Journal of Energy Storage 1 (2015) 7–14

Contents lists available at ScienceDirect

Journal of Energy Storage

jo u r na l h o m e p a g e : www.e l s e v i e r . c o m / l o c a te /e st

Welding techniques for battery cells and resulting electrical contact

resistances
Martin J. Brand a,*, Philipp A. Schmidt b, Michael F. Zaeh b, Andreas Jossen a
a
Institute for Electrical Energy Storage Technology, Technische Universität München, Arcisstr. 21, 80333 Munich, Germany
b
Institute for Machine Tools and Industrial Management, Technische Universität München, Boltzmannstr. 15, 85748 Garching, Germany

A R T I C L E IN F O
A B S T R A C T

Article history:
Received 9 February 2015 In large battery assemblies, which are integrated, for example, in electric vehicles or stationary storage
Received in revised form 14 April 2015 systems, up to several thousand single battery cells are connected together. Every single cell
Accepted 14 April 2015 connection influences the functionality and efficiency of the whole battery system.
Available online 15 May 2015 Resistance spot, ultrasonic or laser beam welding are mostly used for connecting battery cells in the
production of large battery assemblies. Each of these welding techniques has its own characteristics
Keywords: depending on the material properties and contact geometry. Cell casing and terminal dimensions may
Battery assembly constrain possible contact geometries. For example, pouch cells provide detached tabs, whereas cells
Electrical contact resistance with hard casing provide a certain area of the casing for electrical connection. Consequently, it is a
Lithium-ion battery challenging task to find the most suitable welding technique and welding parameters for each
Welding
individual battery assembly.
This paper presents a comprehensive overview on joining battery cells by resistance spot, ultrasonic
and laser beam welding. The specific features, advantages and dependencies of each welding technique
for connecting cells are discussed. In addition, a quantitative analysis on welded test samples reveals
the ultimate tensile strength and heat input into a battery cell. A novel measuring and calculation
method was designed and applied to assess the electrical contact resistance itself. The electrical
contact resistance occurs as a result of the joint and not in the bulk material. Therefore, it is a criterion
which may be transferred to any size of joint partners.
ã 2015 Elsevier Ltd. All rights reserved.

1. Introduction
Large battery assemblies are of particular interest both for the
progressing electrification of mobility and for storing intermittent
renewable energy. Within any battery storage, the smallest energy
storing component is the battery cell or short cell. Whereas for
mobile devices, e.g., laptops, only a few cells are combined, in large
battery assemblies up to several thousand cells have to be
connected. For economic and quality reasons a high degree of
automation of the cell connecting process is needed when
assembling large batteries. Hence, resistance spot welding,
ultrasonic welding and laser beam welding are mostly applied.
Using the example of two battery cells connected in parallel,
Fig. 1 illustrates the influence of the quality of cell connections on
a battery assembly. The higher electrical contact resistance RC,1
generates more heat at the terminal of cell 1. Additionally, the total
* Corresponding author. Tel.: +49 89 289 26979; fax: +49 89 289 26968.
E-mail address: martin.brand@tum.de (M.J. Brand).
current Iges is divided unequally. These uneven loads may lead to
inhomogeneous cell degradations. A higher resistance in a battery
assembly reduces its power capability and efficiency. Furthermore,
defective cell connections are prone to fail suddenly, because of
mechanical stress.
This paper investigates the specific features, advantages and
dependencies of connecting battery cells by resistance spot,
ultrasonic and laser beam welding. The findings are applicable
to all kinds of battery cell casings. Additionally, the three welding
techniques are compared quantitatively in terms of ultimate
tensile strength, heat input into a battery cell caused by the
welding process, and electrical contact resistance. A measuring
and calculation method was developed in order to measure only
the electrical contact resistance at the joint and not the resistance
of the bulk materials.
Brass (CuZn37) test samples are used for the quantitative
comparison of the welding techniques, as this metal can be
processed by all three welding techniques. At the end of the
presented work, the suitability of resistance spot, ultrasonic and
laser beam welding for connecting battery cells is evaluated.

http://dx.doi.org/10.1016/j.est.2015.04.001
2352-152X/ ã 2015 Elsevier Ltd. All rights reserved.

Fig. 1. Two battery cells connected in parallel with uneven thermal and electrical
loads due to different electrical contact resistances (RC,1 > RC,2).
2. Method for measuring electrical contact resistance
As mentioned in Section 1, the electrical contact resistances of
cell connections are of high relevance for the quality of a battery
assembly. To obtain transferable results, the electrical contact
resistance itself, which only occurs as a result of the joint and not
in the bulk material of the conductors, has to be determined. In
order to determine the electrical contact resistance, a measuring
setup and a corresponding calculation method were developed.
According to Fig. 2, each test sample consists of two specimens
which overlap by 15 mm × 15 mm. Three spots for voltage
measurements are marked on every specimen. In Fig. 2, they are
named A1, A2, A3 and B1, B2, B3.
Fig. 3 shows the equivalent resistance network for the two
metal specimens connected as shown in Fig. 2. The serial
resistances RA3,A2 to RA,CN represent the material conductivity of
the specimen A and those from RB3,C1 to RB3,B2 represent the
material conductivity of specimen B.
If the dimensions and the material electrical resistivity rA of
specimen A are exactly known, the resistances from A0 to any point
Ai (with i = 1,2,3) can be calculated by Eq. (1). The equation also
applies to specimen B.
lAi;A0
RAi;A0 ¼ rA
× (1)
h× b
The electrical contact resistance RC is situated at the
interface between the two overlapping metal specimens A and
B. In Fig. 3, the parallel resistances RC1 to RCN represent the
electrical contact resistance. The ideal overlap resistance RA0,B0,id
between A0 and B0 is given for RC1 = . . . = RCN = 0 by Eq. (2).
X
N lA0;B0 lA0;B0
RA0;B0;id ¼ RA;CnjjRB;Cn ¼ rA (2)
n¼1
h × b jj rB
h×b
The actual electrical resistance of the overlap RA0,B0,meas can
be determined by using the measurements of RAi,Bj (with i,j = 1,2,3).
The difference between the measured resistance RA0,B0,meas and
the ideal resistance RA0,B0,id is the electrical contact resistance RC.
Fig. 2. Schematic view of test sample consisting of two connected metal
specimens overlapping by 15 mm × 15 mm.
RC ¼ RA0;B0;meas — RA0;B0;id (3)
The resistances were measured by the four-point probes
method with the BT3562 by Hioki, which provides a 1 kHz AC
current Imeas = 100 mA to the test sample and measures the voltage
drop. Eqs. (1)–(3) apply when the current Imeas is distributed
evenly throughout the cross-section of the specimen. This can be
assumed for the presented measuring setup where the current
input is 7.5 mm away from the first voltage measurement spot.
To narrow the confidence interval, every combination RAi,Bj was
measured five times and averaged. This procedure was followed
for every test sample. The graphs presented in the following
sections are based on series of measurements in which, for every
measuring point, at least five test samples were tested and
averaged. When needed, the number of test samples per
measuring point was increased to get reliable diagrams.
3. Resistance spot welding
The main characteristic of resistance spot welding is that only a
small volume of the work pieces is melted and fused together. The
welding heat is generated by the electrical power Pweld, which
results from the welding current Iweld flowing through the
contact resistance at the interface of the work pieces Rweld (see Fig.
4) [1,2].
2
Pweld ¼ Rweld × Iweld (4)
Fig. 4(a) explains how battery cells with hard casing, i.e.,
cylindrical or prismatic cells, are resistance spot welded with
parallel weld electrodes. Apart from the desired current Iweld, a
stray current Istray flows. Therefore, the power source has to
provide a higher current Ielectrode and electrode sticking may occur.
In Fig. 4(b), spot welding of a pouch cell is illustrated. Assuming
that the external conductor does not exceed a certain thickness,
opposed weld electrodes can be used. For both electrode
arrangements, the electrodes have to be pressed onto the metals.
This pressure influences the contact resistance at the interface of
the work pieces Rweld as well as the contact resistance between the
electrode and the work piece Relectrode. As explained above, the
resistance at the interface of the work pieces Rweld should be
high
enough to form a good weld while the resistance beneath the
electrode Relectrode should be as small as possible to avoid electrode
sticking.
Different types of power supplies are available for providing
the welding power [3]. For this paper, the CD CPM2 power supply
by Sunstone was used. It provides a dual pulse by discharging a
capacitor [4]. The first pulse disrupts the oxides and contami-
nations at the materials’ interface and the second pulse is used to
weld the work pieces [4,5]. After the surface films are disrupted,
the temperature rises and the area affected by the welding heat
increases. The temperatures inside the heat-affected zone reach
the melting point of the material and the materials are fused
together in a weld nugget [6,7]. An exemplary micrograph is
shown in Fig. 5.
3.1. Influence of welding energy on weld quality
In Fig. 6, the electrical contact resistance and the mechanical
strength are shown as a function of the welding energy. The
mechanical strength of each test sample was assessed by
measuring the tensile strength in longitudinal axis of the test
samples with the testing machine Z020 by the Zwick GmbH & Co.,
KG. In Fig. 6, and further figures in this article, the maximum force
achieved for each measuring point is shown instead of the tensile
strength. This is applicable since all tested samples in this work
 Fig. 5. Micrograph of two nickel-plated steel specimens joined by resistance spot
welding.

Fig. 3. Equivalent resistance network of test samples according to Fig. 2.

The heat for the welding process is focused on the weld spots.
have the same dimensions. Hilumin, a nickel-plated steel product Consequently, the heat input into the battery cell is very small, as
by Tata Steel [8], and CuZn37 specimens according to Fig. 2 were discussed in detail in Section 6.
investigated. For both metals a proper weld cannot be achieved
with too little welding energy, in this case less than 200 Ws (see 4. Ultrasonic welding
Fig. 6). More welding power enlarges the weld nugget and leads to
a higher weld quality. In these tests, the optimum for Hilumin was Ultrasonic welding is a solid-state welding technique. The work
reached at 340 Ws and for CuZn37 at 350 Ws. Further increasing pieces are not melted but pressed and scrubbed together [11–13].
the welding energy leads to electrode sticking and significant Fig. 8 illustrates the functional principle of welding battery cells by
expulsion of bulk material [2,5,6,9]. ultrasonic vibration. The sonotrode is pressed on the electrical
The micrographs of the spot-welded CuZn37 test samples conductor with the pressure p and thereby the conductor is forced
showed voids inside the weld nuggets, especially for welding onto the cell terminal beneath. Additionally, an ultrasonic
energies higher than 350 Ws (see Fig. 7). The melting interval of vibration u(t) parallel to the interface between the external
◦ ◦
CuZn37 is between 902 C and 920 C, whereas the boiling point of conductor and the cell terminal is applied. Due to the knurl pattern
◦ of the sonotrode, the vibrations are transferred to the external
zinc is at 907 C. The highest welding temperatures occur in
conductor. The cell terminal beneath must be fixed. For practical
the
reasons, the frequency of the vibration is chosen between 15 kHz
center of the weld nugget, where the boiling temperature of zinc
and 45 kHz [14]. In this work, the Ultraweld L20 by Branson was
was potentially reached [2]. These voids lead to a higher electrical
used, which is operating at 20 kHz and allows adjusting the
contact resistance and to a lower strength of the joint.
amplitude from 0.5 to 49 mm. Fig. 8(a) shows the functional
principle for welding an external conductor to a battery cell’s hard
3.2. Suitability of resistance spot welding for connecting battery cells
casing. When the casing of the cell is fixed, the relative movement
between the external conductor and the cell terminal causes the
The suitability of a metal for resistance spot welding depends
weld. In Fig. 8(b), the ultrasonic welding of pouch cells is shown.
on its thermal conductivity and its melting point [1–3]. Less heat
An anvil with a knurl pattern is used to fix the terminal tab, as the
input and thus less welding power is needed for metals with lower
cell casing cannot guarantee fixation.
thermal conductivity and melting point. Since thermal and
When the welding starts, the oxides and contaminants on the
electrical conductivity correlate for most metals, it is easier to
interfaces are disrupted by the vibrations and areas of pure metal
spot weld steel in contrast to copper.
are uncovered. Uncovered asperities come in contact and
Hence, resistance spot welding is only applicable when the
plastically deform due to the shearing forces. With ongoing
metal sheets do not exceed a certain thickness and conductivity.
vibrations, the contact areas with pure metal grow and the test
For high current applications rather thick and conductive
samples heat up [13]. Up to now, it is not completely understood
conductors are needed, which are more difficult to be spot welded.
which bonding mechanisms join the uncovered areas at the
In particular, welding with parallel electrodes is not suitable for
interface. Mechanical interlocking and grain boundary diffusion
connecting battery cells when high cell currents (>20 A) are
are the two main bonding mechanisms proposed [12,15,16].
requested in the battery assembly. Furthermore, the current flow
Fig. 9 shows a cross-sectional view of an ultrasonically welded
has to be focused on a spot on the interface and stray currents
joint between two CuZn37 specimens. The indents of the
should be minimized. This is difficult for highly conductive
sonostrode’s and anvil’s knurl pattern can be seen. Here material
materials. Slotted external conductors with projections can guide
was transported an extruded due to the applied vibration and
and focus the current flow [10].

Fig. 6. Electrical contact resistance and ultimate tensile force as function of welding
Fig. 4. Functional principle of connecting battery cells by resistance spot welding energy.
with (a) parallel weld electrodes for cells with hard casing or with (b) opposed weld
electrodes for pouch cells.
Fig. 7. Micrograph of spot-welded CuZn37 test sample with void in center of weld
nugget.
pressure [13,17]. The micrograph in Fig. 9 shows that bonding
occurs only in areas with the highest pressure close to the indents.
Longer welding times enlarge these bonded areas but also deepen
the indents. With too long welding times, the material cracked at
some distance from the welded area because of resonance.
4.1. Influence of sonotrode pressure on weld quality
The adjustable welding parameters are, above all, the pressure
of the sonotrode p, the amplitude of the vibration u(t) and the
welding time [18,19]. In Fig. 10, the mechanical welding power is
depicted as a function of the welding time for different pressures p
while the sonotrode area was constantly 13×13 mm2. In order to
avoid cracks in the CuZn37 test samples, the maximum welding
time had to be limited. At the end of each graph in Fig. 10, a steep
drop of the welding power would indicate cracking of the metal.
The amplitude was adjusted for every pressure to get the best
possible weld.
In Fig. 11, the electrical contact resistances and the ultimate
tensile forces depending on the welding pressure are shown. The
graphs in Fig. 10 show that more pressure enables more welding
power. The power input heats up the weld area. Fig. 11 shows that
more power input results in better weld characteristics. This is
also because heat is needed for the diffusion processes [13,20,21].
However, this does not apply for the results at the highest tested
pressure (5 bar). The reduction of the welding power in Fig. 10
occurs probably because the forces of the sonotrode start to
exceed the plasticity limits of the metal [13]. Furthermore, at 5 bar
the material is weakened by excessive extrusion and the joint
shows less mechanical strength and higher contact resistance (see
Fig.11). If the pressure is too high and the travel of the sonotrode
is not limited, the battery cell might be damaged.
4.2. Suitability of ultrasonic welding for connecting battery cells
Ultrasonic welding is a solid-state welding technique and does
not depend on the melting temperatures of the work pieces. The
temperatures of the welding process are low, as will be discussed
in detail in Section 6. Ultrasonic welding is capable of joining
Fig. 8. Functional principle of ultrasonically welding (a) battery cell with hard
casing and (b) pouch cell.
Fig. 9. Cross-sectional view of ultrasonically welded joint between two CuZn37
specimens.
dissimilar materials [21,22]. Furthermore, also multiple metal
layers can be joined [17].
Ultrasonic welding depends on the materials’ hardness and
surface roughness [13,23]. This means that highly electrically
conductive materials, such as copper or aluminum, can be welded,
but some difficulties arise with hard materials such as steel. The
needed welding power depends on the thickness of the metal
sheet vibrated by the sonotrode. The counterpart has to be fixed
but may have any thickness.
It was reported that ultrasonic weld vibrations can damage the
inside of a pouch cell, especially when the conductors inside the
battery cell are also ultrasonically welded. In order to prevent the
propagation of the vibrations into the cell, the terminal tabs need
to be clamped [24].
5. Laser beam welding
Laser beam welding uses the absorption of electromagnetic
waves to heat up the joint partners. The laser beam can be
provided by various laser sources [25]. In this study, the laser
source YLR- 3000-SM by IPG Photonics was used. This single mode
ytterbium- doped fiber laser emits continuous-wave laser
radiation with a maximum output power of 3 kW at a wavelength
of l = 1070 nm. To position the laser beam onto the work
piece, the scanner optics
Remote Welding Elephant by Arges was used. This optics
generates a beam diameter of 50 mm in the focal plane.
In this work, the process of keyhole welding was used to
connect battery cells. The functional principle is shown in the
illustration in Fig. 12(a). The laser beam reaches high power
densities I > 1012 W/m2, which melt and evaporate the metals. The
vapor causes a recoil pressure that opens a capillary [27]. This
capillary is known as the keyhole in which the laser beam is
reflected and absorbed, leading to high absorption rates [26,28].
Laser beam welding offers the opportunity to weld lap (see
Fig. 12(a)) and fillet joints (see Fig. 12(b)) to connect a battery cell
with an external conductor.
5.1. Advantages of superimposed beam oscillation
Fig. 13(a) shows a cross-sectional polish of a lap joint of nickel-
plated steel sheets produced by keyhole welding. The process of
keyhole welding leads to weld seams that are deep compared to
their width (see Fig.13(a)). In order to increase the contact area
and to get a stable welding depth, superimposed beam oscillation
can be applied [29,30]. In Fig. 13(b), beam oscillation with a
frequency f = 200 Hz and an amplitude A = 0.25 mm was
superimposed. The resulting weld seam is twice as wide as
without beam oscillation.
5.2. Suitability of laser beam welding for connecting battery cells
The high power density of the laser beam allows high welding
velocities and can limit the heat input into the battery cell [28,25].
 Fig. 12. Functional principle of keyhole welding for (a) lap joints and (b) fillet joints.

6.1. Electrical contact resistance and corresponding tensile strength

Fig. 10. Welding power depicted over time for different sonotrode pressures p.
The electrical contact resistance was measured as explained in
Section 2. In Fig. 14, the electrical contact resistances of each of the
Temperatures, which arise when connecting cells by laser beam
three investigated welding techniques are shown when the welded
welding, are discussed in Section 6.
area is varied. The optimal geometry of the weld seam for the used
Laser beam welding allows widely arbitrary weld geometries
test samples is given by the edges transversely relative to the
which can be optimized for every application [31]. Furthermore,
longitudinal direction of the overlap [33]. The optimal weld seam
battery tabs or connector bars with a thickness of several
can be realized by laser beam welding resulting in relatively low
millimeters can be joined by keyhole welding [25,28].
electrical contact resistances. For the investigations shown in
Especially for metal surfaces, the reflection of the laser beam is
Fig. 14 the laser weld seam was modified from four independent
problematic, because it can damage objects in close vicinity.
weld seams each 2 mm long to two weld seams each 13.5 mm long.
Furthermore, gaps at the interface of the work pieces must be
The measuring point left in the diagram of Fig. 14 shows two weld
avoided, e.g., by clamping devices. As indicated in Fig. 12(a),
seams each 13.5 mm long but using superimposed beam oscilla-
spatters expulsed from the keyhole can be another unwanted side
tion. For resistance spot welding, the number of weld spots was
effect [25,26,30].
varied from 4 to 8 to 12 weld spots. The way the weld spots were
For battery assemblies, joining of two different metals can be
arranged is the best approach of the optimal weld seam with the
required, e.g., an aluminum cell terminal with a copper external
lowest electrical contact resistances. For ultrasonic welding, four
conductor [31]. When laser beam welding is used, the two molten
sonotrodes with different dimensions were used. In Fig. 14 each
materials are mixed and a metallurgical system is generated,
welding technique has its own abscissa, because the welded areas
which influences the mechanical properties. For example, if the
depend on the joining technique. On the left side of each abscissa,
solubility of the metals is limited, intermetallic phases can be
the values for suboptimal weld geometries are shown whereas the
produced, which weaken the strength of the joint. For welding
best possible weld geometries, in terms of electrical contact
dissimilar materials, elaborate methods have to be applied to
resistance, are placed on the right side of the abscissas.
reduce these intermetallic compounds. For example, a laser-braze-
As already discussed in Section 3, spot welding of
welding process can be used for an aluminum–copper joint [32].
CuZn37 generates voids inside the weld nugget and, therefore,
higher electrical contact resistances as well as lowest ultimate
6. Comparison of the three welding techniques
tensile forces were measured. A possible stray current might pose
another difficulty for further weld spots, when some weld spots
This Section quantitatively compares the three presented
are already made in close vicinity. As discussed in Section 3, a good
welding techniques for connecting battery cells in terms of
weld spot can only be achieved as long as the contact interface
electrical contact resistance, ultimate tensile force and heat input
possesses a relatively high contact resistance. With every pair of
into the cell. In this comparison section, only the results for
weld spots made this contact resistance is reduced and therefore
CuZn37 test samples according to Fig. 2 are discussed, because
the quality of further weld spots might decrease.
CuZn37 can be welded by all three welding techniques.

Fig. 13. Micrographs of nickel-plated steel sheets joined by keyhole welding (a)
Fig. 11. Electrical contact resistances and ultimate tensile forces depending on without beam oscillation and (b) with superimposed beam oscilation.
pressure of sonotrode. (Laser power PL, beam velocity vL, osc. frequency f, osc. amplitude A)

Fig. 14. Electrical contact resistances for CuZn37 test samples depending on the
welding technique and weld area; graphs include 90% confidential interval.
For ultrasonic welding, the micrograph in Fig. 9 revealed very
small spots of actual connection at the surface. Increasing the
welding energy has led to cracks and complete fracture of the test
samples. The measured electrical contact resistances are the
highest for ultrasonic welding.
To evaluate the mechanical strength of the welded test
samples, a tensile strength test was performed, where the samples
were pulled apart in longitudinal direction. Fig. 15 shows the
graphs for the ultimate tensile forces corresponding to the
electrical contact resistances from Fig. 14. No matter which
welding technique was used, all test samples failed at the weld
seam.
The voids inside the weld nuggets for spot-welded CuZn37 test
samples reduce the strength of the joint. But for spot-welded
Hilumin test samples the base material failed and not the weld
spots themselves (see Fig. 6). This was only observed with
resistance spot welding.
The ultrasonically welded test samples produced with the
sonotrode area of 68.25 mm2 and 81.25 mm2 show almost the
same ultimate tensile force. This is because the welded area had
the same width with respect to the tensile direction.
Fig. 15. Comparison of the three welding techniques in terms of ultimate tensile
force depicted as a function of welded area; graphs include 90% confidential
interval.
Fig. 16. CuZn37 sheets welded on negative terminal of 26650 lithium-ion cells;
from left to right: resistance spot weld, ultrasonic weld, laser beam weld.
Because the optimal weld geometry could be realized with laser
beam welding and because of beam oscillation, the highest tensile
strengths as well as the lowest electrical contact resistances were
achieved with laser beam welding.
6.2. Heat input caused by welding process
The three welding techniques were applied to cylindrical
lithium-ion cells of 26650 size. As external conductor a
CuZn37 sheet of 0.2 mm thickness was welded at the negative
pole of the cell. The negative tab of the battery cells is made of
nickel-plated steel. Welding results for the 26650 lithium-ion cells
and the chosen geometries of the weld areas are shown in Fig. 16.
These welding processes were recorded by the thermographic
camera A325sc by Flir. The battery cells, the conductors and the
background were prepared with black chalk spray in order to
reach
an emission coefficient e = 0.95. The hot spot of the welding process
lies beneath the external conductor and, therefore, is not visible
for the thermographic camera. As explained in the previous
sections, with resistance spot welding melting temperatures are
reached and with laser keyhole welding boiling temperatures are
reached. For ultrasonic welding, 30–60% of the melting
temperature of the material is reached, but the exact values
strongly depend on the welding parameters and properties of the
material [13,34]. The highest temperatures occurring in a hot
spot at the external
Fig. 17. Thermographic images of connected 26650 litihum-ion cells; highest
temperatures at cell terminal and cylinder barrel marked.
Table 1
Quantitative comparison for CuZn37 test samples; listed temperatures were measured for joining CuZn37 specimens to negative terminal of 26650 lithium-ion cell made
of nickel-plated steel.
Resistance spot welding
Min. electrical contact resistance 0.167 mV
Max. ultimate tensile force 316.78 N
◦ ◦
Highest temp. at cell terminal 31.0 C
◦ ◦
Highest temp. at cylinder barrel 25.0 C
conductor will not harm a battery cell but the temperatures inside
the casing, which arise at the electrochemically active materials,
are of special interest. In lithium-ion cells, the electrochemically
active materials are in direct contact with the cell casing [35]. For
this reason, the highest temperatures at the cell terminal and the
cylinder barrel are evaluated. These two temperatures are marked
in every thermal image of Fig.17. The thermal images taken
directly after the welding process are presented in the upper row
of Fig. 17. The thermal images, 6 s after the completion of the
welding process, are shown in the lower row. The images after 6 s
indicate how the heat distributes from the welded area into the
battery cell. Accelerated rate calorimetry on various lithium-ion
technologies showed that the decomposition of
electrochemically active
◦
materials starts at about 80 C [36]. Cell manufacturers usually
◦
define 60 C as the maximal temperature.
Directly after the welding process, the areas with increased
temperatures are very focused for each welding technique (see
Fig. 17). For resistance spot welding, the temperature scale in
◦
Fig. 17 is limited to 40 C because the maximum casing
◦
temperature did not exceed 31.0 C. Laser beam welding generates
higher temperatures also because the welded area is larger.
Ultrasonic welding generates the highest measurable temper-
◦
atures with temperatures at the casing exceeding 110 C. On the
one hand this is caused by the weld area, which is larger than the
four weld spots or the circular laser weld seam. On the other hand
the welding time for producing the ultrasonic weld was
significantly longer than for contacting the 26650 cells by
resistance spot or laser beam welding. Nevertheless, the
maximum temperatures occur only on the cell terminal and
◦ References
drop below 70 C only 3 s after the welding process finished.
Thus,
the lithium-ion cell was not damaged.
7. Conclusion
In this study, resistance spot, ultrasonic and laser beam
welding were investigated and compared with regard to
connecting cells in large battery assemblies. The functional
principle was presented and the influences of the adjustable weld
parameters were discussed for each welding technique.
Furthermore, the material properties influencing weld quality
were explained and the suitability of each welding technique for
connecting battery cells was determined, as is summarized
below:
- Resistance spot welding:
Especially welding with parallel electrodes is not suitable for
connecting battery cells, when a high ampacity (>20 A) is required
by the application. Nevertheless, resistance spot welding is
suitable for joint partners with low conductivity, e.g. nickel-plated
steel conductors and 18650 sized cells.
- Ultrasonic welding
Whenever the work pieces are scrubbed together with relative
ultrasonic vibrations, a weld can be produced. Dissimilar and
highly conductive metals can also be welded. Hence, ultrasonic
welding is suitable for nearly any type of battery cell connection.
Without clamping, ultrasonic vibrations can lead to cracks in the
external conductor and damage the battery cell’s inside.
Ultrasonic welding Laser beam welding
0.169 mV 0.130 mV
661.32 N 876.80 N
◦
110.7 C 86.8 C
◦
55.2 C 32.5 C
- Laser beam welding:
Due to the high power density of the laser beam, a large variety
of metals can be welded as well as conductors with several mm of
thicknesses. Superimposed beam oscillation improves the weld
quality. Nearly any battery cell connection can be produced but the
risk of surface reflections and spatters has to be considered.
The three investigated welding techniques were also quantita-
tively compared in terms of ultimate tensile force, electrical
contact resistance and heat input resulting from the welding
process. CuZn37 sheets were used for the comparison, because
they can be welded with each of the mentioned welding
techniques. Table 1 provides an overview of the results.
Lowest electrical contact resistances and highest joint
strengths were obtained with laser beam welding, because the
optimal weld geometry could be realized. Resistance spot welding
caused voids inside the weld nuggets of the CuZn37 test samples.
For ultrasonic welding, the welding energy was reduced to avoid
cracks in the test samples due to resonance. The highest heat input
occurred at ultrasonic welding, but for all welding techniques the
heat was very localized and no damaging temperatures occurred
at the lithium-ion cells.
Acknowledgements
The results presented in this paper were gathered within the
research project EEBatt, funded by the Bavarian Ministry of
Economic Affairs and Media, Energy and Technology. The authors
thankfully acknowledge its financial support. The responsibility
for this publication rests with the authors.
[1] H. Zhang, J. Senkara, Electrothermal Processes of Welding, in Resistance
welding: Fundamentals and Applications, 2nd ed., CRC Press, Boca Raton, FL,
2012, pp. 53–101.
[2] Y. Zhou, P. Gorman, W. Tan, K.J. Ely, Weldability of thin sheet metals during
small-scale resistance spot welding using an alternating-current power
supply, J. Electron. Mater. 29 (May (9)) (2000) 1090–1099, doi:http://dx.doi.
org/10.1007/s11664-004-0270-z.
[3] Y. Zhou, S.J. Dong, K.J. Ely, Weldability of thin sheet metals by small-scale
resistance spot welding using high-frequency inverter and capacitor-
discharge power supplies, J. Electron. Mater. 30 (May (8)) (2001) 1012–
1020, doi:http://dx.doi.org/10.1007/BF02657726.
[4] CD CPM2 Welder – User Manual, SUNSTONE, Engineering, Payson, UT, 2009.
[5] D.W. Dickinson, J.E. Franklin, A. Stanya, Characterization of Spot Welding
Behavior by Dynamical Electrical Parameter Monitoring, Welding Journal,
Welding Research Supplement, pp. 170–176, June 1980.
[6] J.E. Gould, An Examination of Nugget Development during Spot Welding,
Using Both Experimental and Analytical Techniques, Welding Journal, Welding
Research Supplement, pp. 1s–10s, January 1987.
[7] Z. Han J., Orozco, J. E. Indacochea, C. H. Chen, Resistance Spot Welding: A
Heat Transfer Study, Welding Journal, Welding Research Supplement, pp.
363–371, September 1989.
[8] HILUMIN1 nickel-plated steel strip diffusion annealed, Tata Steel Europe Ltd.,
London, UK: [Online], available 04.08.2014: http://www.tatasteeleurope.com/
en/products_and_services/products/flat/electro_plated_steels/hilumin/.
[9] S.J. Dong, G.P. Kelkar, Y. Zhou, Electrode sticking during micro-resistance
welding of thin metal sheets, IEEE Trans Electron Packag. Manuf. 25 (October
(4)) (2002) 355–361, doi:http://dx.doi.org/10.1109/TEPM.2002.807732.
[10] Grundlegende Untersuchung zur Kontaktsituation beim Widerstands-
schweissen von Kupferwerkstoffen, Materials Testing Institute University
Stuttgart, Stuttgart, Germany, Tech. Rep.-February 2010.
[11] L.P. Connor, R.L. O’Brien, W.R. Oates, Ultrasonic Welding of Metals, Welding
handbook Vol. 3 – Welding Processes Part 2, 9th ed., Miami, FL, American
Welding Society, 2007.
[12] J. Devine, Ultrasonic welding, in: D.L. Olson, T.A. Siewert, S. Liu, G.R. Edwards
(Eds.), ASM Handbook, vol. 6, ASM International, Materials Park, OH, 1993, pp.
324–327.
[13] E. de Vries, Mechanics and Mechanisms of Ultrasonic Metal Welding, Ph.D.
dissertation, College of Engineering, Ohio State University, Columbus, OH,
2004.
[14] H.P.C. Daniels, Ultrasonic welding, Ultrasonics 3 (October (4)) (1965) 190–196.
[15] K.C. Joshi, The formation of ultrasonic bonds between metals, Weld. J. 50
(Aug. (12)) (1971) 840–848.
[16] S.S. Lee, T.H. Kim, S.J. Hu, W.W. Cai, J.A. Abell, J. Li, Characterization of joint
quality in ultrasonic welding of battery tabs, J. Manuf. Sci. Eng. 135 (April (2))
(2013) 21004, doi:http://dx.doi.org/10.1115/1.4023364.
[17] S. Lee, Process and Quality Characterization for Ultrasonic Welding of Lithium-
ion Batteries, Ph.D. Dissertation, University of Michigan, Ann Arbor, MI,
2013.
[18] C.Y. Kong, R.C. Soar, P.M. Dickens, Optimum process parameters for ultrasonic
consolidation of 3003 aluminum, J. Mater. Process. Technol. 146 (February (2))
(2004) 181–187, doi:http://dx.doi.org/10.1016/j.jmatprotec.2003.10.016.
[19] S. Elangovan, K. Prakasan, V. Jaiganesh, Optimization of ultrasonic welding
parameters for copper to copper joints using design of experiments, J. Adv.
Manuf. Technol. 51 (April (1–4)) (2010) 163–171, doi:http://dx.doi.org/
10.1007/s00170-010-2627-1.
[20] E.A. Neppiras, Ultrasonic welding of metals, Ultrasonics 3 (July (3)) (1965)
128–135, doi:http://dx.doi.org/10.1016/S0041-624X(65)80003-8.
[21] M. Okada, S. Shin, M. Miyagi, H. Matsuda, Joint mechanism of ultrasonic
welding, Mater. Trans. 4 (4) (1963) 250–255.
[22] M. Shakil, N.H. Tariq, M. Ahmad, M.A. Choudhary, J.I. Akther, S.S. Babu, Effect of
ultrasonic welding parameters on microstructure and mechanical properties
of dissimilar joints, Mater. Des. 55 (March) (2014) 263–273, doi:http://dx.doi.
org/10.1016/j.matdes.2013.09.074.
[23] J. Blau, Friction coefficients of battery metals and the usage in ultrasonic
welding simulations, Presented at 27th EVS, Barcelona, Spain, Nov. 17 –20,
2013, doi:http://dx.doi.org/10.1109/EVS.2013.6914778.
[24] S. Choi, G. Zhang, T.A. Fuhlbrigge, S. Nidamarthi, Vibration Analysis in Robotic
Ultrasonic Welding, in TePRA, Woburn, MA, 2012, pp. 1–6. DOI:
10.1109/CoASE. 2012.6386349.
[25]
S. Katayama, Introduction: fundamentals of laser welding, in: S. Katayama (Ed.),
Handbook of laser welding technologies, Woodhead Publishing, Cambridge, UK, 2013,
pp. 3–17.
[26] S. Postma, Weld Pool Control in Nd:YAG Laser Welding, Ph.D. Dissertation,
University of Twente, Enschede, Netherlands, 2003.
[27] W.M. Steen, J. Mazumder, Laser welding, Laser Material Processing, 4th ed.,
Springer, London, UK, 2010, pp. 199–249, doi:http://dx.doi.org/10.1007/978-1-
84996-062-5_5.
[28] B.L. Luan, G. Campbell, M. Gauthier, X.Y. Liu, I. Davidson, J. Nagata, M. Lé pinay,
F. Bernier, S. Argue, Surface modification and fabrication of Li-ion battery
components for plug-in hybrid electric vehicle, ECS Trans. 25 (November (35))
(2010) 59–71, doi:http://dx.doi.org/10.1149/1.3414004.
[29] B. Mehlmann, A. Olowinsky, M. Thuilot, A. Giller, Spatially modulated laser
beam micro welding of CuSn6 and nickel-plated DC04 steel for battery
applications, Proceedings of 6th LAMP, Niigata, Japan, 2013, pp. 1–8.
[30] M. Schweier, J.F. Heins, M.W. Haubold, M.F. Zaeh, Spatter formation in laser
welding with beam oscillation, Phys. Proc. 41 (2013) 20–30, doi:http://dx.doi.
org/10.1016/j.phpro.2013.03.047.
[31] P. Schmidt, M. Schweier, M.F. Zaeh, Joining of lithium-ion batteries using
laser beam welding: electrical losses of welded aluminum and copper joints,
Proceeding of the 31th ICALEO, Anaheim, CA, 2012, pp. 915–923.
[32] T. Solchenbach, P. Plapper, W. Cai, Electrical performance of laser braze-welded
aluminum–copper interconnects, J. Manuf. Processes 16 (April (2)) (2014) 183–
189, doi:http://dx.doi.org/10.1016/j.jmapro.2013.12.002.
[33] P.A. Schmidt, T. Pauleser, M.F. Zaeh, Optimisation of Weld Seam Configurations
Using a Genetic Algorithm, in 8th DET, Stuttgart, Germany, 2014.
[34] H. Li, H. Choi, C. Ma, J. Zhao, H. Jiang, W. Cai, J.A. Abell, X. Li, Transient
temperature and heat flux measurement in ultrasonic joining of battery tabs
using thin-film microsensors, J. Manuf. Sci. Eng.135 (October (5)) (2013) 51015,
doi:http://dx.doi.org/10.1115/1.4024816.
[35] Y. Inui, Y. Kobayashi, Y. Watanabe, Y. Watase, Y. Kitamura, Simulation of
temperature distribution in cylindrical and prismatic lithium ion secondary
batteries, Energy Convers. Manage. 48 (February (7)) (2007) 2103 –2109, doi:
http://dx.doi.org/10.1016/j.enconman.2006.12.012.
[36] M. Brand, S. Glaeser, J. Geder, S. Menacher, S. Obpacher, A. Jossen, D. Quinger,
Electrical safety of commercial Li-ion cells based on NMC and NCA technology
compared to LFP technology, Presented at the 27th EVS, Barcelona, Spain, Nov.
17–20, 2013, doi:http://dx.doi.org/10.1109/EVS.2013.6914893.