polymers

Article
Fatigue and Wear Performance of Autoclave-Processed and
Vacuum-Infused Carbon Fibre Reinforced Polymer Gears
Zoran Bergant, Roman Šturm *, Damijan Zorko and Borut Černe *

Faculty of Mechanical Engineering, University of Ljubljana, Aškerčeva 6, 1000 Ljubljana, Slovenia

* Correspondence: roman.sturm@fs.uni-lj.si (R.Š.); borut.cerne@fs.uni-lj.si (B.Č.); Tel.: +386-14-771-203 (R.Š.);
+386-14-771-129 (B.Č.)

Abstract: This study focuses on investigating the fatigue and wear behaviour of carbon fibre re-
inforced polymer (CFRP) gears, which have shown promising potential as lightweight and high-
performance alternatives to conventional gears. The gears were fabricated via an autoclave process
using an 8-layer composite made of T300 plain weave carbon fabric and ET445 resin and were tested
in pair with a 42CrMo4 steel pinion and under nominal tooth bending stress ranging from 60 to
150 MPa. In-situ temperature monitoring was performed, using an infrared camera, and wear rates
were regularly assessed. The result of the wear test indicates adhesive wear and three-body abrasion
wear mechanisms between the CFRP gears and the steel counterpart. A finite element analysis was
performed to examine the in-mesh contact and root stress behaviour of both new and worn gears
at various loads and a specified running time. The results point to a substantial divergence from
ideal meshing and stress conditions as the wear level is increased. The fatigue results indicated that
the CFRP gears exhibited superior performance compared to conventional plastic and composite
short-fibrous polymer gears. The described composite gear material was additionally compared with
two other composite configurations, including an autoclave-cured T700S plain weave prepreg with
DT120 toughened resin and a vacuum-impregnated T300 spread plain weave carbon fabric with LG
900 UV resin. The study found that the use of the T700S-DT120 resulted in additional improvements.

Keywords: carbon fibre reinforced polymer; gears; fatigue; wear; finite element analysis; autoclave;
Citation: Bergant, Z.; Šturm, R.;
vacuum infusion
Zorko, D.; Černe, B. Fatigue and Wear
Performance of Autoclave-Processed
and Vacuum-Infused Carbon Fibre
Reinforced Polymer Gears. Polymers
1. Introduction
2023, 15, 1767. https://doi.org/
10.3390/polym15071767
The use of polymer and polymer composite gears has seen a significant rise in re-
cent times, particularly in applications that require low loads where using metal gears
Academic Editors: Seyed R. Koloor may not be cost-efficient. The use of composites is expanding to dynamically loaded
and Geralt Siebert applications, including bearings [1,2] and gears. The growing popularity of polymer ma-
Received: 22 February 2023 terials in gear transmissions is attributed to their superior properties, including reduced
Revised: 24 March 2023 noise and vibration levels, lower weight, enhanced corrosion resistance, and the ability
Accepted: 28 March 2023 to operate in lubrication-free environments. The exceptional fatigue and creep resistance
Published: 1 April 2023 of CFRP, compared to other FRPs, has greatly expanded its use in engineering applica-
tions where the composite is used as the primary load-bearing component in prestressed
applications [3–5]. The use of polymer materials in powertrain systems is experiencing
exponential growth across a range of fields, including household appliances, hand power
Copyright: © 2023 by the authors. tools, medicine, robotics, and beyond. However, a key challenge with currently available
Licensee MDPI, Basel, Switzerland. lightweight materials is that they typically have significantly lower load capacity compared
This article is an open access article
to steel components, often by a factor of 8–10. To address this issue, CFRP composites have
distributed under the terms and
emerged as a class of extremely durable and lightweight materials that can already replace
conditions of the Creative Commons
metals in a variety of engineering applications.
Attribution (CC BY) license (https://
For applications with the most severe operating conditions, lightweight CFRP parts are
creativecommons.org/licenses/by/
joined with parts made of Fe, Al, and Ti alloys, and other materials, to maintain stiffness and
4.0/).

Polymers 2023, 15, 1767. https://doi.org/10.3390/polym15071767 https://www.mdpi.com/journal/polymers

Polymers 2023, 15, 1767 2 of 24

increase wear resistance. These so-called hybrid gears—which feature steel in the teeth and
hub and can reduce weight by as much as 40% compared to standard all-steel gears—are
the subject of several papers [6–8]. Handschuh et al. [6] conducted an experimental study
in which hybrid gears consisting of steel teeth and a steel inner hub were connected by a
composite web. A strong dependence of the operating speed on the vibration level was
found, indicating the importance of the targeted use of composite materials to reduce
vibrations and weight. Similarly, Catera et al. [7] investigated a steel-composite hybrid
drive, where the steel teeth and internal hub are connected by a composite web. In
their studies, a multiscale modelling approach was employed to accurately estimate the
natural frequencies of the hybrid transmission, taking into account the unique properties
of composite materials. In their follow-up study, Catera et al. [8] conducted a numerical
investigation that implemented a thin steel layer to establish a connection between the
steel teeth and the inner hub, successfully addressing the issue of centreing problems.
Additionally, a composite web was incorporated into a steel gear body. A comparative
analysis was conducted between hybrid steel gears with identical macro geometry and
mass. The static transmission errors of the hybrid gear and the lightweight steel gear were
calculated using FEM. The results indicated a considerable decrease in peak-to-peak static
transmission error, leading to an enhanced performance of the hybrid transmission in terms
of noise, vibration, and harshness (NVH). It is worth noting that the composite material
used in these studies served solely as a connection between the steel teeth and the inner
hub, aimed at reducing weight and improving NVH performance.
For low load capacity, the injection moulding process efficiently provides lightweight
polymer gears that do not require lubrication. To increase load capacity and wear resistance,
short reinforcing fibres are embedded in a polymer matrix. Short fibre (SF) composites are
materials that contain fibres that are typically less than 1 mm long. Blais and Toubal [9]
investigated HDPE gears reinforced with short natural fibresand subjected to high-cycle
bending fatigue. Their study shows that injection moulding of HDPE/birch fiber is a
promising process to manufacture eco-friendly machine parts at lower costs. Mohsen-
zadeh et al. [10] investigated the load-bearing characteristics of polyoxymethylene (POM)
with carbon black (CB) CaCO3 ternary nanocomposite gears using a temperature-based
step loading approach with a 30% increase of mechanical properties. Furthermore, the SEM
micrograph of worn surfaces revealed a much smoother worn surface of composite gears,
which indicates beneficial synergistic strengthening effects of CB and nano-CaCO3 compos-
ite gears. Kurokawa et al. [11] tested poly-ether-ether ketone (PEEK) gears reinforced with
15% CF under different test conditions, i.e., dry running and lubricated running, and in
combination with a driven steel gear. For comparison, the PAI-CF30 and PPS -GF30 gears
were also injection-moulded and tested under non-lubricated conditions, with the driving
and driven gears made of the same material. The PEEK + 15% CF gears showed the best
performance compared to the other two high-performance plastics. In their subsequent
study, Kurokawa et al. [12] tested injection-moulded polymer gears to determine their
performance. The gears were made of various types of polyamides (Pas) reinforced with
short carbon fibres, including PA12, PA6, and PA46 with a 15% addition of carbon fibres.
The tests were conducted under grease-lubricated conditions, with the driving and driven
gears made of the same composite material. The results indicated that the PA46 gears
experienced wear in a relatively short period of time while the carbon fibre reinforced PA12
and PA6 gears failed due to fatigue. The gears made of short carbon fibre reinforced PA12
exhibited the best performance, with results comparable to PEEK-CF.
Incorporating reinforcing particles has been demonstrated to enhance wear resistance
significantly. Mao et al. [13] found a significant reduction in specific wear rate and a 50%
increase in load-bearing capacity by introducing 28% glass-reinforcing particles into POM
(polyoxymethylene). Normally, the static tensile strength and modulus of elasticity can be
increased by adding short reinforcing fibres. However, for dynamically loaded components,
polymer gears without short fibre reinforcements can perform better than the composite
gears reinforced with short fibres [14]. While the fibres increase the modulus of elasticity
 can be increased by adding short reinforcing fibres. However, for dynamically loaded
components, polymer gears without short fibre reinforcements can perform better than
the composite gears reinforced with short fibres [14]. While the fibres increase the modu-
Polymers 2023, 15, 1767 3 of 24
lus of elasticity and the quasi-static strength of the composite, the interfaces between the
hard fibres and the soft matrix can become areas of microscopic stress concentration and,
thus, act as crack initiators during fatigue loading [15,16]. On the other hand, there are
and the
many quasi-static
known benefits strength of the composite,
of continuous the interfaces
fibre reinforced composites,between
such the hard fibres
as balance and
of prop-
the soft matrix can become areas of microscopic stress concentration and,
erties, higher stiffness and strength, higher toughness, creep and fatigue resistance, di- thus, act as crack
initiators during
mensional fatigue
stability, etc. loading [15,16]. to
Load, applied Ona the other
fibre hand, there
reinforced are many
composite knownisbenefits
material, trans-
ferred from matrix to fibres by shear stresses along the fibre-matrix interface andstiffness
of continuous fibre reinforced composites, such as balance of properties, higher reaches
aand strength,
maximum higher
when toughness,
fibres creep
are aligned and
in the loadfatigue resistance,
direction; transferdimensional stability,with
efficiency increases etc.
Load, applied to a fibre reinforced composite material, is transferred from
the increase of fibre length-to-diameter ratio [17]. The composite laminate is an assembly matrix to fibres
bylayers
of shear ofstresses
fibrous along the fibre-matrix
composite materials interface
which can and
be reaches
joined toa provide
maximum thewhen fibres
required are
engi-
aligned in the load direction; transfer efficiency increases with the increase
neering properties, including in-plane stiffness, bending stiffness strength, and high mod- of fibre length-
to-diameter ratio [17]. The composite laminate is an assembly of layers of fibrous composite
ulus. Typically, individual layers are orthotropic or transversely isotropic with the lami-
materials which can be joined to provide the required engineering properties, including
nate exhibiting orthotropic or quasi-isotropic properties. The failure types of autoclave-
in-plane stiffness, bending stiffness strength, and high modulus. Typically, individual
cured, woven laminate composite gears and their fatigue properties were investigated by
layers are orthotropic or transversely isotropic with the laminate exhibiting orthotropic or
Zorko et al. [18,19]. The two failure modes were presented: two–three body abrasive wear,
quasi-isotropic properties. The failure types of autoclave-cured, woven laminate composite
and fatigue. A 6-times longer lifespan at the same fatigue load was observed for the woven
gears and their fatigue properties were investigated by Zorko et al. [18,19]. The two failure
CFRP gears compared to the high-performance PEEK gears. In comparison with other
modes were presented: two–three body abrasive wear, and fatigue. A 6-times longer
gears made of polymers or short fibre reinforced polymer composites, the load bearing
lifespan at the same fatigue load was observed for the woven CFRP gears compared to the
capacity was even higher. The fatigue regions between different classes of materials are
high-performance PEEK gears. In comparison with other gears made of polymers or short
shown in Figure 1. The autoclave-cured, woven CFRP gears show excellent potential to
fibre reinforced polymer composites, the load bearing capacity was even higher. The fatigue
close the large performance gap between the polymer and steel gears. Their ability to run
regions between different classes of materials are shown in Figure 1. The autoclave-cured,
withoutCFRP
woven lubrication, excellent
gears show strength-to-weight
excellent potential to closeratio,
theand good
large damping characteristics
performance gap between
make them promising candidates for aerospace applications.
the polymer and steel gears. Their ability to run without lubrication, excellent strength-
to-weight ratio, and good damping characteristics make them promising candidates for
aerospace applications.

Figure
Figure 1.
1. Fatigue region for different classes of materials; woven laminated CFRP gears fill the large
gap
gap between steel
between and plastic
steel and plastic gears
gears and
and composite
composite gears
gears with
with short
short reinforcing
reinforcing fibres.
fibres.

Gear wear
wear and
andfatigue
fatigueperformance
performanceare arehighly
highly dependent
dependent ononthethe tribological
tribological behav-
behaviour
iour
of theofgear
the gear pair
pair at theatmeshing
the meshing contact
contact interface.
interface. BijweBijwe and Sharma
and Sharma [20] conducted
[20] conducted a studya
study
on the on the potential
potential use of
use of CFRP CFRP in tribological
in tribological applications.
applications. They investigated
They investigated how varyinghowthe
varying the carbon
carbon fibre content fibre
ratio content
in CFRPratio
with in CFRP with a polyetherimide
a polyetherimide (PEI)matrix
(PEI) thermoplastic thermoplastic
affected
its mechanical
matrix affectedand tribological and
its mechanical properties. The study
tribological revealed
properties. Thethat an optimal
study revealedcombination
that an op-
of mechanical
timal combinationandoftribological
mechanicalproperties was achieved
and tribological with
properties wasa achieved
CFRP containing 65%
with a CFRP
fibre content. Ramesh and Suresha [21] analysed the frictional and wear
containing 65% fibre content. Ramesh and Suresha [21] analysed the frictional and wear characteristics
of epoxy-impregnated carbon fibre reinforced polymer (CFRP). Their findings indicate
that, when subjected to abrasive wear, carbon fabrics demonstrate superior performance
compared to glass fibres and other types of composite. The researchers observed a further
reduction in surface wear when the CFRP was combined with aluminium oxide (Al2 O3 ) and
molybdenum disulphide (MoS2 ). The findings suggest that, by optimising the carbon fibre
content ratio and incorporating wear-resistant particles, such as Al2 O3 and MoS2 , CFRP has
the potential to enhance its tribological behaviour and minimise surface wear. Moreover,
Polymers 2023, 15, 1767 4 of 24

Černe et al. [22] studied the thermomechanical response of the laminated composite gears,
where they proposed a new iterative method for implicit evaluation of coefficient of friction
(COF) between the woven CFRP gear and its steel counterpart.
Ekoi et al. [23] analysed the static and fatigue behaviours of additive manufactured
continuous carbon fibre reinforced polymer (CFRP) composites, and compared the woven
and non-woven structures. Their investigation also explored the failure mechanisms
associated with fatigue in these composites. The fatigue test results revealed that woven
carbon composites outperformed their unidirectional (non-woven) counterparts when
subjected to 70% of their ultimate strength. In the case of non-woven composites, failure
was observed in multiple regions with long splitting of the composite and fibre breakage.
On the other hand, for woven composites, failure mostly occurred due to fibre pull-out at
the interface between the unit cells, which is attributed to stress concentration.
This study presents a comparative analysis of the fatigue and wear properties of gears
made from carbon fibre reinforced polymer (CFRP) fabricated through different methods,
including autoclave-cured, prepreg, and vacuum-infused techniques. Additionally, the
present study aims to investigate how the durability of gears is affected by different types
of fibres, resin, and areal weights of fabrics.
An autoclave is a type of pressure vessel that uses heat and pressure to cure materials.
The process typically takes place at pressures between 4 and 8 bar and temperatures
between 100 and 250 ◦ C. The parts are typically made from a material called prepreg, which
is a fabric that has already been pre-impregnated with a catalysed resin. Prepreg resins are
typically partially cured or B-staged, meaning that they have already undergone a partial
curing process before they are ready to use. The prepreg is sealed in air-tight plastic bags
and stored at −18 ◦ C before use. Before lamination of the prepreg, it is brought to room
temperature and cut into different shapes to fit the desired mould. The material is then
covered with a peel ply, perforated foil, and a felt or breather (absorbent ply), and sealed
in a vacuum bag. When composite materials are subjected to pressure in an autoclave,
any air bubbles present within the material are forced out of the laminate. The process of
removing air bubbles is called “outgassing”. The air bubbles can escape through any small
openings in the laminate, such as the edges of the mould or through perforations in the
vacuum bagging materials. The vacuum pump also creates a low-pressure environment in
the autoclave chamber which helps to evacuate the remaining air bubbles in the laminate.
In the end, the final composite material is free of bubbles, which improves its mechanical
properties such as strength, stiffness, and toughness. The autoclave manufacturing method
is used to create high-quality structural parts for the aerospace industry [24,25].
Vacuum infusion is a process used to produce composite parts by impregnating the
fibres with a resin. Unlike prepreg resins, vacuum infusion resins are usually fully uncured
or A-staged, and undergo complete curing during the vacuum infusion process. The
process involves laying the fibres inside a form and then creating a vacuum to pull the
resin through the fibres and into the shape. Once the form is filled with resin, it is then
cured under heat and pressure to create a finished composite part. The main advantage
of vacuum infusion over traditional layering techniques is that it allows for more precise
control of the resin-to-fibre ratio and can result in more consistent and higher-quality parts.
The main difference between the two processes is that vacuum infusion impregnates
dry fibres and employs vacuum to pull the resin into the mould, while autoclave curing
requires pre-impregnated fibres, which have to be heated and pressurised to cure the part.
Although the vacuum infusion has many benefits, to completely remove the air from the
dry preform, the available resin choice is limited to non-toughened resin systems. This
research aims to investigate the durability of carbon fibre reinforced polymer (CFRP) gears
made from both toughened and non-toughened resin systems. Toughened epoxy systems
are often used by manufacturers in the production of prepreg materials to improve the
fracture toughness of the final product. However, the specific techniques used for the
toughening of prepreg resins systems are typically not disclosed, although they often in-
Polymers 2023, 15, 1767 5 of 24

volve incorporating reactive oligomers such as carboxyl-terminated butadiene–acrylonitrile

(CTBN) rubber or pre-formed elastomer particles [26,27].
Overall, this paper presents unique insights that have not been widely covered in the
literature, highlighting the potential for woven as well as the continuous 3D printed CFRP
composites to revolutionise the field of powertrain systems. The recognition and adoption
of the findings from this research have the potential to generate positive environmental
effects, such as increasing the durability of powertrain components and reducing wasted
plastic. By incorporating CFRP composites into powertrain systems, these components
can have longer lifetimes, low maintenance costs, and require less frequent replacements,
ultimately reducing waste and the associated environmental impact.
With a suitable processing technique, CFRP gears can exhibit high mechanical strength,
good thermal stability, high thermal conductivity, and favourable tribological properties.
This makes woven laminated CFRP gears an ideal candidate for use in complex technical
systems such as gearboxes and other power transmission components. Opportunities exist
to improve the thermomechanical and tribological properties of composite laminated gears
and achieve superior performance through further research and development.

2. Materials and Methods

2.1. Gears’ Production
For this investigation, we utilised different standard modulus PAN-based carbon fibre
(Toray Industries Inc., Tokyo, Japan) The fibre type, tex, areal density, ply thicknesses, and
mechanical properties are listed in Table 1. The fibre type T300 was used as a baseline
carbon fibre, which is commonly employed in aerospace applications with a tensile strength
of 3530 MPa and an elongation at break of 1.5%. Furthermore, a comparison with a T700S
fabric was carried out, which is a higher-quality fibre known for its strength in high-stress
applications, with a tensile strength of 4900 MPa and an elongation at break of 2.1%. In all
cases, plain weave fabrics with equal properties in warp and weft directions were used.
Table 1. Fibre type and main datasheet parameters of fabrics used in panels.

Fibre Fabrics Areal Density Ply Thickness Tensile Strength Elongation Elastic Modulus
# Tex
Type Designation [g/m2 ] [mm] [MPa] [%] [GPa]
1 T300 CC202 200 (3K) 200 0.26 3530 1.5 230
2 T700S 66090P 67 (1K) 93 0.11 4900 2.1 230
3 T300 Style 469 spread 67 (1K) 93 0.11 3530 1.5 230

Table 2 lists the epoxy resins used in the pre-impregnated (prepreg) fibres and in the
vacuum-infused plate. Prepreg fabrics were impregnated by the selected manufacturers,
whereas the vacuum infusion was performed in the authors’ laboratory. The E445 epoxy-
prepreg resin system (Composite Materials s.l.r., CIT, Legnano, Italy) is a structural epoxy,
specifically designed to achieve an excellent surface finish of the cured part. It can be
polymerised through autoclave, vacuum bag, and hot mould processes. The DT120 epoxy-
prepreg resin DeltaPreg (Delta Tech S.p.A., Toray Group, Lucca, Italy) is a toughened epoxy
from designed for autoclave processing, which is intended for applications that require
higher fracture toughness. Last, the LG 900 UV, the standard 2-component (liquid) epoxy
(GRM Systems s.r.o., Olomouc, Czech Republic), is used for wet-layup, vacuum infusion
(VI or VART), and resin infusion (RTM) processes. It is cured at room or at slightly elevated
temperature, followed by post-curing at a temperature range between 80 and 110 ◦ C.
Table 2. Datasheet parameters of epoxy resins used in panels.

Cure Temperature Glass Transition Tensile Strength Elastic Modulus E Elongation

Resin
[◦ C] T g [◦ C] [MPa] [GPa] [%]
ET445 1 80–150 135 - - -
DT120, toughened 1 100–135 110–125 - - -
LG 900 UV 80–110 60–90 82 3.4 6–6.5
1 Mechanical data for resin in prepreg are not available.
Polymers 2023, 15, 1767 6 of 24

The gears were produced from laminated panels with dimensions of 180 mm × 180 mm.
With a purpose to compare different production processes, fabrics, and matrices, three
different plates were fabricated. The summary of the plate manufacturing and stacking
sequences is given in Table 3.
Table 3. Laminated panels.

Designation Plate Thickness Stacking Total Number Manufacturing

PanelNo.
Fibre Resin [mm] Sequence of Plies Process
1 T300-ET445 2.05 [(45/0)2 ]s 8 Prepreg-AC 1
2 T700S-DT120 2.02 [(0/22.5/45/67.5)2 ]s 18 Prepreg-AC 1
3 T300-LG 900 UV 2.02 [(0/22.5/45/67.5)2 ]s 18 VI-AC 2
1 Prepreg-AC: prepreg vacuum bagging and autoclave. 2 VI-AC: vacuum infusion and autoclave.

The first panel was made using prepreg layup and vacuum bagging, followed by
autoclave consolidation and curing. It comprises 8 layers of 200 g/m2 , 3K plain weave
carbon fibre (CC202, 3K, 200 g/m2 , Torayca T300, Toray Industries, Inc., Tokyo, Japan),
pre-impregnated with resin ET445 (Composite Materials, s.r.l., CIT) with a quasi-isotropic
stacking sequence [(45/0)2 ]S which is a 8-ply laminate [45/0/45/0¦0/−45/0/−45]. In the
used designation, each number represents the longitudinal warp orientation of the ply,
e.g., 45 represents a woven ply oriented with warp (+45◦ ) and weft (−45◦ ) directions. The
laminate was covered with a peel ply and a 3 mm thick non-woven Breatex absorber to
ensure uniform pressure distribution across the entire surface. The produced panel, placed
in the autoclave with attached thermocouple to regulate the curing temperature, is shown
in Figure 2a. The pressure of 400 kPa was applied in an autoclave to consolidate the layers
and ensure proper adhesion. The heating rate was set at 3 ◦ C/min and the panel was cured
Polymers 2023, 15, x FOR PEER REVIEW 7 of 26
at a temperature of 130 ◦ C for 100 min. The mechanical characteristics of the used material
have been thoroughly documented in previous work by the authors [24].

(a) (b)
Figure 2. Production
Figure 2. Production technology
technology of
of CFRP
CFRP gears
gears (a) laminated plate
(a) laminated plate in vacuum before
in vacuum applying aa
before applying
pressure; (b) after clamping and milling operation to produce CFRP gears from laminated plate.
pressure; (b) after clamping and milling operation to produce CFRP gears from laminated plate.

The second panel

The second panel was
wasalso
alsomade
madeusing
usingthe
theautoclave
autoclaveprocess.
process.It It comprises
comprises 1818 layers
layers of
of 1K plain weave carbon fibre (6609P, 93 g/m
2 2, T700S, Toray Industries, Inc., Tokyo, Ja-
1K plain weave carbon fibre (6609P, 93 g/m , T700S, Toray Industries, Inc., Tokyo, Japan)
pan) pre-impregnated
pre-impregnated with with toughened
toughened resinresin
DT120DT120 (Delta-Tech,
(Delta-Tech, Toray
Toray Group,
Group, Lucca,Lucca, It-
Italy).
aly). The production of thin-layered fabric is slow and, therefore, more
The production of thin-layered fabric is slow and, therefore, more expensive than normal expensive than
normal 3K Itfabrics.
3K fabrics. is usedItfor
is obtaining
used for obtaining
the strongestthe and
strongest
densestand densest composites.
composites. To achieve the To
achieve the approximate
approximate thickness of thethickness
previous ofsheet,
the previous
the number sheet, the number
of layers of layers
was increased wasThis
to 18. in-
creased to
prepreg 18. This
material prepreg
is used material
for special is used for where
applications specialhigh
applications
strength in where highdirections
different strength
in
is required. To achieve strength in multiple directions, the fibre orientation changedori-
different directions is required. To achieve strength in multiple directions, the fibre by
22.5◦ increments
entation changed(90 by◦22.5° ◦ ). Therefore,
increments
/4 = 22.5 (90°/4 the
= 22.5°).
fibresTherefore,
were placedthein 0◦ , 22.5
fibres were◦ , 45 ◦ , 67.5in
placed ◦,
0°, 22.5°, 45°, 67.5°, and 90° direction. This method maximised the proportion of 0° layers
in each tooth. The curing procedure of the second panel was identical to that of the first
panel.
The third panel was made using vacuum infusion technology and subsequently
cured in the autoclave. It comprises 18 layers of 1K plain weave carbon fibre, designated
as 469 spread-tow fabrics, 67 tex (C. Cramer GmbH & Co. KG, Heek, Germany), vacuum-
Polymers 2023, 15, 1767 7 of 24

and 90◦ direction. This method maximised the proportion of 0◦ layers in each tooth. The
curing procedure of the second panel was identical to that of the first panel.
The third panel was made using vacuum infusion technology and subsequently cured
in the autoclave. It comprises 18 layers of 1K plain weave carbon fibre, designated as
469 spread-tow fabrics, 67 tex (C. Cramer GmbH & Co. KG, Heek, Germany), vacuum-
infused with laminating resin LG 900 UV with HG120 hardener (GRM Systems, Olomouc,
Czech Republic). The stacking sequence is the same as that used in the second plate. There
are several known technologies to make a spread tow. Spread-tow carbon fibre fabric is
a dry carbon fibre reinforcement made using a spreading technique to flatten the tows of
the carbon fibre, which can be accomplished using a stream of air and the pre-tension [28].
The result is a fabric that is smoother and flatter than conventional woven fabric, with less
crimp that occurs at the intersection of warp and weft: Figure 3. Spread-tow fabrics are
often used to create thin-ply laminates with superior mechanical properties compared to
standard-ply laminates [29]. In thin-ply laminates, more layers per unit thickness allows
for higher variation in ply angle orientations and, therefore, could potentially improve
load-bearing properties. The epoxy LG 900 UV laminating system includes UV inhibitors,
but, unlike the epoxy used in panel 2, it is not toughened. After vacuum infusion, the panel
was
Polymers 2023, 15, x FOR PEER REVIEW
placed into the autoclave, where the pressure of 400 kPa was applied. The cure8cycleof 26
for this resin was 15 h at 40 ◦ C for initial polymerisation, then a ramp of 0.5 ◦ C/min to
80 ◦ C for 3 h, and then a ramp of 0.25 ◦ C/min to 110 ◦ C for 5 h.

Figure3.
Figure 3. Difference
Difference between
between standard
standardand
andspread-tow
spread-towfabric
fabricafter
afterspread
spreadtechnique
technique[28].
[28].

The CFRP
The CFRP gears
gears were
were milled
milled from
frompanels
panelsusing
using aa Sodick
Sodick CNC
CNC machine.
machine. Central
Central holes
holes
(6mm)
(6 mm)were
weremilled
milledfirst;
first;then,
then,the the CFRP
CFRP panels
panels were
were bolted
bolted to an
to an aluminium
aluminium basebase plate
plate for
high-quality geargear
for high-quality edgeedge
cutscuts
as shown
as shown in Figure 2b. 2b.
in Figure
The
Thesteel
steelgears
gearswere milled
were from from
milled tempered steel ENsteel
tempered 42CrMo4 (W.No. 1.7225,
EN 42CrMo4 (W.No. AISI/SAE
1.7225,
4142) which
AISI/SAE haswhich
4142) a chemical
has a composition of 0.41% of 0.41%
chemical composition C, 0.2%ofofC,Si,
0.2%0.75%
of Si,of0.75%
Mn, 01.05%
of Mn,
of Cr and 0.23% of Mo, and then plasma-nitrided. They were treated
01.05% of Cr and 0.23% of Mo, and then plasma-nitrided. They were treated with a super- with a superfinish
to smooth
finish surfaces
to smooth and remove
surfaces and remove sharpsharp
edges. TheThe
edges. flank profiles
flank profileswere measured
were measured with
witha
MarSurf
a MarSurf XC20
XC20conturograph
conturograph andandthe the
surface roughness
surface roughnesswas was
measured
measuredbefore the test
before thewith
test
awith
TesaaRugosurf 90G gauge.
Tesa Rugosurf 90G gauge.The surface roughness
The surface of theofsuperfinished
roughness the superfinished steel steel
gearsgears
was
Ra
was= Ra
0.689 µm and
= 0.689 μm the
andsurface
the surfaceroughness of theofCFRP
roughness gearsgears
the CFRP was Ra was = Ra
0.417 µm. μm.
= 0.417
The
The surface hardness of the steel gears was measured using a standard method
surface hardness of the steel gears was measured using a standard method and
and
an average value of 870
an average value of 870 HV0.2 0.2HV was obtained. A cylindrical spur gear
was obtained. A cylindrical spur gear geometry with param- geometry with
parameters
eters defineddefined in Table
in Table 4 was4usedwas usedin theinmanufacturing
the manufacturing process.
process. AfterAfter manufacturing,
manufacturing, the
the geometry of the carbon fibre reinforced plastic (CFRP) gears
geometry of the carbon fibre reinforced plastic (CFRP) gears was measured with an was measured with
ATOSan
ATOS Compact SCAN 5 M scanner with a nominal accuracy of
Compact SCAN 5 M scanner with a nominal accuracy of ±2 μm. The geometric quality of ± 2 µm. The geometric
quality of was
the gears the gears was in
evaluated evaluated
accordance in accordance
with ISO 1328 with ISO a1328
using using a self-developed
self-developed software for
software
quality control of gears. The results indicated that the pitch quality wasquality
for quality control of gears. The results indicated that the pitch level 7,was
the level
profile7,
the profile quality was level 9, the pitch lead profile quality was level
quality was level 9, the pitch lead profile quality was level 9, and the runout quality was 9, and the runout
quality
level 6.was level 6. Representative
Representative results fromresults from the carried-out
the carried-out gear inspection,
gear inspection, following following
the meth-
the methodology described in Ref. [30], are presented
odology described in Ref. [30], are presented in Figure 4. in Figure 4.
 The steel gears were milled from tempered steel EN 42CrMo4 (W.No. 1.7225,
AISI/SAE 4142) which has a chemical composition of 0.41% of C, 0.2% of Si, 0.75% of Mn,
01.05% of Cr and 0.23% of Mo, and then plasma-nitrided. They were treated with a super-
finish to smooth surfaces and remove sharp edges. The flank profiles were measured with
a MarSurf XC20 conturograph and the surface roughness was measured before the test
Polymers 2023, 15, 1767 8 of 24
with a Tesa Rugosurf 90G gauge. The surface roughness of the superfinished steel gears
was Ra = 0.689 μm and the surface roughness of the CFRP gears was Ra = 0.417 μm.
The surface hardness of the steel gears was measured using a standard method and
Table
an 4. Geometrical
average value of 870parameters of the
HV0.2 was test gears.
obtained. A cylindrical spur gear geometry with param-
eters defined in Table 4 was used in the manufacturing process. After manufacturing, the
Parameter Symbol Value
geometry of the carbon fibre reinforced plastic (CFRP) gears was measured with an ATOS
Profile SCAN 5 M scanner with a nominal
Compact - Involute
accuracy of ±2 μm. The ISO 53Aquality of
geometric
Module m [mm] 1
theNumber
gears was evaluated in accordanceZwith ISO 1328 using a self-developed
of teeth 20
software for
quality control
Pressure angle of gears. The results indicated
◦
α[ ] that the pitch quality
20 was level 7, the profile
quality was level 9, the pitch
Profile shift coefficient—pinion lead profile
x1 quality was level 9, and
0 the runout quality was
Profile shift coefficient—gear x 0
level 6. Representative results from the2 carried-out gear inspection, following the meth-
Transverse
odology contact in
described ratio εα
Ref. [30], are presented in Figure 4. 1.557

Polymers 2023, 15, x FOR PEER REVIEW 9 of 26

(a) (b)

(c) (d)
Figure
Figure4.4.Exemplary
Exemplaryresults
resultsdepicting
depictingthe
theevaluation
evaluationofofgear
gearquality
qualityparameters
parametersaccording
accordingtotoISO
ISO
1328 based on 3D-scanner measurements. (a) cumulative pitch deviation; (b) profile deviations;
1328 based on 3D-scanner measurements. (a) cumulative pitch deviation; (b) profile deviations; (c)
lead profile deviations; (d) gear runout.
(c) lead profile deviations; (d) gear runout.
Table 4. Geometrical parameters
2.2. Thermomechanical of the test
and Frictional gears.
Characteristics of the ET445 CIT Composite
The T300-ET445-200tex plain weave prepreg
Parameter was the benchmarkValue
Symbol composite; therefore.
it has been thoroughly characterised in terms -of its mechanical and
Profile thermal
Involute properties.
ISO 53A
Initially, the mechanical properties in quasi-static load conditions have been evaluated,
Module m [mm] 1
using the procedures thoroughly described in Ref. [24]. The results are shown in Figure 5.
Number of teeth Z 20
As is common for this type of composite, highly variable mechanical properties have been
Pressure angle α [°] 20
identified, depending on the fibre layup configuration and load direction. On the other
Profile shift coefficient—pinion x 1 0
hand, a fairly linear stress–strain relation has been identified for all tested load conditions
Profile shift coefficient—gear
and composite x2
configurations. The measured mechanical 0 the [(45/0) ]s layup
properties of 2
Transverse contact
configuration, whichratio εα
is the one used for the carried-out 1.557 and testing, are
gear production
noted in a more complete form in Table 5.
 ated, using the procedures thoroughly described in Ref.[24]. The results are shown in Fig-
ure 5. As is common for this type of composite, highly variable mechanical properties have
been identified, depending on the fibre layup configuration and load direction. On the
other hand, a fairly linear stress–strain relation has been identified for all tested load con-
ditions and composite configurations. The measured mechanical properties of the
Polymers 2023, 15, 1767 [(45/0)2]s layup configuration, which is the one used for the carried-out gear production 9 of 24
and testing, are noted in a more complete form in Table 5.

Figure 5.
Figure 5. Quasi-static mechanical
Quasi-static characterisation
mechanical resultsresults
characterisation for thefor
produced ET 445 CIT
the produced ET autoclave
445 CIT
prepreg.
autoclave prepreg.

Table5.
Table 5. Mechanical
Mechanicalproperties
propertiesof
ofET
ET445
445CIT
CITprepreg
prepregwith
with[(45/0)
[(45/0)22]]ss configuration and 48%
48% of
of fibre
fibre
volume ratio.
volume ratio.
Mechanical
MechanicalProperty
Property Value
Value Description
Description
E1 = E2 [GPa] 46.9 Axial, transverse in-plane stiffness
E = E2 [GPa] 46.9 Axial, transverse in-plane stiffness
E3 1[GPa] 6 Transverse out-of-plane stiffness
E3 [GPa] 6 Transverse out-of-plane stiffness
𝑛n 12 0.08
0.08 Major Poisson ratio ratio
Major Poisson
12
𝑛n23 = 𝑛 0.3 Minor Poisson ratio ratio
23 = n13 13 0.3 Minor Poisson
𝐺G [GPa]
1212 [GPa] 17
17 In-planeIn-plane
shear stiffness
shear stiffness
𝐺G 2323== 𝐺G1313 [GPa] 3.37
3.37 Out-of-plane shear stiffness
Out-of-plane shear stiffness
𝑇T = 𝑇 T
𝑋1 1= 𝑋2X[MPa]
X 2 [MPa] 292
292 Axial/y-transverse
Axial/y-transverse tensile strength
tensile strength
𝑇 T [MPa] 50
X
𝑋3 3[MPa] 50 Transverse z-tensile strengthstrength
Transverse z-tensile
𝐶 C = 𝐶X C [MPa] −217 Axial/transverse compression
𝑋X 1 1= 𝑋2 [MPa]
2 −217 Axial/transverse compression strength strength
𝐶 C [Mpa] −80
𝑋X 3 3[Mpa] −80 Transverse compressive strengthstrength
Transverse compressive
T T 0.0095 Axial/y-transverse tensile strain limit
𝑒1e𝑇1 ==𝑒2e𝑇2 0.0095 Axial/y-transverse tensile strain limit
T 0.003 Transverse tensile z-strain limit
𝑒3e𝑇3C 0.003 Transverse tensile z-strain limit
e = eC −0.011 Axial/y-transverse compressive strain limit
𝑒1𝐶1C= 𝑒2𝐶2 −0.011 Axial/y-transverse compressive strain limit
e −0.011 Transverse compressive z-strain limit
𝑒3𝐶3 −0.011 Transverse compressive z-strain limit
g12 = g13 0.019 xy and xz shear strain limit
𝑔12 = 𝑔13 0.019 xy and xz
g23 0.014 yzshear
shearstrain
strainlimit
limit
𝑔T 23g [◦ C] 0.014
149 yz shearGlass
straintransition
limit temperature
Tρg [°C]
[g/cm3 ] 149
1.464 Glass transition
Density temperature
ρ [g/cm3] 1.464 Density
The thermal characteristics of the [(45/0)2 ]s ET445 CIT prepreg have additionally been
analysed using a TPS 1500 Hot Disk Thermal Constants Analyser. In the in-plane fabric
direction, a thermal conductivity of 2.416 W/(mK) was measured, along with 2138 J/(kgK)
specific heat capacity. In the traverse, out-of-plane direction, these properties changed to
1.446 W/(mK) and 4106 J/(kgK), respectively, pointing to an orthotropic thermal behaviour
as well.
A key property that can influence the performance and efficiency of a gear pair is the
coefficient of friction (COF). The tribological properties of the [(45/0)2 ]s ET445 CIT prepreg
CFRP have been assessed using a reciprocating cylinder-on-flat tribological test setup, with
a 42CrMo4 cylinder sliding on the flat side surface of the CFRP. Using this type of test
setup, an average COF of 0.34 has been identified, which is comparably slightly higher
than typical values observed for steel thermoplastic pairs such as steel-POM or steel-PA6.
The measured COF value has also been confirmed by applying the implicit gear-pair COF
identification model presented in the previous study [22].
 with a 42CrMo4 cylinder sliding on the flat side surface of the CFRP. Using this type of
test setup, an average COF of 0.34 has been identified, which is comparably slightly higher
than typical values observed for steel thermoplastic pairs such as steel-POM or steel-PA6.
The measured COF value has also been confirmed by applying the implicit gear-pair COF
Polymers 2023, 15, 1767 identification model presented in the previous study [22]. 10 of 24

2.3. Gear Testing Methodology

GearTesting
2.3. Gear tests have been carried out to measure the developed gears’ service life as a
Methodology
function of load and identify
Gear tests have been carried and characterise
out to measurethe predominant
the developed failure modes
gears’ at the
service con-
life as
sidered
a functionloads. The used
of load test rig consists
and identify of two electric
and characterise motors, one failure
the predominant of which serves
modes atasthea
drive motor and the other as a brake electric motor. The test stand allows
considered loads. The used test rig consists of two electric motors, one of which serves asthe torque to be
continuously
a drive motoradjusted via a drive
and the other motorelectric
as a brake and a brake
motor.motor which
The test standareallows
controlled via vari-
the torque to
able frequency drives. The drive motor and the brake motor are four-pole
be continuously adjusted via a drive motor and a brake motor which are controlled via asynchronous
electric motors
variable frequency(Siemens
drives.AG,
TheMunich, Germany)
drive motor and thewith
brakea nominal
motor aretorque of 2.5
four-pole Nm. Power
asynchronous
is transmitted
electric motorsfrom the drive
(Siemens electric motor
AG, Munich, to the
Germany) drive
with shaft oftorque
a nominal the gearbox via aPower
of 2.5 Nm. toothed is
belt that dampens the vibrations of the electric motor. The testing rig has
transmitted from the drive electric motor to the drive shaft of the gearbox via a toothed beltFutek TRS600
torque sensorsthe
that dampens and speed sensors
vibrations of the mounted on theThe
electric motor. drive and rig
testing driven shafts,TRS600
has Futek enabling con-
torque
tinuous torque measurement and closed-loop control. The test rigs
sensors and speed sensors mounted on the drive and driven shafts, enabling continuouswere positioned in a
thermal chamber where
torque measurement andthe ambient temperature
closed-loop and rigs
control. The test humidity were controlled.
were positioned More
in a thermal
details
chamberabout
wherethethetesting
ambient unit can be found
temperature in previous
and humidity werestudies [31,32].
controlled. MoreThedetails
CFRPaboutgear
geometry
the testingandunitdimensions
can be found in ainmesh with driving
previous 42CrMo4
studies [31,32]. Thegear are presented
CFRP gear geometryin Figure
and
6.
dimensions in a mesh with driving 42CrMo4 gear are presented in Figure 6.

Figure 6. Experimental
Figure 6. Experimental set-up
set-up for
for reference gear pair
reference gear pair testing.
testing. (a)
(a) testing
testing rig;
rig; (b) CFRP gear
(b) CFRP gear in
in mesh
mesh
with steel pinion.
with steel pinion.

2.4. Scanning Electron Microscopy (SEM)

The surfaces
surfaces of
of the
theworn
worngears
gearswere
wereexamined
examinedusing
usinga aThermo
ThermoFisher
FisherScientific
Scientific Quat-
Quattro
tro S scanning
S scanning electron
electron microscope
microscope (SEM)(SEM) equipped
equipped with
with an an Everhart–Thornley
Everhart–Thornley detector
detector (ETD)
(ETD) and a circular backscatter detector (CBS). SEM images of the carbon fibre reinforced
and a circular backscatter detector (CBS). SEM images of the carbon fibre reinforced plastic
(CFRP) gear surfaces were taken before and after the tests. The elemental composition of
the worn CFRP surface was investigated using energy-dispersive X-ray spectroscopy (EDS),
and elemental mapping was also carried out to investigate the distribution of elements on
the surface.

2.5. Temperature Measurements

During gear running, a certain temperature rise on the meshing contact interface (flash
temperature) and the gear body (bulk temperature) is always present due to frictional
and hysteresis losses at the contact and in the material structures. An overly elevated
temperature during gear running is generally a strong indicator of unfavourable frictional
characteristics of the used gear pair, or unsuitable thermal conductivity or heat transfer via
the surrounding medium. The bulk temperature of the tested CFRP gear was measured
using a FLIR T 420 thermal imaging camera. The measurement was taken in an area
of 3 × 3 pixels (ROI) in the root area of the gear. To minimise infrared reflection (IR)
from the background surfaces, the surfaces behind the gears were coated with a layer
of Acktar Metal Velvet TM. The camera was calibrated by initially setting the emissivity
to ε = 0.95 and subsequently recalibrating the measurements based on the evaluated
 transfer via the surrounding medium. The bulk temperature of the tested CFRP gear was
measured using a FLIR T 420 thermal imaging camera. The measurement was taken in an
area of 3 × 3 pixels (ROI) in the root area of the gear. To minimise infrared reflection (IR)
from the background surfaces, the surfaces behind the gears were coated with a layer of
Polymers 2023, 15, 1767 Acktar Metal Velvet TM. The camera was calibrated by initially setting the emissivity 11 of 24 to ε
= 0.95 and subsequently recalibrating the measurements based on the evaluated mean
emissivity. The emissivity of the analysed CFRP sample in the temperature range between
30 andemissivity.
mean 72 °C wasThe measured
emissivityto be, on analysed
of the average, εCFRP
t = 0.935; however,
sample a temperature-depend-
in the temperature range
ent behaviour
between 30 andwas ◦
72 C observed with the
was measured emissivity
to be, values
on average, εt = approximately
0.935; however, afollowing the func-
temperature-
tion εt(T) =behaviour
dependent −0.0006 was
× T observed
+ 1.1268 with
(withthetemperature in K).
emissivity values The measurement
approximately followingresults
the are
function (T) = − 0.0006 × T + 1.1268 (with temperature in K). The measurement
shown in Figure 7. The temperature is plotted as the increase above the ambient temper-
ε t results
are shown
ature, which in was
Figure
22 ±7.2 °C
Theduring
temperature is plotted
all carried-out as the
tests. increase for
Evidently, above
the the ambient
selected gear pair
temperature, which was 22 ± 2 ◦ C during all carried-out tests. Evidently, for the selected
configuration and loads, the heat losses are not overly high and do not result in tempera-
gear pair configuration and loads, the heat losses are not overly high and do not result in
tures that could lead to any thermally induced damage mechanisms.
temperatures that could lead to any thermally induced damage mechanisms.

M = 0.4 Nm M = 0.5 Nm
M = 0.6 Nm M = 0.7 Nm
25

20

Temperature rise [°C]

15

10

0
0 1000 2000 3000 4000
Time [s]

Figure7.7.Temperature
Figure Temperature rise
rise above
above ambient
ambient temperature
temperature 2 ◦±C),
(22 ±(22 2 °C), measured
measured belowbelow the root
the tooth tooth root
usingaathermographic
using thermographic camera
camera during
during thethe carried-out
carried-out geargear
tests.tests.

2.6. Wear Measurement

2.6. Wear Measurement
For the wear monitoring tests, the CFRP gears were tested over a period of five
For the wear monitoring tests, the CFRP gears were tested over a period of five inter-
intervals, each lasting 24 h (for a total of 120 h). At each interval, the wear was measured
usingeach
vals, imagelasting 24 h
analysis of(for
the a totalsurfaces.
worn of 120 h). At images
The each interval, the wear
for the wear was were
analysis measured
taken using
imageaanalysis
using Keyence of the worn
VHX-6000 3Dsurfaces. The images
digital microscope at for
50×the wear analysis
magnification. Thewere taken
images using a
were
Keyence
then VHX-6000
imported 3D digitalsoftware
into AutoCAD microscope at 50× magnification.
for analysis. The images
The original image of the were
unwornthen im-
ported
gears into
was AutoCAD
compared software
with the imagefor of
analysis.
the worn The original
gears image of
to determine thethe unworn
amount gears was
of wear.
compared
The with the image
wear measurement of theon
was taken worn
everygears
othertotooth,
determine the of
for a total amount
10 teethofofwear. The wear
each gear.
measurement was taken on every other tooth, for a total of 10 teeth of each gear.
2.7. Numerical Analyses
To analyse the impact of wear on the gear-meshing kinematics and mechanical be-
haviour of the material pair, numerical finite element analyses were performed using both
the theoretical geometry and the worn gear geometries, obtained from microscopy measure-
ments of the tested samples. Samples that had run at loads between 0.4 Nm and 0.7 Nm
for 120 h were used. Figure 8 shows the developed finite element model and the gear
geometries used. The model, built in ANSYS Workbench 2022 R1, employs 2D plane-stress
and quasi-static mechanical behaviour assumptions, and enables a time-dependent contact
analysis of the entire gear-meshing process, using isotropic/orthotropic linear elastic mod-
els for steel and CFRP, respectively (CFRP orthotropic parameters based on data given in
Table 3), along with geometric and contact nonlinear mechanical models and a specified
COF of 0.34 [22].
 Nm for 120 h were used. Figure 8 shows the developed finite element model and the gear
geometries used. The model, built in ANSYS Workbench 2022 R1, employs 2D plane-
stress and quasi-static mechanical behaviour assumptions, and enables a time-dependent
contact analysis of the entire gear-meshing process, using isotropic/orthotropic linear elas-
tic models for steel and CFRP, respectively (CFRP orthotropic parameters based on data
Polymers 2023, 15, 1767 12 of 24
given in Table 3), along with geometric and contact nonlinear mechanical models and a
specified COF of 0.34 [22].

(a) (b)
Figure
Figure 8.
8. Constructed finite element
Constructed finite elementmodel
modelwith
withdepicted
depicted FEM
FEM mesh
mesh and
and applied
applied boundary
boundary condi-
conditions
tions
(BCs) (BCs) and loads.
and loads. (a) FEM(a)model
FEM geometry,
model geometry, mesh,
mesh, BCs BCs and
and loads; (b) loads; (b) considered
considered gear profilegear profile
geometries.
geometries.
3. Results
3. Results
3.1. Wear Characteristics of CFRP Gears
3.1. Wear
The Characteristics
primary causeofofCFRP gear Gears
wear is the abrasion resulting from the friction generated
by the meshing
The primary of cause
the gears. Thewear
of gear meshing
is thecycle for a selected
abrasion resultingpair
from of the
teeth goes from
friction points
generated
A to E (Figure 9). Point A is the initial meshing point; this is point A1
by the meshing of the gears. The meshing cycle for a selected pair of teeth goes from points on the drive gear’s
Aflank
to Eand A2 on
(Figure 9).the driven
Point A isgear’s flank.
the initial Point Bpoint;
meshing is the this
lowest pointA1
is point of single
on thetoothdrivecontact
gear’s
(LPSTC), and point D is the highest point of single tooth contact
flank and A2 on the driven gear’s flank. Point B is the lowest point of single tooth (HPSTC) for the contact
driving
gear. For and
(LPSTC), the driven
point Dgear, the
is the situation
highest is vice
point versa;
of single point
tooth D is the
contact LPSTC,for
(HPSTC) and thepoint B is
driving
the HPSTC.
gear. For theWhen
driven gears
gear,mesh from points
the situation B toversa;
is vice D, thepoint
entireDload is transmitted
is the LPSTC, andvia a single
point B is
tooth pair. When meshing in the regions from A to B and D to E, the load
the HPSTC. When gears mesh from points B to D, the entire load is transmitted via a single is transmitted over
two pairs
tooth pair.ofWhen
teeth.meshing
This phenomenon
in the regionsis known
from A as to
load sharing
B and D to(Figure 6) and
E, the load is extremely
is transmitted
over two pairs of teeth. This phenomenon is known as load sharing (Figure 6)endpoint
important for understanding the load on a single tooth. Finally, point E is the and is ex- of
contact. Spur gears in operation are subjected to torque, which results
tremely important for understanding the load on a single tooth. Finally, point E is the in a normal force F nY
acting in an
endpoint of arbitrary meshing
contact. Spur gearspoint Y between
in operation arethe two teethtointorque,
subjected contact.whichThe normal
results force
in a
Polymers 2023, 15, x FOR PEER REVIEW
FnY canforce
be decomposed into FrY meshing FtY . Inthe the 14 of 26
normal FnY acting in an radial
arbitrary and tangential
point Y force
between involute
two teeth gearinpairs,
contact.
normal force acts along the path of contact. The magnitude of torque applied was selected
The normal force FnY can be decomposed into radial FrY and tangential force FtY. In involute
on the fact that polymer gears have maximum load per gear width of 20 N per 1 mm.
gear pairs, the normal force acts along the path of contact. The magnitude of torque ap-
plied was selected on the fact that polymer gears have maximum load per gear width of
20 N per 1 mm.

(a) (b)
Figure
Figure 9.9. Meshing:
Meshing:(a)(a)theoretical
theoretical meshing
meshing of tested
of the the tested geargeometry;
gear pair pair geometry; (b) theoretical
(b) theoretical load load
sharing for the tested gear geometry.
sharing for the tested gear geometry.

In our previous studies [15,16], the experimental region to test gears was determined
for CFRP ET445 CIT with plain weave carbon fabrics, with an areal mass of 200 g/m². The
upper limit of the torque applied was found to be 0.8 Nm, beyond which failure occurred
almost immediately. For comparison, the maximum transmittable torque for steel gears
 (a) (b)
Polymers 2023, 15, 1767 13 of
Figure 9. Meshing: (a) theoretical meshing of the tested gear pair geometry; (b) theoretical 24
load
sharing for the tested gear geometry.

Inour
In ourprevious
previousstudies
studies[15,16],
[15,16],thetheexperimental
experimentalregion
regionto totest
testgears
gearswaswas determined
determined
forCFRP
for CFRPET445
ET445CIT CITwith
withplain
plainweave
weavecarbon
carbonfabrics,
fabrics,with
withan anareal
arealmass
massofof200200g/m
g/m². The
2 . The
upper limit of the torque applied was found to be 0.8 Nm, beyond which
upper limit of the torque applied was found to be 0.8 Nm, beyond which failure occurred failure occurred
almostimmediately.
almost immediately.For Forcomparison,
comparison,the themaximum
maximum transmittable
transmittable torque
torque forfor steel
steel gearsgears
of
the same size ranges from 1.7 Nm for through-hardened and tempered C45 steel to 3.5toNm
of the same size ranges from 1.7 Nm for through-hardened and tempered C45 steel 3.5
Nmcase-hardened
for for case-hardened 18CrNiMo7-6
18CrNiMo7-6 steel, steel,
with awith a service
service life oflife
107ofload
107 cycles.
load cycles.
Moreover, the wear in spur gears is caused by the
Moreover, the wear in spur gears is caused by the combination of combination of sliding
sliding and
and rolling
rolling
contact. The sliding is most severe at the tip and just above the root of the
contact. The sliding is most severe at the tip and just above the root of the CFRP gear tooth,CFRP gear tooth,
gradually decreasing until only rolling occurs at pitch point C. Pitch
gradually decreasing until only rolling occurs at pitch point C. Pitch point C marks the point C marks the
threshold where the direction of sliding changes as the meshing moves
threshold where the direction of sliding changes as the meshing moves past this point, as past this point, as
illustratedin
illustrated inFigure
Figure10.10.

Figure 10. Meshing (a) sliding and rolling; (b) only rolling; (c) rolling and reversed sliding.

As10.
Figure shown in (a)
Meshing Figure 11,and
sliding wear results
rolling; in anrolling;
(b) only increase in gear
(c) rolling andbacklash, i.e., the clear-
reversed sliding.
ance between the passive tooth flanks of the two meshing gears. It can also significantly
impact Asthe meshing
shown kinematics
in Figure 11, wearand the resulting
results strain-stress
in an increase state in the
in gear backlash, gear
i.e., thestructure.
clearance
Furthermore, the surface
between the passive toothply-delamination is visiblegears.
flanks of the two meshing on the CFRP
It can edge
also of the working
significantly impact
flank. Figure 12
the meshing displays and
kinematics a series
the of images strain-stress
resulting taken after 24state
h intervals during
in the gear testing,Further-
structure. clearly
Polymers 2023, 15, x FOR PEER REVIEW demonstrating theply-delamination
more, the surface gradual degradation of the
is visible onmaterial.
the CFRPThe edgeimages
of thereveal
working typical
flank.wear
15 of 26
Fig-
patterns on the tooth flanks, including the groove at the pitch diameter, increased
ure 12 displays a series of images taken after 24 h intervals during testing, clearly demon- surface
roughness,
strating theedge delaminations,
gradual degradationand extruded/intruded
of the material. The imagesply interfaces.
reveal typical wear patterns
on the tooth flanks, including the groove at the pitch diameter, increased surface rough-
ness, edge delaminations, and extruded/intruded ply interfaces.

(a) (b)
Figure 11.11.
Figure Meshing
Meshingofofthe
thetesting
testing gears, steeldriving
gears, steel driving gear
gear with
with CFP-driven
CFP-driven geargear (a) CFRP
(a) new new CFRP
gear gear
in meshing; (b) worn CFRP gear after 120 h of operation (1 × 10
7 7 cycles) at torque of 0.4 Nm.
in meshing; (b) worn CFRP gear after 120 h of operation (1 × 10 cycles) at torque of 0.4 Nm.
 (a) (b)
Polymers 2023, 15, 1767 14 of 24
Figure 11. Meshing of the testing gears, steel driving gear with CFP-driven gear (a) new CFRP gear
in meshing; (b) worn CFRP gear after 120 h of operation (1 × 107 cycles) at torque of 0.4 Nm.

Figure 12. Wear degradation

Figure 12. of gears
Wear of sampleofCFRP
degradation gear,
gears of tested
sample at torque
CFRP of 0.4
gear, tested at Nm.
torque of 0.4 Nm.

Figure 13 shows the Figure 13 shows the

relationship relationship
between wear between
from thewear
pitch from
linethe pitch
and theline and the number
number
of cycles loading
of cycles for four different for four different
conditions. loading
Eachconditions.
datum point Each datum
is the pointwear
average is thevol-
average wear
volume on 20 gear teeth, with error bars representing 95% confidence intervals. The gears
ume on 20 gear teeth, with error7bars representing 95% confidence intervals. The gears
were tested to 10 cycles, the expected lifespan for common polymer and composite gear
Polymers 2023, were tested
15, x FOR PEERtoREVIEW
107 applications.
cycles, the expected lifespan for common polymer and composite gear
Wear at the pitch diameter is the normal distance from the worn to the unworn 16
applications. Weartooth
at the pitch diameter is the
profile. There is a clear normal distance
correlation fromrate,
between wear the load,
wornandto the un-time. Wear
testing
worn tooth profile.fromThere is diameter
pitch a clear correlation between
increased linearly wear
with rate, load,
increasing cyclesand
andtesting
torque.time.
Wear from pitch diameter increased linearly with increasing cycles and torque.

Figure Wear
Figure13.13. from
Wear pitchpitch
from diameter againstagainst
diameter the number of cycles. of cycles.
the number
3.2. Numerical Analysis
3.2. The
Numerical Analysis
results of the numerical analysis are shown in Figure 14.
The results of the numerical analysis are shown in Figure 14.

1250 120
stress [MPa]
stress [MPa]

1000
80
750
 Figure 13. Wear from pitch diameter against the number of cycles.

Polymers 2023, 15, 1767 15 of 24

3.2. Numerical Analysis
The results of the numerical analysis are shown in Figure 14.

1250 120

Max. principal root stress [MPa]

Peak contact stress [MPa]

1000
80
750

500
40
250

0 0
0 510 15 20 25 30 0 5 10 15 20 25 30
Pinion rotation angle [°] Pinion rotation angle [°]
M = 0.4 Nm M = 0.5 Nm M = 0.4 Nm M = 0.5 Nm
M = 0.6 Nm M = 0.7 Nm M = 0.6 Nm M = 0.7 Nm
M = 0.4 Nm - Th. Geom. M = 0.4 Nm - Th. Geom.
(a) (b)
350
0.8
Max./min.shear cont. stress [MPa]

250

Transmission error [°]

0.6
150

50 0.4
-50
0.2
-150

-250 0
0 5 10 15 20 25 30 0 4 8 12 16 20 24 28 32 36
Pinion rotation angle [°] Pinion rotation angle [°]
M = 0.4 Nm M = 0.5 Nm
M = 0.4 Nm M = 0.5 Nm M = 0.6 Nm M = 0.7 Nm
M = 0.6 Nm M = 0.7 Nm M = 0.4 Nm - Th. Geom.
M = 0.4 Nm - Th. Geom.
(c) (d)
Figure 14. Main FEM analysis results obtained for the theoretical gearing geometry and worn sam-
Figure 14. Main FEM analysis results obtained for the theoretical gearing geometry and worn sample
ple geometries at various loads after 120 h of testing, recreated based on microscopy measurements.
geometries at various
(a) Peak contact loads
pressure; (b)after
Peak 120
root h of testing,
stress on activerecreated
flank side;based
(c) Peakon microscopy
shear measurements.
stress below contact;
(a) Peak contact pressure;
(d) Transmission error. (b) Peak root stress on active flank side; (c) Peak shear stress below contact;
(d) Transmission error.

Evidently, wear has a significant impact on the stresses during gear meshing. Com-
paring the peak contact stress (Figure 14a) between theoretical and worn geometry at
0.4 Nm, we can see that the peak pressure noticeably increased, especially in the second
phase of meshing, i.e., after the pitch point contact is reached. A graphical presentation
of the pitch point contact pressure and overall peak-evaluated pressure for the 0.4 Nm
worn gear geometry load case is additionally shown in Figure 15. The described pattern
is exhibited, and is even more pronounced at higher loads, where the wear level was
higher. A comparison of the root stress results between the theoretical and worn geometries
(Figure 14b) reveals a slight decrease in stress on the worn geometry compared to the
theoretical geometry, with the stress pattern altered and exhibiting stress spikes at the start
and end of the meshing cycle.
Similar to the contact stress, the peak shear stress below the contact area (Figure 14c)
shows an increased level in the second half of the meshing cycle, where the magnitudes
exceed the theoretical geometry values. For a clearer overview, the root and shear stresses
at approximately the pitch point contact C are depicted in Figure 16a,b respectively.
Figure 14d also shows that the transmission error increases substantially compared
to the theoretical profile geometry, even for the lowest wear level, as measured for the
0.4 Nm load case. The error, however, does not increase linearly with the load and wear
levels, pointing to the necessity for a more in-depth investigation of this correlation. The
presented results also point to the fact that wear noticeably decreases the overall length of
the meshing cycle. For the 0.4 Nm load, the duration of the meshing cycle is, e.g., reduced
by 29.3% compared to the theoretical geometry.
 pitch
ical point contact
geometry, pressure
with the stress and overall
pattern peak-evaluated
altered pressure
and exhibiting for theat0.4
stress spikes theNm
startworn
and
gear geometry load case
end of the meshing cycle. is additionally shown in Figure 15. The described pattern is ex-
hibited, and is even more pronounced at higher loads, where the wear level was higher.
A comparison of the root stress results between the theoretical and worn geometries (Fig-
ure 14b) reveals a slight decrease in stress on the worn geometry compared to the theoret-
Polymers 2023, 15, 1767 16 of 24
ical geometry, with the stress pattern altered and exhibiting stress spikes at the start and
end of the meshing cycle.

(a) (b)
Figure 15. Contact pressure evaluated for the 0.4 Nm load worn geometry (after 120 h running) at
(a) approximately the pitch diameter contact C, and (b) at the location of peak pressure during the
simulated meshing cycle.

(a)
Similar to the contact stress, the peak shear stress (b)below the contact area (Figure 14c)
shows
Figure an
Figure15. increased
15.Contact level in
Contact pressure
pressure the second
evaluated
evaluated for the
for half
the of
0.4Nm
0.4 Nmthe meshing
load
load cycle, (after
worngeometry
worn geometry where
(after
120the
120 magnitudes
h running)
h running) at at
(a)
(a) approximately the pitch diameter contact C, and (b) at the location of peak pressure during the the
approximately
exceed the the
theoreticalpitch diameter
geometry contact
values. C,
For and
a (b)
clearerat the location
overview, of
the peak
root pressure
and shearduring
stresses
simulated
atsimulated meshing
approximately cycle.
meshingthe pitch point contact C are depicted in Figure 16a,b respectively.
cycle.

Similar to the contact stress, the peak shear stress below the contact area (Figure 14c)
shows an increased level in the second half of the meshing cycle, where the magnitudes
exceed the theoretical geometry values. For a clearer overview, the root and shear stresses
at approximately the pitch point contact C are depicted in Figure 16a,b respectively.

(a) (b)
Figure
Figure16.
16.FEM
FEM results atat
results 0.40.4
Nm torque,
Nm (a)(a)
torque, thethe
root stress
root onon
stress thethe
active flank
active side
flank evaluated
side as as
evaluated max.
principal
max. principal stress, and (b) the shear stress at and below the contact area. Both results are plottedap-
stress, and (b) the shear stress at and below the contact area. Both results are plotted
proximately at the
ap-proximately at pitch diameter
the pitch point
diameter C. C.
point
(a) (b)
The numerical simulations with the worn gear geometry also showed that the path of
contact
Figure 16.diverges substantially
FEM results from the
at 0.4 Nm torque, theoretically
(a) the root stress predicted linear.
on the active flankA transition
side evaluatedofas
the
max.
contact location
principal was
stress, and (b)observed
the shearfrom
stressthe areabelow
at and near the
the pitch diameter
contact to the
area. Both bottom
results part ofap-
are plotted
proximately at theprofile
the CFRP gear pitch diameter
(near thepoint
baseC.diameter) in the second half of the meshing phase.
This caused a sudden change in the contact path and a significant increase in the contact
pressure on the bottom part of the CFRP tooth flank, which may further speed up the wear
process. During the described transition of the contact location, a short period of double
contact near the pitch and base diameters was observed, which is shown in Figure 17. This
behaviour was observed in all four analysed worn gear geometries, with loads ranging
from 0.4 Nm to 0.7 Nm. The wear level can result in detrimental impacts on the contact
stress state, negatively affecting the performance of the gears. This can also lead to an
increase in noise and vibrations.
 pressure on the bottom part of the CFRP tooth flank, which may further speed up the wear
process. During the described transition of the contact location, a short period of double
contact near the pitch and base diameters was observed, which is shown in Figure 17. This
behaviour was observed in all four analysed worn gear geometries, with loads ranging
Polymers 2023, 15, 1767 from 0.4 Nm to 0.7 Nm. The wear level can result in detrimental impacts on the contact 17 of 24
stress state, negatively affecting the performance of the gears. This can also lead to an
increase in noise and vibrations.

Figure 17.
Figure 17. Transition
Transitionfrom
frompitch
pitcharea
areacontact to to
contact double contact,
double as observed
contact, on all
as observed onworn CFRP
all worn gear gear
CFRP
sample analyses. The depicted case was evaluated for M = 0.7 Nm/120 h sample geometry.
sample analyses. The depicted case was evaluated for M = 0.7 Nm/120 h sample geometry.

3.3. Fatigue Properties of CFRP

One objective of the study was to improve the fatigue properties of CFRP gears beyond
the achievements of our previous research [19]. To achieve this, higher quality fibres, using
thinner plies,
3.3. Fatigue additional
Properties fibre orientations, and a toughened epoxy resin formulation were
of CFRP
used,One
as described in Section
objective of the study 2.1.was
Fortoinstance,
improvereplacing
the fatigue T300 with T700S
properties fibresgears
of CFRP can improve
be-
strength
yond the achievements of our previous research [19]. To achieve this, higher qualityhalf
and potentially reduce wear rates. By utilising plies that are more than as thin
fibres,
(0.11
using thinner plies, additional fibre orientations, and a toughened epoxy resin formula- and
mm instead of 0.26 mm), the load distribution between plies can be improved,
the
tionstress at theasinterface
were used, described regions can2.1.
in Section be reduced.
For instance,Increasing
replacingtheT300number of plies
with T700S for the
fibres
same thickness of gears allows for a wider range of fibre orientations,
can improve strength and potentially reduce wear rates. By utilising plies that are more which maximises
the
thancontent of parallel
half as thin (0.11 mm fibres in each
instead tooth.
of 0.26 mm),Inthe
addition to 0◦ , 45◦ ,between
load distribution and 90◦pliesplies,
can webeused
inter-plies ◦ ◦
improved, of and22.5
the and
stress67.5 to interface
at the decrease regions
the anglecanincrement
be reduced.between plies.the
Increasing Additionally,
number a
toughened epoxy
of plies for the sameformulation
thickness of can gearsdecrease wear
allows for rates,
a wider enhance
range fracture
of fibre toughness
orientations, which at the
interface
maximisesregions, and of
the content reduce thefibres
parallel crackininitiation at the
each tooth. fibre–matrix
In addition to 0°, interfaces.
45°, and 90° plies,
Polymers 2023, 15, x FOR PEER REVIEW 19 of 26
Theinter-plies
we used fatigue properties
of 22.5° and of 67.5°
CFRPs to made from
decrease the the
angle autoclave-processed
increment between plies. and vacuum-
Ad-
infused panel
ditionally, are presented
a toughened epoxyin Figure 18. Thecan
formulation graph features
decrease wearmultiple
rates, scales
enhance the y-axis to
on fracture
facilitate
toughnessa at thorough evaluation
the interface regions, of and
the material’s
reduce thefatigue characteristics.
crack initiation The primaryin-
at the fibre–matrix y-axis
axis displays the tooth bending stress, which is a commonly used metric in gear fatigue
terfaces.
displays the tooth bending stress, which is a commonly used metric in gear fatigue testing.
testing. The secondary y-axis shows the load per gear width, which provides information
The fatiguey-axis
The secondary properties
shows ofthe
CFRPs
loadmade fromwidth,
per gear the autoclave-processed and vacuum-
which provides information on the
on the loading conditions of the gears during testing. The third supplementary y-axis dis-
infused panel
loading are presented
conditions of the in Figure
gears during 18.testing.
The graph
The features
third multiple scalesy-axis
supplementary on the y-axis the
displays
plays the corresponding torque.
to facilitate a thorough
corresponding torque.evaluation of the material’s fatigue characteristics. The primary y-

Figure18.18.
Figure Maximum
Maximum tooth
tooth bending
bending stressstress against
against the number
the number of cyclesof
forcycles for the
the CFRP CFRP
gears gears and in
and in
comparison
comparison with PEEK
with gears.
PEEK gears.

The VDI equation 2736 [33] is a widely accepted and recognised standard for calcu-
lating tooth bending stress in gear fatigue testing. It utilises a mathematical model that
considers the rigid gear geometry and the isotropic linear elastic behaviour of the material.
The equation accounts for the geometric properties of the gear, such as tooth profile and
backlash, as well as the material properties, such as elastic modulus and Poisson’s ratio,
Polymers 2023, 15, 1767 18 of 24

The VDI equation 2736 [33] is a widely accepted and recognised standard for calcu-
lating tooth bending stress in gear fatigue testing. It utilises a mathematical model that
considers the rigid gear geometry and the isotropic linear elastic behaviour of the material.
The equation accounts for the geometric properties of the gear, such as tooth profile and
backlash, as well as the material properties, such as elastic modulus and Poisson’s ratio, in
order to predict the tooth bending stress under a given loading condition. Tooth bending
stress is calculated as:
Ft
σF = KF × YFa × YSa × Yε × Yβ × (1)
b × mn

where the following factor values were used for the tested gear geometry: KF = 1.0;
YFa = 3.01, YSa = 1.51; and Yε = 0.732; Y β = 1. To evaluate the performance of the gear and
to compare results with data from the literature, the load per gear face width, also known
as Fw , is plotted on the second y-axis. Load per gear width Fw is calculated as:

FT
Fw = (2)
b
where FT is tangential load, acting on the gear’s reference diameter d, and b is the gear’s
face width with unit [N/mm]. The relationship between values of torque, load per gear
width, and root bending stress can be found in Table 6.
Table 6. Relations between torque, load per gear width, and root bending stress VDI.

Torque Load per Gear Width Root Bending Stress VDI

[Nm] [N/mm] [MPa]
0.4 20 66
0.5 25 83
0.6 30 100
0.7 35 119
0.8 40 133
0.9 45 150

The experimental fatigue data were fitted using the Basquin power law, which is
widely accepted as a model for the stress versus number of cycles relationship [34]:

∆σ = C × Nfb , ∆σ = σF at R = 0 (3)

σF = C × Nfb (4)
where ∆σ is the stress range from minimum to maximum, σF is the maximum tooth bending
stress, Nf is number of cycles to failure, and C and b are fitting coefficients. Tests for the
T300-ET455 material were also conducted in the high-cycle regime, subjecting it to up to
7 × 107 cycles. However, a transition point was observed for the T300-ET455 material,
where the slope of the curve shifted downward (Figure 18). To account for this transition,
experimental data were fitted separately for cycles up to 1.2 × 107 (low-cycle fatigue) and
cycles between 1.2 × 107 and 7 × 107 (high-cycle fatigue). From the fitted data, we can
see fairly good agreement for the fitted coefficients, which are all higher than 0.9 except
for the second part of the T300-ET445 curve, where the R-square of 0.86 indicates slightly
worse fitting due to the scatter of fatigue data, which is typical behaviour of composite
materials [35]. The fitted coefficients C and b and R-square values are given in Table 7.
The results of the gear fatigue testing indicate that the T700S-DT120 material with
18 layers of 93 g/m2 and toughened resin exhibits superior durability characteristics. On
average, this material shows a 12% increase in peak load capacity and 45-times higher
fatigue endurance when compared to gears fabricated from T300-ET445. The fatigue life
Polymers 2023, 15, 1767 19 of 24

of T300-LG900 UG with thinner 93 g/m2 spread fabrics and resin infusion and wet laying
techniques is slightly inferior when compared to gears fabricated from T300-ET445 laminate.
Table 7. Coefficients C and b for regression-fitted Basquin equation to experimental data.

Material C b R2
T700S-DT120, 93 g/m2 224.7 −0.033 0.94
T300-ET445, 200 g/m2 —up to 1.2 × 107 199.6 −0.034 0.92
T300-ET445, 200 g/m2 —from 1 × 2 × 107 to 7 × 107 14,220 −0.294 0.86
T300-LG900 UV, 93 g/m2 189.1 −0.031 0.91

In the tests, we observed a distinct limit value of the maximum torque that can be
applied without immediate failure. Once the torque limit is exceeded, the material becomes
overstressed, leading to a quick failure shortly after the start of the test. This rapid failure
indicates the involvement of static failure mechanisms. The torque limit is significantly
higher in the T700S-DT120, at 0.9 Nm, compared to the T300-ET445 and T300-LG900UV
materials which both had a torque limit of 0.8 Nm.
For comparison, a PEEK fatigue curve from the research of Zorko et al. [36], which
was conducted on the same testing platform, is also plotted in Figure 18, demonstrating
that the PEEK material exhibits inferior durability characteristics when compared to the
evaluated CFRP gears. The fatigue results of a study by Kurokawa et al. [11] were plotted
for comparison as well. The fatigue results of PEEK with the addition of 15% short carbon
fibres showed a performance that was even worse than pure PEEK gears. It appears that
the short carbon fibres in that case probably acted as stress raisers, promoting fatigue cracks
and reducing the lifetime.
Based on the experimental data obtained from the gear fatigue testing, several key
findings can be deduced. First, the use of prepreg with incorporated toughened resin
has been found to significantly enhance the durability of gears. This improvement can
be attributed to the toughener’s ability to increase the resistance to crack initiation and
propagation in the resin matrix during fatigue. Second, the use of thinner plies and a higher
number of plies for the same thickness has been found to improve the fatigue life of the
gears. There are several studies which state that thin plies can substantially enhance the
strength and design flexibility of the composites [37,38]. This can be explained by the fact
that thinner plies allow for a more homogeneous distribution of stress throughout the gear.
Additionally, a higher number of plies results in a more complex load-sharing network,
further reducing the likelihood of failure.

3.4. Analysis of Backscatter SEM and EDS

The combination of backscattered scanning electron microscopy (BS-SEM) and energy-
dispersive X-ray spectroscopy (EDS) was used to identify the composition and distribution
of elements in gears after meshing.
In Figure 19, the BS-SEM of a tooth is shown from a top-view angle. On the worn
surface, shallow peaks are visible along the sliding direction; the tops of these peaks show
white areas that indicate a more Fe-rich conductive material. The rough asperities of the
steel pinion can cause abrasive wear and, later, the debris from both CFRP and metal can
contribute to three-body abrasion and increase the wear rate. The typical edge delamination
is visible in Figure 19b. The outer surfaces of the CFRP composite material are susceptible
to delamination failure due to lower adhesion between the composite plies at the edge.
Typically, when two bodies are in contact, the non-restrained edge deforms more than the
interior of the material, which is constrained by the surrounding material [25].
Figure 20 displays evidence of material transfer from the steel pinion onto the CFRP
tooth surface. This is an example of adhesive wear, characterised by material transfer from
one surface to another. As visible in Figure 21, the matrix has been crushed and separated
from the fibres in the sliding direction, resulting in metal debris accumulation in cavities.
 been found to significantly enhance the durability of gears. This improvement can be at-
tributed to the toughener’s ability to increase the resistance to crack initiation and propa-
gation in the resin matrix during fatigue. Second, the use of thinner plies and a higher
number of plies for the same thickness has been found to improve the fatigue life of the
gears. There are several studies which state that thin plies can substantially enhance the
Polymers 2023, 15, 1767 20 of 24
strength and design flexibility of the composites [37,38]. This can be explained by the fact
that thinner plies allow for a more homogeneous distribution of stress throughout the
gear. Additionally, a higher number of plies results in a more complex load-sharing net-
During wear, reducing
work, further continuously formed broken
the likelihood fibre and matrix particles act as an abrasive
of failure.
which causes three-body abrasion.
Materialoftransfer
3.4. Analysis fromSEM
Backscatter steeland
pinion
EDSonto the CFRP tooth was confirmed by EDS analysis
of Spectrum 1 shown in Figure 22a, indicating the high Fe content and other metal elements
The combination of backscattered scanning electron microscopy (BS-SEM) and en-
which are present in the 42CrMo4 steel driving gear.
ergy-dispersive X-ray spectroscopy (EDS) was used to identify the composition and dis-
The BS-SEM images from Figures 19–21 show white areas on the surface of the CFRP
tribution of elements in gears after meshing.
gear teeth, indicating the coating of adhesive debris composed mainly of Fe, Si, Cr, Mo, and
In Figure 19, the BS-SEM of a tooth is shown from a top-view angle. On the worn
Mn, which are components of the 42CrMo4 steel pinion. The EDS analyses of Spectra 1–4
surface, shallow peaks are visible along the sliding direction; the tops of these peaks show
are presented in Figure 22. Chemical analysis of Spectra 1–3, given in Figure 22a–c, reveal
white areas
similar that indicate
compositions with airon
more Fe-rich
content conductive
range material.
55.6–62.6%, oxygenThe rough asperities
28.8–38.3, of the
silicon 2.1–6.8%,
steel pinion can cause abrasive wear and, later, the debris from both
chromium 0.62–0.66%, manganese 0.4–0.64%, and molybdenum 0.48–0.57%. The metal CFRP and metal can
contribute to three-body abrasion and increase the wear rate. The typical edge
chemical elements, found on the surface of the CFRP, are the elements of the 42CrMo4 steel delamina-
tion isgear.
drive visible
Thein42CrMo4
Figure 19b. The
steel outer
(Wr. No.surfaces of the CFRP
1.7227) nominally composite
consists material are
of 0.38–0.45% sus-
carbon,
ceptible
0.4% to delamination
silicon, failure due
0.6–0.9 of manganese, to lower
0.9–1.2 adhesion between
of chromium, the composite
and 0.15–0.3 plies atThis
of molybdenum. the
edge. Typically, when two bodies are in contact, the non-restrained edge
implies that the steel gear was also wearing during the meshing, transferring the metal deforms more
than the into
material interior
the of the material,
surface which
of the CFRP is constrained by the surrounding material [25].
gears.

Polymers 2023, 15, x FOR PEER REVIEW 22 of 26

(a) (b)
Material transfer from steel pinion onto the CFRP tooth was confirmed by EDS anal-
Figure
ysis 19.
of19.
Figure SEM backscatter
Spectrum
SEM backscatter images
1 shownimages of
in Figure the22a,
of the worn
worn surface
surface of
indicating CFRP
of the
CFRP gear
high (a)content
gearFe(a) flank surface
flank surface at the
and other
at the metal
top of
top of
the tooth,
elements 65× magnification;
which (b)
are present(b) detail
indetail of the
the 42CrMo4 crack at
steelthe edge.
driving
the tooth, 65 × magnification; of the crack at the edge. gear.

Figure 20 displays evidence of material transfer from the steel pinion onto the CFRP
tooth surface. This is an example of adhesive wear, characterised by material transfer from
one surface to another. As visible in Figure 21, the matrix has been crushed and separated
from the fibres in the sliding direction, resulting in metal debris accumulation in cavities.
During wear, continuously formed broken fibre and matrix particles act as an abrasive
which causes three-body abrasion.

(a) (b)
Figure 20. SEM
Figure 20. SEM backscatter
backscatter images
images of
of the
the worn
worn surface
surface of
of CFRP
CFRP gear
gear (a) flank surface
(a) flank surface at the top
at the top of
of
the tooth with the adhered layer of metal material from the steel pinion; (b) detail of cracks around
the tooth with the adhered layer of metal material from the steel pinion; (b) detail of cracks around
the layer of iron.
the layer of iron.
 (a) (b)

Polymers 2023, 15, 1767 Figure 20. SEM backscatter images of the worn surface of CFRP gear (a) flank surface at the21 top
of of
24
the tooth with the adhered layer of metal material from the steel pinion; (b) detail of cracks around
the layer of iron.

Polymers 2023, 15, x FOR PEER REVIEW 23 of 26

(a) (b)
Figure 21. SEM
gear. The
Figure 21. SEM backscatter
oxides can alsoimages
backscatter resultof
images the
the worn
from
of surface
galvanic
worn of
of CFRP
corrosion
surface gear
gear (a)
CFRPwear flank by
caused
(a) flank surface at the top
top of
the presence
surface at the of
of
the tooth, 100×
electrolytes
the tooth, 100× magnification;
due to humidity.
magnification;(b) detail of microcracks between fibres and matrix.
(b) detail of microcracks between fibres and matrix.

The BS-SEM images from Figures 19–21 show white areas on the surface of the CFRP
gear teeth, indicating the coating of adhesive debris composed mainly of Fe, Si, Cr, Mo,
and Mn, which are components of the 42CrMo4 steel pinion. The EDS analyses of Spectra
1–4 are presented in Figure 22. Chemical analysis of Spectra 1–3, given in Figure 22a–c,
reveal similar compositions with iron content range 55.6–62.6%, oxygen 28.8–38.3, silicon
2.1–6.8%, chromium 0.62–0.66 %, manganese 0.4–0.64%, and molybdenum 0.48–0.57%.
The metal chemical elements, found on the surface of the CFRP, are the elements of the
42CrMo4 steel drive gear. The 42CrMo4 steel (Wr. No. 1.7227) nominally consists of 0.38–
0.45% carbon, 0.4 % silicon, 0.6–0.9 of manganese, 0.9–1.2 of chromium, and 0.15–0.3 of
molybdenum. This implies that the steel gear was also wearing during the meshing, trans-
ferring the metal material into the surface of the CFRP gears.
In Spectrum 4, shown in Figure 22d, only the presence of carbon was found on the
carbon fibre surface. High oxygen content, detected in Spectra 1–3, is likely from the epoxy
matrix or from iron oxide (FeO) formed as a result of oxidative wear on the steel drive

Figure 22. EDS analysis Spectra 1–4 representing worn surfaces at different points as presented in
Figures 20 and 21.
Figure 22. EDS analysis Spectra 1–4 representing worn surfaces at different points as presented in
Figures
In 20 and 21. 4, shown in Figure 22d, only the presence of carbon was found on the
Spectrum
carbon fibre surface. High oxygen content, detected in Spectra 1–3, is likely from the epoxy
4. Conclusions
matrix or from iron oxide (FeO) formed as a result of oxidative wear on the steel drive
gear. The
Theanalysis of the
oxides can lifespan
also result of gears
from made of
galvanic laminated
corrosion composites
wear compared
caused by to plas-
the presence of
tic and metal
electrolytes gears
due has shown that they have the potential to be used for high-strength
to humidity.
applications. However, their limited utilisation in industry is attributed to the high costs
associated with this technology. Despite this, as laminated composite gears have been
found to have superior performance over conventional plastic injection-moulded gears
with the addition of short fibres, it is likely that this trend will change with advancements
in technology and cost reductions. The emergence of 3D-printed continuous fibre technol-
ogy presents a significant opportunity for the efficient and cost-effective production of
Polymers 2023, 15, 1767 22 of 24

4. Conclusions
The analysis of the lifespan of gears made of laminated composites compared to
plastic and metal gears has shown that they have the potential to be used for high-strength
applications. However, their limited utilisation in industry is attributed to the high costs
associated with this technology. Despite this, as laminated composite gears have been
found to have superior performance over conventional plastic injection-moulded gears
with the addition of short fibres, it is likely that this trend will change with advancements in
technology and cost reductions. The emergence of 3D-printed continuous fibre technology
presents a significant opportunity for the efficient and cost-effective production of high-
quality gears. The findings of this study allow the following conclusions to be drawn:
• The use of T700S-DT120 prepreg with incorporated toughened resin has been found
to significantly enhance the durability of gears over T300-ET445 and T300-LG-900UV.
This can be attributed to the higher strength fibres and toughener’s ability to improve
the resistance to crack initiation and propagation during fatigue and wear. Second, the
use of thinner plies results in a more complex load-sharing network, further reducing
the likelihood of failure.
• The results of experimental wear study indicate nearly linear correlation between wear
volume, wear pitch distance, and duration of testing. Furthermore, a consistent in-
crease in the slope of the linear function was observed with increasing loading torque.
• The tooth profile was altered by wear, particularly in the root area, where the shape of
the profile changed from an outer arched form to an inner arched shape. Numerical
analysis showed that the worn out profile of CFRP gears led to a double contact
with the steel pinion, heightened backlash, and elevated transmission error, negatively
impacting the performance of the gears, and causing an increase in noise and vibrations
and decreasing the efficiency.
• BS-SEM and EDS analysis revealed that the degradation of CFRP is due to severe
adhesion and three-body wear, resulting in edge delamination caused by limited
adhesion strength between the matrix and fibres. Fatigue in CFRP is typically caused
by cyclic stress–strain loading, which leads to the formation of microcracks in the
matrix material and eventual delamination or fibre breakage.
These findings provide valuable insights for the optimisation of CFRP gears in terms
of material selection, resin systems, and layup strategies. Carbon fibre reinforced plastic
(CFRP) gears have the potential to be used as a lightweight and durable alternative to
traditional steel gears in various applications.
Nevertheless, when meshed with steel pinion gears, the wear of CFRP gears is a
concern that needs to be addressed in the future to improve the wear resistance. This may
involve testing a range of different fibres and abrasion-resistant matrix materials, as well
as varying the layup and manufacturing processes to determine the best combination for
reducing wear.

5. Patents
The patent resulting from the work reported in this manuscript is as follows:
ZORKO, Damijan, ČERNE, Borut, BERGANT, Zoran. Zobnik iz kompozitnega ma-
teriala s kontinuirnimi vlakni: patent SI 26212 A, 30 December 2022. Ljubljana: Urad
Republike Slovenije za intelektualno lastnino, 2022. 10 f., 3 f. pril., ilustr. http://www3
.uil-sipo.si/PublicationServer/documentpdf.jsp?iDocId=52003&iepatch=.pdf, accessed on
30 December 2022.

Author Contributions: Conceptualization, Z.B. and D.Z.; methodology, R.Š., Z.B., B.Č. and D.Z.;
validation, Z.B., B.Č. and D.Z.; formal analysis, Z.B., B.Č. and D.Z.; investigation, Z.B., B.Č. and D.Z.;
resources, Z.B., B.Č. and D.Z.; data cu-ration, Z.B., B.Č. and D.Z.; writing—original draft preparation,
Z.B., B.Č. and D.Z.; writing—review and editing, Z.B., B.Č. and R.Š.; visualization, Z.B., B.Č. and
D.Z.; supervision, Z.B., B.Č., D.Z. and R.Š.; project administration, R.Š.; funding acquisition, R.Š. All
authors have read and agreed to the published version of the manuscript.
Polymers 2023, 15, 1767 23 of 24

Funding: This work was supported by the Slovenian Research Agency programme No. P2-0270 and
post-doctoral projects Z2-3207 and Z2-3208.
Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.
Data Availability Statement: Data are contained within the article.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Kim, M.T.; Rhee, K.Y.; Lee, J.H.; Hui, D.; Lau, A.K.T. Property Enhancement of a Carbon Fiber/Epoxy Composite by Using
Carbon Nanotubes. Compos. Part B Eng. 2011, 42, 1257–1261. [CrossRef]
2. Kim, B.C.; Park, D.C.; Kim, B.J.; Lee, D.G. Through-Thickness Compressive Strength of a Carbon/Epoxy Composite Laminate.
Compos. Struct. 2010, 92, 480–487. [CrossRef]
3. Guo, R.; Xian, G.; Li, C.; Hong, B. Effect of Fiber Hybrid Mode on the Tension–Tension Fatigue Performance for the Pultruded
Carbon/Glass Fiber Reinforced Polymer Composite Rod. Eng. Fract. Mech. 2022, 260, 108208. [CrossRef]
4. Jiang, Z.; Fang, Z.; Fang, C.; Li, Q.; Wang, Z. Experimental Investigation on High-Temperature Creep Behaviour of Carbon Fiber
Reinforced Polymer Cable. Compos. Struct. 2022, 291, 115533. [CrossRef]
5. Xian, G.; Guo, R.; Li, C.; Wang, Y. Mechanical Performance Evolution and Life Prediction of Prestressed CFRP Plate Exposed to
Hygrothermal and Freeze-Thaw Environments. Compos. Struct. 2022, 293, 115719. [CrossRef]
6. Handschuh, R.F.; LaBerge, K.E.; DeLuca, S.; Pelagalli, R. Vibration and Operational Characteristics of a Composite-Steel (Hybrid) Gear;
NASA Center for Aerospace Information: Hanover, MD, USA, 2014.
7. Catera, P.G.; Gagliardi, F.; Mundo, D.; De Napoli, L.; Matveeva, A.; Farkas, L. Multi-Scale Modeling of Triaxial Braided Composites
for FE-Based Modal Analysis of Hybrid Metal-Composite Gears. Compos. Struct. 2017, 182, 116–123. [CrossRef]
8. Catera, P.G.; Mundo, D.; Treviso, A.; Gagliardi, F.; Visrolia, A. On the Design and Simulation of Hybrid Metal-Composite Gears.
Appl. Compos. Mater. 2019, 26, 817–833. [CrossRef]
9. Blais, P.; Toubal, L. Single-Gear-Tooth Bending Fatigue of HDPE Reinforced with Short Natural Fiber. Int. J. Fatigue 2020,
141, 105857. [CrossRef]
10. Mohsenzadeh, R.; Soudmand, B.H.; Shelesh-Nezhad, K. Load-Bearing Analysis of Polymer Nanocomposite Gears Using a
Temperature-Based Step Loading Technique: Experimental and Numerical Study. Wear 2023, 514, 204595. [CrossRef]
11. Kurokawa, M.; Uchiyama, Y.; Nagai, S. Performance of Plastic Gear Made of Carbon Fiber Reinforced Poly-Ether-Ether-Ketone:
Part 2. Tribol. Int. 2000, 33, 715–721. [CrossRef]
12. Kurokawa, M.; Uchiyama, Y.; Iwai, T.; Nagai, S. Performance of Plastic Gear Made of Carbon Fiber Reinforced Polyamide 12.
Wear 2003, 254, 468–473. [CrossRef]
13. Mao, K.; Greenwood, D.; Ramakrishnan, R.; Goodship, V.; Shrouti, C.; Chetwynd, D.; Langlois+, P. The Wear Resistance
Improvement of Fibre Reinforced Polymer Composite Gears. Wear 2019, 426, 1033–1039. [CrossRef]
14. Tavčar, J.; Grkman, G.; Duhovnik, J. Accelerated Lifetime Testing of Reinforced Polymer Gears. JAMDSM 2018, 12, JAMDSM0006.
[CrossRef]
15. Krairi, A.; Doghri, I.; Robert, G. Multiscale High Cycle Fatigue Models for Neat and Short Fiber Reinforced Thermoplastic
Polymers. Int. J. Fatigue 2016, 92, 179–192. [CrossRef]
16. Rolland, H.; Saintier, N.; Raphael, I.; Lenoir, N.; King, A.; Robert, G. Fatigue Damage Mechanisms of Short Fiber Reinforced PA66
as Observed by In-Situ Synchrotron X-Ray Microtomography. Compos. Part B Eng. 2018, 143, 217–229. [CrossRef]
17. Migneault, S.; Koubaa, A.; Erchiqui, F.; Chaala, A.; Englund, K.; Krause, C.; Wolcott, M. Effect of Fiber Length on Processing and
Properties of Extruded Wood-Fiber/HDPE Composites. J. Appl. Polym. Sci. 2008, 110, 1085–1092. [CrossRef]
18. Zorko, D.; Tavčar, J.; Šturm, R.; Bergant, Z. Investigation of the Durability and Performance of Autoclave-Cured, Woven Carbon
Fiber-Reinforced Polymer Composite Gears in Mesh with a Steel Pinion. Compos. Struct. 2021, 273, 114250. [CrossRef]
19. Zorko, D.; Tavčar, J.; Bizjak, M.; Šturm, R.; Bergant, Z. High Cycle Fatigue Behaviour of Autoclave-Cured Woven Carbon
Fibre-Reinforced Polymer Composite Gears. Polym. Test. 2021, 102, 107339. [CrossRef]
20. Bijwe, J.; Sharma, M. Carbon Fabric-Reinforced Polymer Composites and Parameters Controlling Tribological Performance. In
Wear of Advanced Materials; Davim, J.P., Ed.; John Wiley & Sons, Ltd.: Hoboken, NJ, USA, 2013; Volume 1, pp. 1–60.
21. Ramesh, B.N.; Suresha, B. Optimization of Tribological Parameters in Abrasive Wear Mode of Carbon-Epoxy Hybrid Composites.
Mater. Des. 2014, 59, 38–49. [CrossRef]
22. Černe, B.; Bergant, Z.; Šturm, R.; Tavčar, J.; Zorko, D. Experimental and Numerical Analysis of Laminated Carbon Fibre-Reinforced
Polymer Gears with Implicit Model for Coefficient-of-Friction Evaluation. J. Comput. Des. Eng. 2022, 9, 246–262. [CrossRef]
23. Ekoi, E.J.; Dickson, A.N.; Dowling, D.P. Investigating the Fatigue and Mechanical Behaviour of 3D Printed Woven and Nonwoven
Continuous Carbon Fibre Reinforced Polymer (CFRP) Composites. Compos. Part B Eng. 2021, 212, 108704. [CrossRef]
24. Bergant, Z.; Savin, A.; Grum, J. Effects of Manufacturing Technology on Static, Multi-Frequency Dynamic Mechanical Analysis
and Fracture Energy of Cross-Ply and Quasi-Isotropic Carbon/Epoxy Laminates. Polym. Polym. Compos. 2018, 26, 358–370.
[CrossRef]
Polymers 2023, 15, 1767 24 of 24

25. Hoa, S.V. Principles of the Manufacturing of Composite Materials; DEStech Publications: Lancaster, UK, 2009; ISBN 978-1-932078-26-8.
26. Bagheri, R.; Marouf, B.T.; Pearson, R.A. Rubber-Toughened Epoxies: A Critical Review. Polym. Rev. 2009, 49, 201–225. [CrossRef]
27. Garg, A.C.; Mai, Y.-W. Failure Mechanisms in Toughened Epoxy Resins—A Review. Compos. Sci. Technol. 1988, 31, 179–223.
[CrossRef]
28. Sihn, S.; Kim, R.; Kawabe, K.; Tsai, S. Experimental Studies of Thin-Ply Laminated Composites. Compos. Sci. Technol. 2007, 67,
996–1008. [CrossRef]
29. Galos, J. Thin-Ply Composite Laminates: A Review. Compos. Struct. 2020, 236, 111920. [CrossRef]
30. Urbas, U.; Zorko, D.; Černe, B.; Tavčar, J.; Vukašinović, N. A Method for Enhanced Spur Gear Inspection Based on 3D Optical
Metrology. Measurement 2020, 169, 108584. [CrossRef]
31. Zorko, D.; Kulovec, S.; Tavčar, J.; Duhovnik, J. Different Teeth Profile Shapes of Polymer Gears and Comparison of Their
Performance. JAMDSM 2017, 11, JAMDSM0083. [CrossRef]
32. Černe, B.; Petkovšek, M.; Duhovnik, J.; Tavčar, J. Thermo-Mechanical Modeling of Polymer Spur Gears with Experimental
Validation Using High-Speed Infrared Thermography. Mech. Mach. Theory 2020, 146, 103734. [CrossRef]
33. DIN 3990; Calculation of Load Capacity of Cylindrical Gears. German National Standard: Berlin, Germany, 1987.
34. Deng, H.; Li, W.; Sakai, T.; Sun, Z. Very High Cycle Fatigue Failure Analysis and Life Prediction of Cr-Ni-W Gear Steel Based on
Crack Initiation and Growth Behaviours. Materials 2015, 8, 8338–8354. [CrossRef]
35. Yadav, I.N.; Thapa, K.B. Fatigue Damage Model of Woven Glass-Epoxy Fabric Composite Materials. J. Mater. Res. Technol. 2020, 9,
301–306. [CrossRef]
36. Zorko, D.; Kulovec, S.; Duhovnik, J.; Tavčar, J. Durability and Design Parameters of a Steel/PEEK Gear Pair. Mech. Mach. Theory
2019, 140, 825–846. [CrossRef]
37. Basha, M.; Wagih, A.; Khan, T.; Lubineau, G.; Sebaey, T.A. On the Benefit of Thin Plies on Flexural Response of CFRP Composites
Aged at Elevated Temperature. Compos. Part A Appl. Sci. Manuf. 2023, 166, 107393. [CrossRef]
38. Arteiro, A.; Furtado, C.; Catalanotti, G.; Linde, P.; Camanho, P.P. Thin-Ply Polymer Composite Materials: A Review. Compos. Part
A Appl. Sci. Manuf. 2020, 132, 105777. [CrossRef]

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content.