1. Introduction
The usage of lithium-ion batteries (LIBs) is of high importance in current, as well as future, electric vehicles (EV) due to a fast increasing market, which is accelerated by ambitious goals against climate change [
1]. Besides the continuous improvement of LIBs’ performance in terms of rising energy densities [
2], the improvement of their safety performance is also of great importance. Damage and deformations of EV battery systems in a crash situation can cause multiple safety issues [
3,
4]. For example, external short circuits (ESC) can result in the deformation of electrical conducting components such as cell tabs, high-voltage busbars or cables, and the resulting failure of their insulation layers as described in [
5,
6,
7]. On the other hand, a mechanical, electrical, or thermal abuse of the battery cell itself can lead to an internal short circuit (ISC) caused by separator failure followed by a thermal runaway (TR) [
5,
8]. Cell deformations can trigger TRs directly by deformation and failure of the separator [
8]. The TR mechanisms for a commercial NCM/graphite pouch cell, such as investigated in this study, can be described as a chain of chemical reactions. After the final failure of the separator layer during this process, a high-electric-energy release due to ISC takes place. This can lead to burning electrolytes and, in the worst case, explosions [
8].
A basic requirement for preventing TRs is a comprehensive understanding of the processes taking place in the battery cell triggered by an electrical, thermal, or mechanical overload. For example, the behavior of LIBs under mechanical loading has already been studied in a large number of publications. Different cell conditions and influencing factors such as strain rate effects [
9,
10,
11,
12,
13] and state of charge (SOC) dependence [
14,
15,
16,
17,
18] were investigated with a focus on safety assessment. Zhu et al. [
11] tested dry and wet jelly roll samples and showed that the strain-rate-dependent hardening effect of LIBs is a result of the electrolyte’s flow. Furthermore, it was shown that high strain rates can decrease the maximum force before failure under mechanical loads [
12,
13]. The SOC dependence of LIBs was mostly analyzed for cylindrical cells [
14,
16,
17,
18]. One main conclusion is that increasing SOC levels lead to a higher stiffness of cylindrical cells. This behavior was justified by an induced internal stress by anode expansion due to lithium insertion [
14,
16]. It was also shown that with increasing SOC, higher failure stresses can occur for cylindrical cells [
18]. For pouch cells, scant research is available. Li et al. [
16] discovered that the stiffness of pouch cells is higher only for increasing SOC when the expansion is prevented by geometric constraints [
16]. Deng et al. [
19] also indicated an SOC independence for non-pretensioned pouch cells under mechanical loads in their studies. In order to better understand the behavior of lithium-ion cells under mechanical loads, cell components were also characterized mechanically. Many researchers have focused on analyzing the separator [
20,
21,
22,
23,
24,
25], but some research for the anode and cathode behavior has also been published [
26,
27]. The aforementioned studies of LIBs’ mechanical behavior focused almost exclusively on fresh cells.
Further publications prove that, beside these mostly investigated influences, degradation effects of electrical aging, also leading to a capacity fade of LIBs [
28], must be considered in the safety assessment. These can affect the cells’ TR behavior [
29] and mechanical properties [
18,
22,
30,
31,
32,
33]. In addition, some literature is available with respect to the influence of aging on safety. For example, the effects of lithium plating, usually provoked by fast charging profiles and aggravated by low temperatures [
34], can have a high impact on the safety of LIBs [
29,
34,
35,
36]. Thus, a drastic decrease of the triggering temperature leads to a TR and an increase of maximum temperature can occur during TR [
34]. However, the thickening of the solid electrolyte interface (SEI) improves the TR performance by higher thermal stability [
29].
Although there is a high importance of electrical aging on the mechanical behavior of LIBs, to the authors’ knowledge, few studies [
18,
22,
30,
31,
32,
33] have been published considering the difference between fresh and electrically aged cells under mechanical loads. Wu et al. [
30] investigated electrical aged anode, cathode, and separator materials of a NCM pouch cell. In their studies, only the anode showed significant changes in terms of lower mechanical strength after electrical aging. These mechanical changes of the materials corresponded with significant visual changes of the aged anode in terms of deposition and degradation spots [
30]. Fink et al. [
31] noticed a similar influence of electrical aging on the anode’s mechanical properties. Additionally, a significant impact on the cathode’s mechanical properties in terms of lower Young’s modulus and tensile strength was mentioned [
31]. Zhang et al. [
22] showed that electrical aging of an elliptical cell changed the trilayer separator’s (PP/PE/PP) material properties to a lower mechanical strength. On a cell level, a lower strength and failure displacement was noticed. Chemical reaction products in the form of particles, which were stored in the separator pores and might have created stress concentrations that potentially led to earlier failure, were noticed in scanning electron microscopy (SEM) analyses. They also mentioned the physical degradation of the polymer chains as a possible reason for a lower failure strength [
22]. Another work investigating electrical aging effects on the mechanical behavior of automotive LIBs was published by Kovachev, et al. [
32]. The authors investigated fresh, and at 60
and 1C, electrically aged cells under quasi-static mechanical loads at 100% SOC. Aged cells showed a small decrease in stiffness and a right-shifted force–displacement curve with a higher failure force compared to fresh cells. The softer mechanical response was justified by an additional compression of a thicker SEI layer grown due to side reactions with lost active lithium at the anode’s surface graphite particles. Furthermore, a higher maximum force was noticed before failure. This was interpreted as the effect of a dry separator due to electrolyte decomposition and thus a hardening mechanical behavior [
32]. Additionally, Liu et al. [
33] investigated the mechanical response of pouch cells at four aging states focusing on low-temperature aging at 0
. They also indicated right-shifted force–displacement curves under mechanical indentation and higher force peaks for aged cells. Cell component tests indicated that aged anodes show a similar behavior. Additionally, a faster energy release after cell failure was detected on a cell level for 20% capacity faded cells, which led to a stronger voltage drop and higher temperature increase at the cell surface [
33].
Beside these valuable insights, important influencing factors were neglected in the mentioned studies. The difference of the mechanical behavior’s SOC dependence between fresh and aged cells, as well as the the impact of electrical aging under EV battery module conditions, were not investigated in detail until yet. Therefore, aged cells, which were electrically cycled in EV battery modules with an aging strategy focusing on real customer behavior, were analyzed in this paper. For example, multiscale mechanical tests were performed to investigate the influence of electrical aging and the SOC on the crash behavior of LIBs. Furthermore, several microscopy analyses were chosen to better understand the impact of electrical aging on the mechanical behavior of LIBs and their relevant safety properties.
3. Results
First, tensile and compression tests of the cell components, including anode, cathode, and separator, are presented as they are of high importance to understand the influence of electrical aging on the mechanical properties of lithium-ion cells. In a next step, mechanical abuse tests of fresh and electrically aged cells are conducted. Finally, the results of the microscopic analyses including SEM, visual inspection, and ICP-OES are presented. All curves in the following figures show the calculated average curves of each investigated parameter. In order to consider the spread of experimental data, the minimum and maximum values are marked additionally in the diagrams in terms of error bars.
3.1. Cell Component Tests
The tensile test results of the cell components, containing the stress–strain curves, values of Young’s modulus, and tensile strength of the anode and cathode material, are presented in
Figure 3. Besides the aging state, the cell directions
u (long side) and
v (short side) are shown to analyze if the investigated cell components show an anisotropic behavior under tension. Analyzing
Figure 3a,b, a significant dependency of the anode material’s mechanical behavior under tension due to aging effects can be noticed. The aged anode material shows a lower mechanical strength and softer response under mechanical loading. Analyzing the tensile strength, approximately 25% lower values can be noticed compared to the fresh material. In addition, a 50% lower failure strain and 30% lower Young’s modulus can be observed. The failure of the material for both aging states originated from a straight crack in the middle of the specimen.
These aging effects could be justified by a changed morphological structure of the copper current collector based on high temperatures during electrical cycling. Such an effect could occur as the temperatures would increase the material’s structural transformation temperature. This influence can be excluded by considering the maximum temperature of 36
that EVs are subjected to under typical climate conditions [
45]. It has to be mentioned that cell temperatures inside the battery probably reached higher values. Nevertheless, considering LIBs’ maximum temperature development during operation of up to 54
for an ambient temperature of 40
[
46], any influence from temperature effects can be excluded. On the other hand, pitting holes due to strong oxidation effects from the interaction of copper with the electrolyte [
47] could result in a lower mechanical strength. In this study, no oxidation effects were observed by visual inspection. Liu et al. [
33] tested copper and graphite as single materials in their study. They concluded that electrical aging has a low influence on the copper current collector’s mechanical behavior. For graphite, a high impact was noticed under tension. In summary, changes of the graphite’s morphological structure are the most probable reason.
As mentioned before, besides the aging effects, the mechanical anisotropic behavior was also investigated by analyzing the
u and
v direction of the cell. By considering the indicated error bar of the curves, it can be seen that a low difference of 5% can be observed at the maximum stress values for the fresh anode material in both directions. The aged anode shows very similar behavior, indicating a low anisotropy as well. Wang et al. [
26] concluded this behavior for fresh anodes. Comparing the two-side-coated aluminum collector with NCM in
Figure 3c,d with the anode, a lower influence of electrical aging on the mechanical behavior of the cathode can be determined. The fresh and aged stress–strain curves show a very similar characteristic tensile strength and failure strain. An 8% lower tensile strength is observed for the aged cathode in the
v direction. The Young’s modulus shows a lower mean value for the aged material of 20% in the
u direction. In this case, it has to be mentioned that there is a wide statistical spread, which has to be taken into account and therefore indicates a low aging effect on the mechanical behavior of the cell’s cathode material. This behavior was also noticed by Liu et al. [
33]. Considering the direction-dependent properties, the NCM–aluminum cathode can be described as an isotropic material. This cathode behavior is different compared to the mentioned anisotropic behavior of lithium nickel cobalt aluminum oxide (NCA) in the work of Wang et al. [
26]. A different Young’s modulus of about 44%, depending on the cell direction, was observed under tension [
26].
Figure 4 visualizes the experimental results of the separator’s tensile tests. The PE separator shows a significant dependency on the aging state. The determined maximum tensile strength has values of up to 50% lower. Unlike the anode and cathode, the separator shows necking behavior before mechanical failure under tension. In addition, an anisotropic behavior can be noticed with a different tensile strength of up to 17% for the fresh material.
Possible reasons for the changed material behavior during electrical aging could be the induced mechanical stress from the cell’s volume expansion, while charging/discharging. This can lead to damage of the separator’s lamellar crystals and also pore closure, according to [
48,
49]. The influence of the cell’s operating temperature spread can be excluded due to polyethylene’s low glass transition temperature of up to
[
50]. Furthermore, the wet condition of the separator in the pouch cell could have influenced the degradation and also the mechanical properties. While opening the cell for specimen preparation, most of the electrolyte was noticed at the separator.
Figure 5 shows the compression test’s stress–strain curves in the main diagram, and additionally the first range between 0% and 9% engineering strain in the sub-diagram. In addition, three different impactor velocities were tested to analyze strain rate effects on the mechanical behavior. All three materials show a strain-rate dependence under compression. This behavior can be explained by a reduced reaction time for the materials response in terms of displacement [
51]. As a result, the stress shows increased values, especially in the beginning. A neglecting of the compressible behavior of the materials can be noticed, which can play a significant role for dynamic load cases on lithium-ion cells in future investigations.
By analyzing the beginning of the curve, the aged anode material shows a softer mechanical response at a strain rate of
1/s and 1 1/s compared to the fresh one. This behavior can be explained by the growth of an additional layer on the SEI layer in terms of decomposition products through electrical aging. Similar effects were noticed in the work of Kovachev et al. [
32] on cell level under quasi-static indentation loads. Therefore, further investigations in terms of SEM analyses are presented in
Section 3.4, which confirm this argumentation. In addition, fresh and aged anodes show a force drop at about 50% engineering strain for strain rates of
1/s and 1 1/s. At this point, a brittle cracking behavior of the graphite layer was observed. This behavior was also seen in the work of Wu et al. [
30].
The cathode material shows very similar curve characteristics in both aging states for each strain rate. No significant difference can be observed except low differences for the stress values. This was expected after analyzing the tensile test results.
Compared with the results under tension, the separator’s mechanical behavior under compression is less dependent on electrical aging. The variance for fresh and aged stress–strain curves is low. The different impact of electrical aging on the mechanical properties for tension and compression can be explained by the morphological structure of the separator. In order to allow lithium ions to pass during the charging/discharging process, the separator is deliberately porous. Under tension, stress is applied along the weakest point, in this case the pores, and these crack more easily. In an aged cell, this material is more brittle and has lower strength. In contrast, load is applied in the thickness direction and not along the pores under compression, so there is no failure and also no lower strength. Despite the reduced aging influence under compression, it has to be mentioned that the aged material shows slightly increased stress values. Zhang et al. [
22] showed that decomposition products, caused by electrical aging, can close separator pores. Thus, the higher strength of the aged material could be explained by closed pores under compression.
3.2. Tests on Cell Level
After analyzing the test results on cell component level, the test results of the whole lithium-ion pouch cells are presented.
Figure 6 visualizes the fresh and aged cells’ force–displacement curves for (a) cylindrical indentation and (b) three-point bending at 0% SOC and 100% SOC. The voltage vs. displacement is also illustrated to show the point of occurring ISC based on a voltage drop. Through analyzing the test results in
Figure 6a, a simultaneous force and voltage drop under cylindrical indentation can be noticed, which indicates a direct correlation between mechanical cell failure and occurring ISC. This correlation was already described in different studies [
52,
53]. Furthermore, a high influence of electrical aging on the mechanical properties at cell level can be mentioned. Considering the test results of the anode material under compression in
Section 3.1, a similar influence on the mechanical properties by electrical aging can be noticed. The aged cells show a softer mechanical response indicated by a right-shifted force–displacement curve, which was also noticed in the work of Kovachev et al. [
32] for pouch cells. In addition, the mechanical behavior of aged cells shows a higher dependence on the investigated SOC levels compared to fresh cells. While fresh cells have a different maximum force of up to 4% depending on their SOC, the aged cells at 0% SOC show a lower maximum force of about 30% compared to the fully charged condition. That also represents a 29% lower force compared to fresh cells for 0% SOC, which is a different behavior compared to the literature [
32,
33]. In both studies, a higher failure force for aged cells was observed at 0% SOC. Furthermore, the spread of failure displacement for fresh cells in contrast to aged cells dependent on their SOC is very low. Another aspect which has to be taken into account is the different voltage drop after occurring ISC between fresh and aged cells. For both SOC conditions, aged cells show a stronger voltage drop after cell failure.
The test results of three-point bending confirm the fact that electrical aging leads to lower mechanical properties at cell level. As mentioned previously, under cylindrical indentation, a high SOC dependence of aged cells can be reported. The maximum force at fully charged condition is about 25% higher for electrical aged cells under three-point bending compared to 0% SOC. For fresh cells, a similar behavior can be noticed. By comparing the fresh and aged cells at the same SOC, it can be seen that, similar to cylindrical indentation, fresh cells show a higher maximum force of up to 22%. It has to be mentioned that, in contrast to cylindrical indentation, no ISC was detected for three-point bending up to displacements of 50 mm for all investigated cell conditions. This can be explained by the low compression of the cell components inside the cell during deformation. The main deformation is caused by bending without a local concentrated fracture zone. Under cylindrical indentation, the machine bed restricts the deformation of the cell in the w direction, which results in high local stress concentrations in the cell under the contacting impactor area. This can lead to a higher local plastic compression of the cell components inside the cell with a higher interaction between the single layers. Thus, a fracture of the cell components occurs, which leads to an ISC.
In order to compare fresh and aged cells with a focus on safety behavior in crash situations, it is not only the mechanical behavior that must be taken into account. In addition, the temperature evaluation after ISC can be used to investigate potential risks caused by TR. As already announced, two thermocouples (T
, T
) were installed on the cell surface at a distance of 80 mm for 100% SOC cylindrical indentation tests, as shown in
Figure 7b. Additionally,
Figure 7a illustrates the average temperature increase for each thermocouple after cell failure and ISC (t = 0 s), which indicates a TR. The observed cell behavior can be categorized in four stages. First, the ISC occurs, followed by the formation of smoke at 2 s. In a next step, at 5 s after ISC, a bloating of the cell can be noticed through gas exposition. In the last step, the pouch foil mechanically fails, which leads to a strong exothermic reaction and, as a result, to fire.
Analyzing the safety risks of fresh and aged cell conditions, it has to be mentioned that only one out of three fresh cells started to burn after cell failure in this experiment. In conclusion, the aged cells indicated a higher affinity to burn with two of three cells. By analyzing the temperature profile in
Figure 7a, it can be seen that the fire affinity of aged cells is higher. Furthermore, aged cells show a higher reactivity, which leads to a faster temperature increase in the first 5 s. In addition, the determined maximum temperatures show higher values of up to 75% compared to fresh cells and indicate a higher safety risk of aged cells after cell failure. Considering the findings in the work of Kovachev et al. [
54], a lower thermal conductivity of aged cells could have led to heat accumulation and, as a result, to higher temperatures [
54]. Another reason could be the deposition of lithium at the anode surface by lithium plating during charging and discharging. A temperature increase at oxygen could already be described in
Section 2.5. Further investigations in terms of ICP-OES analyses are shown in
Section 3.5, which confirm the existence of increased lithium deposits on the aged anode surface.
3.3. Thickness Measurement
Table 3 shows the thickness of fresh and aged cells, which was measured before testing to compare the aged cells’ growth due to electrical aging compared to fresh cells. Eight predefined geometric points on the cell surface per cell were evaluated. The mean thickness represents the average value of all tested cells per state. It can be observed that aged cells with 0% SOC are up to
% thicker than fresh cells. Furthermore, the volume expansion due to charging is higher for electrical aged cells, which indicates a higher SOC influence. Aged cells at 100% SOC show a
% increased cell thickness compared to fresh cells.
Compared to the literature [
32,
33], in the present work, a similar cell thickness growth for 0% SOC at a higher SOH level is reached. Kovachev et al. [
32] and Liu et al. [
33] noticed a cell thickness increase of about 6% and 8% for cells with an approximate SOH of 70% and 0% SOC. Both studies used an aging strategy with a constant ambient temperature and a low charging/discharging power. In contrast, in the present work, a high-performance aging strategy with an ambient temperature window between
and 31
was used. Higher charging rates and temperature variation could have directly caused stress concentrations within the anode material, which could have led to stronger rupture of the SEI layer. As a result, an accumulative reaction between the electrolyte and the active material may have occurred, leading to a faster increase in cell thickness [
55].
3.4. SEM Analysis
As already shown, the changes of the cell’s mechanical properties and characteristic behavior due to electrical aging are very similar to the changes of the electrical aged anode material.
Figure 8 shows different microscopic magnifications for the fresh and aged anode material.
At first glance, no difference in the anode’s morphological structure can be noticed at the lowest magnification. In addition, no change in the graphite particle size can be determined due to aging. The main difference, which can be detected by electrical aging, is the appearance of an additional layer on the graphite particles’ SEI layer. The generated layer can be traced back to the decomposition of different products from the interaction between the electrolyte and the anode’s surface during the battery operation. A similar aging effect was described in the work of Kovachev et al. [
32]. The softer mechanical response for aged lithium-ion cells under mechanical abuse was justified with the grown layer on the SEI [
32]. Liu et al. [
33] also noticed a similar additional layer on the graphite particles, which was described as a “mossy material”. It was assumed that the grown layer could be lithium dendrites [
33]. The aged anode’s softer mechanical behavior, mentioned in both studies, could be confirmed by the compression tests performed in the present work (see
Section 3.1). On a cell level, a right-shifted force-displacement curve was also observed. As described in the work of Kovachev et al. [
32], this can be explained with a higher elasticity of the additional layer compared to the graphite particles, which is compressed under mechanical loading. Furthermore, it has to be mentioned that these decomposition products show a nonhomogeneous structure. Additionally, there was no equal distribution on the anode surface for these aging phenomena noticed in this study. This could have led to a more rough and nonplanar anode surface compared to fresh cells. Analyzing the observed cell failure at lower force levels for aged cells at 0% SOC in
Section 3.1, these noticeable spots on the anode surface could have caused local stress concentrations inside the cell, which resulted in an earlier failure of the separator and ISC compared to fresh cells. The 50% lower mechanical strength and higher brittleness of the aged separator under tension, which was noticed in
Section 3.1, could have increased this effect.
3.5. Optical Assessment and ICP-OES Analysis
Figure 9a shows the anode’s results of ICP-OES analyses and
Figure 9b,c of the visual inspection. Neither optical inspection nor ICP-OES analysis revealed any unexpected changes for the cathode material, which is why they are not shown in this section. For a review of the cathode’s ICP-OES results, the reader is referred to the
Appendix A Figure A1.
In contrast, the comparison between the fresh and aged anode surface shows some significant differences. The aged anode exhibits several white colored spots in a ring shape (red marked), which indicate lithium plating according to Cannarella et al. [
44]. These local spots can be caused by pore-closure effects of the separator and thus high current concentration during the charging/discharging process [
44]. The results of ICP-OES analysis in
Figure 9a confirm the expectation of increased lithium deposits on the electrical aged anode surface. The amount of lithium increased by 355% compared to the fresh sample. In contrast, the weight percentages of all other elements examined are unchanged. These results can explain the elevated temperatures of the anode’s surface at oxygen, which were described in
Section 2.5 for the disassembling process. Furthermore, the cell’s temperature increase after ISC for 100% SOC cells under mechanical loading can be explained by a higher exothermic reaction of lithium deposits on the anode surface. The first phase of TR, describing the exothermic decomposition of the SEI layer [
8], could be accelerated due to faster temperature increase by reacting lithium depositions at the aged anode surface. Furthermore, the lithium deposition results in a more probable contact of lithium and electrolyte for aged cells after failure. This can be a reason for higher maximum temperatures after occurring fire. Analyzing the potential risk of this insight, the elevated temperatures of aged cells could also lead to an easier thermal propagation between neighboring cells in an EV battery module.
4. Summary and Conclusions
In this paper, fresh and electrical aged lithium-ion pouch cells were compared with a focus on their mechanical behavior in crash situations. The aged cells were electrically cycled under EV battery module conditions to 90% SOH. In order to analyze the main changes of aged cells relevant for crash safety, multiscale investigations in terms of mechanical cell component and cell level tests at 0% and 100% SOC were performed. Additionally, a detailed post-mortem analysis was conducted to investigate the main degradation effects leading to changed material behavior and potential safety issues.
It was shown that electrical aging influences the mechanical behavior of the anode and separator material. For both materials, a significantly lower failure strength and failure strain was noticed under tension compared to fresh samples. The separator material also showed increased anisotropic behavior with different failure strains and tensile strengths for both aging states. In contrast, no significant effect of electrical aging or anisotropy was found for the NCM cathode’s mechanical properties. In a next step, stack compression tests were performed under three different strain rates for fresh and aged cell components. A high strain-rate dependence was noticed for anode, cathode, and separator, indicated by increased stress values at higher strain rates and a neglect of the material-specific compressible behavior. Overall, a minor effect of aging on the mechanical properties was observed in the compression tests. Only the aged anode showed a slight right-shifted stress–strain curve compared to the fresh material. The results in terms of SEM analyses confirmed the appearance of grown decomposition products on the aged anode’s SEI layer by electrical aging. This can explain the softer mechanical response in the compression tests. The softening behavior was more significant at cell-level tests. The anode showed a great influence under cylindrical indentation loads due to high cell component compression, which was also reported in different publications. Furthermore, it was shown that cylindrical indentation is the critical load case in crash situations compared to three-point bending due to occurring internal short circuits for fresh and aged cells. Despite a simultaneous drop in force and voltage for both aging states, a stronger voltage drop for aged cells was noticed. In addition, lower failure forces of about 29% were determined for aged cells at 0% SOC, which was explained by local induced stress concentrations due to decomposition products and a lower mechanical strength of the aged separator. In addition, a stronger SOC dependence for aged cells was noticed. Furthermore, the 100% SOC cylindrical indentation tests indicated a more critical TR event for aged cells with increased maximum temperatures of up to 75% and a faster TR. This behavior was justified by ICP-OES analyses, where increased lithium deposits of about 355% were detected at the aged anode’s surface compared to fresh cells. In the authors’ opinion, this, among other effects, could have led to stronger reactions during exothermic decomposition of the SEI layer in the TR event and, consequently, to higher temperatures. An increased temperature development during TR caused by lithium plating effects could also be reported in several studies.
The mentioned insights are particularly important to consider for future battery design, as electrical aging can affect the mechanical response and TR behavior of the entire battery in crash situations. Based on the obtained results in this paper, the following conclusions were defined:
The investigated NCM pouch cell shows a simultaneous force drop and occurring internal short circuit, which indicates a direct relation between mechanical failure and internal short circuit.
Aged cells show a right-shifted force–displacement curve under cylindrical indentation, which seems to be a result of an additional layer on the anode’s SEI.
The mechanical behavior of aged cells shows a greater dependence on SOC compared to fresh cells.
High deformation of aged cells can result in lower failure forces of about 29%, which can be explained as an effect of local induced stress concentrations by products of decomposition and lower mechanical strength of aged separator materials.
Aged cells can show a stronger voltage drop and higher temperatures after internal short circuit compared to fresh cells, which indicates a higher safety risk in EVs.
When analyzing thermal runaway and thermal propagation behavior, special degradation mechanisms such as lithium plating caused by certain electrical cycling profiles should be considered, as they can change the temperature development after cell failure.