF additions caused no meaningful difference in the MFI value of ABS, which means that compounding CF with ABS can be done with no obvious problems in the case of composite processing.DS-CF containing composites showed lower MFI values, but higher CF loadings extended the melt flow behavior of the polymer phase, [29] One of the key pieces of information that can be obtained from (Differential scanning calorimetry) DSC analysis is the material's glass transition temperature (Tg).Tg is the temperature at which the material transitions from a rigid, glassy state to a more fluid, rubbery state.This transition is characterized by a change in the material's heat capacity, which is reflected in the DSC data as a sharp peak or step.More information can be found in references [30,31] .In previous research studies, Ozkoc et al., [32] studied the properties of 30 wt.% short glass fiber (SGF) reinforced ABS terpolymer and polyamide 6 (PA6) blends by using the interfacial adhesion approach.The results showed that, APS is the best coupling agent for glass fibers due to its chemical compatibility with PA6.Also, Arsad et al., [33] improved the in- The main objective of this paper is to 3D print gears with glass fiber reinforced ABS composites and to show their performance experimentally, for the first time.Glass fibers with various weight ratios (0%, 5%, 10%, and 15%) are added to ABS and filament rolls are produced via an extrusion procedure.Glass fiber reinforced ABS gears are fabricated by FDM 3D printing.Hardness and compression strength tests are then carried out on the composite gears, in order to obtain their mechanical properties.Thermo-mechanical testing of the 3D printed gears is performed inside a specialized gearbox designed and built for this test.Mechanical behaviors are analyzed in high and environment temperatures under various lubrication states (with or without lubricant), for both driving and driven gears.The newly developed composite gears show a high level of performance detailed in the results and discussion section.Due to the absence of similar concept and results in the specialized literature, this paper is likely to fill a gap in the state-of-the-art gears 3D printing and to be instrumental for designing and 3D printing composite greats with a high performance.

Parent Material
The material of the filaments used in this paper was ABS150 produced in Tabriz petrochemical company with density of 1.07 g cc −1 and Melt Flow Index of 230/3.8C kg −1 .The reinforcing glass fibers were E-glass with 8 standard filaments pro-  duced by Qinhuangdao Guangyu company.To produce filaments reinforced with glass fiber, ABS granules were first dehumidified and dried in 80 °C to prevent any transformation in the filament.Moreover, to achieve a better combination and performance, E glass fibers were crushed by planetary mill device in 300 rpm and room temperature to reach a 25 micron mesh as shown in Figure 1.The chopped glass fibers are dried at 100 °C for 2 h.The ABS granule was mixed with crushed glass fibers in various weight percent (0%, 5%, 10%, and 15%) for 1 h in the room temperature at 150 rpm.Then the mixture was uniformly combined inside a Mardon, melted, extruded under pressure, and pushed out of the nozzle at a certain diameter.The used extruder was single screw with rotating speed of 20 rpm so that the material is extruded in the 200 °C temperature.The filament reinforced with glass fiber was then passed through a hot and cold-water tank, left in the environment to reach temperature balance, and wrapped into a roll.Figure 2 shows the filament roll reinforced with 5% and 10% of glass fiber.The more hardness of the gear lead to an increase of wear resistance, therefore, more accuracy in power transmission will be achieved.One of the aims of adding glass fibers to filament was to increase the hardness of the material and gears' teeth.
DSC was a commonly used technique to investigate the thermal behavior of materials.DSC analysis measured the heat flow that occurs in a sample as it is heated or cooled and provides valuable information about the physical and chemical properties of the material.Therefore, to determine the proper temperature for the nozzle while printing the samples, a thermal analysis test was carried out on 4 types of produced filaments.This test was carried out by Mettler Toledo machine for temperatures up to a maximum of 280 °C, at a rate of 5 °C min −1 in the air atmosphere.Increasing nozzle temperature according to the thermal properties of the printed samples at the time of printing can cause an increase in the mechanical strength and cohesion of the part layers. [34,35]

Production of Gear and Gearbox
To carry out gear performance tests, a gearbox was designed according to Figure 3, with the ability to rotate at various speeds and torques.With the existing gear in mind, a simple gear with the size specifications in Figure 4 was selected for design, modeling, and manufacturing through 3D printing, with 18 teeth, 16 internal teeth, and a pressure angle of 20°of an involute gear.
The design file was imported to the slicing software and used to 3D print by an FDM machine with 0.4 mm nozzle diameter.Process parameters for producing gears include velocity equal to 25 mm s −1 and nozzle temperature obtained from test DSC results.For production gears, a laboratory printer was used and while production, the temperature of the gear printing environment was ≈30 °C.Nozzle printing parameter was 230C according to the DSC analysis.The thickness of the print layer was set 0.12 and 0.4 mm in the vertical direction and transverse distance, respectively.The infill density was 100% and the alignment of the layers was without angle.Figure 5 shows the execution process by the 3D printer, and one of the 3D-printed samples (from multiple sides).Once the printing was done, the garble and temporary bases were cut from the samples and polished using abrasion file and sandpaper.One of the main problems of the fiber-reinforced filament printing was the possibility of clogging of the printer nozzle due to the adhesion of fibers in its opening.It resulted in extra effort/time for cleaning or replacing the nozzle.However, printing with a higher nozzle temperature could reduce this problem.

Hardness Test
Durometer or Shore durometer hardness test was carried out to assess the resistance of 3D-printed gears to plastic transformation, in accordance with the ASTM D2240-15 standard by durometer Test Stand brand and the model of Teclock GS-612. [36]To create uniform conditions for all gears, they were kept in standard conditions for 24 h inside laboratory environment, at 23 °C temperature, and 53% moisture ratio.After the gears were adapted to the environment, the test was performed using a conical intrusive needle with a spherical tip with a reproducibility of five repetitions on different teeth.The more hardness of the gear leads to an increase of wear resistance, and a higher accuracy in power transmission will be achieved.One of the aims of adding glass fibers to filament to increase the hardness of the material and gears' teeth.

Teeth Compressive Test
To investigate the bending strength of gear teeth, and their fracture form under bending forces, an experimental setup was designed.According to Figure 6 empirical test fixture jaws exert a bending force on the two teeth, and prevent the gear's rotation.In this experiment, gear teeth undergo a compressive force, which was delivered continuously and incrementally, until the teeth fracture.To carry out this experiment, and ensure an accurate force exertion in the proper direction on a gear tooth, the gear was placed on specialized jaws as shown in Figure 6a.Upper and lower jaws are presented in Figure 6b and their assembly on the pressure test device is also visible in part (c).
After preparation and assembly of the produced jaws on the compression test device, each of the samples were separately placed on it, and the pressure test was carried out by exerting

Performance Test
In this section, 3D-printed gears were installed on the existing shaft inside the gearbox, once as a driven, and another time as a driving gear.A performance test was carried out at 1900 rpm with a variable torque, the value for which was controlled by 1.7 kg weight placed on the output shaft.The test was carried out in dry and lubricated conditions, as well as high, environment, and 100 °C temperatures.The shape of the gear installed on the shaft, and the location of the shaft inside the gearbox is shown in Figure 8.The performance test was chosen according to the gears' conditions in the gearbox like power transmission and rotation, various temperature and lubrication conditions.The silk diagram for bet-  ter understanding is shown in Figure 9.For conclusion a gear was printed for the hardness test and two gears were printed with the aim of doing the teeth bending test.For each performance test also a gear was printed, it means a gear for test that a gear was installed as a driver and three gear were printed for tests that gears were installed as a driven and various condition, with lubricant at room and high temperature, two tests, and without lubricant, one test.

Behavior Analysis
To evaluate the gear's microstructure, SEM images are taken on G5 sample, at a 100 micron scale, with a 90× and 300× magnification, as depicted in Figure 10.In the less magnified figure, printed layers are visibly cohesive to an acceptable extent.It means that, the selected distance between layers, which is 0.4 mm, is appropriate and there are not any holes or lamination between layers.Also it can be concluded that the chosen temperature for printing was appropriate due to the good adhesion between layers.The more continuity and less holes lead to more strength and hardness and at last better performance in the gearbox.Further magnification reveals the milled glass fibers residing inside the base filament material and shows good distribution and bonding between fibers and base material.DSC diagrams show that the filament reinforced by glass fibers absorbs more thermal energy.The possible cause of this phenomenon is the increase in the interface between the surface of the fibers and the ABS matrix and the heat transfer between them.Thermal Analysis test is performed on four types of filaments with various fiber ratios, and their output DSC chart is obtained, as shown in Figure 11.Note that ABS has no specified melting temperature, due to its amorphous structure, and because ABS is a co-polymer, and the temperature at which it becomes glassy (glass transition temperature) can be determined empirically.Results indicate that an increase in the ratio of glass reinforcing fibers, leads to an increase in the aforementioned glass transition temperature.The increase of reinforcing glass fibers will limit the movement of polymer chains, and as a result, the glass transition temperature will increase, [37,38] The addition of glass fibers leaves a visible effect on the ABS polymer motion, and decreases the material fluidity.The slope of changes in the DCS curves of all four types of tested filaments is almost the same at temperatures lower than the glass transition temperature.This can lead to a reduction in the material heat transfer due to thermal conductivity.Therefore, with the addition of glass fibers, the material enters glassy phase at a higher temperature.Glassy phase is ≈120 °C for pure ABS and 100 °C for the sample reinforced with 15% glass fibers.According to the DSC chart, obtained from the tests, nozzle temperature for printing gears is considered 230 °C in all cases.

Hardness Test
The next test performed on the 3D-printed samples is the hardness test.Based on the aforementioned standard, the test is repetitively five times performed on the 3D-printed reinforced gears.The results are shown in Table 2 and Figure 12.
An increasing and decreasing trend in gear hardness is observed in the results with increments in the ratio of glass fibers.The lowest hardness is obtained for the G0 sample as 67.8 Shore D. An addition of 5% glass fibers increases hardness by 11%, such that the highest hardness obtained for the G5 sample is 75.8D.This can be attributed to be the proper presence and distribution of glass fibers.Moreover, glass fibers have properly bonded with the base filament material, causing a diversion in the exerted force, preventing penetration, and further resisting plastic transformation, leading to an increment in hardness.The experiment continues by increasing glass fiber ratio in the samples to 10 and 15 in the G10 and G15 samples.As a result, their hardness reduces and the results are obtained for their hardness as 73.2 shore D and 69.8 shore D. It could be due to the higher presence of glass fibers inside the base filament material, as it prevents proper distribution and causes coagulation of the fibers, leaving cavities between the fibers and the material.This will reduce the resistance to plastic transformations and thus the obtained hardness.G15 sample is more porous than this, and its hardness is thus 4.8% less than G10 sample.

Teeth Bending Test
Bending test on the gear teeth is performed to obtain the teeth force and fracture stress.After installing the jaws in the device and placing the 3D-printed gears between the jaws, force exertion is started at a speed of 1 mm min −1 and continued to increase until the fracture of teeth of 3D-printed gear.The test was done for twice and the average of the obtained results for the teeth bending test are presented in Table 3 and Figure 13.
The results show that G0 sample has the least force tolerance at about 82.71 kg, and thus the least strain at about 5.1% is observed for this sample.On the other hand, the highest force and fracture stress are 148.7 kg and 2.69 kg mm −2 for G5 sample, respectively.The result signifies 80% performance improvement and growth in terms of teeth strength after the addition of 5% glass fibers to the filament material.Moreover, the highest strain for G5 sample is 7.2%, which is 41% more than the strain observed in the G0 sample.This can be attributed to improper distribution, fibers aggregation, and porosity between the fibers and the filament base material.Figure 14 illustrates the fracture pattern for great teeth with various fiber ratios after the teeth compressive test in which the fracture location is shown using red arrows.In all samples, the fracture occurs at the root of the teeth, under stress concentration and the highest bending stress.Growth of the crack is in the peripheral direction and considering the incomplete sep-aration of the tooth from the body, this is a soft fracture.SEM images from the root fracture are illustrated in Figure 15 where the fracture location for the samples reinforced with glass fibers, the holes containing glass fibers and their dislocation after the fracture are displayed.It can be seen that the fracture behavior of  the samples with more strain is near to ductile fracture and on the contrary, the less strain led to brittle fracture in the samples.As it mentioned in the Figure 15 in G5 with highest strain, more and finer cracks have emerged and the flaking will be decreased which indicates a transformation from the ductile fracture to brittle fracture in G0 with least strain.The lowest strain is obtained in G0 sample because there are no reinforcing fibers inside it, and the filaments inside it have more and larger tears as seen in Figure 15a.The force-deflection graph obtained from the teeth compressive test is presented in Figure 16.

Performance Testing of 3D-Printed Gears
Performance tests of the manufactured gears are carried out inside the gearbox in various conditions.The objective is to calculate the amount of wear and weight loss for the gears after an hour of rotation inside the gearbox.As the weight of these gears is separately measured for each test by means of their production method, the results are also presented separately in each section.

Driver, Without Lubricant and Room Temperature
During the first test, the 3D-printed gears are installed on the moving shaft, in contact with the steel gears.The test is carried out with 1900 rpm rotational velocity with 1.7 kg weight as torque on the output shaft.The test continues for an hour, at room temperature, and the results are presented in Table 4 and Figure 17.
Installation of the gears in the driving role raises the probability of receiving more force.The results for weight reduction shows the highest reduction for G0 sample, by 6.2%, which is 61% more than G5 sample in which glass fibers are being used as reinforcement.It demonstrates the effectiveness of glass fiber presence on strength and wear resistance enhancement, during engagement with steel gears.Among gears reinforced with glass fibers, the highest rate of wear is found for G15, by 3.9%, which is 7.6% more than the abrasion for G10.Considering the weight loss, G15 sample has the lowest weight among the reinforced gears, due to the presence of more holes and pores between the fibers and matrix material, which will lead to reduced strength and increased abrasion.Comparison between the weight of reinforced gears shows 7.7% weight loss, as a result of 10% increase in glass fibers between G5 and G15 samples.To evaluate the extent of abrasion on the surface and the teeth, one must note that the worn surface in one area of G5 is in contact with a different gear.In Figure 18, the gear installed as driver, and indicated with a red arrow, can be seen after many rotations, with visible abrasion effects on the steel gear.Considering the rotation and abrasion pattern of gears inside the gearbox, pictures and SEM images of the G5 gear's worn and intact areas are presented in Figure 19.The damage to the printed layers due to engagement with the metal gear is clearly visible.

Driven, Without Lubricant and Room Temperature
The 3D-printed samples are separately installed on the outer shaft, as the driven, and the test is carried out at room temperature without lubrication and at 1900 rpm and 1.7 kg weight is used as an external torque.Figure 20 demonstrates a gear installed on the shaft and rotation, and test results are presented in Table 5.
By installing the gears in the driven role, their weight loss rate decreases compared to driver gears.The obtained results show that G0 sample has a 3.4% weight loss.This amount of reduction is more than that of G15 sample, in which weight after the test is reduced by 0.2 g.The least abrasion is observed on G5 sample, at ≈0.7% equivalent to 0.06 g weight loss in the gear.Improvement in the abrasion behavior of other gears reinforced with glass fibers can be attributed to the effectiveness of glass fibers as reinforcement against force and plastic transformation of the gears during performance tests.SEM images captured from the worn surfaces are depicted in Figure 21.Abrasion rate is lower for G5 gear, due to the presence of glass fibers inside the base material, and its proper bond with the base material, as well as its higher hardness compared to G0 sample.The image captured from G0 sample shows mechanical abrasion and gear tooth head surface layer peel-off, while the image depicting G5 sample indicates the side-surface of the tooth, where mechanical abrasion of the printed layer is wiped out.

Driven, with Lubricant and at Room Temperature
The next test is similar to the previous one,except gears performed a driven role, under similar conditions, that is, at a rotational velocity of 1900 rpm, under a torque generating weight of 1.7 kg, and at room temperature.The difference is in the simultaneous application of oil on the engagement area with a viscosity of 95 cs on the rotating gears, using a pump residing outside the gearbox, [15] The oil pumping cycle is kept uninterrupted, and its performance technique is presented in Figure 22.
Results for the test are also presented in Table 6.It must be noted that, considering the possibility of oil absorption by the   filament and glass fibers, their weight distribution is performed after a while, to ensure all the oil has left the place, and it has completely dried.As the gears are installed in the driven role and the lubricant is applied simultaneously to reduce the direct contact surface, the rate of weight loss for 3D-printed gears drops, compared to that of the previous test.The lowest abrasion rate for G5 sample was 0.5% which is 68% less than that of the unreinforced gear, G0 sample.Among the glass fiber reinforced gears, the highest abrasion rate is found for G15 sample, which is 22% less than the weight loss rate of the unreinforced gear, G0.Comparison between the two experimental situations with and without a lubricant shows that the highest lubrication effect on the abrasion rate is for the situation where glass fibers have not been added to the filament.Weight loss rate for G0 sample in the unlubricated condition is 3.4% which could reach 1.6% with lubrication, an almost halfway drop.As for G15 sample, the values reach to 2.2% from 2.6%.This signifies the negative effect of glass fibers on lubrication and avoiding the creation of a boundary layer between the engaged surface of gears, as the glass fibers reduce the base ABS material's ability to keep oil.

Driven, With Lubricant and High Temperature
As stated, inside the gearbox, under the shafts, there is an industrial element installed which can be used to keep the lubricant oil temperature up, and constant.In the present test, 3D-printed gears performing a driven role are installed inside the gearbox, and lubrication is performed at the same time.But the lubricant temperature is kept in the ≈50-60 °C range.The gear rotates at a 1900 rpm velocity under a 1.7 kg weight.Test procedure and lubricant temperature are demonstrated in Figure 23.Test results are presented in Table 7.Note that gear weight measurement is   performed during time and after the oil has fully departed and dried.
Results indicate that the lowest weight loss for the G5 sample was 0.2%, which is 81% and 71% less than weight loss in G0 and G15 samples, respectively.As the gears are lubricated overall, their engagement surface, and loss rate lower.Compared to the previous test, and considering the rise in temperature, weight reduction is lower, which can be attributed to phase changes inside the filament during temperature rise, strength increments, and improved performance of glass fibers during engagement with the filament. [39,40]Temperature increase causes the base material to soften, increasing plasticity.Thus, during engagement among the teeth, less cracks appear on the engaged surfaces and less material is separated.On the other hand, gear geometry will transform, and teeth dimension accuracy drops faster.Samples G0 and G5 teeth undergo performance test in high-temperature lubrication and their transformation are compared in SEM photos of Figure 24.As can be seen, teeth profile experiences higher transformation at high temperatures.In sample G0, glass fiber protrusion and their extreme plastic transformation is visible against plastic transformation of G5 sample tooth, the side surface of which has been destroyed due to the softening of printed layers.
Figure 25 displays the cross-section of a tooth from G15 gear, photographed using a profile projector.In Figure 25a, the crosssection of the initially printed tooth is shown before engagement inside the gearbox, where the body curvature is clearly visible.Figure 25 b shows the cross-section of the tooth after engagement, while dried.It has been deformed by mechanical erosion of the tooth surface and also lost some material.The tooth shown in Figure 25 c is displayed after engagement at high temperature in the lubrication situation.As a result of ABS softening at higher temperature (100 °C, according to DSC diagram, considering the reduction of the material's glass transition temperature) less mechanical abrasion has taken place.However, the tooth surface has turned flat, and the profile has transformed compared to the initial state.In fact, during temperature increments, destructive compression of the tooth is the most prevalent surface damage mechanism.It should be noted that the measurement of surface roughness of the gears before and after the performance test was not the aim of this research and it can be done for further research.The analysis of the gears' surface was investigated by SEM images and profile projector.
For a better comparison, the results of experiments where gears are tested for performance in driven, are presented in  and at room temperature.With 96% reduction, the highest decrease in weight in G5 sample is obtained at high-temperature lubrication.In weight loss trend slows down with lubrication and increase in temperature.Moreover, increasing the glass fiber ratio to %5, inside the base filament material, similarly slows down the weight loss, due to the abundance of glass fibers, their entanglement, and lack of proper dispersion and bonding between glass fibers and filaments, as well as the extra presence of porosity.

Conclusion
The main aim of this paper was to introduce glass fiber reinforced ABS composites for FDM 3D printing of involute gears for the first time.ABS was reinforced by milled E-glass fibers with various weight ratios (0%, 5%, 10%, and 15%) and then used to make composite filaments.Glass fiber reinforced ABS composite gears were fabricated by FDM 3D printing technology.A set of experimental tests were performed to analyze thermo-mechanical properties of the composite gears in terms of DSC thermal test, hardness and teeth bending strength tests.Performance tests were carried out by installing the gears inside a gearbox, rotating at 1900 rpm with 1.7 kg weight with and without lubrication at normal and high temperatures.The following main results can be concluded: 1) Adding glass fibers to ABS material increases Tg of the composite material.By increasing the percentage of the glass fiber, this temperature increases due to the bonding between glass fibers and ABS material.
2) The minimum hardness belongs to G0 and the highest hardness is obtained 75.8 Shore D for G5 due to the presence of glass fiber and bonding with ABS base material.By increasing the glass fiber percentage, the hardness decreases due to the agglomeration of glass fibers.3) A higher bending force is achieved for G5 equal to 147.8 kg which is 44% more than that ofG0 sample.That is because of the proper bonding between glass fibers in ABS base material which increases the strength of the teeth of the gear.4) The highest weight loss of the installed gear in the gearbox as a driver is achieved as 6.2% for G0 sample which is 61% more than that of G5 sample.The presence of glass fibers causes the strength and abrasive resistance to increase.By increasing the glass fibers, the weight loss increases due to the presence of more holes and agglomeration.5) The results reveal the highest weight loss for G0 sample that is 3.4% in conditions with no lubrication and room temperature.The least weight loss with 0.2% belongs to G5 sample in the high temperature and lubrication condition.6) Overall, the weight loss trend reduces as lubrication and temperature increases.Moreover, an increase of glass fiber ratio in the filament material to 5% turns the weight trend downward, and further increase in the glass filament ratio leads to less weight loss.This is due to the existence of more glass fibers, their entanglement, lack of proper distribution, a shortfall of proper bonding between the glass and the filament, and the presence of more holes.

Figure 2 .
Figure 2. Reinforced filament with a) 5% of glass fibers and with b) 10% of glass fibers.

Figure 3 .
Figure 3. a) 3D design of gearbox, b) produced gearbox attached to motor, c) produced gearbox.

Figure 4 .
Figure 4. a) Original gear, b) 3D design of involute gear, c) dimensional features of designed gear.

Figure 7 .
Figure 7. Installed gears, a) on the fixture jaws, b) on the device.

Figure 8 . 7 Figure 9 .
Figure 8. a) Gearbox shaft, b,c) installed 3D-printed gear on shaft, d) the view of gearbox and motor.

Figure 10 .
Figure 10.SEM images of G5 sample at two magnifications.

Figure 12 .
Figure 12.The average results of hardness testing of 3D-printed samples.

Figure 13 .
Figure 13.The average results of bending testing of 3D-printed samples.

Figure 15 .
Figure 15.SEM image of fracture cross-section of 3D-printed sample.

Figure 16 .
Figure 16.The force-deflection response of the compressive test of 3Dprinted sample.

Figure 17 .
Figure 17.The obtained results of weight loss of driver, without lubricant and at room temperature.

Figure 18 .
Figure 18.a) Installed 3D-printed gear, b) the filament remnant on the steel gear.

Figure 19 .
Figure 19.Images of teeth surface of G5 gear installed as a driver in the performance test.

Figure 20 .
Figure 20.a) Installed 3D-printed driven gear, b) the installed gear while rotating.

Figure 24 .
Figure 24.SEM image of worn surface of 3D-printed gear.

Figure 26 .
Figure25 displays the cross-section of a tooth from G15 gear, photographed using a profile projector.In Figure25a, the crosssection of the initially printed tooth is shown before engagement inside the gearbox, where the body curvature is clearly visible.Figure25 bshows the cross-section of the tooth after engagement, while dried.It has been deformed by mechanical erosion of the tooth surface and also lost some material.The tooth shown in Figure25c is displayed after engagement at high temperature in the lubrication situation.As a result of ABS softening at higher temperature (100 °C, according to DSC diagram, considering the reduction of the material's glass transition temperature) less mechanical abrasion has taken place.However, the tooth surface has turned flat, and the profile has transformed compared to the initial state.In fact, during temperature increments, destructive compression of the tooth is the most prevalent surface damage mechanism.It should be noted that the measurement of surface roughness of the gears before and after the performance test was not the aim of this research and it can be done for further research.The analysis of the gears' surface was investigated by SEM images and profile projector.For a better comparison, the results of experiments where gears are tested for performance in driven, are presented in Figure26.According to the diagram, the most severe weight loss occurs, by 6.2%, for G0 sample, in the absence of lubrication,

Figure 25 .
Figure 25.G15 sample in, a) primary situation, b) after test at room temperature, c) after test at room temperature and lubrication.

Figure 26 .
Figure 26.The comparison graph of the weight loss of the driven gear.

Table 1 .
The comparison between research studies.

Table 2 .
The hardness test results of 3D-printed gear.

Table 3 .
The teeth bending test results of 3D-printed composite gear.

Table 4 .
The results of performance test of driver, without lubricant and at room temperature.

Table 5 .
The results of performance test for the driven case, without lubricant and at room temperature.

Table 6 .
The results of performance test for the driven case, with lubricant and at room temperature.

Table 7 .
The results of performance test for the driven case, with lubricant and at high temperature.