- Email: [email protected]
Strain rate and anisotropy effects on the tensile failure characteristics of human skin
journal of the mechanical behavior of biomedical materials 41 (2015) 241–250
Available online at www.sciencedirect.com
www.elsevier.com/locate/jmbbm
Research Paper
Strain rate and anisotropy effects on the tensile failure characteristics of human skin Me´lanie Ottenioa,b,c,n, Doris Trana,b,c, Aisling Nı´ Annaidhd, Michael D. Gilchristd, Karine Bruye`rea,b,c a
Université de Lyon, F-69622, Lyon, France Université Claude Bernard Lyon 1, Villeurbanne, France c IFSTTAR, UMR_T9406, LBMC Laboratoire de Biomécanique et Mécanique des Chocs, F69675, Bron, France d School of Mechanical & Materials Engineering, University College Dublin, Belfield, Dublin 4, Ireland b
art i cle i nfo
ab st rac t
Article history:
The anisotropic failure characteristics of human skin are relatively unknown at strain rates
Received 6 May 2014
typical in impact biomechanics. This study reports the results of an experimental protocol
Received in revised form
to quantify the effect of dynamic strain rates and the effect of sample orientation with
6 October 2014
respect to the Langer lines. Uniaxial tensile tests were carried out at three strain rates
Accepted 8 October 2014
(0.06 s 1, 53 s 1, and 167 s 1) on 33 test samples excised from the back of a fresh cadaver.
Available online 16 October 2014
The mean ultimate tensile stress, mean elastic modulus and mean strain energy increased
Keywords:
with increasing strain rates. While the stretch ratio at ultimate tensile stress was not
Human skin
affected by the strain rate, it was influenced by the orientation of the samples (parallel and
Ex-vivo tensile test
perpendicular to the Langer lines. The orientation of the sample also had a strong
Dynamic strain-rate
influence on the ultimate tensile stress, with a mean value of 28.075.7 MPa for parallel
Anisotropy
samples, and 15.675.2 MPa for perpendicular samples, and on the elastic modulus, with
Langer lines
corresponding mean values of 160.8 MPa753.2 MPa and 70.6 MPa759.5 MPa. The study also pointed out the difficulties in controlling the effective applied strain rate in dynamic characterization of soft tissue and the resulting abnormal stress–strain relationships. Finally, data collected in this study can be used to develop constitutive models where high loading rates are of primary interest. & 2014 Elsevier Ltd. All rights reserved.
1.
Introduction
Skin is the largest organ of the human body with an area of about 1.8 m2 and an average thickness of 2 mm. In medical terms, skin can be referred to as an integument, derived from
the latin tegument meaning cover. Indeed, skin covers the whole human body and one of its primary functions is to protect from foreign aggressions. Human skin is subjected to a variety of loading conditions during our daily motions, where it is stretched, sheared and pinched. Skin however,
n Corresponding author at: Laboratoire de Biomécanique et Mécanique des Chocs UMR_T 9406 (IFSTTAR-UCBL), Site Ifsttar Lyon-Bron, 25 Avenue François Mitterrand, Cité des Mobilités, 69675 Bron, France. Tel.: þ33 4 72 14 23 85. E-mail address: [email protected] (M. Ottenio).
http://dx.doi.org/10.1016/j.jmbbm.2014.10.006 1751-6161/& 2014 Elsevier Ltd. All rights reserved.
242
journal of the mechanical behavior of biomedical materials 41 (2015) 241 –250
tends to return to its initial state due to its elastic properties. Moreover, its high mechanical strength ensures it is not easily torn, and therefore, our skin protects us from tetanus, open hemorrhages, or from direct damage to underlying organs. Skin also fails due to trauma such as planned surgical incisions, gunshot wounds and penetrative wounds from stabbing attacks or can even originate internally when a bone pierces the skin. In 1861, Langer was interested in the geometry changes of open wounds. By puncturing a cadaver’s skin with a circular device, he noted that the wound transformed into an elliptic form. By joining the major axes of these ellipses, he drew a pattern of tension lines on the body: these lines are now known as the Langer lines (Langer, 1978). These early experiments illustrated the anisotropic nature of skin, but did not explain fully the mechanism involved. Indeed, the correlation between the Langer lines, the macroscopic mechanical properties of human skin and its structural basis is still under investigation (Ridge and Wright, 1966; Ní Annaidh et al., 2012a; Gąsior-Głogowska et al., 2013). Recently, Ní Annaidh et al. (2012b) have shown through the use of quantitative structural data, that the Langer lines have a structural basis. A point which had previously been alluded to but not quantitatively assessed. The current study takes place in the general context of trauma biomechanics and at strain rates relevant in automotive collisions and in sharp force injury. The collection of experimental data on human skin is an essential step in the development of a numerical model capable of predicting the rupture of skin in sharp force injury. In order to obtain experimental data relevant for this application, it is necessary to understand how skin behaves when it is dynamically loaded above its physiological range (up to rupture). Consideration has previously been given to the effect of strain rate on the properties of skin. Dynamic testing often involves the study of the response of the skin subjected to an oscillatory solicitation (Boyer et al., 2009; Dawes-Higgs et al., 2004; Lamers et al., 2013) or to wave propagation through it (Liang and Boppart, 2010; Lim et al., 2011). These methodologies are useful to examine the in vivo response of skin, or to access the viscoelastic properties of the skin, but not to study the failure properties of the skin. Previous studies which do examine the failure properties of skin, have confined their experiments to low strain rates. Zhou et al. (2010) investigated the coupling of strain rates and temperature for tensile tests on pig skin, but strain rates did not exceed 0.1 s 1. Similarly, Ní Annaidh et al. (2012a) conducted tensile experiments of human skin but did not exceed a strain rate of 0.01 s 1. For intermediate strain rates, valid for automotive collisions and sharp force injury, Haut (1989), Dombi et al. (1993), Arumugam et al. (1994), or Khatam et al. (2014) performed tests on animal skins only, with Haut (1989) and Dombi et al. (1993) having also considered the orientation of
samples. Shergold et al. (2006) performed tests in the range 0.004 s 1 to 4000 s 1 but these experiments were in compression only. To the best of the authors’ knowledge, only one study has tested human skin in the intermediate range of velocities: Jacquemoud et al. (2007) used a customized tensile device based on a drop test machine and Digital Image Correlation to investigate local strain levels. Nevertheless, a clear comparison of the mechanical properties of human skin at quasi-static and intermediate velocities which considers the effect of anisotropy is still wanting in the literature. The present paper aims to provide new material data for human skin via dynamic uniaxial tensile testing with respect to the Langer lines, which can be applied to constitutive models in areas such as impact biomechanics and forensic science.
2.
Materials and method
2.1.
Sample preparation
All tests and procedures were in line with the French ethical rules and the law that allows experiments involving post mortem human subjects (PMHS) for biomedical research under the control of a medical school. The desired location and orientation of test samples was marked on the back of a PMHS with a custom made ink stamp (Fig. 1). The skin tissue was then excised in one piece from the back of the PMHS one day after death and stored at 4 1C in gauze soaked with saline solution. The skin sample included the epidermis, dermis, hypodermis and the underlying adipose tissue. Within two weeks, dogbone shape samples were cut with a custom die, following the previously marked locations. In total, 33 samples were collected. The hypodermis and underlying adipose tissue were then removed from each test sample with a scalpel. As the dermal layer appears visually distinct from the hypodermis, the removal is straight forward to perform. The mean thickness of the samples was 2.370.4 mm. They were stored in gauze soaked with saline solution in a 4 1C storage room prior to testing. Each specimen was grouped into one of three categories: parallel, perpendicular or at 451 to the Langer lines. The orientation of the samples was determined based on anatomy sketches of Langer lines reported in the literature and on the photographs of the sample locations prior to excision. Samples were clamped using custom designed anti-slip grips. They were cautiously positioned in the grips to limit axial preload and bending prestress. Table 1 provides the characteristic measurements of samples before testing.
2.2.
Tensile tests
Tensile tests were performed using a servo-hydraulic machine (Instron 8802, High Wycombe, England). The tensile
Table 1 – Mean values of the specimen characteristics with standard deviation given in brackets.
Long samples (n ¼11) dynamic testing Small samples (n ¼10) dynamic testing Small samples (n ¼11) quasi-static testing
Total length (cm)
Working length (cm)
Width (mm)
Thickness (mm)
11.8 (0.2) 4.5 (0.1) 4.6 (0.3)
7.1 (10.9) 1.2 (0.1) 1.3 (0.1)
5.2 (0.3) 3.6 (0.7) 3.5 (1.0)
2.5 (0.3) 2.0 (0.3) 2.4 (0.5)
journal of the mechanical behavior of biomedical materials 41 (2015) 241 –250
load was measured with a 1 kN load cell. The load cell and lower grip were set on a separate table, independent of the machine frame so as to avoid vibration related issues (Fig. 2). In addition, an accelerometer was fixed to the machine to monitor the level of vibrations generated. Small samples were pre-loaded at 2 N and long samples at 1 N. All samples underwent 10 cycles of 0.5 mm amplitude at 1 Hz as pre-conditioning. As illustrated in Fig. 1, one set of small samples was tested at 50 mm/min (0.06 s 1), the other set was tested at 2 m/s (167 s 1) and the long sample set at 2 m/s (53 s 1). The test speed of 2 m/s was chosen as it is comparable to the experimental impact speed of a blade on the skin during a stabbing motion (Miller and Jones, 1996; Chadwick et al., 1999; Horsfall et al., 1999). Four synchronized SA3 Photron black and white video cameras (Tokyo, Japan) recorded the experiments. They were equipped with 50 mm Zeiss lenses (Oberkochen, Germany). One pair of cameras focused on the sample gauge length while the second pair had a larger field of view which captured the grips during testing. This set up allowed for the detection of slippage of samples and undesirable vibrations during high speed testing. The resolution of the focused cameras was 256 by 768 pixel corresponding to approximately 20 pixel per mm in the region of interest. The acquisition
243
frequency was 5000 frame/s for the dynamic tests and 50 frame/s for the quasi-static tests. Before testing, the epidermal surface was prepared for the stereo-correlation. It was covered with white make-up to create a plain background. A random speckle pattern was then applied using black spray paint. In addition, two black points were painted on the sample so that the displacement could still be tracked in the event of the image correlation failing.
2.3.
Data processing
Video recordings were analyzed using VIC3D stereocorrelation software (CorrelatedSolution, South Carolina, USA). Stretch ratio (current length divided by the initial length) was computed based on Digital Image Correlation (DIC). The speckle pattern painted on the specimen allowed for the determination of the 3D geometry and 3D displacement field. If the correlation was unsuccessful, the stretch ratio was computed based on the displacement of the two painted points. The nominal stress was calculated by dividing the force by the undeformed cross sectional area of the specimen. Stretch ratios versus nominal stress curves were plotted and the following parameters were determined (Fig. 3):
Ultimate tensile stress (UTS), Stretch ratio at UTS, Stretch ratio at failure. Here, the failure is defined as the
point at which a discontinuity appears in the Digital Image Correlation data. This point corresponds to a superficial tear visible on the epidermal surface. Strain energy, the energy per unit volume which is represented by the area under the stress strain curve, E2, slope of the linear portion of the curve.
2.4.
Fig. 1 – Location of samples on the back.
Fig. 2 – Experimental set up.
Statistical analysis
Statistical analyses were performed to determine the significance of the influence of the strain rate, orientation relative to the Langer lines and symmetry with respect to the dorsal line. A Wilcoxon non-parametric test for paired samples was used with po0.05 indicating a significant difference (1-tailed test).
Fig. 3 – Typical nominal stress versus stretch ratio curve.
244
3.
journal of the mechanical behavior of biomedical materials 41 (2015) 241 –250
Results
4.
Of the 33 tests, 6 samples slipped in the grips. In these cases, testing was stopped once the slipping was detected. These samples were retested a second time until rupture and the results were included for further analysis. One sample was excluded from the results because it had become dehydrated following a delay in re-arranging the grips. Table 2 summarises the mean values for the calculated parameters.
3.1.
The profiles of the stress–strain curves presented here all adopt the characteristic J-shape of soft biological tissues subjected to ex-vivo tensile tests (Daly, 1982). This J-shaped curved is generally described as three phases as follows: (i) skin is at first very compliant because the load is borne primarily by elastic fibers, (ii) a toe-region phase, at which point collagen fibers begin to elongate and align themselves in the direction of applied load, (iii) a strain-stiffening phase, where the stiffness of the skin increases rapidly as the unfolded collagen fibers align themselves in the direction of the applied load. While interpreting the data obtained in this study it should be considered that the biomechanical parameters of skin change with respect to the tested species, the gender and the localization of the samples (Yamada, 1970; Cua et al., 1990; Malm et al., 1995; Diridollou et al., 2000). The mode of conservation, method of preconditioning, the nature of the grips and the control of humidity also affect the experimental data (Foutz et al., 1992; Edsberg et al. 1999; Marangoni et al., 1966). The age of the skin (90 years) must also be considered (Pailler-Mattei et al., 2013) as it leads to an overall reduction in extensibility in the toe-region phase (Leveque et al., 1980), and an increase of stiffening at higher strains (Alexander and Cook, 2006). Temperature is another factor which plays a role in the variability of the mechanical properties of skin: subjected to high temperatures, collagen fibers develop into a gel-like material due to the breakdown of molecular crosslinks, and skin becomes less stiff in the strain-hardening phase of tensile tests (Xu et al., 2007). Moreover, the control of tensile tests in the case of soft biological tissues is difficult (Holzapfel, 2006) and often experiments are quite far from the ideal case. First, inserting the samples in the set-up can transfer an axial load to it. Second, the composition of the sample can vary across its thickness. Lastly, the choice of fixation can also be difficult and cumbersome as it must minimize slippage as well as damage in the samples. These remarks are especially true
Influence of strain rate
The strain rate was found to have a statistically significant effect on the parameters shown in Fig. 4. All parameters were significantly different at strains rates of 0.06 s 1 and 53 s 1. When comparing the tests at 53 s 1 and 167 s 1, however, only strain energy (p ¼0.005) and E2 were significantly different (p¼ 0.0005). When comparing the tests at 0.06 s 1 and 167 s 1, UTS (p ¼0.001), strain energy (p¼ 0.001) and E2 (0.001) were significantly different. Fig. 5 displays an example of stress–strain relationships for one location on the back at different strain rates.
3.2.
Influence of orientation relative to the Langer lines
The orientation relative to the Langer lines has a significant influence on the parameters (Fig. 6). In particular for the parallel and perpendicular orientation where all parameters were significantly different.
3.3.
Discussion
Effect of symmetry
There were no significant difference of parameters between the left and right samples. This result is indicative of a line of symmetry about the spine.
Table 2 – Mean values of the calculated parameters with standard deviation given in brackets.
Strain rate
Orientation
Parallel
451
Perpendicular
Features
UTS (MPa)
Stretch ratio at UTS
Stretch ratio at failure
Strain energy (MJ/m3)
E2 (MPa)
0.06 s 1 (n¼ 10) 53 s 1 (n ¼ 10) 167 s 1 (n ¼ 11) Parallel (n ¼8) 451 (n ¼14) Perpendicular (n ¼9) 0.06 s 1 (n¼ 3) 53 s 1 (n ¼ 3) 167 s 1 (n ¼ 3) 0.06 s 1 (n¼ 4) 53 s 1 (n ¼ 5) 167 s 1 (n ¼ 5) 0.06 s 1 (n¼ 3) 53 s 1 (n ¼ 3) 167 s 1 (n ¼ 3)
15.9 24.1 25.8 28.0 22.5 15.6 20.9 32.1 30.9 16.4 23.7 26.3 10.4 16.7 19.7
1.3 1.3 1.3 1.2 1.3 1.3 1.3 1.2 1.2 1.3 1.3 1.3 1.4 1.3 1.3
1.5 1.3 1.5 1.4 1.4 1.5 1.5 1.3 1.3 1.4 1.3 1.5 1.5 1.4 1.5
4.3 4.5 8.2 7.1 5.4 4.9 6.5 5.6 9.2 3.4 4.1 8.3 3.4 4.2 7.0
76.7 104.4 169.1 160.8 121.0 70.6 108.1 152.4 221.9 83.9 105.4 166.2 35.7 54.9 121.2
(5.7) (7.1) (8.2) (5.7) (8.2) (5.2) (1.3) (1.5) (3.7) (6.4) (5.4) 10.0) (1.3) (1.5) (2.0)
(0.1) (0.1) (0.1) (0.0) (0.0) (0.1) (0.0) (0.0) (0.0) (0.0) (0.0) (0.0) (0.0) (0.0) (0.0)
(0.1) (0.1) (0.1) (0.1) (0.1) (0.1) (0.1) (0.0) (0.1) (0.1) (0.1) (0.2) (0.1) (0.0) (0.1)
(1.8) (1.3) (3.5) (2.8) (3.5) (1.8) (1.8) (1.5) (4.0) (0.9) (1.2) (4.5) (1.8) (1.5) (4.0)
(40.3) (44.7) (70.5) (53.2) (58.4) (59.5) (25.9) (22.1) (16.6) (40.6) (33.4) (67.1) (25.9) (22.1) (16.6)
journal of the mechanical behavior of biomedical materials 41 (2015) 241 –250
245
Fig. 5 – Example of stress–strain relationships for one location on the back at different strain rates.
Fig. 4 – Wilcoxon statistical analysis of the strain rate influence with n indicating po0.05 and nnpo0.01. Comparisons between 0.06 s 1 versus 53 s 1 was with n ¼10, 53 s 1 versus 167 s 1 was with n¼ 11, and 0.06 s 1 versus 167 s 1 with n ¼10. (n¼ number of pairs).
for dynamic testing where samples are exposed to higher loading and stresses. In the present study, rupture usually occurs in the center of the sample, far from the grips, which provides some confidence that sample have not been damaged excessively in the grips. Another difficult parameter to control is the loading rates. According to Lim et al. (2011), who used a modified split Hopkinson bar for testing pig skin at very high loading rates, there must be a compromise between using a short gage length for the samples to facilitate the dynamic equilibrium and longer specimen to minimize end effects. Long samples were tested so as to be consistent with previous studies conducted by the authors using the same protocol at quasi-statics strain rates (seven subjects, Ní Annaidh et al., 2012a), and dynamics (three subjects, Gallagher et al., 2012). However, following initial testing at dynamic strain rates (Gallagher et al., 2012), some questions remained unanswered as to unusual stress–strain profiles which had been observed. For this reason, the experimental protocol was modified slightly in the present case. The sudden protrusion shown in Fig. 7a in the evolution of stress with respect to time (red crosses) is an unusual result, but this same effect was observed in Gallagher et al. (2012). The use of Digital Image Correlation to measure the stretch ratio reduces the occurrence and magnitude of the protrusions in the stress profiles (shown in Fig. 7b) but further investigation was still required. In order to check if any slippage occurred, the tracking of two targets located on the moving grip was superimposed on the tracking of two targets painted on the sample. In neither case, was the displacement linear with time. This illustrated that the velocity transmitted to the sample was not equal to the input velocity (see the blue squares in Fig. 7a). For a dynamic test, the machine actuator first accelerates alone until the set speed is reached, then it makes contact with the moving grip block, which grips the sample (Fig. 8). The contact phase between the piston and the moving grip block may explain the difference in observed velocities. The velocity applied to almost all long sample decreased progressively, followed by a sudden acceleration corresponding to the protrusion on the stress strain curve. This feature is present for almost all the long samples. To the contrary, no such protrusions were present in the mechanical
246
journal of the mechanical behavior of biomedical materials 41 (2015) 241 –250
Fig. 6 – Wilcoxon statistical analysis of the orientation influence with n indicating po0.05 and nnpo0.01. For the comparisons between parallel versus 451, 451 versus perpendicular, and parallel versus perpendicular, the number of pairs were n¼ 9 in each comparison.
response of the short samples (Fig. 7d). The velocity profiles of the moving grips still do not correspond exactly to the input velocity, but velocity profiles are more constant for the short samples than for the long samples (Fig. 7). The mechanical response with an initial high loading rate decreasing with time can be of interest, but a sudden acceleration is not desirable, and this is the reason why no statistics have been included with the cited previous works (Gallagher et al., 2012). The results obtained for long samples in the present study should be analyzed with this remark in mind. These difficulties associated with dynamic experimentation highlighted here, go some way to explaining the dearth of reliable data on this topic. Great care has been taken in this study to eliminate the effects of both inter-subject variability and intra-subject variability: all samples were obtained from the same subject and samples were taken immediately adjacent to each other as shown in Fig. 1. By removing the inter-subject variability, this strategy simplifies the task of determining the effect of the strain rate and anisotropy on the failure characteristics of skin. Furthermore, the comparison of results coupled by symmetry with respect to spine does not reveal any statistical difference, supporting the integrity of the experimental data. In agreement with other studies (Haut, 1989; Dombi et al., 1993; Lim et al., 2011), the current study clearly reports an increase of the ultimate tensile stress with increased loading rate. Only Arumugam et al. (1994) were unsure of the effect of the strain rate but this may be explained by the narrow and low range of strain rates tested in their study (from 5% min 1 to 1000% min 1). The parameter E2 (slope of the linear portion of the curve) was also found to be very rate-dependent in the present study, synonymous with a higher stress stiffening effect in dynamics as also reported by Nie et al. (2011) for porcine muscle. In addition to the effect of the strain rate alone, the present study investigates the influence of the orientation of the samples with respect to Langer lines. Table 2 shows that samples perpendicular to Langer lines always break at lower ultimate stress than parallel samples, regardless of the strain rate. Other studies, conducted at both quasi-static and dynamic strain rates, confirm this trend (Stark, 1977; Dombi et al., 1993; Gąsior-Głogowska et al., 2013). Examining more closely the interaction between strain rate and sample orientation, it can be seen that increasing the average input strain rate from 0.06 s 1 to 167 s 1 raised the ultimate tensile stress by 50%, 60% and 90% for parallel, 451 and perpendicular samples respectively. Whereas the same increases in strain rates led to a slight decrease in stretch at rupture from 1.570.1 to 1.370.1 in parallel samples, but for 451 and perpendicular samples no difference was detected. If parallel samples rupture at a lower stretch value than perpendicular samples at high strain rates, it may explain why, as blunt trauma studies have shown in the past, that skin lacerates more often along Langer lines (Leung et al., 1977). The increase of the stress-stiffening effect with the loading rate, evident through the increase in E2, is also intensified with the orientation of the samples. Indeed, E2 is 200% greater for parallel and 451 samples at high strain rate than low strain rate, 200% greater for 451 samples, and 340% times greater for perpendicular samples.
journal of the mechanical behavior of biomedical materials 41 (2015) 241 –250
247
Fig. 7 – (a) Long sample: nominal stress with respect to time (red crosses), input velocity with respect to time (blue squares); (b) long sample: nominal stress with respect to stretch ratio (red crosses), velocity transmitted to the sample with respect to stretch ratio (blue squares); (c) small sample: nominal stress with respect to time (red crosses), input velocity with respect to time (blue squares); (d) small sample: nominal stress with respect to stretch ratio (red crosses), velocity transmitted to the sample with respect to stretch ratio (blue squares).
Fig. 8 – Illustration of the principle of high loading rate operation.
To understand why different stresses are obtained at the same value of stretch ratio for quasi-static and dynamic tests, it is necessary to consider the microscopic structure of skin. The present tests include two main layers of the skin (epidermis and dermis), but the dermis is usually believed to play the most important role in the mechanical strength of the organ. The dermis is a conjunctive tissue, whose extracellular matrix is made of elastic fibers and collagen fibers, both embedded in a ground substance (mostly water and glycoproteins). The collagen fibers become denser and thicker with increasing depth (diameter 10–40 mm), and consist of a gathering of collagen fibrils (diameter 60–100 nm) (Silver et al., 2001). The cohesion of the structure is ensured by chemical bonds, which define the degree of crosslinking within the collagen network, while other chemical bonds bind the collagen network and the ground substance. It seems that the role of the ground substance is less significant than the degree of collagen crosslinking, which determines the strength of the skin at
248
journal of the mechanical behavior of biomedical materials 41 (2015) 241 –250
quasi-static strain rates. The more the network is crosslinked, the more numerous the chemical bonds, the higher the strength is. In dynamic experiments, the effect of the interaction of the collagen fibers with the ground substance is added, as well as the relative sliding of fibers past each other. High friction and viscoelastic forces are created with respect to the strain rate. A simple indicator of these forces can be provided here by the ratio between the mechanical parameters at high and low speed (Dombi et al., 1993). The stretch ratio at failure and the stretch ratio at UTS do not appear to differ significantly in our study. However, this result depends largely on the definition of rupture. Rupture has been defined here as superficial tear on the epidermal surface as detected by the DIC system. In the example given in Fig. 9, a tear is evident in the third image. Using our definition, this is the moment failure occurs. It can be seen however, that beyond this point, a single filament still connects the two pieces of skin at the end of the test. Clearly, an alternative definition of the failure would have resulted in a more significant difference between the stretch ratios at UTS and stretch ratio at failure. The influence of the strain rates on the stretch ratio at ultimate stress or at failure is not very definitive in the study.
The stretch ratio at ultimate tensile stress does not seem to be affected by the loading rates as Vogel (1972) or Haut (1989) has already reported. This suggests that stretch ratio at failure may be a suitable parameter to be used as a local criterion to predict skin failure (tearing). This, however, is in contrast to the findings of Lim et al. (2011) who noted a strong reduction of the strain between quasi-static and dynamic strain rates. The observed reduction of the stretch ratio at rupture for parallel samples already mentioned here is not definitive given the large standard deviations. Moreover, the values of stretch ratio in our study are very low with extensibility of only 30%. This fact, coupled with the low sensitivity of strain rates to stretch ratio at rupture can probably be attributed to the donor age. There is a destruction of elastic fibers with age (Silver et al., 2001) and the degradation of the elastic fibers may explain the low stretch values found in the current study. The collagen network becomes less dense after 30 years of age (Vitellaro-Zuccarello et al., 1994), which would contribute to the diminution of the strength of the skin. The composition of the ground substance is also more aqueous with age (Agache, 2000). It is assumed here that the nature of the ground substance for an elderly person results in lower viscous and
Fig. 9 – Pictures of skin surface at different steps after UTS for a dynamic test.
journal of the mechanical behavior of biomedical materials 41 (2015) 241 –250
friction effects in dynamic loading as discussed previously. However, a clarification of the role of the ground substance at high loading rates is still needed. The trend of the stress–strain relationships obtained in this study shows progressive decreasing stress with strain beyond the ultimate tensile stress. Fig. 9 shows images of skin for this phase. Damage to the material occurs, and microfractures or microtearing are sometimes visible on the surface, but they usually occur internally. Arumugam et al. (1994) suggest that defibrillation rupture is always present, but it usually presents with very thin filaments for quasistatic speeds and thicker filaments for dynamic speeds. Depending on this defibrillation process, the stress does not vanish abruptly, for either quasi-static or dynamic tests. Comparison with literature is always a delicate subject in the context of soft biological tissues. The studies are not numerous and the test conditions vary from study to study. At quasi-static strain rates, the UTS of the current study is found to be in the same range as Ní Annaidh et al. (2012a); Gallagher et al. (2012), Dunn and Silver (1983) and Ankerson et al. (1999). Our elastic modulus of 76.7740.3 MPa is however higher than Gąsior-Głogowska et al. (2013) for the thigh with 2–12 MPa or 18.8 MPa for the abdominal skin found by Silver et al. (2001). Elastic moduli from in vivo studies are evaluated at very low strain range and cannot be compared here (Zahouani et al., 2009; Pailler-Mattei et al., 2008). Very small failure strains are obtained in our quasi-static tests in contrast with the 207% maximal strain found by Jansen and Rottier (1958). At dynamic strain rates, Jacquemoud et al. (2007) performed tests with forehead and arm skin. Their UTS values were approximately 50% less than presented here (i.e. 25.878.2 MPa) and the elastic moduli were in the lower limit of our estimation (i.e. 169.1770.5 MPa). Other studies at dynamic strain rates involved animal skins with UTS values ranging from 0.1 MPa to 30 MPa (Lim et al., 2011; Haut, 1989). However, the test conditions varied greatly (difference of species, age, location, orientation, loading rate, test equipment, etc.) and no direct comparison can be made. There are a number of limitations associated with this study. First, the analysis is based on an assumption about the direction of Langer lines. Ideally, as already noted in Brown (1973), Langer lines should be identified for each individual patient to be really pertinent, this process, however, is destructive and could not be performed here. A clear discussion of the effect of speed and orientation of the samples at very low strains is not really possible due to the choice of the load sensor, optimized for the force levels at rupture. Moreover, the initial acceleration applied to the sample at the beginning of the dynamic tensile test is not well controlled, leading to different strain rates being applied in this elastic region. The separation of the effect of the orientation and the effect of the strain rates in this loading region is therefore difficult. Some studies report the disappearance of the toeregion with high loading rate. In our study, we observe a shortening of this toe-region, a shift towards lower strain, but a toe region was still present. It is possible that the high frequency of acquisition for strain and stress data has facilitated the capture of this feature where it was not previously possible. Finally, the sample size is low, but by
249
eliminating the effects of inter and intra subject variability the data obtained can be used effectively.
5.
Conclusion
This study provides data on the combined effect of strain rate and anisotropy of human skin (90 year old). Both anisotropy and strain rate strongly affect the ultimate tensile stress (UTS), and the strain-hardening effect (E2). The stretch ratio at ultimate tensile stress was found to be constant with strain rate, but statistically dependent on the orientation of the sample. Now that the control of dynamic strain rate has been improved, further tests are planned to clearly distinguish between the role of the collagen and ground substance within the dermis under dynamic loading. A better understanding of the different deformation mechanisms of parallel and perpendicular samples at high loading rates will be the starting point to propose analytical models in future.
r e f e r e nc e s
Agache, P., 2000. Physiologie de la peau et explorations fonctionnelles cutane´es. Edition Me´dicales internationales, Cachan, France. Ankerson, J., Birkbeck, A.E., Thomson, R.D., Vanezls, P., 1999. Puncture resistance and tensile strength of skin stimulants. J. Eng. Med. 213, 493–501. Alexander, H., Cook, T., 2006. Variations with age in the mechanical properties of human skin in vivo. J. Tissue Viability 16 (3), 6–11. Arumugam, V., Naresh, M., Sanjeevi, R., 1994. Effect of strain rate on the fracture behaviour of skin. J. Biosci. 19 (3), 307–313. Boyer, G., Laquie`ze, L., Le Bot, A., Laquie`ze, S., Zahouani, H., 2009. Dynamic indentation on human skin in vitro: ageing effects. Skin Res. Technol. 15, 55–67. Brown, I., 1973. A scanning electron microscope study of the effects of uniaxial tension on human skin. Br. J. Dermatol. 89, 383–393. Chadwick, E.K.J., Nicol, A.C., Lane, J.V., Gray, T.G.F., 1999. Biomechanics of knife stab attacks. Forensic Sci. Int. 105 (1), 35–44. Cua, A., Wilhelm, K.-P., Maibach, H., 1990. Elastic properties of human skin: relation to age, sex, and anatomical region. Arch. Dermatol. Res. 282, 283–288. Daly, C., 1982. Biomechanical properties of dermis. J. Invest. Dermatol. 79 (1), 17–20. Dawes-Higgs, E., Swain, M., Higgs, R., Appleyard, R., Kossard, S., 2004. Accuracy and reliability of a dynamic biomechanical skin measurement probe for the analysis of stiffness and viscoelasticity. Physiol. Meas. 25, 97–105. Diridollou, S., Black, D., Lagarde, J., Gall, Y., Berson, M., Vabre, V., Patat, F., Vaillant, L., 2000. Sex- and site-dependent variations in the thickness and mechanical properties of human skin in vivo. Int. J. Cosmet. Sci. 22, 421–435. Dombi, G., Haut, R., Sullivan, S., 1993. Correlation of high-speed tensile strength with collagen content in control and lathyritic rat skin. J. Surg. Res. 54, 21–28. Dunn, M.G., Silver, F.H., 1983. Viscoelastic behavior of human connective tissues: relative contribution of viscous and elastic components. Connect. Tissue Res. 12, 59–70.
250
journal of the mechanical behavior of biomedical materials 41 (2015) 241 –250
Edsberg, L., Mates, R., Baier, R., Lauren, M., 1999. Mechanical characteristics of human skin subjected to static versus cyclic normal pressures. J. Rehabil. Res. Dev. 36 (2), 1–12. Foutz, T., Stone, E., Abrams, C.J., 1992. Effects of freezing on mechanical properties of rat skin. Am. J. Vet. Res. 53 (5), 788–792. Gallagher, A., Nı´ Anniadh, A., Bruye`re, K., Otte´nio, M., Xie, H., Gilchrist, M., 2012. Dynamic tensile properties of human skin. In: Proceedings of the International Research Council on the Biomechanics of Injury Conference, vol. 40, pp. 494–502. Ga˛sior-Głogowska, M., Komorowska, M., Hanuza, J., Ma˛czka, M., Zaja˛c, A., Ptak, M., Be˛dzin´ski, R., Kobielarz, M., Maksymowicz, K., Kuropka, P., Szotek, S., 2013. FT-Raman spectroscopy study of human skin subjected to uniaxial stress. J. Mech. Behav. Biomed. Mater. 18, 240–252. Haut, R., 1989. The effect of orientation and location on the strength of dorsal rat skin in high and low speed tensile failure experiments. J. Biomech. Eng. 111, 136–140. Holzapfel, G.A., 2006. Determination of material models for arterial walls from uniaxial extension tests and histological structure. J. Theor. Biol. 238 (2), 290–302. Horsfall, I., Prosser, P.D., Watson, C.H., Champion, S.M., 1999. An assessment of human performance in stabbing. Forensic Sci. Int. 102 (2-3), 79–89. Jacquemoud, C., Bruye`re-Garnier, K., Coret, M., 2007. Methodology to determine failure characteristics of planar soft tissues using a dynamic tensile test. J. Biomech. 40, 468–475. Jansen, L., Rottier, P., 1958. Some mechanical properties of human abdominal skin measured on excised strips. Dermatologica 117, 65–83. Khatam, H., Liu, Q., Ravi-Chandar, K., 2014. Dynamic tensile characterization of pig skin. Acta Mech. Sin. 30 (2), 125–132. Lamers, E., van Kempen, T., Baaijens, F., Peters, G., Oomens, C., 2013. Large amplitude oscillatory shear properties of human skin. J. Mech. Behav. Biomed. Mater. 28, 462–470. Langer, K., 1978. On the anatomy and physiology of the skin—I. The cleavability of cutis. Br. J. Plast. Surg. 31, 3–8. Leung, Y.C., Lopat, E., Fayon, A., Banzet, P., Tarriere, C., 1977. Lacerative properties of the human skin during impact. In: Proceedings of the Third International Research Council on the Biomechanics of Injury Conference, Berlin, Germany, pp. 399–411. Leveque, J.L., De Rigal, J., Agache, P.G., Monneur, C., 1980. Influence of ageing on the in vivo extensibility of human skin at a low stress. Arch. Dermatol. Res. 269 (2), 127–135. Liang, X., Boppart, S., 2010. Biomechanical properties of in vivo human skin from dynamic optical coherence elastography. IEEE Trans. Biomed. Eng. 57 (4), 953–959. Lim, J., Hong, J., Chen, W., Weerasooriya, T., 2011. Mechanical response of pig skin under dynamic tensile loading. Int. J. Impact Eng. 38, 130–135. Malm, M., Samman, M., Serup, J., 1995. In vivo skin elasticity of 22 anatomical sites. Skin Res. Technol. 1, 61–67.
Marangoni, R., Glaser, A., Must, J., Brody, G., Beckwith, T., Walker, G., White, W., 1966. Effect of storage and handling techniques on skin tissue properties. Ann. N.Y. Acad. Sci. 136, 441–453. Miller, S.A., Jones, M.D., 1996. Kinematics of four methods of stabbing: a preliminary study. Forensic Sci. Int. 82 (2), 183–190. Nı´ Annaidh, A., Bruye`re, K., Destrade, M., Gilchrist, M., Otte´nio, M., 2012a. Characterization of the anisotropic mechanical properties of excised human skin. J. Mech. Behav. Biomed. Mater. 5, 139–148. Nı´ Annaidh, A., Bruye`re, K., Destrade, M., Gilchrist, M., Maurini, C., Otte´nio, M., Saccomandi, G., 2012b. Automated estimation of collagen fibre dispersion in the dermis and its contribution to the anisotropic behaviour of skin. Ann. Biomed. Eng. 40, 1666–1678. Nie, X., Cheng, J., Chen, W., Weerasooriya, T., 2011. Dynamic tensile response of porcine muscle. J. Appl. Mech. 78 (2), 1–5. Pailler-Mattei, C., Bec, S., Zahouani, H., 2008. In vivo measurements of the elastic mechanical properties of human skin by indentation tests. Med. Eng. Phys. 30 (5), 599–606. Pailler-Mattei, C., Debret, R., Vargiolu, R., Sommer, P., Zahouani, H., 2013. In vivo skin biophysical behaviour and surface topography as a function of ageing. J. Mech. Behav. Biomed. Mater. 28, 474–483. Ridge, M., Wright, M., 1966. The directional effects of skin. J. Invest. Dermatol. 46 (4), 341–346. Shergold, O., Fleck, N., Radford, D., 2006. The uniaxial stress versus strain response of pig skin and silicon rubber at low and high strain rates. Int. J. Impact Eng. 32, 1384–1402. Silver, F., Freeman, J., DeVore, D., 2001. Viscoelastic properties of human skin and processed dermis. Skin Res. Technol. 7, 18–23. Stark, H.L., 1977. Directional variations in the extensibility of human skin. Br. J. Plast. Surg. 30 (2), 105–114. Vitellaro-Zuccarello, L., Cappelletti, S., Dal Pozzo Rossi, V., SariGorla, M., 1994. Stereological analysis of collagen and elastic fibers in normal human dermis: variability with age, sex and body region. Anat. Rec. 238, 153–162. Vogel, H.G., 1972. Influence of age, treatment with corticosteroids and strain rate on mechanical properties of rat skin. Biochimica et Biophysica Acta 286, 79–83. Xu, F., Wen, T., Seffen, K., Lu, T., 2007. Characterization of thermomechanical behaviour of skin tissue—I. Tensile and compressive behaviours. In: Proceedings of the World Congress on Engineering, London, UK, vol. 2, pp. 1446–1450. Yamada, H., 1970. Strength of Biological Materials. Edition F. Gaynor Evans, MI, USA. Zahouani, H., Pailler-Mattei, C., Sohm, B., Vargiolu, R., Cenizo, V., Debret, R., 2009. Characterization of the mechanical properties of a dermal equivalent compared with human skin in vivo by indentation and static friction tests. Skin Res. Technol. 15 (1), 68–76. Zhou, B., Xu, F., Chen, C., Lu, T., 2010. Strain rate sensitivity of skin tissue under thermomechanical loading. Philos. Trans. R. Soc. London, Ser. A 368, 679–690.
Available online at www.sciencedirect.com
www.elsevier.com/locate/jmbbm
Research Paper
Strain rate and anisotropy effects on the tensile failure characteristics of human skin Me´lanie Ottenioa,b,c,n, Doris Trana,b,c, Aisling Nı´ Annaidhd, Michael D. Gilchristd, Karine Bruye`rea,b,c a
Université de Lyon, F-69622, Lyon, France Université Claude Bernard Lyon 1, Villeurbanne, France c IFSTTAR, UMR_T9406, LBMC Laboratoire de Biomécanique et Mécanique des Chocs, F69675, Bron, France d School of Mechanical & Materials Engineering, University College Dublin, Belfield, Dublin 4, Ireland b
art i cle i nfo
ab st rac t
Article history:
The anisotropic failure characteristics of human skin are relatively unknown at strain rates
Received 6 May 2014
typical in impact biomechanics. This study reports the results of an experimental protocol
Received in revised form
to quantify the effect of dynamic strain rates and the effect of sample orientation with
6 October 2014
respect to the Langer lines. Uniaxial tensile tests were carried out at three strain rates
Accepted 8 October 2014
(0.06 s 1, 53 s 1, and 167 s 1) on 33 test samples excised from the back of a fresh cadaver.
Available online 16 October 2014
The mean ultimate tensile stress, mean elastic modulus and mean strain energy increased
Keywords:
with increasing strain rates. While the stretch ratio at ultimate tensile stress was not
Human skin
affected by the strain rate, it was influenced by the orientation of the samples (parallel and
Ex-vivo tensile test
perpendicular to the Langer lines. The orientation of the sample also had a strong
Dynamic strain-rate
influence on the ultimate tensile stress, with a mean value of 28.075.7 MPa for parallel
Anisotropy
samples, and 15.675.2 MPa for perpendicular samples, and on the elastic modulus, with
Langer lines
corresponding mean values of 160.8 MPa753.2 MPa and 70.6 MPa759.5 MPa. The study also pointed out the difficulties in controlling the effective applied strain rate in dynamic characterization of soft tissue and the resulting abnormal stress–strain relationships. Finally, data collected in this study can be used to develop constitutive models where high loading rates are of primary interest. & 2014 Elsevier Ltd. All rights reserved.
1.
Introduction
Skin is the largest organ of the human body with an area of about 1.8 m2 and an average thickness of 2 mm. In medical terms, skin can be referred to as an integument, derived from
the latin tegument meaning cover. Indeed, skin covers the whole human body and one of its primary functions is to protect from foreign aggressions. Human skin is subjected to a variety of loading conditions during our daily motions, where it is stretched, sheared and pinched. Skin however,
n Corresponding author at: Laboratoire de Biomécanique et Mécanique des Chocs UMR_T 9406 (IFSTTAR-UCBL), Site Ifsttar Lyon-Bron, 25 Avenue François Mitterrand, Cité des Mobilités, 69675 Bron, France. Tel.: þ33 4 72 14 23 85. E-mail address: [email protected] (M. Ottenio).
http://dx.doi.org/10.1016/j.jmbbm.2014.10.006 1751-6161/& 2014 Elsevier Ltd. All rights reserved.
242
journal of the mechanical behavior of biomedical materials 41 (2015) 241 –250
tends to return to its initial state due to its elastic properties. Moreover, its high mechanical strength ensures it is not easily torn, and therefore, our skin protects us from tetanus, open hemorrhages, or from direct damage to underlying organs. Skin also fails due to trauma such as planned surgical incisions, gunshot wounds and penetrative wounds from stabbing attacks or can even originate internally when a bone pierces the skin. In 1861, Langer was interested in the geometry changes of open wounds. By puncturing a cadaver’s skin with a circular device, he noted that the wound transformed into an elliptic form. By joining the major axes of these ellipses, he drew a pattern of tension lines on the body: these lines are now known as the Langer lines (Langer, 1978). These early experiments illustrated the anisotropic nature of skin, but did not explain fully the mechanism involved. Indeed, the correlation between the Langer lines, the macroscopic mechanical properties of human skin and its structural basis is still under investigation (Ridge and Wright, 1966; Ní Annaidh et al., 2012a; Gąsior-Głogowska et al., 2013). Recently, Ní Annaidh et al. (2012b) have shown through the use of quantitative structural data, that the Langer lines have a structural basis. A point which had previously been alluded to but not quantitatively assessed. The current study takes place in the general context of trauma biomechanics and at strain rates relevant in automotive collisions and in sharp force injury. The collection of experimental data on human skin is an essential step in the development of a numerical model capable of predicting the rupture of skin in sharp force injury. In order to obtain experimental data relevant for this application, it is necessary to understand how skin behaves when it is dynamically loaded above its physiological range (up to rupture). Consideration has previously been given to the effect of strain rate on the properties of skin. Dynamic testing often involves the study of the response of the skin subjected to an oscillatory solicitation (Boyer et al., 2009; Dawes-Higgs et al., 2004; Lamers et al., 2013) or to wave propagation through it (Liang and Boppart, 2010; Lim et al., 2011). These methodologies are useful to examine the in vivo response of skin, or to access the viscoelastic properties of the skin, but not to study the failure properties of the skin. Previous studies which do examine the failure properties of skin, have confined their experiments to low strain rates. Zhou et al. (2010) investigated the coupling of strain rates and temperature for tensile tests on pig skin, but strain rates did not exceed 0.1 s 1. Similarly, Ní Annaidh et al. (2012a) conducted tensile experiments of human skin but did not exceed a strain rate of 0.01 s 1. For intermediate strain rates, valid for automotive collisions and sharp force injury, Haut (1989), Dombi et al. (1993), Arumugam et al. (1994), or Khatam et al. (2014) performed tests on animal skins only, with Haut (1989) and Dombi et al. (1993) having also considered the orientation of
samples. Shergold et al. (2006) performed tests in the range 0.004 s 1 to 4000 s 1 but these experiments were in compression only. To the best of the authors’ knowledge, only one study has tested human skin in the intermediate range of velocities: Jacquemoud et al. (2007) used a customized tensile device based on a drop test machine and Digital Image Correlation to investigate local strain levels. Nevertheless, a clear comparison of the mechanical properties of human skin at quasi-static and intermediate velocities which considers the effect of anisotropy is still wanting in the literature. The present paper aims to provide new material data for human skin via dynamic uniaxial tensile testing with respect to the Langer lines, which can be applied to constitutive models in areas such as impact biomechanics and forensic science.
2.
Materials and method
2.1.
Sample preparation
All tests and procedures were in line with the French ethical rules and the law that allows experiments involving post mortem human subjects (PMHS) for biomedical research under the control of a medical school. The desired location and orientation of test samples was marked on the back of a PMHS with a custom made ink stamp (Fig. 1). The skin tissue was then excised in one piece from the back of the PMHS one day after death and stored at 4 1C in gauze soaked with saline solution. The skin sample included the epidermis, dermis, hypodermis and the underlying adipose tissue. Within two weeks, dogbone shape samples were cut with a custom die, following the previously marked locations. In total, 33 samples were collected. The hypodermis and underlying adipose tissue were then removed from each test sample with a scalpel. As the dermal layer appears visually distinct from the hypodermis, the removal is straight forward to perform. The mean thickness of the samples was 2.370.4 mm. They were stored in gauze soaked with saline solution in a 4 1C storage room prior to testing. Each specimen was grouped into one of three categories: parallel, perpendicular or at 451 to the Langer lines. The orientation of the samples was determined based on anatomy sketches of Langer lines reported in the literature and on the photographs of the sample locations prior to excision. Samples were clamped using custom designed anti-slip grips. They were cautiously positioned in the grips to limit axial preload and bending prestress. Table 1 provides the characteristic measurements of samples before testing.
2.2.
Tensile tests
Tensile tests were performed using a servo-hydraulic machine (Instron 8802, High Wycombe, England). The tensile
Table 1 – Mean values of the specimen characteristics with standard deviation given in brackets.
Long samples (n ¼11) dynamic testing Small samples (n ¼10) dynamic testing Small samples (n ¼11) quasi-static testing
Total length (cm)
Working length (cm)
Width (mm)
Thickness (mm)
11.8 (0.2) 4.5 (0.1) 4.6 (0.3)
7.1 (10.9) 1.2 (0.1) 1.3 (0.1)
5.2 (0.3) 3.6 (0.7) 3.5 (1.0)
2.5 (0.3) 2.0 (0.3) 2.4 (0.5)
journal of the mechanical behavior of biomedical materials 41 (2015) 241 –250
load was measured with a 1 kN load cell. The load cell and lower grip were set on a separate table, independent of the machine frame so as to avoid vibration related issues (Fig. 2). In addition, an accelerometer was fixed to the machine to monitor the level of vibrations generated. Small samples were pre-loaded at 2 N and long samples at 1 N. All samples underwent 10 cycles of 0.5 mm amplitude at 1 Hz as pre-conditioning. As illustrated in Fig. 1, one set of small samples was tested at 50 mm/min (0.06 s 1), the other set was tested at 2 m/s (167 s 1) and the long sample set at 2 m/s (53 s 1). The test speed of 2 m/s was chosen as it is comparable to the experimental impact speed of a blade on the skin during a stabbing motion (Miller and Jones, 1996; Chadwick et al., 1999; Horsfall et al., 1999). Four synchronized SA3 Photron black and white video cameras (Tokyo, Japan) recorded the experiments. They were equipped with 50 mm Zeiss lenses (Oberkochen, Germany). One pair of cameras focused on the sample gauge length while the second pair had a larger field of view which captured the grips during testing. This set up allowed for the detection of slippage of samples and undesirable vibrations during high speed testing. The resolution of the focused cameras was 256 by 768 pixel corresponding to approximately 20 pixel per mm in the region of interest. The acquisition
243
frequency was 5000 frame/s for the dynamic tests and 50 frame/s for the quasi-static tests. Before testing, the epidermal surface was prepared for the stereo-correlation. It was covered with white make-up to create a plain background. A random speckle pattern was then applied using black spray paint. In addition, two black points were painted on the sample so that the displacement could still be tracked in the event of the image correlation failing.
2.3.
Data processing
Video recordings were analyzed using VIC3D stereocorrelation software (CorrelatedSolution, South Carolina, USA). Stretch ratio (current length divided by the initial length) was computed based on Digital Image Correlation (DIC). The speckle pattern painted on the specimen allowed for the determination of the 3D geometry and 3D displacement field. If the correlation was unsuccessful, the stretch ratio was computed based on the displacement of the two painted points. The nominal stress was calculated by dividing the force by the undeformed cross sectional area of the specimen. Stretch ratios versus nominal stress curves were plotted and the following parameters were determined (Fig. 3):
Ultimate tensile stress (UTS), Stretch ratio at UTS, Stretch ratio at failure. Here, the failure is defined as the
point at which a discontinuity appears in the Digital Image Correlation data. This point corresponds to a superficial tear visible on the epidermal surface. Strain energy, the energy per unit volume which is represented by the area under the stress strain curve, E2, slope of the linear portion of the curve.
2.4.
Fig. 1 – Location of samples on the back.
Fig. 2 – Experimental set up.
Statistical analysis
Statistical analyses were performed to determine the significance of the influence of the strain rate, orientation relative to the Langer lines and symmetry with respect to the dorsal line. A Wilcoxon non-parametric test for paired samples was used with po0.05 indicating a significant difference (1-tailed test).
Fig. 3 – Typical nominal stress versus stretch ratio curve.
244
3.
journal of the mechanical behavior of biomedical materials 41 (2015) 241 –250
Results
4.
Of the 33 tests, 6 samples slipped in the grips. In these cases, testing was stopped once the slipping was detected. These samples were retested a second time until rupture and the results were included for further analysis. One sample was excluded from the results because it had become dehydrated following a delay in re-arranging the grips. Table 2 summarises the mean values for the calculated parameters.
3.1.
The profiles of the stress–strain curves presented here all adopt the characteristic J-shape of soft biological tissues subjected to ex-vivo tensile tests (Daly, 1982). This J-shaped curved is generally described as three phases as follows: (i) skin is at first very compliant because the load is borne primarily by elastic fibers, (ii) a toe-region phase, at which point collagen fibers begin to elongate and align themselves in the direction of applied load, (iii) a strain-stiffening phase, where the stiffness of the skin increases rapidly as the unfolded collagen fibers align themselves in the direction of the applied load. While interpreting the data obtained in this study it should be considered that the biomechanical parameters of skin change with respect to the tested species, the gender and the localization of the samples (Yamada, 1970; Cua et al., 1990; Malm et al., 1995; Diridollou et al., 2000). The mode of conservation, method of preconditioning, the nature of the grips and the control of humidity also affect the experimental data (Foutz et al., 1992; Edsberg et al. 1999; Marangoni et al., 1966). The age of the skin (90 years) must also be considered (Pailler-Mattei et al., 2013) as it leads to an overall reduction in extensibility in the toe-region phase (Leveque et al., 1980), and an increase of stiffening at higher strains (Alexander and Cook, 2006). Temperature is another factor which plays a role in the variability of the mechanical properties of skin: subjected to high temperatures, collagen fibers develop into a gel-like material due to the breakdown of molecular crosslinks, and skin becomes less stiff in the strain-hardening phase of tensile tests (Xu et al., 2007). Moreover, the control of tensile tests in the case of soft biological tissues is difficult (Holzapfel, 2006) and often experiments are quite far from the ideal case. First, inserting the samples in the set-up can transfer an axial load to it. Second, the composition of the sample can vary across its thickness. Lastly, the choice of fixation can also be difficult and cumbersome as it must minimize slippage as well as damage in the samples. These remarks are especially true
Influence of strain rate
The strain rate was found to have a statistically significant effect on the parameters shown in Fig. 4. All parameters were significantly different at strains rates of 0.06 s 1 and 53 s 1. When comparing the tests at 53 s 1 and 167 s 1, however, only strain energy (p ¼0.005) and E2 were significantly different (p¼ 0.0005). When comparing the tests at 0.06 s 1 and 167 s 1, UTS (p ¼0.001), strain energy (p¼ 0.001) and E2 (0.001) were significantly different. Fig. 5 displays an example of stress–strain relationships for one location on the back at different strain rates.
3.2.
Influence of orientation relative to the Langer lines
The orientation relative to the Langer lines has a significant influence on the parameters (Fig. 6). In particular for the parallel and perpendicular orientation where all parameters were significantly different.
3.3.
Discussion
Effect of symmetry
There were no significant difference of parameters between the left and right samples. This result is indicative of a line of symmetry about the spine.
Table 2 – Mean values of the calculated parameters with standard deviation given in brackets.
Strain rate
Orientation
Parallel
451
Perpendicular
Features
UTS (MPa)
Stretch ratio at UTS
Stretch ratio at failure
Strain energy (MJ/m3)
E2 (MPa)
0.06 s 1 (n¼ 10) 53 s 1 (n ¼ 10) 167 s 1 (n ¼ 11) Parallel (n ¼8) 451 (n ¼14) Perpendicular (n ¼9) 0.06 s 1 (n¼ 3) 53 s 1 (n ¼ 3) 167 s 1 (n ¼ 3) 0.06 s 1 (n¼ 4) 53 s 1 (n ¼ 5) 167 s 1 (n ¼ 5) 0.06 s 1 (n¼ 3) 53 s 1 (n ¼ 3) 167 s 1 (n ¼ 3)
15.9 24.1 25.8 28.0 22.5 15.6 20.9 32.1 30.9 16.4 23.7 26.3 10.4 16.7 19.7
1.3 1.3 1.3 1.2 1.3 1.3 1.3 1.2 1.2 1.3 1.3 1.3 1.4 1.3 1.3
1.5 1.3 1.5 1.4 1.4 1.5 1.5 1.3 1.3 1.4 1.3 1.5 1.5 1.4 1.5
4.3 4.5 8.2 7.1 5.4 4.9 6.5 5.6 9.2 3.4 4.1 8.3 3.4 4.2 7.0
76.7 104.4 169.1 160.8 121.0 70.6 108.1 152.4 221.9 83.9 105.4 166.2 35.7 54.9 121.2
(5.7) (7.1) (8.2) (5.7) (8.2) (5.2) (1.3) (1.5) (3.7) (6.4) (5.4) 10.0) (1.3) (1.5) (2.0)
(0.1) (0.1) (0.1) (0.0) (0.0) (0.1) (0.0) (0.0) (0.0) (0.0) (0.0) (0.0) (0.0) (0.0) (0.0)
(0.1) (0.1) (0.1) (0.1) (0.1) (0.1) (0.1) (0.0) (0.1) (0.1) (0.1) (0.2) (0.1) (0.0) (0.1)
(1.8) (1.3) (3.5) (2.8) (3.5) (1.8) (1.8) (1.5) (4.0) (0.9) (1.2) (4.5) (1.8) (1.5) (4.0)
(40.3) (44.7) (70.5) (53.2) (58.4) (59.5) (25.9) (22.1) (16.6) (40.6) (33.4) (67.1) (25.9) (22.1) (16.6)
journal of the mechanical behavior of biomedical materials 41 (2015) 241 –250
245
Fig. 5 – Example of stress–strain relationships for one location on the back at different strain rates.
Fig. 4 – Wilcoxon statistical analysis of the strain rate influence with n indicating po0.05 and nnpo0.01. Comparisons between 0.06 s 1 versus 53 s 1 was with n ¼10, 53 s 1 versus 167 s 1 was with n¼ 11, and 0.06 s 1 versus 167 s 1 with n ¼10. (n¼ number of pairs).
for dynamic testing where samples are exposed to higher loading and stresses. In the present study, rupture usually occurs in the center of the sample, far from the grips, which provides some confidence that sample have not been damaged excessively in the grips. Another difficult parameter to control is the loading rates. According to Lim et al. (2011), who used a modified split Hopkinson bar for testing pig skin at very high loading rates, there must be a compromise between using a short gage length for the samples to facilitate the dynamic equilibrium and longer specimen to minimize end effects. Long samples were tested so as to be consistent with previous studies conducted by the authors using the same protocol at quasi-statics strain rates (seven subjects, Ní Annaidh et al., 2012a), and dynamics (three subjects, Gallagher et al., 2012). However, following initial testing at dynamic strain rates (Gallagher et al., 2012), some questions remained unanswered as to unusual stress–strain profiles which had been observed. For this reason, the experimental protocol was modified slightly in the present case. The sudden protrusion shown in Fig. 7a in the evolution of stress with respect to time (red crosses) is an unusual result, but this same effect was observed in Gallagher et al. (2012). The use of Digital Image Correlation to measure the stretch ratio reduces the occurrence and magnitude of the protrusions in the stress profiles (shown in Fig. 7b) but further investigation was still required. In order to check if any slippage occurred, the tracking of two targets located on the moving grip was superimposed on the tracking of two targets painted on the sample. In neither case, was the displacement linear with time. This illustrated that the velocity transmitted to the sample was not equal to the input velocity (see the blue squares in Fig. 7a). For a dynamic test, the machine actuator first accelerates alone until the set speed is reached, then it makes contact with the moving grip block, which grips the sample (Fig. 8). The contact phase between the piston and the moving grip block may explain the difference in observed velocities. The velocity applied to almost all long sample decreased progressively, followed by a sudden acceleration corresponding to the protrusion on the stress strain curve. This feature is present for almost all the long samples. To the contrary, no such protrusions were present in the mechanical
246
journal of the mechanical behavior of biomedical materials 41 (2015) 241 –250
Fig. 6 – Wilcoxon statistical analysis of the orientation influence with n indicating po0.05 and nnpo0.01. For the comparisons between parallel versus 451, 451 versus perpendicular, and parallel versus perpendicular, the number of pairs were n¼ 9 in each comparison.
response of the short samples (Fig. 7d). The velocity profiles of the moving grips still do not correspond exactly to the input velocity, but velocity profiles are more constant for the short samples than for the long samples (Fig. 7). The mechanical response with an initial high loading rate decreasing with time can be of interest, but a sudden acceleration is not desirable, and this is the reason why no statistics have been included with the cited previous works (Gallagher et al., 2012). The results obtained for long samples in the present study should be analyzed with this remark in mind. These difficulties associated with dynamic experimentation highlighted here, go some way to explaining the dearth of reliable data on this topic. Great care has been taken in this study to eliminate the effects of both inter-subject variability and intra-subject variability: all samples were obtained from the same subject and samples were taken immediately adjacent to each other as shown in Fig. 1. By removing the inter-subject variability, this strategy simplifies the task of determining the effect of the strain rate and anisotropy on the failure characteristics of skin. Furthermore, the comparison of results coupled by symmetry with respect to spine does not reveal any statistical difference, supporting the integrity of the experimental data. In agreement with other studies (Haut, 1989; Dombi et al., 1993; Lim et al., 2011), the current study clearly reports an increase of the ultimate tensile stress with increased loading rate. Only Arumugam et al. (1994) were unsure of the effect of the strain rate but this may be explained by the narrow and low range of strain rates tested in their study (from 5% min 1 to 1000% min 1). The parameter E2 (slope of the linear portion of the curve) was also found to be very rate-dependent in the present study, synonymous with a higher stress stiffening effect in dynamics as also reported by Nie et al. (2011) for porcine muscle. In addition to the effect of the strain rate alone, the present study investigates the influence of the orientation of the samples with respect to Langer lines. Table 2 shows that samples perpendicular to Langer lines always break at lower ultimate stress than parallel samples, regardless of the strain rate. Other studies, conducted at both quasi-static and dynamic strain rates, confirm this trend (Stark, 1977; Dombi et al., 1993; Gąsior-Głogowska et al., 2013). Examining more closely the interaction between strain rate and sample orientation, it can be seen that increasing the average input strain rate from 0.06 s 1 to 167 s 1 raised the ultimate tensile stress by 50%, 60% and 90% for parallel, 451 and perpendicular samples respectively. Whereas the same increases in strain rates led to a slight decrease in stretch at rupture from 1.570.1 to 1.370.1 in parallel samples, but for 451 and perpendicular samples no difference was detected. If parallel samples rupture at a lower stretch value than perpendicular samples at high strain rates, it may explain why, as blunt trauma studies have shown in the past, that skin lacerates more often along Langer lines (Leung et al., 1977). The increase of the stress-stiffening effect with the loading rate, evident through the increase in E2, is also intensified with the orientation of the samples. Indeed, E2 is 200% greater for parallel and 451 samples at high strain rate than low strain rate, 200% greater for 451 samples, and 340% times greater for perpendicular samples.
journal of the mechanical behavior of biomedical materials 41 (2015) 241 –250
247
Fig. 7 – (a) Long sample: nominal stress with respect to time (red crosses), input velocity with respect to time (blue squares); (b) long sample: nominal stress with respect to stretch ratio (red crosses), velocity transmitted to the sample with respect to stretch ratio (blue squares); (c) small sample: nominal stress with respect to time (red crosses), input velocity with respect to time (blue squares); (d) small sample: nominal stress with respect to stretch ratio (red crosses), velocity transmitted to the sample with respect to stretch ratio (blue squares).
Fig. 8 – Illustration of the principle of high loading rate operation.
To understand why different stresses are obtained at the same value of stretch ratio for quasi-static and dynamic tests, it is necessary to consider the microscopic structure of skin. The present tests include two main layers of the skin (epidermis and dermis), but the dermis is usually believed to play the most important role in the mechanical strength of the organ. The dermis is a conjunctive tissue, whose extracellular matrix is made of elastic fibers and collagen fibers, both embedded in a ground substance (mostly water and glycoproteins). The collagen fibers become denser and thicker with increasing depth (diameter 10–40 mm), and consist of a gathering of collagen fibrils (diameter 60–100 nm) (Silver et al., 2001). The cohesion of the structure is ensured by chemical bonds, which define the degree of crosslinking within the collagen network, while other chemical bonds bind the collagen network and the ground substance. It seems that the role of the ground substance is less significant than the degree of collagen crosslinking, which determines the strength of the skin at
248
journal of the mechanical behavior of biomedical materials 41 (2015) 241 –250
quasi-static strain rates. The more the network is crosslinked, the more numerous the chemical bonds, the higher the strength is. In dynamic experiments, the effect of the interaction of the collagen fibers with the ground substance is added, as well as the relative sliding of fibers past each other. High friction and viscoelastic forces are created with respect to the strain rate. A simple indicator of these forces can be provided here by the ratio between the mechanical parameters at high and low speed (Dombi et al., 1993). The stretch ratio at failure and the stretch ratio at UTS do not appear to differ significantly in our study. However, this result depends largely on the definition of rupture. Rupture has been defined here as superficial tear on the epidermal surface as detected by the DIC system. In the example given in Fig. 9, a tear is evident in the third image. Using our definition, this is the moment failure occurs. It can be seen however, that beyond this point, a single filament still connects the two pieces of skin at the end of the test. Clearly, an alternative definition of the failure would have resulted in a more significant difference between the stretch ratios at UTS and stretch ratio at failure. The influence of the strain rates on the stretch ratio at ultimate stress or at failure is not very definitive in the study.
The stretch ratio at ultimate tensile stress does not seem to be affected by the loading rates as Vogel (1972) or Haut (1989) has already reported. This suggests that stretch ratio at failure may be a suitable parameter to be used as a local criterion to predict skin failure (tearing). This, however, is in contrast to the findings of Lim et al. (2011) who noted a strong reduction of the strain between quasi-static and dynamic strain rates. The observed reduction of the stretch ratio at rupture for parallel samples already mentioned here is not definitive given the large standard deviations. Moreover, the values of stretch ratio in our study are very low with extensibility of only 30%. This fact, coupled with the low sensitivity of strain rates to stretch ratio at rupture can probably be attributed to the donor age. There is a destruction of elastic fibers with age (Silver et al., 2001) and the degradation of the elastic fibers may explain the low stretch values found in the current study. The collagen network becomes less dense after 30 years of age (Vitellaro-Zuccarello et al., 1994), which would contribute to the diminution of the strength of the skin. The composition of the ground substance is also more aqueous with age (Agache, 2000). It is assumed here that the nature of the ground substance for an elderly person results in lower viscous and
Fig. 9 – Pictures of skin surface at different steps after UTS for a dynamic test.
journal of the mechanical behavior of biomedical materials 41 (2015) 241 –250
friction effects in dynamic loading as discussed previously. However, a clarification of the role of the ground substance at high loading rates is still needed. The trend of the stress–strain relationships obtained in this study shows progressive decreasing stress with strain beyond the ultimate tensile stress. Fig. 9 shows images of skin for this phase. Damage to the material occurs, and microfractures or microtearing are sometimes visible on the surface, but they usually occur internally. Arumugam et al. (1994) suggest that defibrillation rupture is always present, but it usually presents with very thin filaments for quasistatic speeds and thicker filaments for dynamic speeds. Depending on this defibrillation process, the stress does not vanish abruptly, for either quasi-static or dynamic tests. Comparison with literature is always a delicate subject in the context of soft biological tissues. The studies are not numerous and the test conditions vary from study to study. At quasi-static strain rates, the UTS of the current study is found to be in the same range as Ní Annaidh et al. (2012a); Gallagher et al. (2012), Dunn and Silver (1983) and Ankerson et al. (1999). Our elastic modulus of 76.7740.3 MPa is however higher than Gąsior-Głogowska et al. (2013) for the thigh with 2–12 MPa or 18.8 MPa for the abdominal skin found by Silver et al. (2001). Elastic moduli from in vivo studies are evaluated at very low strain range and cannot be compared here (Zahouani et al., 2009; Pailler-Mattei et al., 2008). Very small failure strains are obtained in our quasi-static tests in contrast with the 207% maximal strain found by Jansen and Rottier (1958). At dynamic strain rates, Jacquemoud et al. (2007) performed tests with forehead and arm skin. Their UTS values were approximately 50% less than presented here (i.e. 25.878.2 MPa) and the elastic moduli were in the lower limit of our estimation (i.e. 169.1770.5 MPa). Other studies at dynamic strain rates involved animal skins with UTS values ranging from 0.1 MPa to 30 MPa (Lim et al., 2011; Haut, 1989). However, the test conditions varied greatly (difference of species, age, location, orientation, loading rate, test equipment, etc.) and no direct comparison can be made. There are a number of limitations associated with this study. First, the analysis is based on an assumption about the direction of Langer lines. Ideally, as already noted in Brown (1973), Langer lines should be identified for each individual patient to be really pertinent, this process, however, is destructive and could not be performed here. A clear discussion of the effect of speed and orientation of the samples at very low strains is not really possible due to the choice of the load sensor, optimized for the force levels at rupture. Moreover, the initial acceleration applied to the sample at the beginning of the dynamic tensile test is not well controlled, leading to different strain rates being applied in this elastic region. The separation of the effect of the orientation and the effect of the strain rates in this loading region is therefore difficult. Some studies report the disappearance of the toeregion with high loading rate. In our study, we observe a shortening of this toe-region, a shift towards lower strain, but a toe region was still present. It is possible that the high frequency of acquisition for strain and stress data has facilitated the capture of this feature where it was not previously possible. Finally, the sample size is low, but by
249
eliminating the effects of inter and intra subject variability the data obtained can be used effectively.
5.
Conclusion
This study provides data on the combined effect of strain rate and anisotropy of human skin (90 year old). Both anisotropy and strain rate strongly affect the ultimate tensile stress (UTS), and the strain-hardening effect (E2). The stretch ratio at ultimate tensile stress was found to be constant with strain rate, but statistically dependent on the orientation of the sample. Now that the control of dynamic strain rate has been improved, further tests are planned to clearly distinguish between the role of the collagen and ground substance within the dermis under dynamic loading. A better understanding of the different deformation mechanisms of parallel and perpendicular samples at high loading rates will be the starting point to propose analytical models in future.
r e f e r e nc e s
Agache, P., 2000. Physiologie de la peau et explorations fonctionnelles cutane´es. Edition Me´dicales internationales, Cachan, France. Ankerson, J., Birkbeck, A.E., Thomson, R.D., Vanezls, P., 1999. Puncture resistance and tensile strength of skin stimulants. J. Eng. Med. 213, 493–501. Alexander, H., Cook, T., 2006. Variations with age in the mechanical properties of human skin in vivo. J. Tissue Viability 16 (3), 6–11. Arumugam, V., Naresh, M., Sanjeevi, R., 1994. Effect of strain rate on the fracture behaviour of skin. J. Biosci. 19 (3), 307–313. Boyer, G., Laquie`ze, L., Le Bot, A., Laquie`ze, S., Zahouani, H., 2009. Dynamic indentation on human skin in vitro: ageing effects. Skin Res. Technol. 15, 55–67. Brown, I., 1973. A scanning electron microscope study of the effects of uniaxial tension on human skin. Br. J. Dermatol. 89, 383–393. Chadwick, E.K.J., Nicol, A.C., Lane, J.V., Gray, T.G.F., 1999. Biomechanics of knife stab attacks. Forensic Sci. Int. 105 (1), 35–44. Cua, A., Wilhelm, K.-P., Maibach, H., 1990. Elastic properties of human skin: relation to age, sex, and anatomical region. Arch. Dermatol. Res. 282, 283–288. Daly, C., 1982. Biomechanical properties of dermis. J. Invest. Dermatol. 79 (1), 17–20. Dawes-Higgs, E., Swain, M., Higgs, R., Appleyard, R., Kossard, S., 2004. Accuracy and reliability of a dynamic biomechanical skin measurement probe for the analysis of stiffness and viscoelasticity. Physiol. Meas. 25, 97–105. Diridollou, S., Black, D., Lagarde, J., Gall, Y., Berson, M., Vabre, V., Patat, F., Vaillant, L., 2000. Sex- and site-dependent variations in the thickness and mechanical properties of human skin in vivo. Int. J. Cosmet. Sci. 22, 421–435. Dombi, G., Haut, R., Sullivan, S., 1993. Correlation of high-speed tensile strength with collagen content in control and lathyritic rat skin. J. Surg. Res. 54, 21–28. Dunn, M.G., Silver, F.H., 1983. Viscoelastic behavior of human connective tissues: relative contribution of viscous and elastic components. Connect. Tissue Res. 12, 59–70.
250
journal of the mechanical behavior of biomedical materials 41 (2015) 241 –250
Edsberg, L., Mates, R., Baier, R., Lauren, M., 1999. Mechanical characteristics of human skin subjected to static versus cyclic normal pressures. J. Rehabil. Res. Dev. 36 (2), 1–12. Foutz, T., Stone, E., Abrams, C.J., 1992. Effects of freezing on mechanical properties of rat skin. Am. J. Vet. Res. 53 (5), 788–792. Gallagher, A., Nı´ Anniadh, A., Bruye`re, K., Otte´nio, M., Xie, H., Gilchrist, M., 2012. Dynamic tensile properties of human skin. In: Proceedings of the International Research Council on the Biomechanics of Injury Conference, vol. 40, pp. 494–502. Ga˛sior-Głogowska, M., Komorowska, M., Hanuza, J., Ma˛czka, M., Zaja˛c, A., Ptak, M., Be˛dzin´ski, R., Kobielarz, M., Maksymowicz, K., Kuropka, P., Szotek, S., 2013. FT-Raman spectroscopy study of human skin subjected to uniaxial stress. J. Mech. Behav. Biomed. Mater. 18, 240–252. Haut, R., 1989. The effect of orientation and location on the strength of dorsal rat skin in high and low speed tensile failure experiments. J. Biomech. Eng. 111, 136–140. Holzapfel, G.A., 2006. Determination of material models for arterial walls from uniaxial extension tests and histological structure. J. Theor. Biol. 238 (2), 290–302. Horsfall, I., Prosser, P.D., Watson, C.H., Champion, S.M., 1999. An assessment of human performance in stabbing. Forensic Sci. Int. 102 (2-3), 79–89. Jacquemoud, C., Bruye`re-Garnier, K., Coret, M., 2007. Methodology to determine failure characteristics of planar soft tissues using a dynamic tensile test. J. Biomech. 40, 468–475. Jansen, L., Rottier, P., 1958. Some mechanical properties of human abdominal skin measured on excised strips. Dermatologica 117, 65–83. Khatam, H., Liu, Q., Ravi-Chandar, K., 2014. Dynamic tensile characterization of pig skin. Acta Mech. Sin. 30 (2), 125–132. Lamers, E., van Kempen, T., Baaijens, F., Peters, G., Oomens, C., 2013. Large amplitude oscillatory shear properties of human skin. J. Mech. Behav. Biomed. Mater. 28, 462–470. Langer, K., 1978. On the anatomy and physiology of the skin—I. The cleavability of cutis. Br. J. Plast. Surg. 31, 3–8. Leung, Y.C., Lopat, E., Fayon, A., Banzet, P., Tarriere, C., 1977. Lacerative properties of the human skin during impact. In: Proceedings of the Third International Research Council on the Biomechanics of Injury Conference, Berlin, Germany, pp. 399–411. Leveque, J.L., De Rigal, J., Agache, P.G., Monneur, C., 1980. Influence of ageing on the in vivo extensibility of human skin at a low stress. Arch. Dermatol. Res. 269 (2), 127–135. Liang, X., Boppart, S., 2010. Biomechanical properties of in vivo human skin from dynamic optical coherence elastography. IEEE Trans. Biomed. Eng. 57 (4), 953–959. Lim, J., Hong, J., Chen, W., Weerasooriya, T., 2011. Mechanical response of pig skin under dynamic tensile loading. Int. J. Impact Eng. 38, 130–135. Malm, M., Samman, M., Serup, J., 1995. In vivo skin elasticity of 22 anatomical sites. Skin Res. Technol. 1, 61–67.
Marangoni, R., Glaser, A., Must, J., Brody, G., Beckwith, T., Walker, G., White, W., 1966. Effect of storage and handling techniques on skin tissue properties. Ann. N.Y. Acad. Sci. 136, 441–453. Miller, S.A., Jones, M.D., 1996. Kinematics of four methods of stabbing: a preliminary study. Forensic Sci. Int. 82 (2), 183–190. Nı´ Annaidh, A., Bruye`re, K., Destrade, M., Gilchrist, M., Otte´nio, M., 2012a. Characterization of the anisotropic mechanical properties of excised human skin. J. Mech. Behav. Biomed. Mater. 5, 139–148. Nı´ Annaidh, A., Bruye`re, K., Destrade, M., Gilchrist, M., Maurini, C., Otte´nio, M., Saccomandi, G., 2012b. Automated estimation of collagen fibre dispersion in the dermis and its contribution to the anisotropic behaviour of skin. Ann. Biomed. Eng. 40, 1666–1678. Nie, X., Cheng, J., Chen, W., Weerasooriya, T., 2011. Dynamic tensile response of porcine muscle. J. Appl. Mech. 78 (2), 1–5. Pailler-Mattei, C., Bec, S., Zahouani, H., 2008. In vivo measurements of the elastic mechanical properties of human skin by indentation tests. Med. Eng. Phys. 30 (5), 599–606. Pailler-Mattei, C., Debret, R., Vargiolu, R., Sommer, P., Zahouani, H., 2013. In vivo skin biophysical behaviour and surface topography as a function of ageing. J. Mech. Behav. Biomed. Mater. 28, 474–483. Ridge, M., Wright, M., 1966. The directional effects of skin. J. Invest. Dermatol. 46 (4), 341–346. Shergold, O., Fleck, N., Radford, D., 2006. The uniaxial stress versus strain response of pig skin and silicon rubber at low and high strain rates. Int. J. Impact Eng. 32, 1384–1402. Silver, F., Freeman, J., DeVore, D., 2001. Viscoelastic properties of human skin and processed dermis. Skin Res. Technol. 7, 18–23. Stark, H.L., 1977. Directional variations in the extensibility of human skin. Br. J. Plast. Surg. 30 (2), 105–114. Vitellaro-Zuccarello, L., Cappelletti, S., Dal Pozzo Rossi, V., SariGorla, M., 1994. Stereological analysis of collagen and elastic fibers in normal human dermis: variability with age, sex and body region. Anat. Rec. 238, 153–162. Vogel, H.G., 1972. Influence of age, treatment with corticosteroids and strain rate on mechanical properties of rat skin. Biochimica et Biophysica Acta 286, 79–83. Xu, F., Wen, T., Seffen, K., Lu, T., 2007. Characterization of thermomechanical behaviour of skin tissue—I. Tensile and compressive behaviours. In: Proceedings of the World Congress on Engineering, London, UK, vol. 2, pp. 1446–1450. Yamada, H., 1970. Strength of Biological Materials. Edition F. Gaynor Evans, MI, USA. Zahouani, H., Pailler-Mattei, C., Sohm, B., Vargiolu, R., Cenizo, V., Debret, R., 2009. Characterization of the mechanical properties of a dermal equivalent compared with human skin in vivo by indentation and static friction tests. Skin Res. Technol. 15 (1), 68–76. Zhou, B., Xu, F., Chen, C., Lu, T., 2010. Strain rate sensitivity of skin tissue under thermomechanical loading. Philos. Trans. R. Soc. London, Ser. A 368, 679–690.