Performance Analysis of Polymer Additive Manufactured Gear Bearings

Featured Application

Polymer bearings.

Abstract

Bearings in general, and in this case rotational bearings, are important elements in many machines. The main objective of this study was to find out the load-bearing capacity of 3D-printed gear bearings under various rotational speeds, test the bearings to failure and estimate their lifetime. An interesting note on the gear bearing is its uncommon geometric configuration, because the rolling elements are gears, which allows for a rolling-sliding motion between the constituting elements, minimizing the sliding effect. The material used is PLA (poly lactic acid), a common thermoplastic polyester, and the printing technology was FDM (fused deposition modeling). Considering the PLA’s temperature sensitivity, this was also monitored but had no influence on the failure of the bearing, as experiments show. The rotational speed range for the experiment is 250–1500 RPM (revolutions per minute) in increments of 250 RPM, and the loads are 18 N (Newton) and 45 N for a gear bearing with 51 mm diameter and 15 mm thickness. The results of this study can be used as a reference for application limitations or to design gear bearings using 3D printing methods.

1. Introduction

Over the past decade, additive manufacturing (AM) has gained increasing popularity within the engineering community. This manufacturing technique allows for the production of complex shapes, primarily employed as structural or design components in machinery. However, when it comes to bearing components, particularly kinematic joints, traditional metallic machined parts are still predominantly utilized.

The primary objective of this research paper is to investigate the viability of employing 3D-printed gear bearings under various conditions of radial loading values and rotational speeds for machinery applications. In the course of this study, we introduce a failure criterion and provide an estimation of the operational lifespan of such bearings within the specified loading conditions. It is important to note that this gear bearing is not assembled but rather created as a single, integral piece directly by the 3D printer.

This paper addresses a relatively unexplored area within machinery construction, involving the utilization of polymer additive manufacturing techniques for bearings. Traditionally, to make rotational or translational joints for 3D-printed elements, metallic bearings are used because of their endurance and low friction characteristics, and these are inserted into the 3D-printed elements. This is a good system if metallic bearings are available, but an alternative could be to use 3D-printed bearings in the situation in which access to classical bearings is difficult, or the application areas require non-metallic components, such as inside a magnetic resonance imaging device.

Gear bearings are a rather new concept, but their main function is to allow for rotating motion between two elements with low energy loss due to friction, similar to traditional plain and rolling element bearings. From the point of view of geometrical configuration, gear bearings are rolling element bearings in the sense that the gear elements, similar to balls or rollers, have rolling motion relative to the central gear or the outer geared cage, but their behavior is different when manufactured using the FDM process. Our experience with 3D-printed rollers or ball bearings is that they tend to get stuck and stop rolling due to manufacturing imprecisions, thus transforming the rolling motion into a sliding motion. This is the case also when a cage is present in the roller bearings. Gear bearings, on the other hand, are less sensitive to manufacturing imperfections, and the rolling motion is not affected. Also, they also do not need a cage for gear spacing, since the distance between gears is retained due to the teeth meshing. In this paper, for the manufacturing process, we used the common FDM process, and the resulting gears do not have a very smooth surface. Still, the performance at low rotational speeds and loads is promising.

The origins of gear bearings come from developments at NASA in the 2000s [1,2]. An example gear bearing is presented in Figure 1. It is similar to a planetary gear setup, where the gears have taken the place of the balls in a ball bearing or in a cylindrical bearing. The gears have a herringbone shape, and the whole bearing is printed in one piece together. This configuration does not allow the gears to slide past each other; thus, the gear bearing can take limited axial loads. Another feature is that it does not require a cage to keep the rollers spread apart, because the gears keep them spread apart. An early pioneer of this type of setup is Emmet Lalish, who in 2013 published a video on YouTube in which he presented a 3D-printed gear bearing. An extension of this gear bearing is that one can put two such devices side by side, connect the planet gears together and obtain a very high transmission ratio and a robust planetary gearbox, with applications in robotics [1,3], medicine [4], aeronautics or space applications [5].

There are a number of parameters that influence the mechanical properties of additive manufactured parts, especially additive manufactured gears. In [6], the authors investigate the effects of infill volume while maintaining infill type, shell thickness and manufacturing speed. Also, in this paper, the service life of gears was tested, and failure types were characterized. Another paper that explores the infill percentage effects on part strength is [7]. The authors also present and test three distinct types of infill patterns. It is to be noted that in the tests conducted in this paper, the standard 15% infill on the printed gear was used, as will be shown in the following sections. Papers addressing the behavior of polymer gears are also [8] on durability, ref. [9] on loading conditions, ref. [10] on thermal behavior, ref. [11] on elastic tooth deflection and [12] on wear.

The idea of using additive manufacturing for bearings is not new, and experiments with 3D-printed ball bearings were conducted by Lee et al. in [13], in which the friction performance of the ball bearing was evaluated to verify its feasibility as an engineering part. Compared to our paper, in which we use FDM, they used multi-jet printing (MJP) as a method to prepare the ball bearing specimen, as it provides the advantage of high surface quality. They designed a test rig and constructed experiments to evaluate the friction performance under various rotating speeds (500 rpm and 700 rpm) and radial load conditions (50 N, 70 N, 100 N), and their conclusion was that 3D-printed bearings can be used under a constant load and rotational speed.

Another type of bearing is the herringbone grooved journal bearing. These are well known for their reliability and high rotor dynamic stability thresholds. In [14], research is conducted to determine whether current technology limits plastic 3D-printed parts’ lubricant flow due to layer thickness, or whether 3D-printed parts can be used as an alternative in manufacturing the journal bearing. It is to note that this type of bearing uses hydrodynamic effects, and in comparison, the gear bearing does not use any lubrication, or at least this is the case in our experiments.

In the field of polymer printed bearings, the Igus company provides custom 3D-printed plain bearing solutions for rotational motion, Figure 2. They use their proprietarily developed high-performance polymers that, according to their tests, last up to 50 times longer than standard 3D printing materials inside moving applications. Another important mention is that these plain bearings are dry running and maintenance-free [15].

In the literature, there are also studies on tribological aspects of 3D-printed parts. The study in [16] shows the impact of layer thickness, infill angle, infill pattern and orientation of the deposition and indicates the correlation between printing parameters and wear.

Our experiments were designed to allow for the direct performance evaluation of the bearing mounted on the shaft, subject to different rotating speed and loads. The material used for the gear bearing manufacture is PLA, but an interesting and promising expansion of this paper could be the use of composite materials. An interesting study in this di-rection is presented in [17], in which a bronze-filled PLA-based composite filament is analyzed for use in FDM and shows improved mechanical characteristics, which in turn could increase the gear bearing’s life.

2. Materials and Methods

2.1. Gear Bearings: An Introduction—The Involute

The general definition of gears is that they are machine parts that have teeth, which mesh with one another and have the function of converting torque or speed. A bearing is defined as a machine part that allows for motion between two parts while reducing friction. For the case of gear bearings, the main function of the assembly is to allow for rotation between parts while reducing friction by means of meshing gears. The assembly configuration is presented in Figure 1, and it is similar to a planetary gear setup.

For the gears to be able to mesh, two conditions need to be satisfied simultaneously. The first condition is that the teeth of wheel one need to fit in the space between the teeth of wheel two. The second condition is that the transmission ratio needs to be constant during the meshing of the gears. To have a constant transmission ratio, the most popular gear tooth profile is generated using the involute curve. The involute curve can be obtained theoretically by tracing the path of a fixed point on a line, and the line is rolling (without sliding) on a circle as shown in Figure 3. The literature provides mathematical formulation, such as in [18], for the involute that can easily be integrated into CAD software, like CREO Parametric 5, to generate teeth. For our gear bearings, we used a module of 1.5 and a gear bearing outer diameter of 51 mm.

The design choice of using herringbone gears is justified, since this type of gear suppresses axial shifting and axial loads on gears when in motion. This situation is convenient, since there is no need for further machine parts to hold the assembly in place, and the gear bearing can be printed in one piece.

2.2. The Test Stand

The testing platform enables the assembly of a gear bearing and the evaluation of its performance under varying loads and speeds. The kinematic arrangement is depicted in Figure 4, featuring key components such as the electric motor (1), the motor shaft (2), fixed supports (3,5) and the gear bearing (4), with “F” representing the applied load.

The platform is based on the Gunt PT 500 experimental stand, with additional 3D-printed parts in order to allow for the experiments that were designed.

In the conventional machinery setup, the load is typically applied by the shaft to a bearing mounted within the housing. However, in our specific case, we have implemented a reverse configuration, depicted in Figure 4, in which the gear loading system applies a load to a bearing placed on a shaft that is supported by two steel ball bearings. In this arrangement, the gear bearing undergoes the influence of a radial force on the shaft, effectively yielding loading characteristics equivalent to the traditional setup. A similar methodology is discussed in [13].

In Figure 5, the force-applying mechanism is shown on the gear bearing. The gear bearing is inserted into the sliding part, and an external force is applied. The external force is created by masses of different values (1.8 kg and 4.5 kg), generating 18 N and 45 N (considering g = 10 m/s/s).

2.3. Printer Settings

To build the gear bearings (Figure 1) and the gear loading system (Figure 5), we used an Ender-5 Pro printer and 1.75 mm Everfil PLA-N.01 filament provided by a local supplier, 3D KORDO, and the printing was executed using standard quality settings at 0.2 mm. All the settings are presented in

Table 1. Settings for printing the gear bearing.

Quality	Layer height	0.2 mm
	Initial layer height	0.2 mm
	Line width	0.4 mm
Walls	Wall thickness	0.8 mm
	Wall line count	2
Top/bottom	Top/bottom thickness	0.8 mm
	Top/bottom layers	4
Infill	Infill density	15%
	Infill pattern	Grid
	Infill overlap percent	30%
	Infill overlap	0.12 mm
Material	Printing temp	210
	Build plate temp	60
	Flow	100%
Build plate adhesion		Skirt

.

As slicer software, we used Ultimaker Cura v5.0.0, and a screenshot is presented in Figure 6. One can observe the infill pattern for the gear bearing. The outer ring diameter is 51 mm, and the printing speed is 80 mm/s for the infill and 40 mm/s for walls.

3. Results

3.1. Lifetime

Using the above printer settings and test rig, experiments were conducted to determine the duration of the gear bearing to failure. The moment of failure is considered when the gear bearing stops functioning (rotating). A sequence of tests was conducted with increasing rotational speeds, starting at 250 rpm and ending with 1500 rpm in steps of 250 rpm. For each rotational speed, two types of loads were applied, 18 N and 45 N, and the functioning time was measured. Three specimens were used for every case. In

Table 2. Time of running for a load of 18 N.

	250 rpm	500 rpm	750 rpm	1000 rpm	1250 rpm	1500 rpm
Test 1 (h)	59	18	0.3	0.2	0.1	0.1
Test 2 (h)	58	16	0.1	0.1	0.1	0.1
Test 3 (h)	55	14	0.1	0.1	0.1	0.1
Average (h)	57.3	16	0.16	0.13	0.1	0.1

, results with an 18 N load are presented, and in

Table 3. Time of running for a load of 45 N.

	250 rpm	500 rpm	750 rpm	1000 rpm	1250 rpm	1500 rpm
Test 1 (h)	1.1	0.2	0.2	0.1	0.1	0.1
Test 2 (h)	1.1	0.2	0.1	0.1	0.1	0.1
Test 3 (h)	1	0.1	0.1	0.1	0.1	0.1
Average (h)	1.06	0.16	0.13	0.1	0.1	0.1

, results with a 45 N load are shown. Figure 7 graphically shows the decrease in lifetime in correlation with the increase in rotational speed.

3.2. Temperature

We found and have shown in Figure 8 that the functioning temperature of the gear bearing reaches a steady equilibrium state at around 37 degrees Celsius. This is far lower than the critical melting temperature. Temperature measurements were conducted using a FLIR B200 thermography device at an hourly sampling rate, and the room temperature was 20 degrees.

Relevant experiments for temperature analysis were considered only for 18 N-250 rpm and 18 N-500 rpm. For the other experiments, temperature measurements were conducted only at the time of failure, measuring the gear bearing’s final temperature, because of the short duration of the experiment. The observations were that the temperature rise was insignificant, with temperatures not rising above 40 degrees Celsius at a room temperature of 20 degrees Celsius.

3.3. Weight

The specimens were weighed, and their starting weight was 22 g. No significant weight was lost during the experiments, except for the situation with 18 N-250 rpm, where the final weight was 21 g, resulting in a loss of 1 g of material.

3.4. Radial Play

Radial play appears due to abrasive wear of gears. The most significant case for radial play is in the case of the sample that lasted for 57 h—18 N-250 rpm. In this case, the play evolution was measured using a comparator gauge, and the graphic evolution of the play is shown in Figure 9. It should be noted that measurements were taken every 5 h. The value for the radial play after 55 h was 1320 μm, and the final play after 57 h is not relevant, since the failure mode of the gear bearing is gears jamming, which in turn produces no play at all.

3.5. Failure Modes

The detected failure modes were of two types:

• Gear jamming (J).
• Gear failure as a part (F).

Gear jamming was observed in most scenarios, as shown in

Table 4. Failure modes for tested gear bearings.

18 N-250 rpm	18 N-500 rpm	18 N-750 rpm	18 N-1000 rpm	18 N-1250 rpm	18 N-1500 rpm
J J J	J J F	J J J	J J J	J J J	J J J
45N-250 rpm	45 N-500 rpm	45 N-750 rpm	45 N-1000 rpm	45 N-1250 rpm	45 N-1500 rpm
J J J	J J J	J J F	J J J	J J J	J J J

.

The first mode of failure of gear bearings is gear jamming This happens due to misalignments between the gears while rotating. The misalignments are caused due to imperfections in the shape of meshing gears and relatively high play between the gears. Since they are built built-in-one, this results in the gears skipping one tooth and jamming.

This case is presented in Figure 10a, where one can observe, indicated by a red arrow, a relatively large gap between the satellite gears that appeared when the satellite gear skipped and contacted the neighboring gear, causing the mechanism to fail. In Figure 10b,c, two pictures of the same central gear are presented, with an indication (a red arrow) as to where the satellite gear was stuck, which caused the whole mechanism to fail.

In the second failure mode, gears fail as a part due to poor layer adhesion when manufacturing the part. The parts that failed in this way were subjected to rotational speeds of 500 and 750 rpm, respectively, as shown in Table 4. Note that the gear bearing shown in Figure 11 was functional even after one gear started falling apart, and the bearing failed completely only after multiple gears were broken.

3.6. Material Abrasion and Buildup

Material abrasion and buildup is another phenomenon that was observed during the experiments. In Figure 12, images of a gear bearing 18 N-250 rpm are shown after being tested for 1 h (Figure 12a,c) and after 57 h (Figure 12b,d), and they show the wear of the parts. The images are shown in different magnification ratios, with Figure 12c,d being shot at 2× and Figure 12a,b at 20× magnification. In Figure 12a,c, the relatively intact layers of the 3D-printed part can be observed, in comparison with Figure 12b,d, in which the wear on the gears is visible. In addition to the wear, a supplementary effect is present: dust particles resulting from abrasion are forced in between the layers and generate a rather smooth and continuous surface that can take loads and maintain the functioning of the bearing for extended periods of time.

In Figure 13, a gear bearing 18 N-250 rpm after 57 h of functioning is shown. It failed through mode 1—gear jamming. One can observe here the PLA dust debris particles generated through abrasion during the 57 h of functioning.

3.7. Vibration

Vibration also appears during the functioning of the gear bearing because of three main reasons. First, there is an imprecision in the gear shape manufacturing due to the limitations of 3D printing technology, which can assure manufacturing precision of 0.2 mm. Secondly, the gear bearings are printed in one piece, putting the gears in contact at the time of manufacturing, but in order to spin the gear bearing, the contact needs to be broken, leaving behind imperfections on the gears. The third factor contributing to vibrations is the elastic deformation of the gear bearing. This deformation arises from the continuous load and the shifting contact point caused by the satellite gears as they rotate in relation to the outer ring gear. It is also to note that at the end of the experiment with 18 N-250 rpm, the radial play and vibration amplitude was 1300 μm.

4. Conclusions

The purpose of this study was to determine the feasibility and limitations of additively manufactured polymer gear bearings. The gear bearings were manufactured using PLA with an infill of 15%, and the experiments were designed such that the gear bearings were subjected to loads of 18 N and 45 N and rotational speeds from 250 rpm to 1500 rpm in increments of 250 rpm. PLA was used with an infill of 15% as a lower limit or baseline, and we consider that selecting a material with better mechanical properties combined with an optimal 3D printing technology and an optimized infill pattern and density can only improve the bearing’s performance.

Although PLA is not a prevalent choice for bearing materials, we embarked on exploring how this type of 3D-printed bearing performs. Common selections for FDM 3D printing include ABS (acrylonitrile butadiene styrene), PET (polyethylene terephthalate) and PEEK (poly-ether-ketone), but these materials are also not typically employed for bearings. In the existing literature, it is evident that polymer bearings are fairly widespread, with numerous manufacturers; however, our focus was on the utilization of 3D printing technology.

It is possible that PLA may not become the primary manufacturing choice in the future, but the unique geometry we employed may still have applications. Consequently, our aim was to present our findings, not only from the perspective of using 3D printing technology but also considering the bearing’s geometric design.

The rapid evolution of additive manufacturing techniques has opened up new ways for the production of complex mechanical components with interesting material properties. In this study, we explore the feasibility of employing 3D-printed polymer materials for the creation of bearing systems. Our research sheds light on the potential of polymer-based bearings to broaden available solutions to industries in which friction reduction and cost-effective manufacturing are concerns.

The experiments have shown that at low rpm, the bearing is a feasible alternative for short periods of time, with the longest average lifetime of a gear bearing being 57 h at 250 rpm for a load of 18 N, leading to a number of 855.000 loading cycles.

Another factor to take into consideration are the vibrations encountered during testing. These appear because of poor precision in the manufacturing process, as well as dimensional and geometrical tolerances. This issue can be addressed by using other additive manufacturing technologies that can provide better performance or using other types of polymers.

An optimization study can address the situation in which the number of gear teeth can improve the lifetime, with more teeth in contact meaning a greater contact surface. Also, an increase in infill density should have a positive impact on lifetime and bearing capacity. Since we now have a baseline after this study, we can extend and optimize designs.

The applications of PLA can be chosen based on its advantages: it is non-toxic, environmentally friendly, non-magnetic and can be used in simple 3D printers. On the other hand, its mechanical properties are rather low when compared to steel, ceramics or other polymers. Taking these into consideration, some industries that might benefit from PLA’s properties are medical devices, such as custom surgical guides being 3D printed for patient-specific cases (dental, orthopedic or tumor resections), prosthetics, braces, wearable sensors or drug delivery systems for one-time use that can be environmentally disposed of. Also, in the context of MRI-magnetic resonance imaging devices, the non-critical elements involved could use such bearings, for example, adjustable mounts, brackets, fixtures for holding instruments and accessories.

Other application industries might be food and beverage, prototyping, education and hobbies, artistic works or the toy industry, which could benefit from the environmental friendliness of PLA.

Future studies might address the problem of the short lifetime or bearing capacity of the gear bearing by increasing the infill amount or by changing the manufacturing process to another additive manufacturing technology.