Fatigue evaluation of fiber samples

Abstract

Apparatuses, methods, and related computer programs for assessing fatigue of a fiber sample including hair fibers are disclosed. The method includes receiving fatigue test data for a plurality of loading cycles for a fiber sample, wherein the fatigue test data includes force or displacement-based stress or strain data; determining, for at least a subset of the cycles, a loading energy based on data from the fatigue test data; and determining an indication of fatigue in the fiber sample based on the loading energy for at least the subset of the cycles.

Description

Technical Field

[0001] This disclosure relates to a method for evaluating the fatigue of fiber samples, as well as related computer program code and apparatus. In particular, the method may involve evaluating the fatigue of human hair samples. Background Technology

[0002] Mechanical testing, particularly tensile testing and fatigue analysis, is used in a wide range of industries to analyze the properties of products and materials, allowing engineers to improve designs and chemical formulations.

[0003] In the rapidly evolving consumer goods sector, the development of hair care products necessitates testing the product's impact on hair fiber properties. This allows for evaluation of the product's effectiveness as a chemical treatment agent and comparison with existing formulations. Typically, this analysis is performed by comparing the mechanical properties of fibers treated with and without the product. A common mechanical parameter used for these comparisons is Young's modulus. This can be measured using a simple tensile test (measuring force vs. strain, or stress vs. strain). In some instances, approximately 50 fibers of each type undergo repeated tensile tests (treated vs. untreated). This type of test can be termed a "classical tensile test." Overall, this experiment is relatively quick and can be used to screen formulations to determine their effectiveness in influencing the physical properties of fiber samples in the desired manner.

[0004] However, compared to stress / strain measurements, fiber fatigue testing is more relevant to the performance of some materials, including human hair, under real-world conditions. In some instances, fatigue measurements can be performed by repeatedly cyclically straining and recovering the sample. In hair care product testing, this analysis is more representative of the mechanical damage experienced by hair fibers during combing than classical tensile testing. Typically, this cyclic process can be repeated until the fiber breaks. This type of test can be called a "single-fiber fatigue test." The average number of "cycles to break" for multiple fibers (e.g., 30 or 50 fibers or more) can be recorded. The literature also reports that the differences between hair types in fatigue measurements can be much larger than in conventional measurements.

[0005] However, using fatigue measurements to screen active compositions is currently impractical because statistically significant fatigue analysis is extremely time-consuming. A limiting factor is the need for repeated measurements (thousands of cycles) on each of multiple fibers until fiber failure. For example, a study involving 30 fibers could potentially take many weeks.

[0006] It is possible to accelerate fiber breakage by altering experimental parameters. This could include changing the amount of stress, force, or strain used in the experiment. Similarly, humidity or other environmental parameters used in the experiment could be adjusted to affect the number of cycles required to break the fibers. However, significant alterations to these physical parameters can cause the test to deviate significantly from real-world, consumer-relevant conditions, thus reducing the evidentiary value of the assessment.

[0007] Some aspects of this disclosure relate to methods for accelerating hair fiber fatigue measurements under desired consumer-related conditions, making formulation screening feasible.

[0008] Other aspects of this disclosure relate to identifying the initial physical properties of the fiber sample being tested. Summary of the Invention

[0009] According to embodiments of this disclosure, a method for evaluating the fatigue of one or more fiber samples is provided, comprising:

[0010] Fatigue test data for fiber samples received after multiple loading cycles;

[0011] For at least a subset of the cycles, the applied energy is determined based on data from the fatigue test;

[0012] Fatigue indicators are determined based on the loading energy of at least a subset of the cycle.

[0013] Fatigue testing can be a single-fiber fatigue test.

[0014] In one or more embodiments, determining the fatigue rate includes determining the rate of change of the applied energy relative to the number of cycles.

[0015] In one or more embodiments, the rate of change of the loaded energy is based on a smoothed value of the loaded energy over multiple cycles.

[0016] In one or more implementations, the fatigue rate indication is determined based on the initial loading cycle.

[0017] In one or more embodiments, the fatigue test data includes multiple cycles of fewer than one of the following: 10,000, 20,000, or 50,000 cycles.

[0018] In one or more embodiments, the method further includes an indication of the initial condition of the fiber sample based on a plurality of maximum values ​​of the applied energy relative to the number of cycles.

[0019] In one or more embodiments, fatigue test data includes force or displacement-based stress or strain data. As a further alternative to using force or tensile displacement, fatigue test data may include bending or torsional force or stress as a function of displacement or strain.

[0020] In one or more embodiments, the loading energy of a cycle is determined by integrating the force or / or displacement-based stress or / strain of the loaded portion of the cycle.

[0021] In one or more embodiments, the loading energy of a cycle is determined by integrating the loading / unloading hysteresis internal forces or displacement-based stress or strain of that cycle. This method can be computer-implemented.

[0022] In one or more embodiments, determining fatigue indication based on cyclic loading energy includes integrating the force or displacement-based stress or strain of the loaded portion of the cycle.

[0023] In one or more embodiments, determining fatigue indications further includes integrating the force or displacement-based stress or strain of the unloading portion of the cycle to determine the energy loss within the cycle.

[0024] In one or more embodiments, the method further includes an automated test plan that stops the sample in response to determining that a fatigue indication meets a predetermined value.

[0025] In one or more embodiments, one or more fiber samples include human or animal hair fibers.

[0026] In one or more embodiments, the method further includes performing fatigue tests over multiple loading cycles, which can be executed as an automated test plan.

[0027] According to a further embodiment of this disclosure, a method is provided, comprising:

[0028] Fatigue test data for fiber samples received after multiple loading cycles;

[0029] For at least a subset of the cycle, the applied energy is determined based on data from fatigue testing;

[0030] An indication of the initial condition of the fiber is determined based on the loading energy characteristics as a function of the cycle number.

[0031] According to another aspect, this disclosure provides a computer program that can be distributed via electronic data transmission, comprising computer program code means adapted to cause the computer to perform any of the methods described herein when the program is loaded onto the computer; or this disclosure provides a computer program product comprising a computer-readable medium having computer program code means thereon adapted to cause the computer to perform any of the methods described herein when the program is loaded onto the computer.

[0032] According to a further embodiment of this disclosure, a fatigue testing device is provided, which includes a controller configured to perform one or more methods described herein during use. Attached Figure Description

[0033] One or more embodiments will now be described by way of example only with reference to the accompanying drawings, wherein:

[0034] Figure 1 A method for evaluating the fatigue of fiber samples is shown;

[0035] Figure 2 The force-strain hysteresis plot for the fiber loading cycle is shown;

[0036] Figure 3 A graph showing the measured loaded energy as a function of the cycle number is displayed;

[0037] Figure 4 It shows that for those with a ratio Figure 3 The sample showed significantly more localized fracture points, as measured by the plot of the applied energy as a function of the cycle number; and

[0038] Figure 5 The results of a “breakage cycle” experiment comparing raw hair samples and heat-treated hair samples are shown.

[0039] Figures 6A to 6C Showing previous references Figures 1 to 4 The results of the described analytical methods were compared between raw hair samples and heat-treated hair samples.

[0040] Figure 7 A schematic block diagram of a system for performing a method to evaluate the fatigue of fiber samples is shown; and

[0041] Figure 8 It displays a computer-readable medium containing computer program code. Detailed Implementation

[0042] This disclosure relates to an improved method for determining fatigue in fiber samples. In some cases, the fiber sample may contain human or animal hair fibers to allow for effective comparative testing of damage treatments (such as heat treatment) or chemical treatments (such as different shampoos or other consumer product formulations). Alternatively, the fiber sample may contain synthetic or natural fibers.

[0043] Figure 1An example method 100 for evaluating the fatigue of fiber samples is outlined. This method includes receiving fatigue test data 102 of the fiber sample over multiple loading cycles. The fatigue test data may include force or displacement-based stress or strain data. In some examples, the method may include performing fatigue tests to obtain fatigue test data, and it should be understood that in practice, multiple different samples may be tested under various conditions to allow for comparisons between the performance of one or more products.

[0044] For at least a subset of the cycle, the loading energy 104 is determined based on data from fatigue testing.

[0045] Then, fatigue indicators 106 are determined based on the loading energy of at least a subset of the cycle.

[0046] The claimed method 100 provides a way to predict fiber fracture during testing without waiting for an actual fracture event to occur. This can allow for increased efficiency in evaluating the effectiveness of various treatment formulations and can allow for screening a much larger number of formulations over a given time period. In particular, it has been found that the use of loading energy provides a predictor that allows for useful comparisons of fatigue performance between fiber samples without testing the fiber samples until failure. Therefore, the fatigue test data described herein do not necessarily include testing of fiber samples until failure.

[0047] The following is for reference. Figures 2 to 4 The various aspects of this method are discussed in further detail, including the determination of fatigue indicators. In particular, refer to... Figure 3 The first indicator of fatigue is described, and references are made. Figure 4 This describes a second indicator of fatigue.

[0048] As will be understood from the discussion below, various aspects of this method allow for the study of the evolution of specific parameters as the number of fatigue cycles applied varies. In some examples, the results for each sample set can be correlated with conventional fatigue measurements of the number of fracture cycles.

[0049] Figure 2 The force-strain hysteresis curve for an exemplary loading cycle 200 of a single loading cycle is shown. The X-axis represents the strain 210 experienced by the fiber sample, and the Y-axis represents the force 208 applied to the sample. Alternatively, the Y-axis can display stress, and the X-axis can display strain, producing similar results. The force applied to the fiber sample induces stress in the test material. The data captured for the selected cycle includes the complete loading portion 202 (upper curve) of cycle 200 for the sample and the complete unloading portion 204 (lower curve) of cycle 200.

[0050] The experiment is conducted using a predetermined stress. The cyclic loading ends when this stress is reached. Other embodiments of the experiment can use a fixed force or a fixed strain.

[0051] The area under the force / strain loading portion 202 of cycle 200 provides a measure of the loading energy. The loading energy for a given cycle can be determined by integrating the force / displacement-based stress / strain of the loading portion 202 for that cycle. The loading energy varies with the number of loading cycles applied to the fiber sample. The loading energy parameter can be used to determine the number of cycles required for the fiber sample to fracture. Therefore, monitoring the loading energy can be used to provide a predictable assessment of the fatigue of the fiber sample. For example, while the analysis may not necessarily allow fracture prediction to be performed on a fiber-by-fiber basis, fatigue performance can be compared between fiber samples based on determined values. Fiber samples with a larger magnitude of the rate of change of loading energy relative to the number of cycles indicate a higher fatigue rate compared to fiber samples with a lower rate of change.

[0052] Fatigue test data can be acquired using existing commercially available instruments using standard data acquisition parameters and techniques. Fatigue data can be obtained using the Dia-Stron CYC801. In conventional experiments, only the number of cycles leading to fracture is recorded. However, in the disclosed method, the loading energy is determined for at least a subset of cycles based on data from fatigue testing. In the illustrated data, force / strain curves are recorded during selected cycles of the fatigue test (every 200th cycle in this example).

[0053] The area 206 between the loading section 202 and the unloading section 204 indicates the amount of energy lost during the loading cycle 200. The energy loss indicated by area 206 is equal to the area under the loading curve 202 (loading energy) minus the area under the unloading curve 204 (unloading energy).

[0054] Additional parameters derived from the loading energy can also be used to determine an indication of fatigue. The energy lost through the fiber sample during the cycle can be determined using the area under the loading portion 202 of cycle 200, or the area under the unloading portion 204 of cycle 200, or the area 206 within the hysteresis curve. The energy lost during the cycle can be used as a parameter to provide a predictive indication of fiber breakage. More specifically, variations in the shape of the loading portion 202 of cycle 200, the unloading portion 202 of cycle 200, and the area 206 within the hysteresis curve determined for each selected cycle can allow for the determination of indicators of the likelihood of fiber breakage or fatigue.

[0055] The following is for reference. Figure 3 The first aspect of discussing methods for determining fatigue indices.

[0056] Figure 3The time series distribution 300 of selected data samples 302, including physical samples (or fiber samples), is displayed. Distribution 300 shows the loaded area 304 for each selected sample 302 relative to the cycle number 306. Figure 3 Each point in the table represents the loaded energy of the selected cycle, which can be used, for example, from a previous reference. Figure 2 The measurements described are used to determine each point. Alternatively, each point may correspond to parameters derived from the loading energy, such as the energy lost during the cycle (loading energy minus unloading energy).

[0057] Fatigue has been demonstrated by the variation of the loading curve across multiple loading cycles. It should be understood that the term "loading cycle" is used in its conventional sense here: a cycle in which loads are applied and removed. Therefore, the evolution of the area under the loading curve can be used as a predictive parameter for failure, as well as as an indicator of fatigue. Figure 3 Trends in the data can be used to provide an indication / prediction of fiber breakage before the fiber undergoes complete breakage. For example, such as Figure 3 As shown, initially, the loading energy (as indicated by the loading area 304) decreases as the number of cycles 306 increases. In particular, the loading energy decreases with the logarithm of the number of cycles 306.

[0058] As the number of cycles 306 increases, the fiber is weakened, and therefore the amount of energy lost during each cycle (i.e., the loading area 304 under the loading curve) also decreases. This can be interpreted as the fiber sample becoming more easily deformed by the applied force.

[0059] Indicating fatigue rate can include determining the rate of change of the applied energy relative to the number of cycles. In particular, the initial slope of the applied energy relative to the number of cycles has been found to provide a useful fatigue indicator. The fatigue rate can be determined based on the initial loading cycles. This can be determined by comparing energy loss associated with the initial state of the fiber sample, for example, when the loaded area has decreased by a threshold level (e.g., 10% or 20%). Alternatively, the initial loading cycles can include, for example, a fixed number of cycles, such as 5,000, 10,000, 20,000, or 50,000 cycles. As another alternative, the number of loading cycles providing the initial loading cycles can be based on determining when the rate of change of the initial slope of the applied energy relative to the number of cycles reaches a threshold.

[0060] The rate of change of loading energy can be based on a smoothed value of loading energy 304 over multiple cycles 306. The application of a standard smoothing function is useful for removing sample-to-sample variability and improving the consistency of the rate of change calculated based on the smoothing function with the overall trend. In other words, smoothing the first derivative of the data provides an effective reduction rate of loading area relative to the number of cycles.

[0061] like Figure 3 As shown, the data also includes unexpected spikes 308 in the curves. These spikes are interpreted as corresponding to partial fiber breakage events.

[0062] To enable the prediction of fiber failure using collected data, it is preferable to consider the temporal evolution of the area under the loading curve without being affected by partial fracture events. Conventional techniques can be used to filter out outliers or data samples affected by partial fracture events, such as determining when the loaded area of ​​a sample reaches a threshold level compared to the loaded area of ​​an ongoing sample (or a group of data samples). Alternatively, outliers can be filtered out by detecting local maxima in the data and excluding those data samples.

[0063] It has been found that by employing a pattern based on the first derivative of the data, the analysis can be less affected by spikes associated with small / partial fracture events. Other averaging methods can also be used; however, additional signal processing may be required to eliminate the influence of partial fractures.

[0064] In this way, Figure 3 The first derivative of the initial segment of the data shown can be considered to provide a fatigue indication of the fiber sample tested according to the first aspect of this method. This method has been found to identify differences between fiber samples that have undergone different treatments (e.g., untreated hair fibers vs. heat-treated hair fibers) after the 50th data sampling point, which corresponds to the 10,000th cycle (i.e., 50 * 200 cycles) for every 200th cycle of data capture. Therefore, only 20,000 cycles are needed in total to compare the two treatments. That is, in this example, only the initial segment of data from the first 10,000 cycles for each of the two samples is considered. This is approximately an order of magnitude fewer cycles than the number required using a test-to-failure method. Therefore, the method described above allows for a significant improvement in sample testing and may allow, for example, the efficient testing of more treatments or the repeated testing of the same treatment to verify the results.

[0065] First-order differentials have been found to provide a significantly improved fatigue metric compared to many other candidate metrics, including time series analysis (SARIMAX), log / linear fitting, and derivatives of autocorrelation and log / linear data combined with various integrals.

[0066] The following describes a specific example implementation of the first aspect of the method. (See also: Regarding...) Figure 1 In practice, the method discussed may also include fatigue testing over multiple loading cycles. In a specific example, the data processing stages required to predict fiber fracture include:

[0067] For each fiber sample:

[0068] 1. Data acquisition is performed using a commercial fatigue testing instrument (such as the Dia-Stron CYC801 or other similar systems).

[0069] 2. Transfer the data to a suitable system, such as a laptop;

[0070] 3. Use Python for data processing in Jupyter notebooks or any other data processing system;

[0071] 3.1 Data visualization (optional - analysis can be run in batch mode without visualization) and cleanup.

[0072] 3.2 Prepare the data for each loop:

[0073] Find the threshold in the force / strain data.

[0074] Find the area under the loaded portion of the force / strain curve, which is equal to the "loaded area".

[0075] 3.3 Response Build: Load Area vs. Loop Count

[0076] 3.4 The response data is smoothed and a local first derivative is taken at each data point to give the decay constant. This is achieved using the Savitzky-Golay filter (Signal.savgol_filter Python package, version 1.6.3, The SciPyCommunity, https: / / docs.scipy.org / doc / scipy / reference / signal.html).

[0077] 3.5 The model for calculating the attenuation constant value based on the data of this fiber (see Figure 6 below for a more detailed discussion of the model calculation example).

[0078] 3.6 For each dataset:

[0079] One of the datasets includes 40 values ​​for each fiber treatment, and there are at least two fiber treatments.

[0080] The Mann-Whitney U test (scipy.stats.mannwhitneyu Python package, version 1.6.3, The SciPy Community, https: / / docs.scipy.org / doc / scipy / reference / generated / scipy.stats.mannwhitneyu.html) was used to compare the data between fiber processing pairs.

[0081] The data analysis described above may involve thousands of data points for each of the thousands of cycles.

[0082] The following is for reference. Figure 4 The second aspect of the discussion method includes determining the initial condition of the fiber sample.

[0083] Figure 4 Showing comparison with previous references Figure 3 Another example of a similar type of time series distribution 400 is described, with a loading area of ​​404 relative to a cycle number of 406. Figure 4 In, with Figure 3 Compared to the sample tests in the previous tests, the data showed many unexpected protrusions 408. As mentioned earlier, these protrusions or outliers are interpreted as corresponding to partial fracture events.

[0084] exist Figure 4 The analysis also includes many erroneous data points 410 that have no physical meaning. For example, these data points can be filtered out from the analysis by thresholding.

[0085] The difference in the number of partial fracture events detected in the distribution 400 of the fiber samples is interpreted as indicating that the fiber samples are in a different initial state or condition compared to other fiber samples. The initial state refers to the state of the fiber prior to the fatigue test. Therefore, the initial state includes the fiber's total past history, including the effects of all combing events and treatments prior to the measurement. Aspects of the initial state may include, for example, fiber strength, integrity, physical damage (such as splitting within the fiber), and moisture concentration. The initial state can be representative of how much damage the fiber sample has already suffered before the fatigue test. Initially more intact fibers may be able to withstand more partial fracture events before complete breakage. Therefore, a greater number of unexpected protrusions 408 can indicate a fiber with a less damaged initial state, where the fiber was initially more intact than a fiber that only exhibits a few protrusions.

[0086] In other words, fibers exhibiting more 408 protrusions during data acquisition can be considered to be in a better initial state than fibers with fewer protrusions. Therefore, fibers in a better initial state can withstand more small / partial breakage events during the experiment before the entire fiber breaks or weakens to the point where the loaded energy reaches a threshold level.

[0087] The observation of the number of partial fracture events in a fiber provides an indicator of the fiber's initial state, which is not available from conventional testing. In other words, the initial state of a fiber sample cannot be taken into account in conventional testing, and this can lead to a certain range of fatigue measurements of the fiber sample.

[0088] Therefore, by counting the number of partial breakage events (reaching failure or up to a predetermined number of cycles, such as 100,000), a measure of existing fiber damage that defines the initial state of a fiber sample can be provided.

[0089] In this way, Figure 3 or Figure 4 The number of partial fracture events detected in the data can be considered to provide an indication of the initial state of the fiber sample being tested according to the second aspect of the method.

[0090] Figure 5 The experiment using conventional fatigue measurement methods was compared with Figure 6. Figure 1 The data was obtained using a new method. The samples used in the experiment were from the following hair types:

[0091] a) A tuft of original mixed medium brown Caucasian hair, and

[0092] b) A strand of mixed medium brown Caucasian hair that has been heat-treated (using a hair styling curling iron at 450°F 30 times).

[0093] Raw hair samples were provided by International Hair Importers. For both hair types, samples were taken from the tips of the fibers. Therefore, two sample types were available.

[0094] Fatigue tests were conducted using a cyclic testing instrument (Dia-Stron CYC801) in "constant stress mode" at 80% relative humidity (80% RH). The fatigue rate was set to 60 mm / s, and the number of fracture cycles was measured using a target stress of 105 MPa at 80% relative humidity. Note that both force and stress (stress = force / cross-sectional area) were recorded in these experiments. However, because the fatigue process will change the effective cross-section of the fiber (the active cross-section of the fiber will decrease as a function of crack propagation), the analysis is performed as a force / strain curve. This also eliminates the need to correct the data using the total fiber cross-sectional dimensions. This is reasonable in this application because we are investigating the evolution of parameters between cycles for the same fiber. To calculate the stress, the fiber dimensions were measured at 60% RH (at five points along the fiber).

[0095] Approximately 40 fibers were used in each experiment. Instrumentation issues resulted in smaller datasets for some units of the experiment. Furthermore, due to limitations on the amount of data that could be saved during a single experiment, some fibers (those requiring very large breakage cycles) had to be rerun, resulting in more than one data file per fiber.

[0096] Figure 5The results of a “breakage cycle” experiment comparing raw hair samples with heat-treated hair samples are shown. In both cases, hair was taken from the tip of the hair tuft.

[0097] Figure 6 shows the previous reference. Figures 1 to 4 The described analytical method is applied to the results of A) the complete dataset (all broken cycle data), B) the first 100 data points, and C) the first 50 data points.

[0098] from Figure 5 A comparison with 6 shows that, using both classical and new analytical methods, the fatigue properties of hair exhibit similar qualitative changes under both heated and unheated conditions.

[0099] Figures 6A to 6C The results also show that good differentiation between samples can be achieved for all three datasets. Therefore, satisfactory results can be obtained, for example, using only the first 50 data points. This indicates a good difference compared to the reference dataset. Figure 5 Compared to conventional fatigue measurements, the obtained Figure 6C The amount of data required to produce the experimental results shown has been reduced by a factor of 7.

[0100] The data processing steps used to prepare Figure 6 are discussed below. As mentioned earlier, the first derivative has been calculated for each time point. The resulting data is formed as a histogram of counts versus first derivative values. In this example, 100 bins are used for the values ​​of the first derivative, although this can vary. The bin with the largest number of counts has been determined, and the midpoint of that bin has been assigned to the decay constant. This process has been performed for each fiber in the set to give a pattern distribution of the decay constant for each fiber set. The histogram is obtained using the matplotlib.pyplot.hist Python package (see https: / / matplotlib.org / stable / api / _as_gen / matplotlib.pyplot.hist.html). The maximum value of the histogram is found using the numpy.argmax Python package (see https: / / numpy.org / doc / stable / reference / generated / numpy.argmax.html).

[0101] Figure 7A schematic block diagram illustrating a system 700 that can be used to implement the previously described methods is provided. System 700 includes one or more processors 702 communicating with a memory 704. The memory 704 is an example of a computer-readable storage medium. The one or more processors 702 also communicate with one or more input devices 706 and one or more output devices 708. Various components of system 700 can be implemented using common methods known in the art for computing. For example, input device 706 may include a keyboard or mouse, and output device 708 may include a monitor or display, as well as audio output devices such as speakers. Additionally, system 700 includes fatigue testing equipment, which may be conventional, such that fatigue testing equipment 710 is at least partially under the control of one or more processors 702.

[0102] Figure 8 A non-transitory computer-readable storage medium 800 is disclosed. The non-transitory computer-readable storage medium includes computer program code configured to cause a processor to perform the methods disclosed herein.

[0103] Using a computer for calculations allows for faster computation than a corresponding cycle-to-fracture experiment, enabling more rapid testing of fiber samples. It also allows for real-time processing—that is, processing data from each fatigue cycle immediately after acquisition. This allows for quality control during the experiment. Once data indicates that some predefined levels of quality and consistency have been achieved, this can also lead to the intelligent termination of the experiment. This approach can further save time. For example, the method could also include stopping the automated test program for a specific sample in response to determining that fatigue indicators meet predetermined values. In some examples, stopping the test program includes stopping any additional loading cycles applied to the fiber sample.

Claims

1. A method for evaluating the fatigue of a hair fiber sample, the method comprising: For multiple loading cycles, fatigue test data of fiber samples are received, wherein the fatigue test data includes force or displacement-based stress or strain data; For at least a subset of the cycle, the applied energy is determined based on data from the fatigue test data; and The fatigue indication of the fiber sample is determined based on the loading energy of at least a subset of the cycles.

2. The method of claim 1, wherein determining the fatigue rate indicator includes determining the rate of change of the applied energy relative to the number of cycles.

3. The method of claim 2, wherein the rate of change of the loaded energy is based on a smoothed value of the loaded energy of the plurality of cycles.

4. The method of claim 2 or 3, wherein the indication of fatigue rate is determined based on an initial loading cycle, wherein the initial loading cycle comprises fewer than one of the following: 10,000, 20,000, or 50,000 cycles.

5. The method of claim 2 or 3, wherein the indication of fatigue rate is determined based on an initial loading cycle, wherein the number of loading cycles used to provide the initial loading cycle is determined by the rate of change of the loading energy relative to the initial slope of the number of cycles reaching a threshold.

6. The method of claim 1, further comprising an indication of the initial condition of the fiber sample based on a plurality of maximum values ​​of the loading energy relative to the number of cycles.

7. The method as described in any of the preceding claims, wherein the cyclic loading energy is determined by integrating the force / displacement-based stress / strain of the loaded portion of the cycle.

8. The method of any one of claims 1 to 6, wherein determining the fatigue indication based on the cyclic loading energy comprises determining it by integrating the force or displacement-based stress or strain of the loaded portion of the cycle.

9. The method of claim 8, wherein determining the fatigue indicator further comprises integrating the force or displacement-based stress or strain of the unloading portion of the cycle to determine the energy loss within the cycle.

10. The method as claimed in any of the preceding claims, comprising an automated test plan for stopping the sample in response to determining that the fatigue indication meets a predetermined value.

11. The method of any of the preceding claims, wherein the hair fiber sample comprises human or animal hair fibers.

12. The method as described in any of the preceding claims further includes performing fatigue testing on the plurality of loading cycles.

13. A computer-readable storage medium comprising computer program code configured to cause a processor to perform the method of any one of claims 1 to 10 when in use.

14. A fatigue testing apparatus including a controller, the controller being configured to perform the method of claim 12 when in use.

Images