The relentless pursuit of higher energy density and enhanced safety in energy storage systems has positioned the lithium-ion (li-ion) battery as a cornerstone technology for modern portable electronics, electric vehicles, and grid-scale storage. The complex electrochemical performance and long-term durability of a li-ion battery are governed by the synergistic interactions between its core components: the cathode, anode, electrolyte, and separator. While significant research efforts are dedicated to advancing electrode and electrolyte materials, the mechanical integrity of the separator is increasingly recognized as a critical, yet sometimes underestimated, factor influencing both performance and safety.
Functionally, the separator in a li-ion battery serves a dual purpose: it must be a permeable membrane to facilitate ionic transport between electrodes and an impeccable electronic insulator to prevent physical contact between the anode and cathode, thereby averting internal short circuits. This role becomes critically demanding during operation. Repeated lithium intercalation and de-intercalation during charge-discharge cycles induce significant volume changes in electrode materials, particularly high-capacity ones like silicon. These dimensional changes exert mechanical stresses on the separator. Furthermore, during fast charging or under abusive conditions like overcharge, the growth of lithium dendrites poses a direct penetration threat. A separator with insufficient mechanical strength can be easily breached by these metallic protrusions, leading to catastrophic thermal runaway. Therefore, a comprehensive understanding and precise quantification of a separator’s mechanical properties are paramount for predicting its behavior in a working li-ion battery and for designing safer, more robust cells.
Existing literature reveals that the mechanical response of separators is highly dependent on their material composition and manufacturing process. Widely used polyolefin-based separators, such as those made from polyethylene (PE) or polypropylene (PP) manufactured via dry or wet processes, often exhibit pronounced in-plane anisotropy. Their mechanical properties—elastic modulus, tensile strength, and Poisson’s ratio—can differ significantly between the machine direction (MD) and transverse direction (TD). This anisotropy stems from the molecular and microstructural alignment induced during the film-stretching stage of production. For instance, studies have reported elastic modulus differences up to a factor of 5 and strength differences up to a factor of 10 between the MD and TD for certain dry-process separators. Furthermore, these materials frequently show strong strain-rate sensitivity, where properties like modulus and yield stress increase with loading rate, a behavior typical of viscoelastic polymers. In contrast, non-woven or ceramic-coated separators have been reported to demonstrate more isotropic behavior.
This study focuses on the detailed mechanical characterization of a specific separator: a polyethylene (PE) base membrane coated on both sides with a thin layer of polyvinylidene fluoride (PVDF). PVDF coating is often applied to enhance electrolyte wettability and thermal stability. The primary objective is to determine its fundamental mechanical parameters—elastic modulus, ultimate tensile strength, and Poisson’s ratio—in both principal in-plane directions. A secondary goal is to investigate the influence of tensile strain rate on these properties, which simulates different mechanical loading conditions a li-ion battery might experience. The experimental methodology combines conventional uniaxial tensile testing with full-field strain measurement using Digital Image Correlation (DIC) technology. This integrated approach allows for the accurate determination of elastic constants and provides insights into the deformation homogeneity and failure mechanisms. Contrary to the typical anisotropy found in many commercial separators, our findings indicate that this particular PVDF-coated PE separator exhibits remarkably isotropic in-plane mechanical properties, coupled with high modulus and strength, presenting a potentially advantageous profile for li-ion battery applications where uniform mechanical response is desired.
1. Experimental Methodology
1.1 Material and Specimen Preparation
The separator under investigation was supplied in roll form. The core material is polyethylene (PE) with a nominal base thickness of 8 μm. A layer of polyvinylidene fluoride (PVDF) is coated on both surfaces, with each coating layer having an approximate thickness of 1 μm, resulting in a total separator thickness of about 10 μm. The overall web width was 1300 mm.
Specimens for mechanical testing were prepared with meticulous care to ensure consistency and avoid edge defects that could prematurely initiate failure. The primary testing orientations were the Machine Direction (MD) and the Transverse Direction (TD). As a supplementary check, specimens oriented at ±45° relative to the MD were also prepared. For each orientation, a batch of 20 specimens was fabricated to ensure statistical relevance.
The sampling procedure was designed to obtain representative material. The outer three layers of the membrane roll were discarded to avoid any surface damage or contamination. Square sheets of 500 mm × 500 mm were first cut from the roll, ensuring one edge was parallel to the MD (deviation less than 5°). From each square, four tensile coupons were carefully cut using a sharp surgical blade against a straight edge. The coupon geometry was a standard dog-bone shape with a overall length of 250 mm, a width of 15 mm, and a central reduced-section gauge length of 100 mm. All cutting was performed on a flat, clean surface under controlled environmental conditions of (23 ± 2)°C temperature and (50 ± 10)% relative humidity. Operators wore gloves to prevent contamination. The quality of each specimen was verified using optical microscopy (Zeiss Smartzoom 5) to confirm the absence of nicks, tears, scratches, or visible imperfections along the edges and within the gauge section.
1.2 Uniaxial Tensile Testing for Modulus and Strength
Quasi-static uniaxial tensile tests were conducted using an LS5 universal testing machine (AMETEK, USA) with a 5000 N load cell. The machine was placed on a vibration-isolated optical table to minimize external disturbances. Specimens were mounted in the wedge-action grips with an initial gauge length (distance between grips) set to (100 ± 5) mm. Care was taken to align the specimen’s longitudinal axis with the loading axis of the machine. The grip pressure was adjusted to be sufficiently high to prevent slippage during testing but not so high as to cause crushing or premature failure at the grip edges.
The tensile tests were performed at four different crosshead displacement rates: 1 mm/min, 25 mm/min, 100 mm/min, and 250 mm/min. The rate of 250 mm/min was selected as the primary reference rate, as it aligns with standard test methods for determining the elastic modulus of plastic films. The other rates were used to systematically investigate the strain-rate dependence of the separator’s mechanical properties. During each test, the machine recorded the applied force (F) and the crosshead displacement (Δl) with high resolution. The engineering stress (σ) and engineering strain (ε) were calculated post-test using the following fundamental relations:
$$ \sigma = \frac{F}{A_0} $$
$$ \varepsilon = \frac{\Delta l}{L_0} $$
where \( A_0 \) is the original cross-sectional area (width × thickness) and \( L_0 \) is the original gauge length (100 mm). The elastic modulus (E) was determined from the slope of the initial linear portion of the engineering stress-strain curve. According to standard practice, the slope was calculated using a linear least-squares fit over a strain interval from 0.05% to 0.25%. The ultimate tensile strength (σuts) was simply taken as the maximum engineering stress recorded before specimen fracture.
1.3 Poisson’s Ratio Measurement via Digital Image Correlation (DIC)
To measure Poisson’s ratio, which requires simultaneous measurement of axial and transverse strains, a non-contact optical technique was employed. A 2D Digital Image Correlation (DIC) system was integrated with the tensile testing machine. Prior to testing, the surface of the specimen’s gauge section was prepared with a high-contrast, random speckle pattern. This was achieved by first cleaning the surface and then applying a fine mist of matte black paint over a white background.
The DIC setup consisted of a high-resolution digital camera mounted on a stable tripod, positioned normal to the specimen surface. Uniform white-light illumination was provided to ensure consistent lighting. The camera was focused on the gauge section, and its settings (aperture, exposure time) were adjusted to obtain a sharp, well-contrasted image. The image acquisition software was synchronized with the tensile testing machine to capture images at a fixed frequency throughout the loading process.
For the Poisson’s ratio tests, a constant, slow crosshead speed of 1 mm/min was used. This rate was chosen to provide sufficient time for image capture and to align with standard recommendations for measuring Poisson’s ratio, which suggest a strain rate on the order of 1% per minute.
After the test, the sequence of images was processed using DIC software. The software tracks the movement of the speckle pattern subset by subset between consecutive images. By comparing the deformed images to the reference (undeformed) image, the software computes the full two-dimensional displacement field (\(u_x, u_y\)) over the region of interest. From these displacement fields, the in-plane strain components—axial strain (εxx along the loading direction) and transverse strain (εyy perpendicular to loading)—are calculated. Poisson’s ratio (ν) is defined as the negative ratio of the transverse strain to the axial strain in the linear elastic region:
$$ \nu = -\frac{\varepsilon_{yy}}{\varepsilon_{xx}} $$
In practice, the slope (K1) of the transverse strain vs. time curve and the slope (K2) of the axial strain vs. time curve were calculated over the strain interval from 0.3% to 0.8% axial strain. Poisson’s ratio was then computed as:
$$ \nu = -\frac{K_1}{K_2} $$
This method provides a robust and direct measurement of the Poisson’s ratio for the separator material.
2. Results and Discussion
2.1 Elastic Modulus and Ultimate Tensile Strength at Reference Rate
The engineering stress-strain curves for both TD and MD specimens tested at the reference strain rate of 250 mm/min are shown below (representative curves). The curves exhibit typical polymeric film behavior: an initial linear elastic region, followed by a nonlinear regime leading to yielding and plastic deformation, and finally fracture. The repeatability in the elastic region was excellent across multiple specimens.
The elastic modulus (E) and ultimate tensile strength (σuts) were extracted as described. The results for five specimens in each direction are summarized in Table 1.
| Direction | Sample | Elastic Modulus (MPa) | Ultimate Strength (MPa) |
|---|---|---|---|
| TD | 1 | 2015.4 | 164.9 |
| 2 | 2091.0 | 182.2 | |
| 3 | 2073.8 | 181.8 | |
| 4 | 2043.3 | 188.0 | |
| 5 | 2217.0 | 178.9 | |
| Average | 2025.7 | 179.1 | |
| MD | 1 | 1886.6 | 176.0 |
| 2 | 1894.7 | 182.8 | |
| 3 | 2131.5 | 164.1 | |
| 4 | 2259.7 | 161.3 | |
| 5 | 1820.9 | 163.2 | |
| Average | 1998.7 | 169.5 |
The average elastic modulus is 2025.7 MPa for TD and 1998.7 MPa for MD, a difference of only about 1.3%. The average ultimate strength is 179.1 MPa for TD and 169.5 MPa for MD, a difference of approximately 5.7%. Both differences are within the typical scatter of polymeric film testing and are significantly less than the 10% threshold often used to indicate a meaningful difference. Therefore, we can conclude that, at this strain rate, the PVDF-coated PE separator exhibits essentially isotropic in-plane elastic modulus and tensile strength. This is a notable finding, as it contrasts with the strong anisotropy reported for many commercially available dry-process polyolefin separators used in li-ion batteries. The high values of modulus (~2 GPa) and strength (~175 MPa) are advantageous for a li-ion battery separator, as they indicate a robust membrane capable of withstanding significant internal stresses from electrode expansion and resisting penetration attempts from lithium dendrites.
An important observation related to failure mode was recorded. Specimens that fractured within the central gauge section, away from the grips, yielded consistent and reproducible data. Specimens that failed near the grip edges showed greater data scatter and often lower strength values, likely due to stress concentration induced by grip pressure or minor edge damage. All data reported in this study are from specimens that exhibited the desired “gauge failure” mode.
2.2 Influence of Tensile Strain Rate
The mechanical behavior of polymeric materials, including those used in li-ion battery separators, is inherently viscoelastic. Therefore, properties like stiffness are expected to depend on the rate of loading. To characterize this effect, tests were conducted at additional crosshead speeds: 100 mm/min, 25 mm/min, and 1 mm/min. Representative stress-strain curves at 100 mm/min and 25 mm/min showed similar shapes to those at 250 mm/min, but with different slopes in the elastic region. The calculated average elastic moduli and ultimate strengths are compiled in Table 2. For the slowest rate (1 mm/min), only the modulus is reported, as the test duration to failure was excessively long.
| Tensile Rate | Direction | Avg. Elastic Modulus (MPa) | Change vs. 250 mm/min | Avg. Ultimate Strength (MPa) | Change vs. 250 mm/min |
|---|---|---|---|---|---|
| 100 mm/min | TD | 1702.8 | -16.0% | 166.8 | -6.9% |
| MD | 1644.8 | -17.7% | 158.6 | -6.4% | |
| 25 mm/min | TD | 1567.2 | -22.6% | 162.8 | -9.1% |
| MD | 1532.3 | -23.3% | 153.0 | -9.7% | |
| 1 mm/min | TD | 903.8 | -55.4% | — | — |
| MD | 851.5 | -57.4% | — | — |
The data reveals a strong and systematic strain-rate dependence of the elastic modulus. As the tensile rate decreases from 250 mm/min to 1 mm/min, the measured modulus drops by over 55%. This is a classic viscoelastic response: at higher rates, the polymer chains have less time to undergo rearrangement and slip, resulting in a stiffer apparent response. At very low rates, the molecular segments have sufficient time to relax and reconfigure under stress, leading to a significantly lower modulus. This has important implications for li-ion battery modeling, as the separator’s stiffness perceived during a sudden impact (high rate) will be much higher than during slow, long-term expansion from electrode cycling (low rate).
In contrast, the ultimate tensile strength shows a much weaker dependence on strain rate. The decrease from the 250 mm/min value is less than 10% even at 25 mm/min. This suggests that while the initial elastic resistance is highly rate-sensitive, the final failure strength, which involves large plastic deformation and chain disentanglement or rupture, is a more intrinsic material property with less pronounced rate effects within this range. The isotropic nature of the properties is maintained across all tested strain rates, as the values for TD and MD remain closely matched at each rate.
2.3 Poisson’s Ratio Determined by DIC
The DIC system provided full-field strain maps and precise strain-time histories for both the axial (loading direction, sign2) and transverse (sign1) directions. Representative strain-time curves for a TD specimen tested at 1 mm/min are shown conceptually below. The axial strain increases linearly with time initially, while the transverse strain decreases linearly. Poisson’s ratio was calculated from the negative ratio of the slopes of these linear portions in the strain interval of 0.3% to 0.8% axial strain.
The results for five specimens in each orientation are presented in Table 3.
| Direction | Sample 1 | Sample 2 | Sample 3 | Sample 4 | Sample 5 | Average |
|---|---|---|---|---|---|---|
| TD | 0.468 | 0.438 | 0.413 | 0.475 | 0.411 | 0.441 |
| MD | 0.458 | 0.438 | 0.430 | 0.410 | 0.432 | 0.434 |
The average Poisson’s ratio is 0.441 for the TD direction and 0.434 for the MD direction. The difference is a negligible 1.6%, confirming that this property is also isotropic for the tested separator. The value of approximately 0.44 is within the typical range reported for polymeric materials and is comparable to values found for other commercial li-ion battery separators. This isotropic Poisson’s ratio further reinforces the conclusion that the PVDF-coated PE separator behaves as an in-plane isotropic material. From a battery safety perspective, an isotropic separator with a well-characterized Poisson’s ratio is advantageous. It ensures predictable and uniform deformation in all in-plane directions when subjected to external pressure or internal electrode expansion, reducing the risk of localized stress concentrations that could lead to tearing or failure. A separator that contracts uniformly in the transverse direction when stretched aids in maintaining complete coverage between electrodes, a crucial factor in preventing short circuits.
3. Conclusion
This work presents a comprehensive experimental investigation into the mechanical properties of a specific polyethylene-based separator with dual-sided PVDF coating for li-ion battery applications. By integrating standard uniaxial tensile tests with advanced Digital Image Correlation (DIC) strain mapping, a complete set of fundamental mechanical parameters was accurately determined.
The key findings are as follows:
- In-Plane Isotropy: Contrary to the pronounced anisotropy commonly reported for many dry-process polyolefin separators, the tested PVDF-coated PE separator exhibits essentially isotropic behavior in the plane of the membrane. Both the elastic modulus and ultimate tensile strength showed no significant directional dependence between the Machine Direction (MD) and Transverse Direction (TD). The Poisson’s ratio, measured directly via DIC, was also identical within experimental error for both directions.
- High Stiffness and Strength: The separator demonstrates relatively high mechanical properties, with an elastic modulus of approximately 2 GPa and an ultimate tensile strength of about 175 MPa when tested at a standard rate of 250 mm/min. These values indicate a robust membrane capable of withstanding significant mechanical challenges within a li-ion battery, such as electrode volume changes and potential dendrite growth.
- Pronounced Strain-Rate Sensitivity: The elastic modulus of the separator is highly dependent on the tensile strain rate, a characteristic viscoelastic response. The modulus decreased by over 55% when the testing rate was lowered from 250 mm/min to 1 mm/min. This strong rate dependence must be accounted for in mechanical modeling of li-ion batteries, as the operational conditions can range from slow, long-term creep to rapid impact events.
- Rate-Independent Ultimate Strength: In contrast to the modulus, the ultimate tensile strength showed only a mild dependence on strain rate, varying by less than 10% across the rates from 25 mm/min to 250 mm/min. This suggests that the failure strength is a more intrinsic material property for this separator.
The combination of high strength, isotropic response, and well-characterized rate-dependent stiffness makes this PVDF-coated PE separator a promising candidate for advanced li-ion battery designs where uniform mechanical performance and reliability under diverse loading conditions are critical. The experimental data and methodology provided here serve as a essential foundation for engineers and researchers to model separator behavior, optimize battery pack safety designs, and develop next-generation separators with tailored mechanical properties for specific li-ion battery applications.