Engineering Failure Analysis 110 (2020) 104411

Contents lists available at ScienceDirect

Engineering Failure Analysis

journal homepage: www.elsevier.com/locate/engfailanal

Failure analysis of 3D-printed steel gears

T
Tugce Tezel , Eyup Sabri Topal, Volkan Kovan
⁎

Akdeniz University, Department of Mechanical Engineering, Antalya, Turkey

ARTICLE INFO ABSTRACT

Keywords: The basic principle of additive manufacturing technology is that the model can be manufactured
3d printing directly via computer-aided design system. This technology is quite different from traditional
DMLS methods and allows production to avoid wasting material.
Wear This study features 420 steel gears produced using both additive and conventional manu-
Spur gear
facturing. The gears were subjected to identical test processes and compared to one another. The
Failure
effects of rotational speed, torque, production technique, and post-production surface treatment
were investigated. The researchers performed wear loss, efficiency analysis, oil analysis, damage
analysis, optical, and scanning microscope surface analysis. It was revealed that the density and
hardness of the gears produced by additive manufacturing (Direct Metal Laser Sintering – DMLS)
were quite close to those produced by conventional methods (hobbing machine); however, the
surface properties were different. Therefore, the types of damage that occur under the same
operating conditions differ. As a result, it was found that gears produced by additive manu-
facturing wear more, but the weight loss (which is a measure of wear) from wear was mostly
caused by non-sintered powders. However, under the same operating conditions, it was found
that production technique has little effect on efficiency. In some conditions, the performance of
the gears produced by additive manufacturing (with a smoother tooth profile) was better from
the computer-aided design. It was also seen that surface treatment after additive manufacturing
has a preventive effect on gear wear and damage.

1. Introduction

Gears are a machine element required in almost all machine designs, given that they move and transmit mechanical power
between shafts. As a result of developments in the field of machine design and manufacturing, gears are sought after because they are
inexpensive, light, and quiet, with a high degree of power transmission. Gear design must continuously be adapted to keep up with
the latest developments in advanced technology [1]. Early detection of gear failure is important to prevent increasing costs and time
loss caused by damage to machines such as surface fatigue (pitting), tooth breakage, and lubrication errors (scoring). It is necessary to
know in advance the damage mechanisms and strengths to use gears with small modules practically. Hobbing machines are generally
used in the manufacturing of gears, which can be produced using cutting dies, extrusion, and sintering. Producing high-quality gears
with new methods will offer alternatives by overcoming the limitations of conventional processes. Recently, additive manufacturing
has become an effective production method and started to gain a place in the manufacturing sector as technology develops.
Additive manufacturing is the combining of materials by overlapping layers to make objects from three-dimensional model data.
This additive process contrasts with subtractive methods such as conventional machine processing [2].
The researchers of this study produced gears with similar properties using either additive or traditional manufacturing, then

⁎
Corresponding author.
E-mail address: [email protected] (T. Tezel).

https://doi.org/10.1016/j.engfailanal.2020.104411
Received 23 September 2019; Received in revised form 25 December 2019; Accepted 24 January 2020
Available online 30 January 2020
1350-6307/ © 2020 Elsevier Ltd. All rights reserved.
T. Tezel, et al. Engineering Failure Analysis 110 (2020) 104411

subjected them to identical test processes. The gears’ mechanical behaviours were then compared. The applied test processes were
formed based on operating parameters and determined in accordance with the service conditions of the gears. Results from this work
may suggest operating parameters in which additive can replace traditional manufacturing for gears – important machine elements
used in multiple fields.
No studies of a similar nature currently exist in the literature. Existing studies investigate the mechanical properties of additive
manufacturing test specimens alongside analysis of the additive manufacturing of parts from various materials. Research mainly
concentrates on the mechanical properties of steel samples produced by additive manufacturing. Khaing et al. [3] produced metal test
samples using Direct Metal Laser Sintering (DMLS). Surface roughness, dimensional accuracy, toughness and hardness of the parts
were measured and analysed under an electron microscope. Tay et al. [4] applied a variety of techniques to improve the common
limitations of DMLS-manufactured parts including low surface quality, wear-resistance and soft and porous structures. Doyle et al. [5]
emphasized the importance of considering the effect of manufacturing parameters in additive manufacturing. This study looked at the
effect of various parameters on material properties such as density, tensile and yield strength during the production process. Ren et al.
[6] used a 3D gel printing method to produce stainless steel gears, after which they examined surface roughness and strength for
different raw materials. Pal et al. [7] observed how post-production processes impact DMLS stainless steel and revealed that post-
production processing may be required for the product to carry certain mechanical and surface properties. Pal et al. [7] performed
tensile tests on stainless martensitic steel samples fabricated using direct metal laser melting and found that post-production may be
required for the mechanical and surface properties to meet a certain standard.

2. Material & methods

Various tests were used to analyse the behaviour of gears subject to variable torque and rotation speeds during operation. The
gears used in this study were operated on a wear test machine.

2.1. Test gears

The modulus of 420 steel gear was 1, whereas the number of teeth were 20. In addition, the pressure angle was 20°. Fig. 1 shows
the technical drawing of the test gear.

2.2. Production methods of the gears

Production techniques, details and nomenclatures are given in Table 1.

2.2.1. Conventional manufacturing

Four hundred twenty steel cold-drawn rods were shaved down to appropriate diameter using a lathe. The hob was threaded over
the entire length to achieve the desired tolerance and quality. The gears were machined to suitable width, slots were made, and the
rods were then ground down. The terms ‘conventional’ or ‘traditional’ production are used to describe the production technique of
hobbing and machining without computer-aided design [8,9].

Fig. 1. Technical drawing of the test gear (All dimensions are in mm) [8]

2
T. Tezel, et al. Engineering Failure Analysis 110 (2020) 104411

Table 1
Production Methods of the Gears.
Material Manufacturing Methods Nomenclature

420 Hobbing + machining Conventional 420

DMLS Additive DMLS-420 Unpolished
DMLS + polished Additive DMLS-420 Polished

2.2.2. Additive manufacturing

SolidWorks was used to draw 3D solid models based on the technical drawing in Fig. 1. After the design was completed, computer-
aided design models were converted into ‘.stl’ files, thus enabling the 3D model to be converted into G code using auxiliary software.
The data of the ready-to-manufacture model was then transferred to an additive manufacturing machine. Production was made using
DMLS production steps, which are shown in Fig. 2.
The gears produced by additive manufacturing were positioned horizontally on the device table during production. The reason for
the vertical positioning of the gears in Fig. 2 is for easy expression of production. In addition, conditions such as shrinkage were
considered during production. Stainless powder layers were melded together to produce steel. A special printhead from DMLS moves
forward and backward over the layer as directed by the design file, thus leaving the binding agent on the steel powder layer. After the
sheet is complete and dried using strong heaters, a new layer of powder is emitted, and the process starts over again. Doing so thus
yields a bottom-up designed product. The piece at post-production is quite fragile and is named ‘Green Product’. Finally, bronze is
diffused into the piece, allowing the binder to be replaced. The result steel includes highly resistant metals such as bronze. The ratio
of the composition in the diffused bronze is 90% Cu and 10% Sn. The post-production steel chemical composition consists of 60%
steel and 40% bronze.
After cooling, the model was placed into a mechanical polisher, which is an additional operation that may be performed should a
smooth surface be desired. The DMLS device used was an EOS brand offering a minimum wall thickness of 1 to 3 mm depending on
the dimensions, clearance of 0.8 mm and an accuracy of ± 1% (after 2–3% shrinkage compensation). One drawback of additive
manufacturing is limits on the maximum size of parts that can be produced, which is also applicable to the EOS device. Yet because
the gear size examined for this study was small, no problems were encountered. Table 2 shows the expected properties of 420 steel
materials according to the device’s production parameters.

2.3. Gear wear test rig

Wear testers for a module of 1 mm gears consisted of AC servomotors, torque sensors, a gearbox, couplings, and brake me-
chanisms, along with various other mechanical parts. Components of the construction are shown in Fig. 3.
An AC servomotor is controlled by its driver and provides control at the desired speed. The movement from the AC servomotor is
transmitted first to the input torque meter via the flexible coupling and then to the gearbox by a second flexible coupling. At the
output of the gear unit, the movement is transmitted to the output torque meter by a third flexible coupling, and then to the brake
mechanism representing the load by the fourth flexible coupling. The servomotor drive allows the motor to move at the desired speed
while the brake mechanism adjusts the load level. Similar to the AC servomotor, the brake mechanism is controlled by driver. In the
electronic part of the design, a measurement and automation device (Labjack U12), which provides analogue and digital inputs and
outputs, is used for controlling the AC servomotor, brake mechanism and torque meters. Data was collected from the computer using
software for receiving and controlling data from the measurement device.
A gearbox was used to design and manufacture the gear wear device, whereupon the gears to be tested could be fitted. The gears
were fixed to a Ø6h6 tolerance shaft, the A-type slots in the radial direction, and the shaft ring in the axial direction. Prior to each
test, a fixed amount of the same oil was placed in the gearbox, whereupon the test oil was then collected and used for analysis [8].

Fig. 2. Gear Production using DMLS.

3
T. Tezel, et al. Engineering Failure Analysis 110 (2020) 104411

Table 2
Material properties expected from gears manufactured
using DMLS.
Density (g/cm3) 7.86
Tensile strength (MPa) 682
Breaking elongation (%) 2.3
Young’s modulus (GPa) 147
Yield strength (MPa) 455
Hardness (HRb) 97

Fig. 3. Gear Wear Test Machine [9]

2.4. Data collection

To begin with, the hardness, weight and density values of the gears were determined. For the hardness measurements, an RBOV-
200 hardness tester was used. The side surface of the gears was determined as the hardness measurement point and the measurement
was repeated three times from different points. Radwag AS / 220 / C / 2 device and density kit with a sensitivity of 0.0001 g were
used to determine the weight and density of the gears. Density and weight measurements were measured separately for driven and
rotating gears before and after wear tests. Weight and density measurements also were planned to be performed three times, as were
the hardness measurements. However, due to the differences in density in the gears produced by additive manufacturing, these
measurements were repeated many times. The density of the gears having this production technique was determined by taking the
average of the results obtained from all gears produced by additive manufacturing. Since the gears produced by the conventional
method did not evidence such changes in density, the measurements for the traditional method were repeated only three times. To
better understand the gear surfaces and their damages, the surfaces before and after the wear test were examined with a ZEISS
GeminiSEM 500 scanning electron microscope and a SOIF SZ780-B2/L Trinocular Stereo Zoom microscope. The surface of the SEM
images in the study is shown in Fig. 4.
For the determination of wearing elements, oil analysis was performed. The researchers added the oil to the gearbox, and then
collected it after the test. This oil is mineral-based and consists of an excessive pressure additive. At the end of test, the researchers
determined what elements were present and their quantity as revealed through an [10] standard method.
To determine the efficiency of the gears, instantaneous recorded data were used in the input and output torque meters in the wear

Fig. 4. The surface of the SEM images.

4
T. Tezel, et al. Engineering Failure Analysis 110 (2020) 104411

tester. The ratio of the input torque to the output torque showed the efficiency in which the gears transmit torque.
For the limited number of experiments to be useful to all designers and researchers, normalization was applied to all data obtained
and calculated. Therefore, specific wear rate was calculated and [11] was used during the calculation. In addition, a similar cal-
culation method was used from a comparable study [12]. These calculations consider the amount of wear for a system, sliding
distance, sliding rate, material properties, applied load and more, dependent upon many parameters.
According to [11] standard, volume loss was determined using weight loss. The volumetric loss is then divided into the product of
the applied load and applied travel distance (105rotations). From the torque data stored in the wear device, the forces to the gear can
easily be calculated [9].

2.5. Test parameters

The researchers made their calculations based on gear theories and established a working range for those gears manufactured
using additive versus hobbing machines. The results obtained by numerical analysis were then expanded and ranges increased. The
stresses of the 420 steel gears were examined according to torque. Both the contact pressure and equivalent stress rose when the
applied torque was increased. However, the rate of contact pressure increase was high compared to the rate of equivalent stress
increase. In addition, the contact pressure at all torques was well above the equivalent stress. For this reason, any damage caused by
contact pressure in gears produced from 420 steel was foreseen. For this purpose, the researchers made sure to examine the gear tooth
sliding points [13]. Also, it was easy to explain which variables were the cause of damage that occurred.
A working range was determined according to numerical and theoretical studies. The researchers planned to carry out wear tests
at the torque values of 0.25 and 0.5 Nm at speeds of 250, 500 and 1000 rpm. Two different techniques produced gears operating
under the same conditions: DMLS and the conventional hobbing machine. The researchers also applied a surface polishing process to
the DMLS-produced gears, and assessed how DMLS post-production surface polishing affected many properties of the gears. Also, for
all tests, the gears were driven a total of 105 cycles.

3. Results & discussions

3.1. Mechanical properties of gears

Surfaces, hardness and density of gears produced by the different techniques were examined. Microscope images of gear tooth
surfaces produced using the hobbing machine are shown in Fig. 5a, while the gears produced by additive manufacturing are shown in
Fig. 5b and c. Upon examining Fig. 5, one sees that there are clear differences in the tooth surfaces of the gears dependent upon
production technique. In Fig. 4a, the force applied to the gears produced using a hobbing machine is seen on the tooth surface as
being wear from the production. Also, after production, both tooth surfaces appear to take on similar properties.
The DMLS gear surface is composed of sintered powders and has variable properties across the entire surface (Fig. 5b). Electron
microscope images reveal that there are gaps caused by the sintering process. Surface polishing applied to the additively manu-
factured steel caused the tooth surfaces to change, as can be seen in Fig. 5c. The effect of dust grains that bring the material together
along the surface of the tooth is clearly visible in the images of the polished gears. Even though the gear is made from the same
material, it is observed that their surfaces are very different from one another due to how they were produced.
The properties of the 420 steel gears obtained after the density and hardness measurement tests are given in Table 3. Upon taking
a closer look at the hardness and density values, one sees that that the values obtained from the steels produced by each method are
close to one another.

3.2 wt. & oil analysis results

This section discusses the amount of wear by weight of the gears produced by the different techniques as well as analysis of the
elements obtained from the oil used during operation of the gears. Fig. 6 shows the level of wear obtained from the total weight loss
of the steel gear operated at 250, 500 and 1000 rpm at a torque of 0.25. It also lists the amount of wear produced by additive
manufacturing of both the unpolished and polished surface gear at 250, 500 and 1000 rpm. The lowest wear weight of the gear was
produced by the hobbing machine at 250 rpm and is similar to the results obtained at other rotational speeds. However, at this
rotational speed, one observes that the gears produced by additive manufacturing were more worn out. It is thought that the reason
for the excessive amount of wear produced by unpolished additive manufacturing is the separation of the unsintered powders from
the surface at low rotational speeds.
Increasing the rotation speed almost never affected the wear level of the steel gears produced by the hobbing machine and
polished additive manufacturing. However, the amount of wear of the unpolished gears produced by additive manufacturing in-
creased 2.5-fold when the rotation speed was increased to 500 rpm from 250 rpm. When the rotation speed was increased to
1000 rpm, the amount of wear was not affected, just like the gear produced by other production techniques.
Fig. 7 illustrates the change in the amount of wear resulting from the total weight loss of the 420 steel gear pairs operated at a
torque of 0.25 and 0.50 Nm at 500 rpm. At each torque, the production technique whereby wear was most common was unpolished
additive manufacturing, followed polished additively manufactured steel and hobbed steel gear. Although an increase in torque
affected the level of wear of the steel gears produced by additive manufacturing, it almost never affected the level of wear of the gears
produced by the hobbing machine. The steel gear production technique that was most affected by the increase in torque was the

5
T. Tezel, et al. Engineering Failure Analysis 110 (2020) 104411

a. Produced by hobbing machine

b. Produced by DMLS-unpolished

c. Produced by DMLS polished

Fig. 5. Scanning electron microscope and optical microscope images (before wear test).

Table 3
Properties of 420 steel gears.
Production Technique Hardness (HB10) Density (g/cm3)

Conventional 128 7.52

DMLS-unpolished 115 8
DMLS-polished 132 7.85

surface polished gears produced by additive manufacturing.

Figs. 6 and 7 reveal the most significant wear elements in the gears including the amount of wear resulting from the oil analysis. A
change in the amount of Fe obtained from the oil in which the 420 steel gears operated at 250, 500, and 1000 rpm is given in Fig. 8.
At 250 and 1000 rpm, the amount of Fe in the oil gave similar results for all the production techniques. However, at 500 rpm, the type
of production technique showed a significant impact on the amount of Fe detected in the oil. For all the manufacturing techniques,
the effect of the increased rotation speed on the amount of Fe is also shown in Fig. 8. As a result of the increased rotational speed, the

6
T. Tezel, et al. Engineering Failure Analysis 110 (2020) 104411

Fig. 6. The amount of wear of 420 steel gears depending on the rotational speed at 0.25 Nm torque.

Fig. 7. The amount of wear of steel gears dependent on a torque of 500 rpm.

Fig. 8. The amount of Fe detected in the oil in which the steel gears work with a torque of 0.25 Nm, depending on the rotational speed.

amount of Fe in the oil increased and then decreased in the gears produced using additive manufacturing. However, the amount of Fe
in the oil of the gears produced using the hobbing machine decreased before rotation speed had increased, and then remained almost
constant. These results are not enough to compare either the weight changes or the number of elements obtained from the oil. There
is a difference between the amount of Fe in the gears according to production techniques used.
Bronze impregnation is used in order to prevent fragility after manufacturing by DMLS, resulting in gears that consist of ap-
proximately 60% steel and 40% bronze. Fig. 9 shows a result of the sum of the elements in the oil common to the three production
techniques, which has a great impact on weight. Fig. 9 displays the amount of Fe + Cr & Cu + Sn detected in the oil in which the
steel gears work with a torque of 0.25 Nm, depending on the rotational speed. The sum of the weight of Fe and Cr detected in the oil
using the gears produced using the hobbing machine was similar to change in the amount of wear by weight in Fig. 6. In addition to

7
T. Tezel, et al. Engineering Failure Analysis 110 (2020) 104411

Fig. 9. Amount of Fe + Cr & Cu + Sn detected in the oil in which the steel gears work, depending on the rotational speed and 0.25 Nm torque.

Fe and Cr, Cu and Sn were also examined in the oil where the gears produced by additive manufacturing were used. In other words, it
was investigated whether the amount of element detected in the oil is the same as the amount lost by weight due to wear.
Fig. 9 shows that the investigated elements were denser at 250 and 1000 rpm in the additively manufactured unpolished steel
gears; it also shows these elements being present in the polished gears at 500 rpm. Figs. 6 and 9 reveal that results of the elemental
analysis obtained from the oil analysis and changing of weight are comparable. Both analyses show us that the maximum amount of
wear was observed in additive manufacturing product gears with an unpolished surface. One might consider that if the other elements
in the steel composition were added in Fig. 9, the results would converge to the results obtained in Fig. 6.
Figs. 6, 8 and 9, when evaluated as a whole, were found to be supported by the studies in the literature. The reduction in the wear
occurs due to a decrease in the contact period of a single tooth and an increase in the strain rate at higher rotational speeds. The
relationship between rotational speed, contact period and strain rate [14]
1
Contact period of a tooth =
number of tooth x speed

1 Torque acting on the tooth

Strain rate or rate of loading = .
contact ratio contact period

3.3. Gear damage

When gear damage is encountered, there are three main causes to which it can be attributed: surface or tooth fatigue damage
(pitting), tooth fracture damage (overloading cleavage or fatigue cleavage) or lubrication damage (scoring, abrasion, scuffing) [15].
Types of damage occurring on tooth surfaces were examined in this context. To make the tooth surfaces examined more visible,
Fig. 10 can be used.
Fig. 10 shows the surface images of 420 steel gears after working at a torque of 0.25 Nm. The gear produced by the hobbing
machine at 250 rpm clearly shows fatigue damage of pitting on the surface. It is observed that this damage, especially in the left part
of the addendum region, intensifies at the face of the addendum region of the gear teeth surface. The damage mechanism of the
hobbed gear changed at 500 rpm. At a low rotation speed, the pitting at the addendum of the tooth decreased; however, at the
dedendum, scuffing and pitting were observed to be in the direction of the force of the contacted tooth. Additionally, this figure
shows that pitting at 250 rpm is the initial pitting. The tooth surface profile of a hobbed gear changes at 1000 rpm, and thus reveals to
us that this damage is a plastic deformation.
On polished gears produced by additive manufacturing at 250 rpm, the tooth surface had a small amount of pitting (fatigue
failure), and there also was scuffing damage (lubrication failure) in the direction of the force applied by the contact surface. Damage
to the polished surface was noticeable with increased rotation speed (500 rpm). There was pitting on all tooth faces. In addition, there
was an intense plastic deformation in the direction of the load, which was subject to surface pressure at the addendum. Likewise,
there was scuffing damage with highly enlarged pitting at the root of the teeth. Here, it can be said that the sintered powders are
largely removed from the tooth surface. The regions where the damages occurred were different compared to the other rotational
speeds as a result of the changing rotational speed. The dense plastic deformation, especially at the root of the tooth, almost destroyed
the images of sintered dust. When one compares the damage mechanisms resulting from the production techniques, pitting (fatigue)
damage was observed in both production techniques (DMLS polished/unpolished and traditional) at a low speed. Similarly, the
scuffing damage type (lubrication defect) was intense in both production techniques at 500 rpm and resulted in more dense and
various damages on the gear surfaces produced by additive manufacturing. There also was a plastic deformation in the gears pro-
duced by both production techniques at 1000 rpm.
Fig. 10 also exhibits the effect of the surface treatment applied to the gears produced by the additive manufacturing on the gear
damage mechanisms. Both gear types (produced using DMLS polished/unpolished and traditional) show pitting at 250 rpm. At the

8
T. Tezel, et al. Engineering Failure Analysis 110 (2020) 104411

100μm 100μm 100μm

Produced by hobbing machine Produced by DMLS-polished Produced by DMLS unpolished
a. 250 RPM

100μm 100μm 100μm

Produced by hobbing machine Produced by DMLS-polished Produced by DMLS unpolished
b. 500 RPM

100μm 100μm 100μm

Produced by hobbing machine Produced by DMLS-polished Produced by DMLS unpolished
c. 1000 RPM
Fig. 10. Tooth surfaces examined for damage after working at 0.25 Nm torque.

unpolished gear surface, the scuffing was distributed along the tooth root and directed towards the addendum area. However, in the
polished surface, scuffing was only seen in small amounts in the tooth root. At 500 rpm, the unpolished gear surface showed dense
scuffing on the pitch diameter region of the tooth surface as well as a small amount of plastic deformation. Moreover, pitting in the
root of the teeth, in the pitch diameter area, enlarged regional pitting, and dense plastic deformation on the tooth stand out on the
polished surface. At 1000 rpm, scratches are visible on the surface of the unpolished teeth surface, along with modest pitting at the
addendum and the tooth root in the direction of the contact force. However, plastic deformation and pitting are seen in the dedendum
on the surface of the polished tooth. Thus, the applied surface treatment impacts the damage types and the areas where the damage
occurs.
The effect of increasing the torque on gears fabricated by DMLS is shown in Fig. 11. The formation of a local crack in the untreated
gear is worth noting due to the increase in torque. In fact, it is seen that dust is trapped inside this crack. For gears with post-
production polishing, there was minor plastic deformation at the dedendum area of the teeth, a small amount of pitting and scuffing
at the pitch diameter area, and more intensive pitting at the addendum. Figs. 10 and 11 present the connection between increased
torque and the applied surface treatment.

3.4. Efficiency analysis

The efficiency of 420 steel at 250, 500 and 1000 rpm at 0.25 Nm torque is given in Fig. 12, which also shows us that the values are
similar for additive manufacturing and the traditional method. In particular, the efficiency of gears produced by additive manu-
facturing techniques is closer to one another as the rotation speed increases. The results of the weight change and the oil analysis
reveal that the unpolished gear produced by the additive manufacturing is generally the most worn, however this does not appear to
decrease its efficiency.

9
T. Tezel, et al. Engineering Failure Analysis 110 (2020) 104411

a. DMLS unpolished - 500RPM b. DMLS unpolished - 500 RPM

Fig. 11. 420 steel gears after working at 0.50 Nm torque, dependent on post production method.

Fig. 12. Efficiency of 420 steel gears depending on rotational speed at 0.25 Nm torque.

With increasing rotational speed, the efficiency of the gears produced by additive manufacturing increases compared to those
produced by the conventional method. Despite the increased amount of wear, such a result may be due to a smoother tooth profile
with computer-aided DMLS manufacturing. The straightness of the tooth profile directly affects the efficiency.

3.5. Normalized results

The results obtained from the featured gears were normalized according to ASTM-G99. Thus, the results obtained in this study
proved to be applicable for other gears to be used by the researchers. Figs. 13 and 14 show the results obtained according to ASTM-
G99. Fig. 13 presents changes in the specific wear rate due to the rotation speed of 420 steel gears at 0.25 Nm torque. At 250, 500 and

Fig. 13. A specific wear rate of 420 steel gears depending on the rotational speed at 0.25 Nm.

10
T. Tezel, et al. Engineering Failure Analysis 110 (2020) 104411

Fig. 14. Specific wear rate of 420 steel gears depending on torque at 500 rpm.

1000 rpm, the lowest specific wear rate was obtained from gears produced using the hobbing machine. At 500 and 1000 rpm,
polished and unpolished gears produced using additive manufacturing followed those produced by the hobbing machine. Fig. 14
illustrates the changes in the specific wear rate of 420 steel gears at 500 rpm due to torque. This increase had almost no impact on the
specific wear rate of the polished gear produced using additive manufacturing, whereas it did have a profound impact on the
unpolished gear. The results obtained in Figs. 13 and 14 are nevertheless similar to one another.
When the literature is examined, the specific wear rate varies greatly in the range of 10–15 to 101 mm3/Nm, depending on the
operating conditions and material selection [16]. The specific wear rate values obtained for all materials used in this study confirm
the results as they are in the appropriate range.

4. Conclusion

Additive manufacturing makes it easier for designers to produce a physical product as well as design. Compared to conventional
production processes, additive manufacturing can significantly reduce energy use by using fewer materials and by eliminating several
steps. In this context, the results of the study are as follows:

• The density and hardness of the gears produced by additive manufacturing are very close to those produced by conventional
methods. Also, the density and hardness value of the surface polished gear produced by additive manufacturing is closer than that
obtained by the conventional method compared to the gear produced by the unpolished additive manufacturing. Furthermore, the
surface properties differ due to the production technique used.
• In general, the efficiency of steel gears produced by additive manufacturing remain either the same or are higher than those
produced by the conventional method. Therefore, this can replace gears from the same material produced using conventional
methods.
• The surface polishing process applied in addition to the gears produced by additive manufacturing did not cause any change in
efficiency and usage range. However, it has a preventive effect on surface pressure damage caused by high speeds. Surface
polishing will be beneficial to prevent surface damage.
• Although the steel gears produced using both production techniques (DMLS polished/unpolished and traditional) show different
composition depending on their alloys, the element that exhibited the most wear was Fe.
• One of the biggest advantages of gears produced using additive manufacturing is the correct thread profiles, thanks to computer
aided design, given that they are able transmit high efficiency power to the gears.
• The damage mechanisms are quite different from one another due to the various tooth surface properties stemming from each
technique. While surface deformations caused by shear forces were observed on the gear surfaces produced by machining, non-
uniform tooth surfaces were a result of sintering on the gear surfaces produced by additive manufacturing. The cavities left by dust
particles that are not sufficiently sintered expand with the effect of oil pressure and accelerate the pitting damage in the gears
produced by additive manufacturing.
• Researchers and designers will benefit from this study data when it comes to looking at gears of different sizes with the help of
normalized data.
• Particularly in the case of spare part requirements, it has been demonstrated that additive manufacturing products can be used
under certain conditions in place of machine parts manufactured by conventional methods.
• The type of damage seen in small modulus gears, its formation time and development differ from normal modulus gears. In
normal-sized gears, voids from dust particles that are not sufficiently sintered affect a relatively small volume of the tooth surface
area, whereas in small modulus gears, dust particles which are not sufficiently sintered affect a relatively larger volume of the
tooth. With the effect of oil pressure, the conversion of initial pitting in the micro-modulated gear into progressive pitting (re-
sulting in tooth fracture damage) may develop faster than in normal-sized teeth.

11
T. Tezel, et al. Engineering Failure Analysis 110 (2020) 104411

Declaration of Competing Interest

The authors declared that there is no conflict of interest.

Acknowledgement

This work was supported by The Scientific Research Projects Coordination Unit of Akdeniz University. Project Number: FDK-
2017-2808.

References

[1] S.P. Radzevich, Dudley’s Handbook of Practical Gear Design and Manufacture CRC Press, Taylor & Francis Group, 2016.
[2] ASTM 52900. 2015. Standard Terminology for Additive Manufacturing Technologies-General Principles-Terminology ASTM International, West Conshohocken,
PA, ABD.
[3] M.W. Khaing, J.Y.H. Fuh, L. Lu, Direct metal laser sintering For Rapid Tooling: processing and characterisation of EOS parts, J. Mater. Process. Technol. 113
(2001) 269–272.
[4] F.E.H. Tay, E.A. Haider, Laser sintered rapid tools with improved surface finish and strength using plating technology, J. Mater. Process. Technol. 121 (2002)
318–322.
[5] M. Doyle, K. Agarwal, W. Sealy, K. Schull, Effect of layer thickness and orientation on mechanical behavior of binder jet stainless steel 420+bronze parts,
Procedia Manuf. 1 (2015) 251–262.
[6] X. Ren, H. Shao, T. Lin, H. Zheng, 3D gel-printing—An additive manufacturing method for producing complex shape parts, Mater. Des. 101 (2016) 80–87.
[7] S. Pal, H.R. Tiyyagura, I. Drstvenšek, C.S. Kumar, The effect of post-processing and machining process parameters on properties of stainless steel PH1 product
produced by direct metal laser sintering, Procedia Eng. 149 (2016) 359–365.
[8] T. Tezel, An Experimental Investigation of the Use of Three Dimensional Metal Printed Gears In Place Of Conventional Methods. PhD Thesis, Antalya, Turkey,
2019.
[9] T. Tezel, E.S. Topal, V. Kovan, Characterising the wear behaviour of DMLS-manufactured gears under certain operating conditions, Wear (2019) 440–441.
[10] ASTM D 5185 Standard Test Method for Multielement Determination of Used and Unused Lubricating Oils and Base Oils by Inductively Coupled Plasma Atomic
Emission Spectrometry (ICP-AES).
[11] ASTM G99 Standard Test Method for Wear Testing with a Pin-on-Disk Apparatus.
[12] M. Masjedi, M.M. Khonsari, On the prediction of steady-state wear rate in spur gears, Wear 342–343 (2015) 234–243.
[13] Tezel, T., Kovan, V., Topal E.S. 2018. Effects Of Tooth Profile Modification On Root Stress In 3D Printed Micro Module Spur Gears : A Numerical Analysis,
International Conference On Technology and Science, December, Antalya, Turkey.
[14] P.K. Singh, Singh A.K. Siddhartha, An investigation on the thermal and wear behavior of polymer based spur gears, Tribol. Int. 118 (2018) 264–272.
[15] T.C. Glew, Failure analysis and repair techniques for turbomachinery gears, Texas A&M University. Gas Turbine Laboratories, ABD, 1980.
[16] J.F. Archard, Contact and rubbing of flat surfaces, J. Appl. Phys. 24 (1953) 981–988.

12