applied

sciences
Article
Application of Infrared Thermography in E-Bike Battery Pack
Detail Analysis—Case Study
Hrvoje Glavaš 1 , Tomislav Barić 1 , Mirko Karakašić 2, * and Tomislav Keser 1

1 Faculty of Electrical Engineering, Computer Science and Information Technology Osijek, University of Osijek,
31000 Osijek, Croatia; [email protected] (H.G.); [email protected] (T.B.); [email protected] (T.K.)
2 Mechanical Engineering Faculty, University of Slavonski Brod, 35000 Slavonski Brod, Croatia
* Correspondence: [email protected]; Tel.: +385-98-936-7007

Abstract: The purpose of this research was to analyze the applicability of infrared thermography in
assessing the condition of an e-bike battery. The main challenge was to interpret the surface radiation
detected by a thermal imaging camera. In addition to basic challenges of thermography, such as
determining the emissivity, the angle of image acquisition, the problem of low resolution and thermal
inertia, it was necessary to understand the physical behavior of the elements inside the battery and
battery pack that respectively determine the dynamics and shape of the thermal pattern. E-bike
batteries are mainly composed of 18650 format Li-ion cells, and the topology of the battery cell is
related to the energy profile of the system. With the aim of establishing a link between the surface
radiation patterns and the parameters of the individual 18650 cells that dictate the internal thermal
behavior, a detailed analysis of the battery pack interior, upon its lifetime, was performed. Based on
the results of the capacity analysis of the individual cells and determination of the exact position of
the cells within the structure of the battery, a visualization and comparison with the thermographic
records were performed to determine whether infrared thermography is able to diversify faulty
cells. The conclusion of the study is that it is not possible to make a judgment about the thermal





 pattern expression or the pattern shape based on the estimated capacity values; however, based on
Citation: Glavaš, H.; Barić, T.; the pattern, it is possible to draw a conclusion about the homogeneity of the capacity of the individual
Karakašić, M.; Keser, T. Application battery cells.
of Infrared Thermography in E-Bike
Battery Pack Detail Analysis—Case Keywords: e-bike; Li-ion battery; 18650 cell; infrared thermography; battery pack revival
Study. Appl. Sci. 2022, 12, 3444.
https://doi.org/10.3390/
app12073444

Academic Editor: Young-Kyu Han 1. Introduction

Received: 1 March 2022

The work provides a postmortem analysis of an e-bike battery pack in the form of a
Accepted: 27 March 2022
detailed factorial review that surveys the battery pack end of service and lifetime properties
Published: 28 March 2022
after seven years of its age, five years of use, and hundreds of charge cycles in combination
with 2543 km traveled in total. The analysis was performed on an e-bicycle, PEDELEC
Publisher’s Note: MDPI stays neutral
(pedal electric cycle), Kalkhoff Groove 28” made in 2011, equipped with a front hub motor
with regard to jurisdictional claims in
and Li-ion battery pack of nominal voltage 36 V, capacity 9 Ah (324 Wh). Experiences of
published maps and institutional affil-
other users with the same bike indicate a battery pack lifetime of four years regardless of
iations.
the mileage but in the same period of time, as well as under similar terrain and climatic
conditions [1]. The environmental temperatures that prevailed at the time of the use
significantly contribute to battery life shortening [2], and age [3] contributed to an annual
Copyright: © 2022 by the authors.
capacity drop of 20%. That is supported by an extremely small number of charging cycles
Licensee MDPI, Basel, Switzerland. compared with the usual practice of a cycle analysis [4]. The bicycles were being used on
This article is an open access article flat terrain from 2014 to 2018, without significant vibrational stresses that could reduce the
distributed under the terms and battery life [5]. The average wind speed at the site was 2.1 m/s, and riding speed of the
conditions of the Creative Commons e-bike was in the range of 20 to 22.8 km/h. In comparison, at the same location, the average
Attribution (CC BY) license (https:// car speed was from 38.4 to 39.7 km/h and the classic bicycle from 12.5 to 14.6 km/h [6].
creativecommons.org/licenses/by/ An end-of-life battery problem directly impacts energy policy and transportation
4.0/). energy efficiency improvements, as well as the overall e-bike lifecycle. There are various

Appl. Sci. 2022, 12, 3444. https://doi.org/10.3390/app12073444 https://www.mdpi.com/journal/applsci

Appl. Sci. 2022, 12, 3444 2 of 20

possible causes of battery end of life [7]. Therefore, the article provides a brief history of the
development of electric bicycles, a link between bicycle riding parameters and correspond-
ing energy needs, and a description of the usual daily route to evoke the typical dynamics
of battery usage. The core of the article deals with the usability case and comparative
analysis of infrared thermograms, with the results of the analysis of the individual cells
of the real-life e-bicycle battery pack based on a CAD modeling and spatio-thermal cells
diversification. In order to preserve and prolong the functionality of the real-life exemplary
e-bicycle, an additional analysis and modus operandi of a possible battery lifetime exten-
sion is proposed. It is based on the concept of the battery revival, in this case replacing
20 cells (half of the total number) with cells from other systems (laptops/notebooks). A
condition assessment of scrapped cells and their placing optimization within the battery
pack structure was carried out, thus allowing further unrestricted use of the exemplary
real-life electric bike and giving back its functionality.

2. Electric Bicycles
Bicycles are predominantly the first vehicle we learn to ride. Sometimes it is a way
of living and a favorite recreational choice [8]. The globally accepted trends in the form
of utilizing energy systems efficiently—governed by strategical planning in transport as
well—encourage the use of bicycles and electric bicycles in accordance with the strategies
for reducing greenhouse gas emissions and energy consumption overall [9,10].

2.1. A Brief History of Bicycles

The history of bicycle development and cycling is interesting and rich, making it im-
possible to summarize in a few words. Only the most significant moments in bicycle history
(bicycle technical progress [11] and the phenomenon of electromobility [10]) are elaborated
here. The following years were the key moments in the rich bicycle development history:
1817 Appearance of first bicycle, the Vélocipede, by Karl Freiherr von Drais in Germany [12].
1839 Introduction of pedal and pedaling by Kirkpatrick MacMillan in Scotland [13,14].
1861 Relocation of pedals to the front wheel in “Michaulina” by Pierre Michaux in France [15].
1869 Introduction of solid rubber tires [16].
1864 Appearance of the chain by James Slater [17].
1870 Invention of the tangentially spoked wheel by James Starley [17].
1874 Women’s bicycle with a side seat and one pedal by James Starley [17].
1885 A modern kind of bicycle (Rover) designed by John Starley Kemp [18].
1888 Introduction of a cushion type of pneumatic tires [19,20].
1895 Appearance of the first electric bicycle by Ogden Bolton Jr. [21].
1896 Cycling became a sport at the first Olympic games in the Athens in 1896 [16].
1975 Panasonic made the E-bike driven by 24 V lead-acid car batteries [15].
1989 Sanyo introduced NiCd batteries [22].
1993 Yamaha presented PAS, an electrically power assisted bicycle [23].
2002 EU introduced and set on e-mobility legislation 2002/24/EC [24].
2003 NiMH chemistry prevailed in batteries for electrical driving systems [23].
2004 Li-ion became a standard battery choice [23].
2005 onward, a demand for PEDELECs grows constantly [25].
The written history of the electric bike began in 1881 when the Frenchman Gustave
Truvé put an electric drive on a tricycle [26]. In 1895, Ogden Bolton Jr. submitted patent
US 552 271 A for an electric bike with two wheels. John Schnepf patented the “friction
drive” in 1899 [27]. In 1975, Panasonic produced an electric “pony” bicycle that did not
look much different from today’s models. Furthermore, there have been a lot of different
patents since the 1970s, but we can single out 1989 when Sanyo Enacle introduced NiCd
chemistry-based batteries. In 1993, Yamaha electric bicycles presented PAS, the pedal-assist
system. The same company used lead–acid batteries back in 1993 but switched to the NiCd
one in 1995 and has been using NiMH batteries on some models since 1999. Generally,
NiMH batteries have been a standard battery since 2003, and only a year after 2004, a
 a Li-ion battery took over that title. It can be said that the era of electric bicycles be
the mid-90s with the price reduction in neodymium-based magnets caused by the
growth of the computer industry.

Appl. Sci. 2022, 12, 3444 2.2. Technical and Regulatory Aspects of Electric Cycling 3 of 20

The technical solution of the electric bicycle in the form of PEDELEC (pedal
cycle) [28] is the closest to the initial idea of cycling. A greater demand for PED
Li-ion battery took over that title. It can be said that the era of electric bicycles began in
began in 2005
the mid-90s withprimarily thanks to
the price reduction in the wider use of magnets
neodymium-based Li-ion batteries.
caused by PEDELEC
the rapid is “b
with
growthpedals
of theequipped with one auxiliary electric motor with a maximum continuou
computer industry.
power of 0.25 kW, whose output value is constantly reduced and finally switch
2.2. Technical and Regulatory Aspects of Electric Cycling
when the vehicle reaches a speed of 25 km/h, or sooner if the cyclist stops stepping
The technical solution of the electric bicycle in the form of PEDELEC (pedal electric
pedals”
cycle) [28]and
is theisclosest
given in initial
to the the description
idea of cycling.ofA PEDELEC
greater demand legislature,
for PEDELECs in EU Di
2002/24/EC
began in 2005[24]. Despite
primarily thanks its to
higher
the wider weight
use of(approximately 27 kg and
Li-ion batteries. PEDELEC more) than a c
is “bicycles
with pedals
tional bike equipped
(approximatelywith one 14 auxiliary electric motor
kg), PEDELEC with a an
allows maximum continuous rated
easier movement regardless
power of 0.25 kW, whose output value is constantly reduced and
rain configuration and helps to lower the physical exertion of the driver. European finally switched off when
the vehicle reaches a speed of 25 km/h, or sooner if the cyclist stops stepping on the pedals”
LEC manufacturers for the USA and Canada market, owing to the difference in legis
and is given in the description of PEDELEC legislature, in EU Directive 2002/24/EC [24].
produce
Despite its350 W weight
higher motors(approximately
exclusively.27Furthermore,
kg and more) than thea USA legislature
conventional even allows
bike (approx-
ofimately
electric motors
14 kg), PEDELECup toallows
750 W and Canada
an easier movement upregardless
to 500 W. In Europe,
of terrain Switzerland
configuration
and helps to lower the physical exertion of the driver. European
electric motors up to 500 W, but the speed limit is in accordance with the EU—lim PEDELEC manufacturers
for the USA and Canada market, owing to the difference in legislation, produce 350 W
25 km/h. Switzerland legislature supported their 500 W limitation due to the nee
motors exclusively. Furthermore, the USA legislature even allows the use of electric motors
specifics
up to 750 ofW their predominant
and Canada up to 500terrain
W. In Europe,configuration.
Switzerland Inallows
the end, PEDELEC
electric motors up is subjec
requirements
to 500 W, but the ofspeed
the international
limit is in accordancestandard withrelated to bicycles
the EU—limited EN
to 25 14764
km/h. “City and tr
Switzer-
land legislature supported their 500 W limitation due
bicycles—Safety requirements and test methods”, but in addition, it must to the needs and specifics of their meet
predominant terrain configuration. In the end, PEDELEC is subject to the requirements of
quirements of EN 15194 “Cycles—Electrically power assisted cycles—EPAC bic
the international standard related to bicycles EN 14764 “City and trekking bicycles—Safety
which refers and
requirements to thetestelectrical
methods”,part but in only.
addition, it must meet the requirements of EN
When riding, the biggest problems
15194 “Cycles—Electrically power assisted cycles—EPAC for the rider are hills
bicycles”, andrefers
which strong to winds.
the Th
age riderpart
electrical develops
only. 100 W recreationally and up to 200 W when in a good physical
When riding,
tion. A professional the biggest
rider problems
can normally for the rider
produceare hills
200 and
Wstrong
to 300 winds. The average
W during a non-comp
rider develops 100 W recreationally and up to 200 W when in a good physical condition.
session and in races of up to 400 W continuously over a distance of 14 km. The
A professional rider can normally produce 200 W to 300 W during a non-competitive
requirements for riding
session and in races of up toare 400best described over
W continuously in Figure
a distance1 [29].
of 14The
km. green
The powerinterval in
the average for
requirements operational
riding are bestcycling speed
described that prevails
in Figure 1 [29]. Theduring a cycling
green interval session,
indicates the the o
average operational cycling speed that prevails during a cycling
colored intervals present a maximum value of the power required, and a light-oran session, the orange-colored
isintervals
a power present a maximum value of the power required, and a light-orange one is a
requirement in the event of the mean wind occurrence at the cycling sit
power requirement in the event of the mean wind occurrence at the cycling site.

Figure 1.Power
Figure 1. Powerintervals required
intervals for cycling.
required for cycling.
In order to define, more precisely, the need for cycling power, it is necessary to
In order to describe
deterministically define, such
morea precisely,
process in thetheform
needoffor cycling power,
a deterministic it is necessary
mathematical
terministically describe such a process in the form of a deterministiccycling
relation. The mathematically defined and sufficiently accurately described need for mathematic
powerThe
tion. is given with expressiondefined
mathematically (1). The dependence and influential
and sufficiently behavior
accurately of individual
described need for
variables are in detail described in [14].
power is given with expression (1). The dependence and influential behavior of ind
h i
variables are in detail
W = described
Ka · (V + Vwin [14].
)2 + g · (mr + mb ) · (s + Cr ) · V (1)
 W =  K a ⋅ (V + Vw ) + g ⋅ ( mr + mb ) ⋅ ( s + Cr )  ⋅ V
2
 
where
Appl. Sci. 2022, 12, 3444 W mechanical power at the wheel, [W] (which is slightly less than what the rider
4 of 20
due to losses and coupling efficiency);
Ka drag factor, [kg/m] (usually in interval of 0.1 to 0.3; for a small recumbent rider 0
where upright rider 0.3; typical values 0.2–0.25);
WV mechanical
bike velocity,
power at [m/s];
the wheel, [W] (which is slightly less than what the rider produces due
s to losses and coupling efficiency);
road slope, [%], (e.g., 5% = 0.05);
Ka drag factor, [kg/m] (usually in interval of 0.1 to 0.3; for a small recumbent rider 0.1; large
rolling
Cr upright resistance coefficient [1] (0.02 for racing tires; 0.08 for heavily grooved M
rider 0.3; typical values 0.2–0.25);
V bike 0.03 typically
velocity, [m/s];for road bikes);
sVw road heading wind
slope, [%], (e.g.,velocity,
5% = 0.05);[m/s];
Cmr r rolling resistance
driver mass, [kg]; coefficient [1] (0.02 for racing tires; 0.08 for heavily grooved MTB; and
0.03 typically for road bikes);
mb bike mass, [kg];
V w heading wind velocity, [m/s];
mgr driver
gravitational
mass, [kg]; constant, [m/s ] (taken an average value of 9.81 m/s ).
2 2

mb bike mass,
The [kg]; as a power source is a more critical element of an electric bicycle
battery
git intrinsically
gravitational constant, 2 ] (taken an average value of 9.81 m/s2 ).
poses a [m/s
hazard in the event of overcharging, short circuit, or physic
age The
[10].battery as a power
Batteries that source
are notis acertified
more critical
by element
BATSOof(Battery
an electricSafety
bicycle because
Organization)
itnot
intrinsically poses a hazard in the event of overcharging, short circuit,
be left unattended on the charger. Batteries also have a limited lifespan, so t or physical
damage [10]. Batteries that are not certified by BATSO (Battery Safety Organization) should
ranty is usually limited to two years due to manufacturer-specific application sc
not be left unattended on the charger. Batteries also have a limited lifespan, so the warranty
Since
is 2014,
usually there
limited hasyears
to two beenduea growing movement in
to manufacturer-specific the EU scenarios.
application to standardize
Since batte
charging
2014, systems
there has been a used
growing in movement
LEVs (light electric
in the vehicles), batteries
EU to standardize and there andare plans to im
charging
systems used in LEVs (light electric vehicles), and there are
their requirements in WD 61851, Part 3-3: Requirements for Light Electric plans to implement their Vehicle
requirements in WD 61851, Part 3-3: Requirements for Light Electric Vehicles (LEV) battery
battery swap systems [25].
swap systems [25].
The
The “Bicycle
“Bicycle Kalkhoff
Kalkhoff GrooveGroove 28”,was
28”, which which
used was used as atest
as a real-world real-world
bed for thetest bed
analysis
analysis in in
thisthis article,
article, has used
has been beensince
used since
2014. 2014.
It has It has a replaceable
a replaceable battery locatedbattery
in the locate
vehicle
vehicle above
above the the
rear rear
wheel, Figure Figure
wheel, 2. 2.

Figure2. 2.
Figure “Kalkhoff
“Kalkhoff Groove
Groove 28” bicycle.
28” bicycle.

The bicycle battery, which is the subject of the further analysis, is shown in Figure 3, in
The bicycle battery, which is the subject of the further analysis, is shown in Fig
the form of a CAD modeled object placed in its holder/carrier assembly. The reason for the
the form
CAD of aofCAD
modeling modeled
the Kalkhoff object
battery placed
is closely in itstoholder/carrier
related the need for the assembly. The reaso
further analysis
CAD modeling
involving of the Kalkhoff
the parameterization battery is
of the geometric closelyofrelated
variables tocomponents
the circuit the need for the further
(battery
cells, enclosures, and BMS systems) and the application of the infrared
involving the parameterization of the geometric variables of the circuit components thermography
analysis. In Figure 3, and
cells, enclosures, the leftmost drawing shows
BMS systems) and thea top view of the holder
application of the and the inserted
infrared thermograp
battery, and the mid-left drawing shows a bottom view, while the mid-right and rightmost
ysis. In Figure
drawings 3, the
show what theleftmost drawing
battery itself shows
looks like whenaittop viewout
is pulled of of
the
theholder
holder.and the inse
tery, and the mid-left drawing shows a bottom view, while the mid-right and ri
drawings show what the battery itself looks like when it is pulled out of the holder
 Figure 3. Kalkhoff Groove 28” battery pack model.
Appl.
Appl. Sci.
Sci. 2022,
2022, 12,
12, 3444
3444 5 5ofof 20
20
3. Analysis and Diagnostics of the Battery Pack
3.1. Thermography in the Battery State Assessment and Analysis
Infrared thermography is a contactless method for determining the temperature dis-
tribution on the surface of an observed object by measuring the intensity of radiation in
the infrared region of the electromagnetic spectrum. Thermography is often used as a
maintenance tool. In the field of electric power engineering, thermography is an unavoid-
able test method closely associated with the measurement of electrical parameters [30].
Contrary to popular belief that an infrared thermal imager/camera measures temperature,
it actually registers electromagnetic radiation in the infrared portion of the spectrum, to
which itKalkhoff
Figure assigns temperature values depending on the imaging parameters. Figure 4
Figure 3. 3. Kalkhoff Groove Groove 28” 28” battery
battery pack pack model.
model.
graphically shows the components of the radiation registered by the infrared thermal im-
3.aging
3. Analysis
Analysis camera. and Diagnostics
and Diagnostics of of the
the Battery
Battery Pack
Pack
3.1. Thermography in the Battery State Assessment and Analysis
3.1. Thermography in the Battery State Assessment and Analysis
er-battery = ε Wbattery + (1–ε) Wambience
Infrared thermography is a contactless methodatmosphere for determining the temperature distri-
Infrared thermography is a contactlesssmethod for determining the temperature dis-
bution on theα1surface of an observed object by ameasuring
r
r
the intensity
ε τ Wbattery of radiation in the
tribution on the surface of an observed object by measuring(1–ε)
α1 ≠ α 2 the intensity of radiation in
τ is
Wambience
infrared region of the electromagnetic spectrum. τ Thermography often used as a mainte-
(1–τ) Watmosphere is often used as a
the infrared region of the electromagnetic er -battery spectrum. Thermography
ε Wbattery nance tool.
er-battery In the field of electric power engineering, thermography is an unavoidable test
maintenance tool. In the field of electric power Tatmosphere
engineering, thermography is an unavoid-
(1–ε) Wambience method closely associated with the measurement of electrical parameters [30]. Contrary to
α2
able test method closely associated with the measurement ofInfrared thermal camera
electrical parameters [30].
popular belief that an infrared thermal imager/camera er measures temperature, it actually
at a greater distance
Infrared thermal camera
Contrary to popular belief that an infrared thermal imager/camera measures temperature,
registers electromagnetic radiation in the infrared portion of the spectrum, to which it
near battery itassigns
actually registers electromagnetic
temperature values depending radiation in the infrared aportion
ambience on the imaging parameters.
of
Figurethe spectrum, to
radiation 4 graphically
T ambience εambience =1 r – absorbed
which it assigns temperature values
shows the components of the radiation registered by the infrared depending on the imaging
er – emittedparameters.
thermal
radiation
imaging Figure 4
camera.
sr – scattered radiation
graphically shows the components of the radiation registered by the infrared thermal im-
aging camera.

er-battery = ε Wbattery + (1–ε) Wambience atmosphere

sr ar
α1 ≠ α2 α1 ε τ Wbattery
τ (1–ε) τ Wambience
er -battery (1–τ) Watmosphere
ε Wbattery er-battery Tatmosphere
(1–ε) Wambience
α2 Infrared thermal camera
er at a greater distance
Infrared thermal camera
near battery
ambience ar – absorbed radiation
T ambience εambience = 1
er – emitted radiation
sr – scattered radiation

Figure 4. Components of longwave infrared radiation registered by the infrared thermal camera.
Figure 4.
Figure 4. Components
Components of
of longwave
longwave infrared
infrared radiation
radiation registered
registeredby
bythe
theinfrared
infraredthermal
thermalcamera.
camera.
Given the distance, the atmosphere may have an effect on the registered radiation.
Given the
Given thedistance,
distance,the theatmosphere
atmosphere maymay have
have an effect
anoptical
effect on on registered
the the registered radiation.
radiation. The
The distance of the camera depends mainly on the angle of the camera lens. The
The distance
distance of the ofcamera
the camera
dependsdepends
mainly mainly
on theon the optical
optical angle angle
of the of the camera
camera lens. lens.
The The
closer
closer the camera gets to the object to be analyzed, the weaker the influence of the atmos-
closer
the the camera
camera gets togets to the object
the object to be analyzed,
to be analyzed, the weaker
the weaker the influence
the influence of theof the atmos-
atmosphere
phere becomes, but the possibility of reflection of the operator’s radiation increases. In
phere becomes,
becomes, but the but the possibility
possibility of reflection
of reflection of the operator’s
of the operator’s radiationradiation
increases.increases.
In additionIn
addition to the fact that the power dissipation inside the battery is higher, its radiation
addition
to the facttothat
the the
factpower
that the power dissipation
dissipation inside theinside
batterytheis battery
higher, is itshigher,
radiationits radiation
becomes
becomes dominant, and the influence of the ambient radiation on the overall balance be-
dominant, and the influence
becomes dominant, and theofinfluence
the ambient radiation
of the ambientonradiation
the overall onbalance becomes
the overall smaller.
balance be-
comes smaller. The smaller distance to the object has another effect—namely, the radiation
The
comes smaller
smaller.distance to thedistance
The smaller object has another
to the objecteffect—namely, the radiationthe
has another effect—namely, captured at
radiation
captured at different angles relative to the normal on the surface. The acquisition angle is
different angles relative to the normal on the surface. The acquisition angle is important
important because of the change in emissivity, which can be seen in detail in Figure 5.
because of the change in emissivity, which can be seen in detail in Figure 5. Emissivity is
Emissivity is the most important parameter and can change depending on the viewing
the most important parameter and can change depending on the viewing angle or different
angle or different materials used. The emissivity also depends on the wavelength at which
materials used. The emissivity also depends on the wavelength at which the object is
the object is observed [31]. The emissivity of the plastic, used for the housing, is 0.95. It is
observed [31]. The emissivity of the plastic, used for the housing, is 0.95. It is confirmed
confirmed by sticking a strip of reference tape on the battery housing, which is not visible
by sticking a strip of reference tape on the battery housing, which is not visible on any
on anythermograms,
of the of the thermograms, thus confirming
thus confirming the choice
the choice of theofcorrect
the correct emissivity.
emissivity. In Figure
In Figure 5, in
5, in addition to the emission values at each angle shown with a red
addition to the emission values at each angle shown with a red line, we see (in brown) line, we see (in brown)the
the temperature
temperature valuevalue
thatthat
thethe camera
camera would
would associate
associate with
with thethe
samesame amount
amount of of radiation
radiation at
at different
different angles
angles for for
the the
spotspot measurement
measurement fromfrom
FigureFigure
6a. A6a. A significant
significant error inerror
the in the
choice
choice of emissivity
of emissivity would wouldresult inresult in the display
the display of an incorrect
of an incorrect temperature
temperature value, shown
value, shown in blue
in blue in
in Figure 5.Figure 5.
Appl.
Appl. Sci. 2022, 12,
Sci. 2022, 12, 3444
3444 66of
of 20
20

Appl. Sci. 2022, 12, 3444 Figure 5. Dependence of the emissivity of the battery pack and influence of emissivity changes on
7 of 20
the temperature displayed
the temperature displayed by
by the
the thermal
thermal camera.
camera.

We can conclude that in the case of plastic enclosures, variations in the angle of cap-
ture do not significantly affect the measurement result, especially if we take into account
the ±2 °C accuracy of the thermal imaging cameras used in this study, Table 1. Based on
the details of the thermograms, it is clear that a camera with a higher resolution was used
for the analysis. The higher resolution gives a clearer pattern but not necessarily better
measurement results.

Table 1. Comparison of cameras used for imaging and analysis, FLIR E60bx and E6.

Camera Flir E60bx Flir E6

IR resolution 320 × 240 pixels 160 × 120 pixels
(a) (b)
Thermal sensitivity/NETD <0.045 °C @ + 30 °C/45 mK <0.06 °C/<60 mK
Field of view (FOV) Figure
Figure6.6.Functional
Functional battery
batterypack
25° packcharging
× 19° charging thermal patterns
thermal nearnear
patterns the the
endend
of its
45° life,
×of34° (a) top
its life, (a) view (b)
top view
bottom.
(b) bottom.
Spatial resolution (IFOV) 1.36 mrad 5.2 mrad
Spectral range 7.5–13 μm 7.5–13 μm
Next,
We canirregularities
conclude that in in
thethethermal
case of patterns, shown invariations
plastic enclosures, Figure 6, are a consequence
in the angle of capture of
Object temperature range
changes in the –20 ohmic
internal °C to +resistance
120 °C of individual battery–20 °Castowell
cells +250as°Ctheir capac-
do not significantly affect the measurement result, especially if we take into account the
ity.
±2That ±2 °Cleads
◦ C accuracy
fact or ±2%
ofto
theofthermal
reading,and
improper for spatially
ambient
imaging tem-used±2 thermal
irregular
cameras °C
in or ±2%
this of reading,
dissipation
study, 1. for
Table within ambient
Based theonbat-
the
Accuracy
tery
details perature
packofconfinement 10 °C
space. When
the thermograms, to 35
it is clear °C
the that
battery temperature
startedwith
a camera to show 10 °C
signs resolution
a higher to 35
of reachingwas °Cthe used
end
of
foritsthe
service life, the
analysis. Thebicycles started to shut
higher resolution givesoff/down
a clearerunexpectedly.
pattern but not Soon after noticing
necessarily better
the Thermography
aforementioned
measurement results. easily
first findsthe
signs, its thermographic
application in the batteryofanalysis
analysis the bike[32]. Unfortunately,
battery pack was
the thermographic
performed, and thermal analysis is challenging
hot spots were noticed.in our case due to asymmetry
Additionally, the closed system, mainly
and spatial ir-
because
Table 1. the measurement
Comparison of camerasof electrical
used for parameters
imaging and is not
analysis, possible.
FLIR
regularity of heat dissipation within the pack now was augmented significantly, Figure 8. E60bx The
and consideration
E6. of
using the
Hot spots with thermographic analysis as
a temperature difference Flir a method
of ≥2 for evaluating the condition
°C represent a possible battery of the battery
Camera E60bx Flir E6 defect that
and diagnosing
warrants further faults arose from
investigation, the fact to
according thatNETAthe available range Electrical
(International is constantly decreasing
Testing Asso-
IR resolution 320 × 240 pixels 160 × 120 pixels
after
ciation). each
The full charge
pattern
Thermal sensitivity/NETD cycle.
disturbanceThus, is there
◦
located wasin a
the
<0.045 C @ + 30 C/45 mKneed
◦ for
spatial a non-invasive
center of the ◦ analysis
<0.06 C/<60 mK nor-
battery and method
is
that would not interfere with the ◦ × 19
functionality of◦ the battery under operating ◦ × 34◦conditions.
mal since it represents the core of the entire battery. Furthermore, the 45
Field of view (FOV) 25 comparison of the
Spatial
Figure 6 shows resolution (IFOV) 1.36 mrad 5.2 mrad near the
thermograms Spectralin arange
thermographic
Figures recording leads
7 and 8 inevitably of the battery pack during charging,
7.5–13 µm to the conclusion that a homogeneous
7.5–13 µm
endObject
heat of its temperature
life. The
distribution image
pattern was indicate
may
range taken when
−a ◦ Cthe
20healthy range
to +120 ◦ Cof the bike was
condition or proper significantly
−20 ◦ C to +250 ◦decreas-
battery behavior.
C
ing. Figure 7 shows the same battery ±2 ◦ Cand
or ± 2%
its ±2 ◦ Cduring
of reading, for pattern
thermographic or ±2%charging,
of reading,taken
for
Accuracy ambient temperature 10 ◦ C to ambient temperature 10 ◦ C to
when the battery was at the beginning of its ◦life, long before the images in◦ Figure 6 were
35 C 35 C
taken. The battery in Figure 7 shows a nearly uniform thermal pattern of a fully functional
battery that is in excellent condition. However, this uniform thermal pattern has a slight
Thermography
disturbance in the heat easily finds its application
distribution pattern causedin the battery
by the analysis [32]. Unfortunately,
spatial distribution of heat
the thermographic
sources in the battery analysis
pack oriscells.
challenging in our case due to the closed system, mainly
because the measurement of electrical parameters is not possible. The consideration of
using the thermographic analysis as a method for evaluating the condition of the battery

(a) (b)
 (a) (b)
Figure 6. Functional battery pack charging thermal patterns near the end of its life, (a) top view (b)
bottom.
Appl. Sci. 2022, 12, 3444 7 of 20

Appl. Sci. 2022, 12, 3444 Next, irregularities in the thermal patterns, shown in Figure 6, are a consequence 7 ofof20
changes in the internal ohmic resistance of individual battery cells as well as their capac-
ity. That fact leads to improper and spatially irregular thermal dissipation within the bat-
tery
andpack confinement
diagnosing faults space. When
arose from thethe battery
fact started
that the to show
available signs
range of reaching
is constantly the end
decreasing
ofafter
its service
each fulllife, the bicycles
charge startedthere
cycle. Thus, to shut
wasoff/down
a need for unexpectedly.
a non-invasiveSoon after noticing
analysis method
the
thataforementioned
would not interfere first signs, thefunctionality
with the thermographic analysis
of the batteryof under
the bike battery conditions.
operating pack was
Figure 6 shows
performed, a thermographic
and thermal hot spots recording
were noticed.of the battery packasymmetry
Additionally, during charging, near ir-
and spatial the
end of its of
regularity life.
heatThedissipation
image waswithin
taken when the now
the pack rangewasof the bike was significantly
augmented significantly,decreasing.
Figure 8.
Figure
Hot spots7 shows
with a the same battery
temperature and itsofthermographic
difference ≥2 °C representpattern during
a possible charging,
battery defecttaken
that
when thefurther
warrants batteryinvestigation,
was at the beginning
accordingoftoitsNETA
life, long before the Electrical
(International images inTesting
Figure 6Asso-
were
taken. The
ciation). Thebattery
patternindisturbance
Figure 7 shows a nearly
is located in uniform
the spatialthermal
centerpattern of a fully
of the battery functional
and is nor-
battery
mal sincethat is in (a)
it representsexcellent condition.
the core However,
of the entire this Furthermore,
battery. uniform
(b) thermal the pattern has aofslight
comparison the
disturbance in
thermograms in the heat 7distribution
Figures pattern
and 8 inevitably caused
leads to thebyconclusion
the spatialthat
distribution of heat
a homogeneous
Figure 6. Functional battery pack charging thermal patterns near the end of its life, (a) top view (b)
sources
heat in the battery
distribution patternpack
mayorindicate
cells. a healthy condition or proper battery behavior.
bottom.

Next, irregularities in the thermal patterns, shown in Figure 6, are a consequence of

changes in the internal ohmic resistance of individual battery cells as well as their capac-
ity. That fact leads to improper and spatially irregular thermal dissipation within the bat-
tery pack confinement space. When the battery started to show signs of reaching the end
of its service life, the bicycles started to shut off/down unexpectedly. Soon after noticing
the aforementioned first signs, the thermographic analysis of the bike battery pack was
performed, and thermal hot spots were noticed. Additionally, asymmetry and spatial ir-
regularity of heat dissipation within the pack now was augmented significantly, Figure 8.
Hot spots with a temperature difference of ≥2 °C represent a possible battery defect that
warrants further investigation, according to NETA (International Electrical Testing Asso-
(a)
ciation). The pattern disturbance is located in the spatial center of (b) the battery and is nor-
mal since
Figure 7. it represents
Thermal patternthe
of core
the of the
healthy entire
battery battery.
pack: Furthermore,
inside the bike holder
Figure 7. Thermal pattern of the healthy battery pack: inside the bike holder the(a)
comparison
(a)and outsideof
andoutside the
(b).
(b).
thermograms in Figures 7 and 8 inevitably leads to the conclusion that a homogeneous
Next, irregularities
heat distribution in the
pattern may thermal
indicate patterns,
a healthy shown in
condition orFigure
proper6,battery
are a consequence
behavior. of
changes in the internal ohmic resistance of individual battery cells as well as their capacity.
That fact leads to improper and spatially irregular thermal dissipation within the battery
pack confinement space. When the battery started to show signs of reaching the end of
its service life, the bicycles started to shut off/down unexpectedly. Soon after noticing
the aforementioned first signs, the thermographic analysis of the bike battery pack was
performed, and thermal hot spots were noticed. Additionally, asymmetry and spatial
irregularity of heat dissipation within the pack now was augmented significantly, Figure 8.
Hot spots with a temperature difference of ≥2 ◦ C represent a possible battery defect
that warrants further investigation, according to NETA (International Electrical Testing
Association). The pattern
(a) disturbance is located in the spatial center (b) of the battery and is
normal since it represents the core of the entire battery. Furthermore, the comparison of
Figure 8. Battery pack assembly thermal pattern, at the end of the lifetime, (a) top layer, (b) bottom layer.
the thermograms in (a) Figures 7 and 8 inevitably leads to the conclusion (b) that a homogeneous
heat distribution pattern may indicate a healthy condition or proper battery behavior.
Figure 7. Thermal pattern of the healthy battery pack: inside the bike holder (a) and outside (b).

(a) (b)
Figure 8. 8.
Figure Battery pack
Battery assembly
pack thermal
assembly pattern,
thermal at at
pattern, thethe
end ofof
end thethe
lifetime, (a)(a)
lifetime, top layer,
top (b)(b)
layer, bottom layer.
bottom layer.
 Appl.Sci.
Appl. Sci.2022, 12,3444
2022,12, 3444 8 of20
20
Appl. Sci. 2022, 12, 3444 88 ofof20

Sincethe
Since
Since thebattery
the batterywas
battery wasno
was nolonger
no longersuitable
longer suitablefor
suitable forcycling,
for cycling,itititwas
cycling, wasdisassembled,
was disassembled,and
disassembled, andthe
and the
the
internal circuits
internalcircuits
circuitsandand visual
andvisual condition
visualcondition
conditionof of the
ofthe battery
thebattery cells
batterycells were
cellswere analyzed.
wereanalyzed. Figure
analyzed.Figure 9 depicts
Figure99depicts
depicts
internal
the interior
the interior of
of the subjected
thesubjected battery
batterypack,
subjectedbattery pack,fully
fullydisassembled
disassembled with focus
with onon
focus areas where
areas the
where
the interior of the pack, fully disassembled with focus on areas where
hotspots
thehotspots are
hotspotsare noticed (suspected
arenoticed
noticed(suspected faulty
(suspectedfaulty cells
faultycellsand
cellsanda BMS
andaaBMS circuitry).
BMScircuitry).
circuitry).
the

(a)
(a) (b)
(b) (c)
(c)
Figure9.9.Disassembled
Figure Disassembled batterypack
pack (a),cells
cells thatproduce
produce thehot
hot spots(b),
(b), a BMScircuitry
circuitry (c).
Figure 9. Disassembledbattery
battery pack(a),
(a), cellsthat
that producethe
the hotspots
spots (b),a aBMS
BMS circuitry(c).
(c).

The
The battery
Thebattery pack
batterypack structural
packstructural analysis
structuralanalysis resulted
analysisresulted
resultedin in
inaaacircuit
circuit principal
circuitprincipal diagram,
principaldiagram, shown
diagram,shownshownin in
in
Figure
Figure 10.
10. The
The results
results can
can be
be summed
summed upup as
as follows:
follows: the
the battery
battery
Figure 10. The results can be summed up as follows: the battery pack consists of 40 cells pack
pack consists
consists of
of 40
40 cells
cells
Panasonic
Panasonic CGR18650CG
PanasonicCGR18650CG
CGR18650CGtype, type,cells
type, cellsnominal
cells nominalvoltage
nominal voltageisis3.6
voltage 3.6VVand
andnominal
and nominalcharge
nominal chargecapacity
charge capacity
capacity of
of 2250
of2250
2250mAh, mAh, individual
mAh,individual
individualcell cell
cell weight
weight
weight is isis
4545 45 g [33].
g [33].
g [33]. The The
The electrical
electrical
electrical connections
connections
connections arrangement
arrangement
arrangement was
was
was
made made
madein theinthe
in theform
form form often
of
of ten ten cell
cellcell blocks
blocks
blocks connected
connected
connected ininseries
in seriesseries (10s),
(10s),
(10s), and and
and each
each
each cell
cell
cell block
block
block con-
con-
consists
sists
sists of
of four four
of four parallel
parallel
parallel connected
connected
connected cells
cellscells
(4p)—a (4p)—a
(4p)—a so-called
so-called
so-called the thethe
4p10s 4p10s
4p10s battery
battery
battery cluster.
cluster.
cluster. TheTheThe
waywayway
cells
cells
cells arejoined
are are
joinedjoined together
together
together isiscrucial
crucial
is crucial forforforpacket
packet packet behavior
behavior
behavior [34,35].
[34,35].
[34,35]. TheThe balancing
balancing
The balancing leads
leads
leads connec-
connection
connec-
tion
tion points,
points, forthe
for for
points, the thesake
sake sake ofclarity,
of clarity,
of clarity,
are are aremarked
marked marked with
withwith colored
colored
colored circles
circles atatBMS
at the
circles theBMS
the BMS
andandand
in inbe-
between
in be-
tween cellblocks.
cellblocks. Each Each
balancingbalancing
line has line has
temporally temporally
thermal thermal
overcurrent
tween cellblocks. Each balancing line has temporally thermal overcurrent protection ac- overcurrent
protection protection
accomplished ac-
complished
by Polymerby
complished by
PTC Polymer
Resettable
Polymer PTC
PTCFuse Resettable
of JK-DFuse
Resettable Fuse
Series;of JK-D
of furthermore,Series; furthermore,
the BMS has two
JK-D Series; furthermore, the BMS has
additional
the BMS has
two
two additional
thermal thermal
monitoring
additional thermal and monitoring
protectionand
monitoring and protectionprobes.
probes.
protection probes.

Figure
Figure 10.
Figure10. Principal
10.Principal structural
Principalstructural diagram
structuraldiagram and
diagramand electric
andelectric schematic
schematicofof
electricschematic 4p10s
of4p10s battery
4p10sbattery pack.
batterypack.
pack.

The
The further
Thefurther analysis
furtheranalysis
analysiswas was followed
wasfollowed
followedby by connecting
byconnecting
connectingthe the open
theopen battery
openbattery
battery toto
to aacharger
acharger
charger device,
device,
device,
and
andatat
and the
atthe same
thesame time,
sametime,
time,thethe pack
thepack
packwaswas monitored
wasmonitored
monitoredwith with
withaaathermal
thermal imager
thermalimagerimagerduringduring
duringthe the charging
thecharging
charging
process—charging
process—chargingcycle.
process—charging cycle.
cycle.TheThe thermographic
Thethermographic
thermographicanalysis analysis
analysisof of the
ofthe thermal
thethermal pattern
thermalpatternpatternof of the
ofthe pack’s
thepack’s
pack’s
cells
cells shows
shows that
that the
the first
first two
two pack
pack cells
cells were
were significantly
significantly warmer
warmer
cells shows that the first two pack cells were significantly warmer than the adjacent cells than
than the
the adjacent
adjacent cells
cells
and
and that
that remedial
remedial action
action must
must be
be taken
taken according
according to
to [30].
[30].
and that remedial action must be taken according to [30]. The reason for building up the The
The reason
reason for
for building
building up
up the
the
heat in some is supported by the fact that ohmic resistance of
heat in some cells is supported by the fact that ohmic resistance of all cells in the block isis
heat in some cells is supported by the fact that ohmic resistance all
of cells
all cellsin the
in block
the is
block not
approximately
not
not approximately
approximately equalequalanymore—i.e.,
equal anymore—i.e.,
anymore—i.e., it is it
noticeably different.
itisisnoticeably
noticeably Instead
different.
different. of equally
Instead
Instead distributed
ofequally
of equally dis-
dis-
charge current
tributed
tributed charge among
chargecurrent all
currentamong (four)
among cells
all
all within
(four)
(four) cellsthe
cells block,
within
within the
the the current
block,
block, theis
the unevenly,
current
current reversely
isisunevenly,
unevenly,
proportionally
reversely
reversely distributed
proportionally
proportionally through the
distributed
distributed cells according
through
through thecells
the to their ohmic
cellsaccording
according totheir
to resistance.
their In our case,
ohmicresistance.
ohmic resistance.
only
In our two
case,cells
onlytaketwo all charge
cells takecurrent,
all and
charge other
current, two
and in the
other
In our case, only two cells take all charge current, and other two in the cell block have cell
two block
in thehavecell significantly
block have
higher ohmic
significantly resistance.
higher ohmic Contrary
resistance. to intuition
Contrary with
to infrared
intuition
significantly higher ohmic resistance. Contrary to intuition with infrared thermography, withthermography,
infrared the
thermography, healthy
cells
the
the warmcells
healthy
healthy up while
cells warm
warm the
updefective
up whilethe
while thecells look cooler,
defective
defective Figure
cellslook
cells look 11. Figure
cooler,
cooler, Figure11. 11.
 Appl.
Appl. Sci.
Appl.Sci. 2022,
Sci.2022, 12,
2022,12, 3444
12,3444 9of
9 9of of2020

(a)
(a) (b)
(b)
Figure
Figure11.
Figure 11.Overheated
11. Overheated
Overheated healthy
healthy
healthycells
cellsthat
that
cells conduct
conduct
that conductthe overall
thethe
overallbattery
battery
overall pack
pack
battery current
current
pack (a(aview
current viewfrom dif-
fromfrom
(a view dif-
ferent
ferentperspectives),
perspectives),
different (a)
(a)from
perspectives), from
(a) rear
rear
from side
side
rear and
and
side (b)
(b)from
and from
(b) top
from side
toptop
side
side.

Further,
Further,ininorder
ordertotodetermine
determinethe thecondition
conditionofofotherothercells
cellsinincellblocks,
cellblocks,aadifferent
different
indirect
indirectthermal
thermaland andelectrical
electricalparameter
parameterobservation
observationmethod methodhad hadtotobe beused.
be used.For
used. Forthis
this
purpose,
purpose,aabalancing
balancingresistor
resistorthermal
thermalthrottling
throttlingobservation
observationcombined
combinedwith withthe
thethermal
thermal
imaging
imaginganalysis
analysiswaswasapplied.
applied.The Thethermogram
thermogramshown shownininFigure
Figure12a 12aillustrates
illustratesthatthatthere
there
isisno
no dissipation
no dissipation
dissipationon on the
onthe two
thetwo resistors,
tworesistors, which
resistors,which
whichmeans
means
meansthat
that only
only
that two
only parallel
twotwoparallel connections
parallelconnections
connections are
are
still
stillin
are inthe
still incharging
the process
the charging
charging and
andthat
process
process thatmost
and that of
most ofthe
most parallel
theof connections
the parallel
parallel are
areininaare
connections
connections asimilar state.
in a similar
similar state.
Additionally,
state.
Additionally, that
thatpoints
Additionally, thatto
points tothe
thefact
points that
to the
fact the
fact
that battery
thethat theblocks
battery blocksare
battery not
nothomogeneous
blocks
are are inintheir
not homogeneous
homogeneous their
electrical
in
electrical (resistive/ohmic)
their electrical properties.
(resistive/ohmic)
(resistive/ohmic) properties. Figure
Figure1212Figure
properties. visually 12describes
visually visually
describes the
thedynamics
describes ofofBMS
the dynamics
dynamics BMS
of BMS operation
operation
operation asaswell as
wellasas well as dissipation,
dissipation,
dissipation, which
whichleads which
leads totoanleads
an to aninin
increase
increase increase
the
thepack
packinambient
the pack
ambient ambient
tempera-
tempera-
temperature ◦ C, while the maximum temperature of the active resistor at the time of
of 6 the
ture
tureofof66°C,°C,while
while themaximum
maximumtemperature
temperatureofofthe theactive
activeresistor
resistoratatthethetime
timeofofrecording
recording
recording increases by 19.6 ◦ C.
increases
increasesby by19.6
19.6°C.
°C.

(a)
(a) (b)
(b) (c)
(c)
Figure
Figure12.
Figure 12.BMS
12. BMSthermal
BMS thermalload
thermal loadduring
load duringcharging
during chargingprocess
charging processat:
process at:(a)
at: (a)1:45
(a) 1:45PM;
1:45 PM;(b)
PM; (b)1:58
(b) 1:58PM;
1:58 PM;(c)
PM; (c)2:05
(c) 2:05PM.
2:05 PM.
PM.

3.2.Battery
3.2. BatteryCells
CellsElectric
ElectricParameters
Parametersand andFunctional
FunctionalAnalysis
Analysis
ToTocomplete
completethe theresearch,
the research,further
research, further
further analysis
analysis
analysis steps
steps
steps werewere
were taken
taken
taken torestore
restore
totorestore the
thethe function-
functional-
functional-
ality
ity of
ityofofthe
thethe battery
battery
battery and
andand
make make
make it operational
ititoperational
operational again.
again.
again. TheThe
The stepssteps
steps thatthat
that followed
followed
followed included
included
included the the
there-re-
replacement
placement
placementofofthe of
thethe defective,
defective,
defective, faulty faulty
faulty cell cell
withwith
cellwith a healthy
aahealthy
healthy one. one.
one.Prior
Priorto Prior
tothat,
that,toaathat, a determina-
determination
determination ofof
tion of electrical
electrical
electrical properties
propertiesproperties
ofofeach of
eachcell each
cell hadcell
had totobehad
be to be conducted
conducted
conducted and and correspondingly
andcorrespondingly
correspondingly their their
theirspatial
spatial
spatial
arrangement arrangement
within within
the pack the
had pack
to be had to
planned. be planned.
In this case,In a
arrangement within the pack had to be planned. In this case, a disassembly and the meas- this case,
disassembly a disassembly
and the and
meas-
the
urement measurement
urementofofall of
all4040batteryall 40
batterypack battery
packcells pack
cellswas cells
wascarried was
carriedout; carried
out;their out;
theirstate
stateandtheir state
andelectrical and electrical
electricalcondition
condition
condition
(internal
(internalohmic (internal
ohmic ohmic resistance
resistance
resistance and
andcharge
charge and charge capacity)
capacity)
capacity) were were determined
weredetermined
determined using
usingthe using the uni-
theuniversal
universal
versal
charger charger
chargerJUNS JUNS
JUNSiCharger iCharger
iCharger1010B+1010B+and 1010B+
andterminalsand terminals
terminalswith withfresh,with
fresh,cleanfresh,
cleancontact clean contact
contactsurface surface
surfacetotoachieve
achieve to
achieve
the
theleast the
leastamount least amount
amountofofinfluence of
influenceon influence
onthethecontacton the contact
contactresistance. resistance.
resistance.ItItwaswasaadifficult It was
difficultand a difficult
andaatime-con- and
time-con-
a time-consuming
suming process, but in the interest of the research
suming process, but in the interest of the research and the fact that it was a singlebattery
process, but in the interest of the research and the fact and
that it the
was fact
a that
single itbattery
was a
single
pack,
pack,the battery
thecompletepack, the
completedetailed complete
detailedanalysis detailed
analysiswas analysis
wasnecessary. was necessary.
necessary.InInthe theattempt In the
attempttotoreviveattempt to revive
revivebatteries,
batteries,
batteries,
which
whichinvolves whichthe
involves involves
the managementthe management
management ofoflarger of larger
largerbattery
battery battery quantities,
quantities,
quantities, ititisisnecessary
necessaryit is to
necessary
touse to use
usemethods
methods
methods
ofofcell of
cellselectioncell selection
selectionbased
basedon onthe based on
themeasurement the measurement
measurementofofresistance of resistance
resistanceororX-rays
X-rays[36]. or X-rays
[36].Prior [36].
Priortotothe Prior
theanalysis
analysis to
Appl. Sci. 2022, 12, 3444 10 of 20
Appl.Appl. Sci. 2022,
Sci. 2022, 12, 3444
12, 3444 1010
of 20
of 20

continuation,
the analysiswe
continuation, wenumerically
continuation, designatedand
we numerically
numerically designated andlabeled
labeled
designated allall
and battery
labeled
battery all pack
cellscells
battery
pack pack for
cellseasier
for easier anal-anal-
for
ysiseasier analysis management,
management, Figure 13. Figure 13.
ysis management, Figure 13.

Figure 13. Labels of individual cells in the pack arrangement, an assembly (b); a top cell layer TOP
Figure 13. 13.
Figure Labels
Labels ofofindividual cellsinin
individual cells thethe pack
pack arrangement,
arrangement, an assembly
an assembly (b); a top(b); a topTOP
cell layer cell(a);
layer TOP
(a); a bottom cell layer BOT (c).
(a); aa bottom celllayer
bottom cell layer BOTBOT
(c). (c).

The
Thefirst
first two faultycells,
two faulty cells,marked
marked with
with 1 and1 and 3 in Figure
3 in Figure 13, show
13, do not do not showwithin
voltage voltage
Theexpected,
within first twonormal faulty operating
cells, markedrange. with
The 1 and
cells 3 in Figure
discharged 13, do
during the not show voltage
measurement
expected, normal operating range. The cells discharged during the measurement despite
within
despite expected,
theinternal
the high normal
high internal
resistanceoperating
resistance ofrange.
theAn
of the device. The An
device.
examplecells discharged
example
of an of an
extreme during
is extreme
cell isthe
1 whose cellmeasurement
1 whose
diagram
despite
of discharge, via a voltmeter, is shown in Figure 14a. When cell 1 was removed fromcell
diagram the ofhigh internal
discharge, viaresistance
a voltmeter,of the
is device.
shown in An example
Figure 14a. of
When an extreme
cell 1 was is the1 whose
removed
diagram
from theofcellblock,
cellblock, discharge,
it mechanically viacaused
a voltmeter,
it mechanically caused
a slight is ashown in Figure
slight perforation
perforation of the 14a.
of tab
battery the When
battery
weld cell
tab
[37], 1 weld
was [37],
making removed
it
from unsafe
making ittounsafe
handletofurther.
the cellblock, ithandle During
further.
mechanically the cell recharge,
During
caused the cellcell
a slight 3 received
recharge, cellvoltage
perforation 3of the values
received of tab
voltage
battery 5.76 V,
values
weld [37]
ofwhich
5.76 V,is significantly
which is higher
significantlythan the
higher maximum
than the allowed
maximum values of
allowed 4.2
making it unsafe to handle further. During the cell recharge, cell 3 received voltage values V,
valuesFigure
of 14b.
4.2 V, The
Figure
fact thatfact
there was no inflammation clearly indicates the degree the
of degradation.
of14b.
5.76TheV, which thatis there was no inflammation
significantly higher than clearly indicates
the maximum alloweddegree of degradation.
values of 4.2 V, Figure
14b. The fact that there was no inflammation clearly indicates the degree of degradation.

(a) (b)
Figure
Figure14.14.Cell
CellNo.
No.1 1high
highimpedance
impedance load dischargegraph
load discharge graph(a)(a)and
and overcharged
overcharged cellcell
No.No. 3, temper-
3, tempera-
ature
tureand
andstructurally (a) (b).
stable
structurally stable (b). (b)
Figure 14. Cell
The No. 1 high
behavioral impedance
properties, load dischargecharacteristics,
electro-chemical graph (a) and overcharged cell No. 3, temper
The behavioral properties, electro-chemical characteristics,and andenergy
energy transforma-
transforma-
aturetional
and structurally
mechanisms stable
of (b).
battery cells based on Li-ion chemistry are well described and ex-ex-
tional mechanisms of battery cells based on Li-ion chemistry are well described and
plained in detail in the articles by Xuning et al. in Thermal runaway
plained in detail in the articles by Xuning et al. in Thermal runaway mechanism of lithium-mechanism of lithium-ion
The
ionbatterybehavioral
forfor
battery electric
electric properties,
vehicles
vehicles[38], and
[38], electro-chemical
by Minggao
and by Minggao et al.etcharacteristics,
in and capacity
al.Overcharge-induced
in Overcharge-induced energy transforma-
fading
capacity fad-
tionalanalysis for large format
mechanisms of lithium-ion
battery batteries
cells basedwith
onLiyNi1/3Co1/3Mn1/3O
Li-ion chemistry 2 + LiyMn
are well 2 O composite and ex-
described
4
ing analysis for large format lithium-ion batteries with LiyNi1/3Co1/3Mn1/3O2 + LiyMn2O4 com-
cathode
plained in[39].
detail According
in According to the behavioral
the articles properties of such battery cells formechanism
the measured
posite cathode [39]. tobytheXuning et al.properties
behavioral in Thermal ofrunaway
such battery cells for of the
lithium-
voltage of 5.76 V, a temperature of more than 50 ◦ C is expected to be reached, but the
ionmeasured
battery for voltage of 5.76 V, a temperature of more than 50 °C is expected to be reached, fad-
electric vehicles [38], and by Minggao et al. in Overcharge-induced capacity
cell remained at ambient temperature. That 50 ◦ C temperature marks a breaking point
ingbut
analysis
when cellfor
the the cellslarge
startformat
remained lithium-ion
toatchange
ambient their batteries
temperature.
structure with
That LiyNi1/3Co1/3Mn1/3O
mechanically 50 °Cand
temperature marks
the deformation 2 +a
ofLiyMn cell2O4 com-
breaking
the
point
posite when
cathode
becomes the cells start
[39]. According
visually to
noticeable. The change their
to absence structure
the behavioral
of the above mechanically
properties
indicates of and the deformation
such battery
a significantly cells for
impaired of the
thecapacity
measuredcell becomes
voltage visually
of the cell.of Figure
5.76 V, noticeable.
15adepicts aThe
temperature absence
of more
charge-thermal of dynamic
the
thanabove °Cindicates
50 of an a significantly
isovercharged
expected tocell
beinreached
impaired
but relation capacity
the celltoremained of
its measured the voltage. temperature. That 50 °C temperature marks aover-
cell.
at ambient Figure 15 depicts a charge-thermal dynamic of an breaking
charged cell in relation to its measured voltage.
point when the cells start to change their structure mechanically and the deformation of
the cell becomes visually noticeable. The absence of the above indicates a significantly
impaired capacity of the cell. Figure 15 depicts a charge-thermal dynamic of an over-
charged cell in relation to its measured voltage.
ppl. Sci. 2022, 12, 3444
Appl. Sci. 2022, 12, 3444 11 of 20

Appl. Sci. 2022, 12, 3444 11 of 20

Figure 15. Li-9on cell charging process and stages of overcharge states [38,39].
15.Li-9on
Figure 15.
Figure Li-9oncell cell
charging process process
charging and stagesand
of overcharge
stages ofstates [38,39].
overcharge states [38,39].
The
The measurement
measurement setup,setup, shown
shown in in Figure
Figure 16,
16, consists
consists of of aa universal
universal Li-ion
Li-ion charger
charger
The measurement
(leftmost),
(leftmost), the JUNS iChargersetup,
the JUNS iCharger 1010B+,shown
1010B+, and in Figure
current
and current clamps 16, consists
(middle
clamps (middle image) of aaauniversal
image) with
with clean
clean L
contact
contact surface
surface for
for achieving
achieving the
the lowest
lowest possible
possible clamp
clamp
(leftmost), the JUNS iCharger 1010B+, and current clamps (middle image) contact
contact resistance.
resistance. Furthermore,
Furthermore,
the
the same
same figure
figure (rightmost)
(rightmost) shows
shows the the contents
contents ofof the
the disassembled
disassembled e-bike e-bike battery
battery pack
pack
contact
without2surface
without 2pack-origin fordefective
pack-origin achieving
defective cells
cells the
and lowest
with
and 10 possible
10 replacement
with replacement clamp
cells acquired
cells contact theresistance.
from from
acquired used
the F
the
usedsame
laptop laptop figure
batterybattery (rightmost)
packs. Replacement
packs. Replacement shows
cellscellsthe
from the
from contents
used
the used of the
battery
battery packs disassembled
packs were originallye-bike
were originally
planned to be a replacement source and eventually
without 2 pack-origin defective cells and with 10 replacement cellsof
planned to be a replacement source and eventually to
to replace
replace all
all defective
defective cells
cells ofacquired
the
the f
initial e-bike
initial e-bike battery
battery pack
pack [40].
[40]. Among
Among newlynewly sourced
sourced cells,
cells, we
we found
found twotwo battery
battery cells
cells
laptop
that
battery
that matched in
packs. Replacement
inappearance
appearanceand andlabeled
labeled exactly
cells
as those
exactly
from
as thosein the
the
in original
usede-bike
the original
battery
battery
e-bike
packs we
pack,
battery
planned
so weso
pack, wetoplaced
placed be
themathem
atreplacement
cellatpositions 1 source
cell positions in and
and1 3and the eventually
3 ininitial
the battery
initial topack
pack
battery replace
layout. With
layout. all defectiv
that
With
substitution, the battery pack was restored back to the
initial e-bike battery pack [40]. Among newly sourced cells, we found two
that substitution, the battery pack was restored back to operational
the operational state, thus
state, making
thus making the
further
the investigation
further and analysis
investigation possible
and analysis to continue
possible and theand
to continue battery pack usable
the battery packagain to
that
some
matched
extent.
in appearance and labeled exactly as those in theusable
original e
again to some extent.
pack, so we placed them at cell positions 1 and 3 in the initial battery pack
that substitution, the battery pack was restored back to the operational state,
the further investigation and analysis possible to continue and the battery
again to some extent.

(a) (b) (c)

Figure 16. A system charger (a); battery cell current clamp (b); and disassembled battery pack cells (c).
Figure 16. A system charger (a); battery cell current clamp (b); and disassembled battery pack cells (c).

Table
Table 22 shows
shows the
the discharge
discharge parameters
parameters of of all
all cells
cells that
that were
were previously
previously charged
charged toto
nominal
nominal maximum voltage of
maximum voltage of 4.2
4.2V.
V.The
Thecells
cellswere
weredischarged
dischargedwith with current
current ofof2A2Ato to volt-
voltage
age
of 3ofV3[41].
V [41]. That
That specific
specific valueofofdischarge
value dischargecurrent
currentisisthethevalue
value closest
closest to
to finding
finding the
the
maximum
maximum capacity under real conditions and also the value expected for the maximum
capacity under real conditions and also the value expected for the maximum
battery
battery load
load under
under dynamic
dynamic operating
operating conditions
conditions for
for an
an engine
engine at
at the
the typical
typical speed
speed shown
shown
(a) in
(b) (c)
in Figure
Figure 1. 1. According
Accordingto tothe
thetechnical
technicaldata
data[33],
[33],the
thecut-off
cut-offvoltage
voltageisis2.50
2.50V,V, but
but after
after
reaching
reaching
Figure 16.the voltage
Avoltage
systemofofcharger
33V,
V,the
thecapacity
capacity
(a); change
change
battery isis
cell significantly
significantly
current less
less
clamp noticeable,
noticeable,
(b); which
which fur- battery
further
and disassembled
ther confirms
confirms the guidelines
the guidelines of [41].
of [41].

TableTable 2 pack
2. Battery shows the discharge
cell parameter parameters
values during 2 A discharge.of
all cells that were previousl
nominal maximum voltage of 4.2 V. TheRemaining
Group cells were discharged
Cell with current o
Cell Resistance Capacity Discharge
age of 3 V [41]. That specific value of discharge
Capacity Capacity current
Voltageis the value closest t
No. R (mΩ) C (mAh) Time (min)
maximum capacity under real (mAh) conditions(mAh)
and also U (V)value expected for th
the
1
battery load114 351
under dynamic 1067
operating 15.60
conditions for3.480
an engine 13 at the typical
2 101 119 5.29 3.960 5
in Figure
3 1.114
According350 to the technical data 15.56[33], the cut-off voltage
3.640 15 is 2.50
reaching
4 the101
voltage of247
3 V, the capacity change
10.98 is significantly
3.820 less noticeabl
11
ther confirms the guidelines of [41].
Appl. Sci. 2022, 12, 3444 12 of 20

Table 2. Battery pack cell parameter values during 2 A discharge.

Group Remaining Cell Discharge

Resistance Capacity
Cell No. Capacity Capacity Voltage Time
R (mΩ) C (mAh)
(mAh) (mAh) U (V) (min)
1 114 351 1067 15.60 3.480 13
2 101 119 5.29 3.960 5
3 114 350 15.56 3.640 15
4 101 247 10.98 3.820 11
5 141 268 1105 11.91 3.600 11
6 101 278 12.36 3.670 12
7 103 259 11.51 3.740 11
8 105 300 13.33 3.720 12
9 101 277 934 12.31 3.640 12
10 98 285 12.67 3.710 12
11 122 132 5.87 3.900 6
12 106 240 10.67 3.810 10
13 103 185 1411 8.22 3.880 8
14 90 895 39.78 3.550 29
15 103 185 8.22 3.910 8
16 122 146 6.49 3.990 6
17 103 243 1193 10.80 3.754 10
18 106 248 11.02 3.685 10
19 103 229 10.18 3.743 10
20 101 473 21.02 3.556 17
21 90 1101 1995 48.93 3.398 34
22 103 286 12.71 3.710 12
23 99 292 12.98 3.789 12
24 109 316 14.04 3.689 13
25 95 353 1502 15.69 3.783 14
26 97 315 14.00 3.578 19
27 95 529 23.51 3.607 13
28 106 305 13.56 3.724 12
29 111 229 1175 10.18 3.841 9
30 118 136 6.04 3.929 6
31 91 258 11.47 3.634 11
32 99 552 24.53 3.688 20
33 101 267 762 11.87 3.601 11
34 112 128 5.69 3.776 6
35 112 120 5.33 3.840 5
36 99 247 10.98 3.769 10
37 104 249 1051 11.07 3.860 11
38 97 260 11.56 3.830 11
39 99 319 14.18 3.794 13
40 102 223 9.91 3.778 9

The evaluation of the cells’ condition and the parameter measurements showed that
the cells’ condition in the pack was poor. The best one of all cells had only 48.93% of the
initial and nominal capacity. Despite the fact that the resistance values, read by the charger,
depend significantly on the cell’s electromechanical connection [42], it still can be confirmed
that the cells with a higher capacity had the lowest value of internal resistance [43]. The
current–voltage (C-V) discharging dynamics of the best cell, cell no. 21, and the cell with
the highest capacity, are shown in Figure 17a.
 Appl.
Appl. Sci.Sci.
Sci. 2022,
12, 12,
2022,
2022, 34443444 13 13 of
of 20
of1320 20

6
6 6 1200
Capacity (mAh) 6 1200
Capacity (mAh) Capacity (mAh)
Voltage (V)

5 Capacity (mAh)
Voltage (V)

Voltage (V)
5 5 1000

Voltage (V)
5 1000

Capacity (mAh)
Capacity (mAh)
4

Capacity (mAh)
4 4 800

Capacity (mAh)
4 800
Voltage (V)
Voltage (V)
3 3 600
Current (A)

3 3 600
Current (A)

Current (A)
Voltage (V)

Current (A)
Voltage (V)
2 2 400
2 2 400
Current (A) Current (A)
Current (A) Current (A)
1 1 200
1 1 200
0 0
0 0 0

5m 20s 0
0

40s
20s
00s

20s
00s
40s
1h 09m 20s
1h 12m 00s
40s
20s
00s
40s
20s
00s
40s
20s
00s
40s
20s
00s
40s
20s
00s
40s
20s
00s
40s
0
2m 20s 0

2m 40s

2m00s

56m20s
1h 04m 00s
01m40s
04m20s
06m00s
5m40s
8m20s
16m 00s
13m40s
16m20s
18m00s
21m40s
24m20s
26m00s
34m 40s
32m20s
40m 00s
37m40s
45m 20s
42m00s
45m40s
48m20s
50m00s
58m 40s
30s
40s
50s
00s
10s
20s
30s
40s
50s
00s
10s
20s
30s
40s
50s
00s
10s
20s
30s
32m 40s
33m 50s
10s
20s
30s
40s
50s
00s
10s
20s
0

10m 30s
9m40s
12m 50s
11m00s
12m10s
14m20s
15m30s
18m 40s
17m50s
21m 00s
19m10s
21m20s
24m 30s
23m40s
26m 50s
25m00s
26m10s
28m20s
29m30s
32m 40s
31m50s
1m 10s

1m30s
2m40s
3m50s
7m 00s
5m10s
7m20s

10m

29m

34m

40m

53m

58m
8m

1h 01m

1h1h06m
1h1h09m
1h1h12m
10m
13m

18m
21m
24m
26m
29m
32m

37m

42m

48m
50m
53m
56m
10m

16m

18m

22m

24m

30m
4m

8m
11m

14m
15m
16m
17m

19m

22m
23m

25m

28m
29m
30m
31m

33m
3m
4m
5m

8m
9m

Discharge time h:m:s Charge time h:m:s

Discharge time h:m:s Charge time h:m:s

(a)(a) Discharging
Discharging at 2atA2 rate
A rate (b)(b) Charging
Charging at 1atA1rate
A rate
Figure
Figure 17.17.
Figure 17. Comparison
Comparison
Comparison of
of discharge
of discharge (a) (a)
discharge andand
(a) charge
charge
and (b)(b)
charge current–voltage
current–voltage
(b) (C-V)
(C-V)
current–voltage characteristics
characteristics
(C-V) of
of best
of best
characteristics best
performing
performing cell,
cell,cell, cell
cellcell No.
No.No. 21.
21. 21.
performing

Finally,
Finally,
Finally, discharge-based
discharge-based
discharge-based capacity
capacity
capacity assessment of of
assessment
assessment all
allof cells
all
cells within
cells
within thethe
within pack
the
pack led
pack
led to to
aleda to
vis-a
vis-
ually
ually representative
visually spatial
representative
representative allocation
spatialspatial of
allocation
allocation cell capacities for
of cell capacities
of cell capacities both
for bothfor pack
both
pack assembly layers
pack assembly
assembly layers (TOP (TOP
layers
andand
(TOP BOT),
BOT), and Figure
BOT),
Figure 18.18.
Figure
TheThe
18.results confirm
Theconfirm
results results thethe
confirm poor
poor the state
poor
state of of the
state
the pack
of
pack the assembly and
pack assembly
assembly and aand
a highhigha
level
high of degradation.
level of degradation.
level of degradation.

(a)(a) (b)(b)
Figure
Figure 18.18. Spatial
Spatial distribution
distribution of TOP/upper-
of TOP/upper- (a) (a)
andand BOT/lower-layer
BOT/lower-layer (b)(b) pack
pack cells
cells capacities.
capacities.
Figure 18. Spatial distribution of TOP/upper- (a) and BOT/lower-layer (b) pack cells capacities.

In In
In order
order
order to
to to confirm
confirm
confirm cells’
cells’
cells’ capacities
capacities
capacities determined
determined
determined with
with
with the
the
the discharging
discharging
discharging method,
method, a sec-
a second
sec-
ondond approach
approach
approachwas was employed
wasemployed to
employedtotovalidatevalidate
validatethe the results.
theresults.
results.ThatThat approach
Thatapproach
approachrelied relied
reliedon on
onand and utilized
andutilized
utilizedthe
thethe charge
charge
charge counting
counting
counting method
method
method whilewhile
while the the
thecell cell
cellwas
waswas charged
charged
chargedupuptoupitsto
to its
maximum
its maximum
maximum voltagevoltage
butbut
voltage but
with
with current
current that that
was was
lower lower
in in
relationrelation
to the to the
nominal nominal discharge
discharge current
with current that was lower in relation to the nominal discharge current of 2 A—i.e., current
of 2 of
A—i.e.,2 A—i.e.,
charge
charge
current
charge current
of 1 A
current of
1 A1used.
ofwas A was
was Theused.
used. The
results
The ofresults
resultsthe of of
thethe capacities
capacities determined
capacities determined
in this in
determined wayin this
are
this way
shown
way areare
in
shown in Table
Tablein3.Table 3.
shown 3.

Table
Table 3. Battery
3. Battery pack
pack cells
cells parameter
parameter values
values during
during charge.
1 A1 charge.
A charge.

Resistance Capacity
Resistance Capacity Group
Group Remaining
Group Ca- Cell
RemainingCell Voltage Charge
Cell Charge Time
Charge
Cell
Cell No.
No. Resistance Capacity Remaining Ca- Voltage Time
R (mΩ) Cell No.
C (mAh) Capacity Capacity
(mAh) pacity
pacity Capacity
(mAh) Voltage
U (V) Time
(min)
R (mΩ) C (mAh) Capacity
R (mΩ) (mAh)
C (mAh) (mAh) U (V) (min)
(mAh) (mAh) 4.140U (V) (min)
1 1 113113 115115 937937 5.11
5.11 4.140 16 16
1 113 115 93715.11 5.11 4.1204.140 16
2 2 101101 340340 15.11 4.120 34 34
2 101 340 15.11 4.120 34
3 3 109109 3 277
277
109 277
12.31
12.31 4.120
12.31 4.120 4.120 26 2626
4 4 101101 4 205 205 101 205 9.11
9.11 4.100
9.11 4.100 4.100 23 2323
5 5 101101 5 174
174
101 737737174 7.73
7.73
737
4.090
7.73 4.090 4.090 20 2020
6 6 101101 6 189 189 101 189 8.40
8.40 4.120
8.40 4.120 4.120 23 2323
7 7 103103 7 156 156 103 156 6.93
6.93 4.100
6.93 4.100 4.100 18 1818
8 106 218 9.69 4.130 27
8 8 106106 218218 9.69
9.69 4.130
4.130 27 27
9 9 101101 9 207207 101 668668207 668 9.20
9.20 4.1904.190
9.20 4.190 23 2323
10 98 226 10.04 4.140 25
10 10 98 98 11
226226 117 68
10.04
10.04 3.02
4.140
4.140 4.140
25 259
11 11 117117 12 68
68
111 167
3.02
3.02 4.140
7.42 4.140 4.120 9 920
12 12 111111 167167 7.42
7.42 4.120
4.120 20 20
 Appl. Sci. 2022, 12, 3444 14 of 20

Appl. Sci. 2022, 12, 3444 14 of 20

Table 3. Cont.

13 103 121 Resistance

1131Capacity 5.38
Group 4.130 Cell
Remaining 16Charge
14 95 Cell No.
831 Capacity
36.93 4.110 Voltage
Capacity 61 Time
R (mΩ) C (mAh)
(mAh) (mAh) U (V) (min)
15 103 105 4.67 4.130 12
16 122 13 74 103 121 1131
3.29 5.38 4.130 4.130 11 16
14 95 831 36.93 4.110 61
17 105 15
162 103
899 105
7.20 4.67
4.120 4.130
19 12
18 105 16 163 122 74 7.24 3.29 4.110 4.130 20 11
19 109 17 160 105 162 7.11
899 7.20 4.110 4.120 21 19
20 105 18 414 105 163 18.40 7.24 4.130 4.110 37 20
21 89 191049 109 1693 160 46.62 7.11 4.140 4.110 14 21
22 103 20 203 105 414 9.02 18.40 4.150 4.130 23 37
23 93 21 211 89 1049 1693
9.38 46.62 4.130 4.140 23 14
24 103 22 230 103 203 10.22 9.02 4.130 4.150 25 23
23 93 211 9.38 4.130 23
25 90 24 310 103 1297 230 13.78 10.22 4.130 4.130 32 25
26 106 246 10.93 4.130 25
25 90 310 1297 13.78 4.130 32
27 92 26 520 106 246 23.11 10.93 4.140 4.130 46 25
28 106 27 221 92 520 9.82 23.11 4.110 4.140 25 46
29 117 28 141 106 993 221 6.27 9.82 4.110 4.110 17 25
30 118 29 94 117 141 4.18
993 6.27 4.130 4.110 14 17
31 101 30 193 118 94 8.58 4.18 4.120 4.130 21 14
31 101 193 8.58 4.120
32 100 565 25.11 4.140 48 21
32 100 565 25.11 4.140 48
33 103 207 654 9.20 4.110 24
33 103 207 654 9.20 4.110 24
34 109 84 3.73 4.110 11
34 109 84 3.73 4.110 11
35 109 35 92 109 92 4.09 4.09 4.140 4.140 14 14
36 103 36 271 103 271 12.04 12.04 4.110 4.110 30 30
37 109 37 146 109 760 146 6.49
760 6.49 4.130 4.130 17 17
38 103 38 187 103 187 8.31 8.31 4.110 4.110 21 21
39 103 39 262 103 262 11.64 11.64 4.130 4.130 29 29
40 103 165 7.33 4.140 19
40 103 165 7.33 4.140 19

TheThe results
results in Table
in Table 3, a3,charge-based
a charge-based capacity
capacity assessment
assessment approach,
approach, validate
validate andand
confirm
confirm thethe results
results of the
of the cellcell capacity
capacity determination
determination method
method thatthat utilized
utilized thethe discharging
discharging
approach.
approach. Finally,
Finally, a charge-based
a charge-based approach
approach yielded
yielded similar
similar visually
visually representative
representative cellcell
spatial mapping for both pack assembly layers (TOP and BOT), as shown in Figure 19.19.
spatial mapping for both pack assembly layers (TOP and BOT), as shown in Figure

(a) (b)
Figure 19. TOP/Upper-
Figure (a) and
19. TOP/Upper- (a)BOT/lower-layer (b) cell capacities
and BOT/lower-layer determined
(b) cell capacities with 1 A cells
determined with charg-
1 A cells
ing charging
method. method.

TheThe
results support
results support thethe
confirmation
confirmationofofthe thepoor
poorpack
pack assembly
assembly state and andcells
cellsdegrada-
deg-
radation determined
tion determined bybythethe discharging
discharging method.
method. TheThe comparison
comparison of cells
of cells capacities,
capacities, de-
determined
termined
with the charging and discharging method, shows a capacity deviation in accordanceac-
with the charging and discharging method, shows a capacity deviation in with
cordance with the expectation
the expectation [41,44], as presented
[41,44], as presented in Figure in20.Figure
A cell20. A cell parameters
parameters degra-was
degradation
dation was expected
expected [45,46],
[45,46], but in thebut in the
case caseanalyzed
of the of the analyzed
battery battery
pack, it pack, it occurred
occurred rather
rather quickly.
quickly.
Appl.
Appl.Sci. 2022,
Sci. 12, 3444
2022, 15 ofof2020
Appl.
Appl. Sci.
Sci. 12,12,
2022,
2022, 12,3444
3444
3444 1515of
15 20
of 20

1200
1200
1200
Discharge
Discharge
1000
1000 Discharge
1000 Charge
Charge
Charge
800
800
800 delta R R*10^-1
(mAh)

delta *10^-1
Capacity (mAh)

delta R *10^-1
(mAh)

600
600
600
Capacity
Capacity

400
400
400
200
200
200
0
0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
-200 1 12 23 34 45 56 67 78 89 91010
1111
1212
1313
1414
1515
1616
1717
18181919
2020
2121
2222
2323
2424
2525
2626
2727
2828
2929
3030
3131
3232
3333
3434
3535
3636
3737
3838
3939
4040
-200
-200 Cell number
Cell number
Cell number
Figure 20.20.
Figure
Figure Comparison of of
Comparison cells capacities
cells determined
capacities with
determined the
with charging
the and
charging discharging
and method.
discharging method.
Figure 20.20. Comparison
Comparison ofof cells
cells capacities
capacities determined
determined with
with the
the charging
charging and
and discharging
discharging method.
method.
TheThe
The visualization
visualization
visualization of
ofofthethe
the spatial
spatial
spatial capacity
capacity
capacity distribution
distribution
distribution within
within
within thethe
the battery
battery
battery pack
pack
pack as
asaswellwell
well
The visualization of the spatial capacity distribution within the battery pack as well
asas the
the corresponding
corresponding temperature
temperature distribution
distribution are
are shown
shown in
inin Figures
Figures 21
2121 and
and 22.22. Unlike
Unlike thethe
asas thecorresponding
the correspondingtemperature
previously presented
temperaturedistribution
thermograms,
distributionare
we
areshown
decided
shownin
to export
Figures21
Figures
the
and22.
and
radiometric
22.Unlike
data
Unlikethe
to
the
a more
previously presented
previouslypresented thermograms,
presentedthermograms,
thermograms,we we decided
wedecided to export
decidedtotoexport
exportthethe radiometric
theradiometric
radiometricdatadata to a more
datatotoa amore
more
previously
readable and convenient form for a further analysis—a CSV file—and
readable
readable and
and convenient
convenient form
form for
for a afurther
furtheranalysis—a
analysis—a CSV file—and tototo display
display ititit
in in
aa
readable
differentand convenient
form by form in
processing fora abasic
further analysis—a
spreadsheet CSV file—and to display it inina a
tool.
CSV file—and display
different form
differentform
formbyby processing
byprocessing
processingin inina abasic spreadsheet
basicspreadsheet tool.
spreadsheettool.
tool.
different a basic

(a) (b)
(a)(a) (b)
(b)
Figure 21.21.
Figure
Figure Comparison ofof
Comparison (a)(a)TOP layer
layercapacity
capacity distribution
distribution in
inin(mAh)
(mAh) and
and (b)
(b) measured
measured tempera-
temperature
Figure
ture 21.21.Comparison
values
Comparison
in (°C). ofof(a)(a) TOP
TOP layercapacity
layer capacitydistribution
distributionin (mAh)and
(mAh) and(b)
(b) measured
measured tempera-
tempera-
ture values
values in ( ◦in (°C).
C).
ture values in (°C).

(a) (b)
(a)(a) (b)
(b)
Figure 22.
Figure22. Comparison
22.Comparison ofof(a)
Comparisonof BOT layer
layercapacity
capacitydistribution
distributioninin(mAh)
(mAh)
(mAh)and
and(b) measured tempera-
Figure
Figure
ture values22.inComparison
(°C). (a)(a)
of (a)
BOT
BOT
BOTlayer
layercapacity
capacitydistribution
distributionin
in (mAh) and
(b)measured
and (b)
measuredtempera-
(b) measured
tempera-
temperature
ture
ture values
values in (°C).
values in in (°C).
(◦ C).
Figures
Figures 21 and 22 clearly depict that the thermal distribution pattern and cell state of
Figures
Figures 2121
21and
and 2222clearly
and 22clearly
clearly depict
depict
depict that
that the
thatthe thermal
thermal
the thermal distribution
distribution
distribution pattern
pattern andcell
and
pattern cellstate
and state
cell ofof
state
capacity
capacity cannot
cannot be correlated. The reason is in the inertia of the thermal process [47] and
capacity cannot
of capacity bebecorrelated.
cannot correlated. Thereason
The
be correlated. reason
The isisininthe
reason the
is inertia
inertia
in ofofthe
the inertia theofthermal
thermal process
process
the thermal [47]and
[47]
process and
[47]
ininthe balancing
thebalancing
balancing dynamics
dynamics shown
shownin in Figure
inFigure
Figure 12. Among
12.Among
Among all the
allthe cells
thecells
cellsin the
inthe
theanalyzed
analyzed battery
battery
in and
the in dynamics
the balancing shown
dynamics shown in12.
Figure 12. all
Among allin
the analyzed
cells in the battery
analyzed
pack,
pack, only
only two can
twoonly bebeused
canbe used to make a new package—cells 14 and 21. Attempting to as-
pack, only
battery two
pack, can twoused betomake
can to make
used a anew
to newpackage—cells
make apackage—cells
new package—cells 1414and
and 21.
14 21. Attempting
Attempting
and totoas-
21. Attempting as-
to
semble
semble a abattery
battery from
from these
these cells
cellswould
would bebepointless,
pointless, as
asnone
none ofofthe
the parallel
parallelcells
cells would
would
assemble
semble a battery
a battery fromfrom
thesethese
cellscells
wouldwould be pointless,
be pointless, as none
as none of the
of the parallel
parallel cells
cells would
would
be able to match the capacity of a ahealthy cell, Table 4.
bebe
be able
able toto
able to match
match
match the
thethe capacity
capacity
capacity ofof
of ahealthy
healthy
a healthy cell,
cell,
cell, Table
Table
Table 4.4.
4.
Appl. Sci. 2022, 12, 3444 16 of 20

Table 4. Optimal layout of pack compound in 4p10s configuration arranged from the existing cells
regarding cell capacities (mAh).

10s
1049 831 520 262 565 271 414 277 310 340
68 92 146 226 115 211 162 205 189 167
4p

74 94 156 230 121 218 163 207 193 174

84 105 160 246 141 221 165 207 203 187
sum (mAh) 1275 1122 982 964 942 921 904 896 895 868
∆ to avg.(mAh) 298 145 5 −13 −35 −56 −73 −81 −82 −109
% 30.51 14.85 0.52 −1.32 −3.57 −5.72 −7.46 −8.28 −8.38 −11.15

4. Battery Pack Revitalization, Results, and Discussion

4.1. Revival and Optimization of Battery Assembly
In order to preserve the e-bike function by revitalizing its battery pack, we decided to
use the batteries from Figure 16c (used laptop battery cells) to assemble the new functional
pack. The availability of such cells is a rather limiting factor for making a proper selection,
and in case of a larger quantity, it is necessary to check all the cells and choose the best
performing [48]. Due to the different cells, which are not designed for the currents expected
from e-bike batteries, it was decided to lower the discharge current by 20% to obtain
more accurate capacity values. Table 5 shows cells’ charge states and cells’ capacities after
discharging all cells with current of 1.8 A. The higher cell resistance values are the result of
oxidation of the contacts [42]. The reused cells 21 and 14 show a higher capacity due to
the lower discharge current and the slight overcharge, which can be seen from the voltage
value before the discharge begins. The overcharge is the result of using a simple charger to
shorten the charging time.

Table 5. Reused cell capacities determined with the discharging method (discharged with 1.8 A).

Start Cell Cell Discharge End Charge

Cell No. Voltage Resistance Capacity Time Voltage Time
U s (V) R (mΩ) (mAh) td U e (V) tc (min)
41 4.22 189 2283 1 h 15 m 3.60 16
42 4.28 175 2412 1 h 20 m 3.61 34
43 4.23 165 2326 1 h 17 m 3.60 26
44 4.28 158 2412 1 h 20 m 3.45 23
45 4.25 172 2381 1 h 19 m 3.57 20
46 4.22 162 2255 1 h 14 m 3.53 23
47 4.21 239 820 27 m 28 s 3.78 18
48 4.24 202 840 27 m 59 s 3.71 27
49 4.23 226 1094 36 m 25 s 3.78 23
50 4.25 206 951 31 m 39 s 3.81 25
51 4.17 215 387 12 m 57 s 3.77 9
52 4.21 206 566 18 m 53 s 3.60 20
53 4.24 135 2696 1 h 29 m 3.34 16
54 4.21 185 1328 44 m 09 s 3.70 61
55 4.23 187 953 31 m 43 s 3.45 12
56 4.23 168 2117 1 h 10 m 3.40 11
57 4.22 186 2128 1 h 10 m 3.44 19
58 4.24 141 1009 33 m 43 s 3.60 20
59 (14) 4.24 137 1318 43 m 05 s 3.54 21
60 (21) 4.25 160 1455 48 m 33 s 3.50 37

After an optimal structure of the 2p10s pack was determined, the new pack parameter
assessment was conducted, Table 6. As is shown in Figure 23, a repaired battery pack now
Appl. Sci. 2022, 12, 3444 17 of 20

Appl. Sci. 2022, 12, 3444 17 of 20

After an optimal structure of the 2p10s pack was determined, the new pack parame-
ter assessment was conducted, Table 6. As is shown in Figure 23, a repaired battery pack
now has 35%
has 35% of capacity
of the the capacityof aof
newa new
one.one. However,
However, it still
it still has has three
three times
times thecapacity
the capacityofofa
apack
packthat
thathas
hasreached
reachedits itsend
endof ofservice,
service, when
when itit has
has only 11% of the initial capacitycapacity and
and
two defective
two defective cells
cells in
inaapack
packassembly.
assembly. In In addition
addition to toFigure
Figure 23,23,there
there is
isaathermogram
thermogramthat that
monitors the
monitors the equilibrium
equilibrium of of the
the items
items during
during the
the initial
initial fill
fill that
that shows
shows aa fairly
fairly uniform
uniform
thermal load
thermal load on
on all
all cells.
cells. Such
Suchuniform
uniformdistribution
distributionof ofthe
thethermal
thermalloadloadacross
across the
the cells
cells in
in
the pack is a strong indicator that the battery pack is now recovered
the pack is a strong indicator that the battery pack is now recovered to some degree, as to some degree, as
determined by
determined by the
the characteristics
characteristics of of the
the cells
cells used.
used.

(a) (b)
Figure
Figure 23.
23. Repaired
Repairedand
andrevived
revived battery
battery pack:
pack: an
an assembly
assembly (a)
(a) and
and its
its thermal
thermal footprint
footprint (b).
(b).

Table 6. Optimal layout of pack compound in 2p10s configuration restated from the used cells (con-
Table 6. Optimal layout of pack compound in 2p10s configuration restated from the used cells
sidering its capacities, mAh).
(considering its capacities, mAh).
10s
10s
2117 2326 2255 2283 2412 2128 2381 2696 2412 1455
2p

2117 2326 2255 2283 2412 2128 2381 2696 2412 1455
2p

13181318 951
951 1009
1009 953
953 820820 1094
1094 840
840 387
387 566
566 1328
1328
sum (mAh) 3435 3277 3264 3236 3232 3222 3221 3083 2978 2783
sum (mAh) 3435 3277 3264 3236 3232 3222 3221 3083 2978 2783
Δ to avg.(mAh) 262 104 91 63 59 49 48 −90 −195 −390
∆ to avg.(mAh) 262 104 91 63 59 49 48 −90 −195 −390
% 8.25 3.27 2.86 1.98 1.86 1.54 1.51 −2.84 −6.15 −12.29
% 8.25 3.27 2.86 1.98 1.86 1.54 1.51 −2.84 −6.15 −12.29

4.2. Functional Assessment and Discussion

4.2. Functional Assessment and Discussion
The new battery pack, shown in Figure 23, is only a temporary solution that allows
The new
the bicycle to bebattery pack,ashown
used until in Figure
new battery pack23,isismanufactured
only a temporary solutionacquired.
or otherwise that allows
It
will be restructured by the means of its mechanical construction and coolingacquired.
the bicycle to be used until a new battery pack is manufactured or otherwise system,
It will makes
which be restructured by thepack
the new battery means of itsand
lighter mechanical construction
longer lasting. and cooling
The previous system,
experience of
which makes the new battery pack lighter and longer lasting. The previous
modeling systems, based on the principle of air cooling [49,50], as well as analyzing vari-experience of
modeling systems, based on the principle of air cooling [49,50], as well as analyzing
ous forms of battery packs [51], will take into account the models of individual cells [52], various
forms of battery packs [51], will take into account the models of individual cells [52],
and serve to build a CAD model of the battery pack using a feature-based parametric
and serve to build a CAD model of the battery pack using a feature-based parametric
modeling system. In addition, due to the poor thermal management of the current battery
modeling system. In addition, due to the poor thermal management of the current battery
pack, optimization of the thermal management system of the new battery pack must be
pack, optimization of the thermal management system of the new battery pack must be
performed. The optimization will be based on a computational fluid dynamics analysis
performed. The optimization will be based on a computational fluid dynamics analysis
(CFD) that also takes into account the electrical parameters of the cells, such as the variable
(CFD) that also takes into account the electrical parameters of the cells, such as the variable
cell resistance, the state of cell charge, and ancillary components. The analysis of a new
cell resistance, the state of cell charge, and ancillary components. The analysis of a new
battery pack will rely on CFD simulations of different pack models [53], taking into ac-
battery pack will rely on CFD simulations of different pack models [53], taking into account
count the characteristics and specifics of the electric bicycle application.
the characteristics and specifics of the electric bicycle application.
5.
5. Conclusions
Conclusions
Infrared
Infrared thermography
thermography isis aa practical
practical and
and non-destructive
non-destructive test
test and
and measurement
measurement
method
method forfor battery
battery condition
condition assessment
assessment and
and analysis,
analysis,but
butonly
onlyininaaqualitative
qualitativeway.
way.
Changes
Changes in in aa battery
battery and
and battery
battery cells’
cells’ temperature
temperature values,
values, due
due to
to thermal
thermal capacity
capacity and
and
thermal resistance, can only provide information on the battery condition. On the thermo-
grams of healthy batteries, a temperature difference of 2.1 ◦ C was found at points with
Appl. Sci. 2022, 12, 3444 18 of 20

similar cells. At the end of the battery life, this difference was slightly larger and amounted
to 2.4 ◦ C. The reported values were close to the camera accuracy of ±2 ◦ C and depended
on power dissipation due to current that we could not measure. Therefore, general conclu-
sions based on absolute temperature values are not possible. Owing to the high thermal
sensitivity of the NETD camera, the pattern can be seen. The faulty battery had the largest
temperature difference of 5.5 ◦ C, which, according to a recommendation from the literature,
is an indication of a possible defect (needs to be investigated). The homogeneity of the
thermal pattern can serve as a criterion for assessing the health and service expectancy of
the battery pack. Thermal hotspots within a pack, which occur when one or more battery
cells start to fail, are caused by overloading of healthy cells. Contrary to the common
reasoning, where hotspots represent a place of failing component, the situation here is the
opposite. The hotspot represents a place of a normal operational component that generates
excessive heat due to acting as a charging current sink, having low resistance within a cell-
block. In addition, the thermographic analysis of the balancing circuitry can also provide
information about the state of charge and capacity of individual cells by monitoring the
power dissipation at each balancing resistor over time. Due to the way the e-bike’s battery
pack operates, a failure of the battery pack usually inevitably means a failure of all the cells.
The main factors leading to the battery cell failure are cell aging and operating conditions
associated with high operating and resting temperatures. In order to draw conclusions
based on thermographic records, detailed knowledge of the physical behavior of the object
under study is required. Therefore, it was necessary to perform a detailed analysis of
the battery pack. The main conclusion of the study is that it is not possible to make a
judgment about the appearance of the thermal pattern or the pattern shape based on the
estimated capacity values. However, based on the thermal pattern, it is possible to make a
judgment about the homogeneity of the capacity of individual battery cells. Thus, in its
qualitative form, infrared thermography is an ideal non-destructive method for monitoring
the condition of e-bike batteries. Due to the high thermal sensitivity/NETD of today’s
thermal imaging cameras, the dynamic of thermal pattern during charging and discharging
can provide information about the condition of a battery pack.

Author Contributions: Conceptualization, H.G. and M.K.; methodology, H.G. and T.B.; software,
M.K. and T.K.; validation, T.K. and T.B.; formal analysis, H.G.; investigation, H.G.; resources, H.G.
and M.K.; data curation, H.G., T.B., T.K. and M.K.; writing—original draft preparation, H.G., M.K.
and T.K.; writing—review and editing, M.K. and T.K.; visualization, H.G. and M.K.; supervision, T.B.;
project administration, T.B.; funding acquisition, M.K. and H.G. All authors have read and agreed to
the published version of the manuscript.
Funding: This research received no external funding.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Glavaš, H.; Karakašić, M.; Petrović, I.; Vidaković, D. PEDELEC Li-Ion Battery Pack Lifetime. In Proceedings of the VIII
International Conference Industrial Engineering and Environmental Protection 2018 (IIZS 2018), University of Novi Sad Technical
Faculty “Mihajlo Pupin”, Zrenjanin, Serbia, 11–12 October 2018; pp. 27–33, ISBN 978-86-7672-309-6.
2. Kim, J.; Mallarapu, A.; Finegan, D.P.; Santhanagopalan, S. Modeling cell venting and gas-phase reactions in 18650 lithium-ion
batteries during thermal runaway. J. Power Sources 2021, 489, 229496. [CrossRef]
3. Juarez-Robles, D.; Jeevarajan, J.A.; Mukherjee, P.P. Degradation-Safety Analytics in Lithium-Ion Cells: Part I. Aging under
Charge/Discharge Cycling. J. Electrochem. Soc. 2020, 167, 160510. [CrossRef]
4. Pastor-Fernández, C.; Bruen, T.; Widanage, W.D.; Gama-Valdez, M.A.; Marco, J. A Study of Cell-to-Cell Interactions and
Degradation in Parallel Strings: Implications for the Battery Management System. J. Power Source 2016, 329, 574–585. [CrossRef]
5. Berg, P.; Spielbauer, M.; Tillinger, M.; Merkel, M.; Schoenfuss, M.; Bohlen, O.; Jossen, A. Durability of lithium-ion 18650 cells
under random vibration load with respect to the inner cell design. J. Energy Storage 2020, 31, 101499. [CrossRef]
6. Glavaš, H.; Barić, T.; Keser, T. Pedelec—Pedal Electric Cycle. In Proceedings of the 35th Conference on Transportation Systems
with International Participation Automation in Transportation 2015, Zagreb, Croatia, London, UK, 3–8 November 2015.
7. Kabir, M.M.; Demirocak, D.E. Degradation mechanisms in Li-ion batteries: A state-of-the-art review. Int. J. Energy Res. 2017, 41,
1963–1986. [CrossRef]
Appl. Sci. 2022, 12, 3444 19 of 20

8. EuroVelo. The European Cycle Route Network. Available online: http://www.eurovelo.com/en/eurovelos (accessed on 30
April 2021).
9. Weiss, M.; Dekker, P.; Moro, A.; Scholz, H.; Patel, M.K. On the electrification of road transportation—A review of the environmen-
tal, economic, and social performance of electric two-wheelers. Transp. Res. Part D Transp. Environ. 2015, 41, 348–366. [CrossRef]
[PubMed]
10. GoPedelec! (A Manual). 2015. Available online: www.gopedalec.eu (accessed on 24 July 2015).
11. Hoffmann, U. Das Fahrradbuch, Kauf, Technik, Wartung, Reparaturen E-Bikes und Pedelecs, Stiftung Warentest; Stiftung Warentest:
Berlin, Germany, 2012.
12. Leaserad at Go Pedelec. Available online: http://www.datei.de/public/extraenergy/GoPedelec_Handbuch-DE-V1_0-b.pdf
(accessed on 30 April 2021).
13. Kirkpatrick MacMillan (1812–1878) Inventor of the Pedal Bicycle. 2021. Available online: http://gaukartifact.com/2013/04/05
/kirkpatrick-macmillan-1812-1878-inventor-of-the-pedal-bicycle (accessed on 28 February 2021).
14. Wilson, D.G. Bicycling Science, 3rd ed.; MIT Press Cambridge: London, UK, 2004.
15. Remusat, T. Pierre Michaux and the History of the Motorized Bike, HISTORIA; Librairie Jules Tallandier: Paris, France, 2001; pp. 40–41.
16. Herlihy, D.V. Bicycle: The History; Yale University Press: New Haven, CT, USA, 2006; ISBN 0-300-10418-9.
17. Renold: A Company History. 2021. Available online: https://www.renold.com/company/history (accessed on 30 April 2021).
18. Hadland, T.; Lessing, H.E. Bicycle Design, An Illustrated History; Massachusetts Institute of Technology: Cambridge, MA, USA,
2014; ISBN 978-0-262-02675-8.
19. Kyle, C.R.; Weaver, M.D. Aerodynamics of human-powered vehicles. Proc. Inst. Mech. Eng. Part A J. Power Energy 2004, 218,
141–154. [CrossRef]
20. Clayton, N. The Birth of the Bicycle; Amberley Publishing: Stroud, UK, 2016; ISBN 1445648822.
21. Bolton, O., Jr. Electrical Bicycle. U.S. Patent No. 552,271, 31 December 1895. Available online: https://patents.google.com/
patent/US552271?oq=Ogden+Bolton+Jr (accessed on 21 February 2021).
22. Ron/Spinningmagnets. Electric Bike History, Patents from the 1800’s. 2013. Available online: https://www.electricbike.com/e-
bike-patents-from-the-1800s/ (accessed on 30 April 2021).
23. Power Assisted Bicycle, Two Decades of Electrically Power Assisted Bicycles; Yamaha Motor Co., Ltd.: Iwata, Japan, 2013.
24. European Commission. Directive 2002/24/EC of the European Parliament and of the Council of 18 March 2002 Relating to the
Type-Approval of Two or Three-Wheel Motor Vehicles. Available online: https://eur-lex.europa.eu/legal-content/EN/TXT/
?uri=CELEX:32002L0024 (accessed on 30 April 2021).
25. Neupert, H. Light-Electric-Vehicles Global Market Trends and Standardization; China Cycle Show: Shanghai, China, 2014.
26. Desmond, K. Gustave Trouve: French Electrical Genius (1839–1902); McFarland & Company: Jefferson, NC, USA, 2015;
ISBN 0786497092.
27. Schnepf, J. U.S. Patent US627066A, 1898. Available online: https://patents.google.com/patent/US627066 (accessed on 30
April 2021).
28. eCycleElectric. Available online: http://www.ecycleelectric.com/susanne-bruesch (accessed on 30 April 2021).
29. CROW. Design Manual for Bicycle Traffic; CROW: Ede, The Netherlands, 2016; ISBN 9789066286597.
30. Glavaš, H.; Józsa, L.; Barić, T. Infrared Thermography in Energy Audit of Electrical Installations. Teh. Vjesn. 2016, 23, 1533–1539.
[CrossRef]
31. Baehr, H.D.; Stephan, K. Wärme-und Stoffübertragung; Springer GmbH Deutschland: Wien, Austria, 2019; ISBN 978-3-662-58441-5.
32. Zhang, G.; Tian, H.; Ge, S.; Marple, D.; Sun, F.; Wang, C.Y. Visualization of self-heating of an all-climate battery by infrared
thermography. J. Power Sources 2018, 376, 111–116. [CrossRef]
33. Panasonic. CGR18650CG Datasheet; Panasonic: Osaka, Japan, 2018.
34. Ye, M.; Song, X.; Xiong, R.; Sun, F. A Novel Dynamic Performance Analysis and Evaluation Model of Series-Parallel Connected
Battery Pack for Electric Vehicles. IEEE Access 2019, 7, 14256–14265. [CrossRef]
35. Baronti, F.; Di Rienzo, R.; Papazafiropulos, N.; Roncella, R.; Saletti, R. Investigation of series-parallel connections of multi-module
batteries for electrified vehicles. In Proceedings of the IEEE International Electric Vehicle Conference (IEVC), Florence, Italy, 17–19
December 2014; pp. 1–7. [CrossRef]
36. Chen, H.; Shen, J. A degradation-based sorting method for lithium-ion battery reuse. PLoS ONE 2017, 12, e0185922. [CrossRef]
37. Masomtob, M.; Sukondhasingha, R.; Becker, J.; Sauer, D.U. Parametric Study of Spot Welding between Li-ion Battery Cells and
Sheet Metal Connectors. In Proceedings of the Engineering Journal, 7th Thai Society of Mechanical Engineers-International
Conference on Mechanical Engineering, (TSME-ICoME 2016), Chiang Mai, Thailand, 13–16 December 2016; Volume 21. [CrossRef]
38. Feng, X.; Ouyang, M.; Liu, X.; Lu, L.; Xia, Y.; He, C. Thermal runaway mechanism of lithium-ion battery for electric vehicles: A
review. Energy Storage Mater. 2018, 10, 246–267. [CrossRef]
39. Ouyang, M.; Ren, D.; Lu, L.; Li, J.; Feng, X.; Han, X.; Liu, G. Overcharge-induced capacity fading analysis for large format lithium-
ion batteries with LiyNi1/3Co1/3Mn1/3O2+LiyMn2O4 composite cathode. J. Power Sources 2015, 279, 626–635. [CrossRef]
40. Salinas, F.; Krüger, L.; Neupert, S.; Kowal, J. A second life for li-ion cells rescued from notebook batteries. J. Energy Storage 2019,
24, 100747. [CrossRef]
41. Pop, V.; Bergveld, J.H.G.; Op het Veld, P.P.L.; Regtien, D.D.; Notten, P.H.L. Modeling Battery Behavior for Accurate State-of-Charge
Indication. J. Electrochem. Soc. 2006, 153, A2013–A2022. [CrossRef]
Appl. Sci. 2022, 12, 3444 20 of 20

42. Bolsinger, C.; Zorn, M.; Birke, K.P. Electrical contact resistance measurements of clamped battery cell connectors for cylindrical
18650 battery cells. J. Energy Storage 2017, 12, 29–36. [CrossRef]
43. Ning, G.; Haran, B.; Popov, B.N. Capacity fade study of lithium-ion batteries cycled at high discharge rates. J. Power Sources 2003,
117, 160–169. [CrossRef]
44. Andrea, D. Battery Management Systems for Large Lithium-Ion Battery Packs; Artech House: Norwood, MA, USA, 2010; p. 02062,
ISBN 9781608071050.
45. Baumann, M.; Wildfeuer, L.; Rohr, S.; Lienkamp, M. Parameter variations within Li-Ion battery packs–Theoretical investigations
and experimental quantification. J. Energy Storage 2018, 18, 295–307. [CrossRef]
46. Marcis, V.A.; Praneeth, A.V.J.S.; Patnaik, L.; Williamson, S.S. Analysis of CT-CV Charging Technique for Lithium-ion and NCM
18650 Cells Over Temperature Range. In Proceedings of the IEEE International Conference on Industrial Technology (ICIT),
Buenos Aires, Argentina, 26–28 February 2020; pp. 947–952. [CrossRef]
47. Tang, W.; Tam, W.C.; Yuan, L.; Dubaniewicz, T.; Thomas, R.; Soles, J. Estimation of the critical external heat leading to the failure
of lithium-ion batteries. Appl. Therm. Eng. 2020, 179, 115665. [CrossRef] [PubMed]
48. Kim, J.; Cho, B.H. Screening process-based modeling of the multi-cell battery string in series and parallel connections for high
accuracy state-of-charge estimation. Energy 2013, 57, 581–599. [CrossRef]
49. Saw, L.H.; Ye, Y.; Tay, A.O.; Chong, W.T.; Kuan, S.H.; Yew, M.C. Computational fluid dynamic and thermal analysis of Lithium-ion
battery pack with air cooling. Appl. Energy 2016, 177, 783–792. [CrossRef]
50. Ling, Z.; Wang, F.; Fang, X.; Gao, X.; Zhang, Z. A hybrid thermal management system for lithium-ion batteries combining phase
change materials with forced-air cooling. Appl. Energy 2015, 148, 403–409. [CrossRef]
51. Yoon, C.O.; Lee, P.Y.; Jang, M.; Yoo, K.; Kim, J. Comparison of internal parameters varied by environmental tests between
high-power series/parallel battery packs with different shapes. J. Ind. Eng. Chem. 2019, 71, 260–269. [CrossRef]
52. Yang, X.; Hu, X.; Chen, Z.; Chen, Y. Effect of ambient dissipation condition on thermal behavior of a lithium-ion battery using a
3D multi-partition model. Appl. Therm. Eng. 2020, 178, 115634. [CrossRef]
53. Buidin, T.I.C.; Mariasiu, F. Modeling Approach of an Air-Based Battery Thermal Management System for an Electric Vehicle.
Appl. Sci. 2021, 11, 7089. [CrossRef]