INTERNATIONAL JOURNAL OF SCIENTIFIC & … · This research aimed at a review the types of FE head...

INTERNATIONAL JOURNAL OF SCIENTIFIC & TECHNOLOGY RESEARCH VOLUME 2, ISSUE 7, JULY 2013 ISSN 2277-8616

17

IJSTR©2013 www.ijstr.org

Finite Element (FE) Human Head Models / Literature Review

Hatam samaka, Faris Tarlochan

Abstract: Road accidents claim thousands lives of pedestrians in the world every year, so the subject of pedestrian protection at collision of the most important topics of interest to researchers and modern car designers, so the researchers began to design friendly cars by using FE simulation to perform the tests required for this design and build models for this purpose to reduce the expenses of these tests and get results more quickly and close to

reality. This research aimed at a review the types of FE head models which is considered the most vulnerable part which exposed to the fatal injury and designed by companies, universities are used in the tests required to be achieved design of the front of vehicle.

Index Terms: Pedestrian protection, Friendly cars, EEVC WG17 standard headform models, Head injury ,Friendly cars design ————————————————————

Introduction This paper aims to analyze previous studies on FE human head models, which are used in the medical and engineering fields. This study proposes a technique to protect pedestrians from fatalities and injuries resulting from vehicle-pedestrian accidents and to minimize the annual occurrence of such accidents worldwide. This study also aims to review several head injuries that often occur in daily human life such as work injury, sporting injury, and soldier injury. Approximately four million pedestrians worldwide are killed annually. The classification of pedestrian fatalities varies across low-, middle-, and high-income countries. The burden of road traffic injuries on vulnerable motorists differs substantially across income levels (Table 1) [1].

Table1: Pedestrian fatalities percentage and the number of fatalities

Road traffic accidents (RTAs), falls, and criminal assaults are the most common causes of head injuries based on the rate of head injuries in the three regions shown in Fig. 1. Accidents are the most common cause of severe head injuries leading to death, whereas falls and assaults often lead to non-fatal injuries or hospitalization [2]. Head injuries in the United States (US) kill one person every two and a half minutes [3]. In the US alone, the cost allocated for medicines and hospitalization reaches up to three billion dollars yearly.

Fig.1: Distribution of causes of head injury in the three regions in

the world [4]

In the 1960s and 1970s, most related studies showed that head injuries mostly focused on military aircraft manufacturing. However, the automotive industry now the focus of such studies because of government and customer requests to reduce traffic accidents [5]. Since 1970, studies on pedestrian protection have been conducted, with a focus on the prevention and reduction of pedestrian injuries [6]. Through computer simulation, the trajectory of body parts during a crash was determined. Results showed that the most important factor causing pedestrian death is head injury. The assessment of performance of the pedestrian protection techniques rely mostly on the (European Enhanced Vehicle-Safety Committee Working Group 17 (EEVC17) test procedure. This report focuses on the hood to study head injuries upon front collision with a pedestrian. Studies focused on the crashing of the head against the windshield, the A-pillar, and two front sides of the car [7]. Several studies utilized an EEVC WG17 adult headform FE model (FEM; standard model), whereas others focused on the FEM head designed by using the LS-DYNA program.

Types of FE models used for the human headform in previous studies 1. EEVC WG17 adult headform FE model Most headform impactors are made of aluminum spheres covered by thick vinyl skins. The adult (4.8 kg) and small adult (3.5 kg) headform impactor have an outer diameter of 165 mm, whereas the outer diameter of a child headform (2.5 kg) is 130 mm, Fig. 2 [8,

————————————————

Hatam Mahmood Samaka, Master Degree, University of Technology – Iraq, Uniten university- Malaysia, E-mail: [email protected]

Faris Tarlochan Institution, Assoc. Prof. Ir. Dr. of Mechanical Engineers - Malaysian Branch, Head, Center for Innovation and Design, University Tenaga Nasional., Malaysia, E-mail: [email protected]

mailto:[email protected]

mailto:[email protected]


18


9]. In the numerical model, the aluminum parts are simplified as the rigid parts of the human head with specific mass and inertia properties. The model contains 19572 nodes, 4044 shells, and 16514 solid components.

Fig.2: EEVC WG17 headform models of three heads:

Child – Adult – Small Adult and impactor parts [9] The impactor FE model developed by the commercial company ARUP consists of four parts (Fig. 3); the back plate, sphere, skin, and null shell. The back plate and sphere are rigid components. The skin comprises a layer of vinyl material with thickness of 14 mm. The null shell is a type of foam that encases the sphere. The output data and head injury criterion (HIC) value were calculated based on linear acceleration. This model contains 14045 nodes and 12162 shells [10]. The results were verified through a comparison with practical tests performed on the model using real car hood models to improve the FE-impactors [11].

Fig.3: Developed EEVC WG17 adult headform model [10]

2. Yang et al. 2008 human head FE model This model was designed based on the brain anatomy of a human body according to the specifications of each part of the physical

brain, including the cerebellum, bone, and skull. The HIC value, angular head acceleration, and spinal fluid pressure inside the cranium can be determined by this model (Fig. 4). The model contains 66624 nodes, 11514 shells, and 32 components [10].

Fig.4: Yang et al. 2008 human head FE model [10]

This model represents all features of the fiftieth percentile of a male head. To achieve the required precision and verify accuracy, test results were compared with the practical results obtained by Nahum et al. (1977) and Tosseille et al. (1992) [10].

3. Wayne State University (WSU) Head (Brain) Injury Model (WSUHIM) The WSU developed a group of FEMs to study the human head and brain injury, as well as the WSUBIM model by Ruan et al. [12] The first version underwent developed numerous times, and the final version was developed by Zhang, Hardy, Omori, Yang, and King in 2001, with the number of elements reaching 245000 [13]. The model was developed such that it contains the specifications of different materials such as the gray and white matter (Fig. 6), to increase the accuracy of predicting the occurrence of diffuse axonal injury (DAI) in the areas of contact between gray and white matter. The results were verified through a comparison with the results obtained by Nahum [14]. This model was redeveloped by Al-Bsharat et al. [15] and Zhang et al. [16][17] to be more accurate and to achieve a more detailed simulation of the sliding surface between the brain and cerebrospinal fluid (CSF) to obtain better results with details of the heterogeneous structures of the brain [18]. This model is commonly used in the medical field because it can determine the pressure inside the cranium. This model was developed by Zhang et al. with 3000 elements (Fig. 5). The model was created using the neutral density target technique and bi-planar high speed X-ray system (brain atlas) [19].


19


4. TNO Head FE Model This model, which exhibits a hexagonal form, was created by TNO. The model consists of 22 parts, 13868 elements, and 15637 nodes installed to hold the skull and brain, which are similar in shape (Fig. 7). All parts of the model (cranium visceral, dura mater, falx cerebri, and tentorium cerebellar) were represented by homogeneous isotropic linear elastic material, whereas the brain, brain stem, and the cerebellum used the same viscoelastic material to simplify the model and to simplify the analytical study of the mechanical responses of the brain by mitigating shear strain [22].

Fig.7: TNO FE Head model [22]

The head–ground collision process has been studied using this model with the ground divided into three modes: frontal part of headform collision with the ground, temporary collision with the ground, and occipital bone collision with the ground. The model was established from Hypermesh, and the material properties are shown in Table 2 [23]. The numerical analysis was performed using the LS-DYNA program. The simulation results showed the distribution of intracranial stresses, the growth of stress waves, and changes in intracranial pressure of the brain at the point of collision [22].

Table 2: Material properties of TNO FEM [22]

5. Head-Brain FE model This model consists of [24] three main parts: the skull, brain, and skin. The skull consists of the cortical bone, which is represented by solid elements, and the spongy bone, which is represented by shell elements. The brain is composed of hexagonal solid elements representing each of its parts: the brain, cerebellum, brain stem, gray matter, white matter, and CSF. The dura mater, arachnoid, membrane meningitis, sickle brain, and tentorium (tent cerebellum) are represented by the shell element (Fig. 8).

Fig.8: Head-Brain FEM [8]

The model weighs 4.39 kg. The space between the brain and the skull is filled with CSF, which cushions the brain with a simple connection between the brain and skull over the arachnoid. The representation of CSF elements is hexagonal because of the problem that occurs in analytical studies using the LS-DYNA program in cases where CSF is represented by the specifications of a structural liquid material, such that the hexagonal solid elements will be adopted. The model contains 25200 shells, 49700 elements, and 24100 solids [24]. The representation of the skull utilized an elastic-plastic material with failure (MAT24/105), whereas the brain substance utilized a non-compressible viscoelastic material model (MAT5) [25].

6. ULP FEM of the human head This model was developed at the University of Louis Pasteur in Strasbourg, France and is based on other anatomical designs of the human head (Table 3). HIC and the internal brain pressure can be calculated through this model. The advantages of this model include the capability to validate long term cases over 15 ms as well as the use of compounds that can facilitate the rotational acceleration of the model [26].

Table3: ULP FEM specification [26]


20


The human head injury mechanism was established based on previous studies by Nahum [27] and Willinger [28], Table 4. The table shows that despite the prediction for a specific injury, the victim remained uninjured [9]. If the value of Von Mises shearing stress increases to 18 kpa, neurological injuries occur (diffuse axonal injuries or hemorrhagic DAI injuries). If the value of Von Mises shearing stress increases to 38 kpa, severe neurological injuries occur (SEV DAI)[26][28]. Table 4: Human head injury mechanisms with tolerance limits

for ULP FEM [26]

This model weighs 4.8 kg [26], and the Hypermesh software was used. This model consists of 13208 elements, 10395 brick elements (Venetian tiles) and 2813 shell elements. The results of this model were validated by comparing the simulation results with the experimental results from Nahum [29] and Trosseille [26]. 7. Politecnico di Torino University FEM of the human head This model was built based on computer tomography (CT) scans and magnetic resonance images, taking approximately 160 CT scan images of approximately 1.25 mm from the adult human head to build the internal and external surfaces of the cranium as well as the facial bones. The model consists of 2600 nodes and 55,246 elements. [30] The FEM (Fig.9.a) was obtained by using Hypermesh 5.1. The scalp was represented by brick elements with a 6 mm thickness. The cranial bones were represented by three layers, two for the external layers of compact bone and one for the internal layer of cancellous bone. The dura mater, falx, and tentorium membranes were represented by four-node shell elements. CSF was represented by eight-node brick elements. The brain tissues and ventricles were represented by tetrahedral elements (Fig. 9b). [30]

Fig. 9: Politecnico di Torino University FEM [30], a) Frontal and

perspective view of the head model, b) Coronal and axial section of the head model

The dura mater was designed pursuant to the internal surface of the skull and falx, with the tentorium manually designed depending on anatomical images. CSF was approximately 2 mm from the membrane surface surrounded all membranes and the brain. The mechanical properties for all tissues of this model were estimated based on experimental tests and by considering the literature data. An average value was used for some parameters [30].

8. Harvard Medical School FEM This FEM was built in Harvard Medical School and was developed at the Bioengineering Laboratory of the University of Salerno, Italy. Axial MRIs available in all images of the brain atlas of the Harvard Medical School were used to meh the model [31][32][33] (Fig. 10).

Fig.10: Harvard Medical School brain FEM: mid-sagittal and mid-

coronalsections [31][32]

The model weighs 1.4 kg, the brain size is 1508 cm3 including CSF, and the skull size is 678 cm3 with a mass of 0.8 kg. The model was simulated with a 10-node composite tetrahedral and 39047 tetrahedral composite element [31] [34]. This model is represented different portions of the brain tissue with compounds considering viscoelastic mechanisms (the brain is composed of viscoelastic materials). One-term Ogden functions were utilized extensively in previous studies on head injury simulations, such as those by Zhou et al. [18], Zhang et al. [16],[35], Kleiven [36], Kleiven and von Holst [37], and Horgan and Gilchrist [21]. In this study [31], the short-term shear module reduction by a factor of 1/2.5 was achieved by Zhang et al. [16] for white matter, gray matter and the brain stem to ensure expedience with the short-term brain tissue model implemented by Mendis et al. [38].

University College Dublin Brain Trauma Model (UCDBTM)

9. Three–dimensional (3D) FEM The six types of 3D FEM are the baseline model, Sliding Boundary Model, Grey-White-Ventricular matter model (GWV model), three-element CSF model, projection mesh model, and morphed model. All model designs are based on data provided by the physiological and anatomical details of the human head. These data are simplified for easier processing based on geometry, loading conditions, properties of elements, and foundation configuration. To select the appropriate model type [39] and to determine the degree of complexity required, a simulation was employed. The 3D-FEMs are applied using CT [4][21], and the scanned images of the human head were stored with the CT data to create a polygonal model using vtk software [40]. The model was then reduced for easier implementation [39]. Data were then converted to the IGES format using the commercial software MSC/Patran [39] [41]. This study highlights one type of 3D- FEM, the Baseline Model [39]. An


21


example of this model, the University College Dublin Brain Trauma Model (UCDBTM) [4] and the results obtained in previous studies are shown in Figs. 11 and 12.

Fig. 11: Baseline finite element model (UCDBTM) [39]

Fig.12: UCDBTM finite element model [4]

The Baseline Model consists of the [21][39] cerebrum, cerebellum and brainstem, intracranial membranes (falx and tentorium), pia, CSF layer, dura, a three-layered skull (cortical and trabecular bone layers) with varying thickness, scalp, and the facial bone. The brain consists of 7318 hexahedral elements, whereas the CSF layer consists of 2874 hexahedral elements with one element through the thickness. Earlier FEMs of the brain used linear elastic material to represent and build the model [2][21] [42][43]. Recent studies adopted a linear viscoelastic material associated with the theory of large-deformation of the representation of the model brain tissues [44][45][46]. This material was also selected in this study to represent the brain [21]. The rest of the model utilized the same materials as in previous studies [37][46][47][48]. Table 5 shows the mechanical specifications of materials used in the model. Model weight at 4.01 kg and brain weight at 1.422 kg were supported by the results of 37 experiments conducted by Nahum [14] [21].

Table 5: Material properties of baseline 3-D FEM [21]

Results of tests of FE models

1. Yang et al. 2008 human head FE model The FE Euro NCAP adult headform model was utilized as the main tool in assessing the severity of head injury according to the tests required in the EEVC WG17 in collisions with the hood (Fig. 13). The Yang et al. (2008) adult headform identified HIC with the highest efficiency (bio-fidelity) for being closer to a real human head. The model also determined the baseline on the hood and can also identify the severity and mechanism of the injury and its location on the head based on specific intracranial pressure [10].

Fig.13: Two models head form and Yang et al. (2008) adult

headform human head with reversible hood. [10]

To verify the effectiveness of Yang et al.‘s (2008) adult headform model [10], it was implemented in seven crash tests with a hood using the headform and human headform models to calculate the HIC. The results are shown in Table 6.In cases 3 and 4, the value of HIC decreased. The first reason is the location of these two points, which is near the center of the hood and away from the rear end of hood. The second reason is that these two points are far from the surface of the engine. This case cannot be adjusted in the process of simulation, thus resulting in the difference.

Table 6: HIC values (test and simulation). [10]

2. Wayne State University Head (Brain) Injury Model (WSUHIM) The WSUHIM model [19] was used to simulate a real collision between a pedestrian and a car, proving the capability to predict the DAI in the brain through the knowledge of the principal strain it incurs. Through further research, the model was found incapable of providing DAI models because of an error in building the model, including wrong vehicle speed and contact between the car and the pedestrian‘s head ( Figs. 14 and 15,where two specimens, male and female, were used in the study). The model was used to test the curve acceleration of the head using the actual program (Figs. 16, 17, 18, and 19 for two specimens). Notably, the severity of head


22


injury HIC = 1000 is in accordance with the maximum. The principal strain reached 0.2, and the rotational acceleration caused a significant edge that can injure the brain.

Fig. 16: Linear Head Acceleration for male [19]

Fig. 17: Angular Head Acceleration for male [19]

Fig.18: Linear Head Acceleration for female [19]

Fig.19: Angular Head Acceleration for female [19]

Figs. 25 and 26 [19] show the difference between simulation and reality in terms of the maximum principal strain (MPS) with reference to Gennarelli et al. [49], Thibault et al. [50], Ueno et al. [51], and Eppinger et al. [52]. The maximum amount of principal strain reached 0.15 in this study. Fig. 25 shows an increase in the amount of MPS of more than 0.15 in the male sample, which includes half the brain. Fig. 26 shows an increase of more than

0.15 in the female sample, which includes the whole brain. 3. TNO Head Finite Element Model With the simulation of the collision with the ground at three positions: frontal, temporal (side) and occipital, results show that the lateral and occipital positions increased the amplitude of mechanical response with increasing speed of collision, such that a linear relationship exists [22]. In the study of these three collision conditions, a disorder force initially appeared in the brain tissues, and then a compressive force occurred in the contrecoup position (Fig.20) as well as far from the contrecoup point. The amplitude of the disorder force is constant, whereas the amplitude of the compressive force is larger. The collision in the frontal part of the model showed that the response of the intracranial pressure of brain tissue inside the cranium in contrecoup position formed the main pressure points, either in the occipital collision or in the form of main points of tension associated with the law of collision speed, thus turning into a main pressure point during high-speed collision.

Fig. 20: Coup and contrecoup position (injury) [53]

http://en.wikipedia.org/wiki/File:Contrecoup.svg


23


From the simulation of the collision, Von Mises stress wave rapidly decayed from the skull and brain tissue, Fig. 21 shows the transmission of Von Mises stresses in three cases: front, side, and occipital during collision with a rigid body.

Fig. 21: Von Mises stress distribution at a speed of 15 km/h [22] In this study, the distribution of intracranial pressure among brain tissues within the skull fundus in the frontal collision and the emergence of a turning area do not appear clearly in the temporal collision and disappear completely in the occipital collision (Fig. 22).

Fig. 22: Intracranial pressure distribution in the surface of brain

tissues at a speed of collision of 15 km/h [22]

For the stress waves in the skull surface, the point of impact is considered to be in the center and then spreads to neighboring areas, with the primary concentration region of stresses being in the skull fundus (Fig. 23) [22].

Fig. 23: Stress wave spreading in the skull surface [22]

4. Head-Brain FE model In this study, [24] the Head-Brain FE model was used as the head of a car driver. The study results were validated using the previous study data by Trosseille et al. [29] The simulation conditions include the use an impactor similar to a steering wheel with a weight of 23.4 kg. The direction of impact is in front of the model, and the speed of impact 7 m/s. The results showed that the linear speed was 1000 m/s and 7600 m/s2 for the angular acceleration of the center of gravity of the head (Fig. 24) when the driver is in a sitting position. Intracranial pressure in the brain and acceleration are both measured, and the simulated results are compared in Fig. 24, which shows the curve of the time history with the intracranial pressure of the brain in cases of front and occipital impact, as well as the curve of time to acceleration under the same situations. The results were reasonable compared with practical experiences [24].

Fig. 24: Simulation of Head-Brain FEM [24]

5. ULP FEM of the human head The study [26] used nine tests cases, each of which was calculated by simulating the brain pressure (kpa), brain Von Mises stress (kpa), global strain energy of the brain/skull interface (MJ), and global strain energy of the skull (MJ). The result of each case is shown in Table 7. Pedestrian-generated brain contusions can be divided into three cases. The first case can be predicted by numerical simulation, whereas brain injury I the other two cases cannot be predicted by simulation. The other six cases of brain contusions were not examined. Brain neurological injuries did not occur in one case, but moderate brain neurological injuries occurred in the other cases. Three cases exhibited subdural or subarachnoid hematoma, for which the model cannot predict the specific injury. No subdural or subarachnoid hematoma occurred in the other cases. Four cases exhibited skull fractures. The model predicted the specific injury for two cases but was incapable of predicting the specific injury for the other two cases [24].

6. Politecnico di Torino University FEM of the human head The mechanical properties of the different components of this FEM model are summarized in Table 8 [30]. The scalp mechanical properties are ρ=1200 Kg/m3, E=16.7 Map, and ν=0.42. These values have been used in the WSUBIM FEM [10] by Willinger–Kang–Diaw [54, 55] and in the model by Zhou et al. [18]. To validate the loading condition and the numerical model, the results were compared with experimental tests conducted by Nahum in 1977 [14], which used impactors to impact the head from the front. The validation reference in this study was the impact force,


24


pressure distribution in front of the skull (impact position), and the posterior-fossa subarachnoid space [30].

Table 7: The cases simulations result of ULP FEM. [24]

Table 8: Material characteristics [30]

To validate the loading condition and the numerical model, the results were compared with experimental tests conducted by Nahum in 1977 [14], which used impactors to impact the head from the front. The validation reference in this study was the impact force, pressure distribution in front of the skull (impact position), and the posterior-fossa subarachnoid space [30]. The maximum impact force value was F=7.56 kN. Nahum used a similar value of 7.9 kN [14] and set up the impactor speed at V=7.0 m/s, where the best results were obtained between numerical and experimental impact force behavior (Fig. 25) by using an impactor with mechanical properties shown in Table 9.

Fig. 25: Impact force behavior [30]

Table 9: Impactor mechanical properties [30]

Figs. 26 and 27 show the difference between the numerical and experimental pressure values in the frontal area of the cranium and in the posterior fossa.

Fig. 26: Pressure behavior in the front area [10]


25


Fig. 27: Pressure behavior in the posterior fossa area [30]

Fig. 28 [30] shows the pressure distribution in the brain and the gradual transfer of pressure from the front (impact point) to the occipital zone. This transfer occurred because of the push force of impact in the brain from the front to the occipital area, which generated high stress in the connective tissue between the brain and skull. The brain goes to equilibrium and normal pressure 6.3 ms from the beginning of impact (Fig. 28). The analytical study including the variation of pressure distribution in different moments enables the investigation of the loud transfer mechanism from the point of impact in the skull to the brain. The figure also shows the value of the highest pressure in the skull bone 1.8 ms from the beginning of the impact, with the brain floating in CSF without being damaged. The maximum pressure occurs in the brain after approximately 4.4 ms.

Fig. 28: Pressure distribution in the brain sections during impact

[30]

The result of pressure distribution in the front area of the skull and the occipital does not differ significantly from the results obtained in the case where a complete model of the head was used while considering the effect of mechanical membrane classes [56]. 7. Harvard Medical School FEM Fig. 29 shows the curves of the intracranial pressure of the brain and time histories [31] for each of the impact positions: front, occipital (posterior), and parietal lobes. The results are compared with the practical results of Nahum [14], and a significant convergence was noted between the simulation and practical results, making the model legitimate and proving that it can be used in relevant experiments. Based on the simulation results shown in

Fig. 29, the most significant intracranial pressure (compressive–positive) in the brain was in the frontal area, whereas the lowest intracranial pressure (tensile–negative) in the brain was in the occipital (posterior-fossa) area. Fig. 30 shows the intracranial pressure in the mid-sagittal area of the brain model during frontal impact. The most significant intracranial pressure was observed after 3 ms, which continued to decrease until the end of the pulse at 6.5 ms

Fig. 29: Simulation and experiment for three positions of

impact [31]

Fig. 30: Frontal impact: intracranial pressure contours (Pa) [32]

The waves of stresses transmitted from the coup region to the contrecoup region through the brain are reflected after impact in the internal cranium surface in the occipital and back toward the coup. This condition results in a negative pressure wave (tensile stress),


26


which generates cavitation damage (volumetric plastic strain) in various areas of the brain, especially in the contrecoup, Fig. 31. Cavitations appear when the amount of stress reaches the critical cavitation pressure [57][58][59] and unstable growth occurs in the brain tissue, which is amplified by the release wave region and backlash [31].

Fig. 31: Frontal impact, cavitation damage [31]

Dynamic deformation by shear stress varies in the brain. Fig. 32 shows that the effective shear stress continues to increase even after the maximum value of the wave has been reached, which is supposed to be the highest value in the corpus callosum, thalamus, parietal lobe, and midbrain (cf. Zhang et al. [16, 35], Horgan and Gilchrist [60]), but is still less than the value adopted for plastic deformation. Thus, no distortion occurs in this model.

Fig. 32: Frontal impact, shear stress contours. [31]

8. 3D FEM [The University College Dublin Brain Trauma Model (UCDBTM)] The baseline model and other 3D models support were applied in a simulation based on Trosseille‘s impact test [61] and were verified using the practical results of Nahum [14][62]. Figs. 40 and 41 show a comparison between the experimental and simulation results in the front and occipital pressure zones using the results of Nahum for all 3D models. Fig. 42 shows a comparison between the experimental results and simulation results in the lateral ventricle (baseline and GWV models only). Figs. 33, 34, and 35 show that

the trends in the pressure curves of simulation converged with the practical results at different amount, especially in the occipital lobe.

Fig. 33: Experimental results [61] and simulated predictions of

frontal and occipital pressure.D34: Experimental frontal pressure result, C28: Experimental occipital pressure result [39]

Fig. 34: Comparison of the pressure responseof the three models

versus the Trosseille experimental data [39]


27


Fig. 35: Comparison of experimental and simulated pressures at

the lateral ventricle for the baseline and the GWV models [39] In other studies [19], analysis of the baseline models employed ABAQUS 5.8 [63]. The load was 7000 N at peak as a semi-sinusoidal pulse with duration of 6 ms. Fig. 36 shows the correlation between the simulated and experimental pressure time histories for the impact using the model according to Nahum‘s dimensions. Fig. 37 shows the pressure with the maximum impact force, whereas Fig. 38 shows the Von Mises stress at the same instance.

Fig. 36: Comparison between the experimental intracranial

pressures of Nahum (solid lines) and those predicted by simulation (dashed lines) [21]

Fig. 37: Hydrostatic pressure of brain (Pa) at maximum impact

force [62]

Fig. 38: Von Mises stress in the brain (Pa) at maximum impact

force [62]

Conclusion Human head FEMs were designed according to the purpose of use and tests were conducted to validate the models. Engineering studies of pedestrian or passenger head models require a less complex design than medical studies. Thus, the number of layers constituting the model is more accurate and conforms to the reality of human head parts and materials. The human anatomy is the foundation for the design of the models to achieve the highest possible precision of the head and to obtain results that are closer to reality in terms of materials used, including the layers surrounding the brain, cerebellum, and CSF. The simplified models that obtain approximate results are incomplete, as is the case for the standard EEVCWG17 adult head form FEM. By contrast, other models were developed to obtain accurate results, as is the case of the Harvard Medical School (Fig. 11 and Table 10).


28


Table10: Model specifications

Table 10 shows a comparison among the human head models investigated in this study in terms of shapes, geometry, layers and components, materials, boundary conditions, purpose (engineering or medical field), and calculated results. The development of previous models enabled the obtained results to be close to the results obtained in practical studies. The practical studies by Nahum [14] and Trosseille [29] were used to verify the simulations in the adoption of the models. The EEVC WG17 adult headform was used as the major tool for evaluating the quasistatic and dynamic compression, shear, tension, confined quasistatic compression [9], and HIC in addition to intracranial pressure, which

can indicate the severity of the head injury and detail the location and mechanism of injury [10]. The human head FEM of Yang et al. was the most complex in terms of the number of layers, materials used to represent parts of the brain, and the number of elements used to calculate the intracranial pressure of brain, ventricular pressure, and HIC [10]. The WSUHIM model was used to simulate head impacts in real pedestrian accidents and its capability to predict the occurrence of DAI based on maximum principal strain was demonstrated. This model can calculate intracranial pressure of brain, brain Von Mises stress, maximum principal strain, linear acclamation, MPS contour, DAI map, and HIC [19]. In the TMO Head FEM is a simple model in terms of design and the number of elements and nodes. Results of the simulation in this model include sharp decays of the stress wave during coupling, intracranial pressure mutant in the surface of brain tissue in the skull fundus, spread of the stress wave on the skull surface, intracranial stress-time courses on the surface of brain tissue, intracranial pressure, and HIC [22]. The ULP FEM of the human head, which is a complex model that contains details of the whole brain, cerebellum, CSF, and layers surrounding each of these parts, was used to calculate the intracranial pressure of the brain, brain Von Mises stress, global strain, energy of the brain/skull interface, and HIC [26]. The Politecnico di Torino University FEM of the human head, which is also a complex model, used images obtained by TC or MRI scanners that proved to be fundamental in obtaining a realistic geometry that can be used as a starting point for the numerical model. Satisfactory results were obtained for the impact force and the pressure distribution behavior, but difficulties arose in simulating pressure behavior in the posterior fossa. Intracranial pressure of brain, brain shear stress, DAI, shear stress distribution on sagittal, pressure distribution on a sagittal section during impact, and HIC were obtained using this model [30]. The Harvard Medical School FEM was used for medical purposes to determine clinically relevant injuries, such as DAI and cavitations, which are related to specific mechanical damage modes of brain tissue involving plastic and/or viscous deformation. Brain intracranial pressure, cavitation damage, shear stress contours, and HIC were obtained using this model [31]. The UCDBTM is a simple model that contains three layers with all elements represented by hyperelastic material. The brain tissues were heterogeneous at both the macro and micro scale. The mechanical properties of the brain and CSF will affect the pressure response of the brain. Thus, any model of the intracranial system that is excessively stiff or compliant will not accurately simulate a physical head response. Frontal brain pressure, amount of brain movement, brain shear stress, and HIC were obtained from this model [39]. The results were inaccurate. Thus, this model cannot be used in the medical field. A 3D model of the skull-brain complex was constructed, and the method by which it was built wad reported. The procedure could be used by those who wish to develop separate skull-brain models.


29


References [1]. H Naci, D Chisholm and T D Baker group: a global

comparison Distribution of road traffic deaths by roaduser, http://www.who.int/choice/publications/RTIDistribution Article.pdf

[2]. Watson, W.L. & Ozanne-Smith, J., 2000, Injury

surveillance in Victoria, Australia: Developing comprehensive injury incidence estimates, Accident Analysis and Prevention, 32, pp. 277-286.

[3]. Ommaya, A.K., Thibault, L. & Bandak, F.A., 1994,

Mechanisms of impact head injury.International Journal of Impact Engineering, 15, pp. 535-560.http://www.sciencedirect.com/science/article/pii/0734743X94800336

[4]. Michael D. Gilchrist, ―Modelling and Accident

Reconstruction of Head Impact Injuries‖, Key Engineering Materials Vols. 245-246 (2003) pp. 417-429, http://www.scientific.net

[5]. Svein Kleiven, ―Finite Element Modeling of the

Human Head‖, Department of Aeronautics Royal Institute of Technology,S-100 44 Stockholm, Sweden Report 2002 – 9,ISSN 0280-4646

[6]. Rikard Fredriksson,Yngve Håland, Autoliv Research,

Sweden Jikuang Yang ―Evaluation of Anew Pedestrian Head Injury Protection System With A Sensor In The Bumper And Lifting Of The Bonnets Rear Part‖ http://www.autoliv.com/wps/wcm/connect/5f997a004ce4f3f8ac68eef594aebdee/Pedestrian.pdf? MOD=AJPERES

[7]. Yutaka Okamoto, Tomiji Sugmoto, Koji Enomoto, and

Junichi Kikuchi ―Pedestrian Head Impact Conditions Depending on the Vehicle Front Shape and Its Construction - Full Model Simulation‖. http://www.tandfonline.com/doi/pdf/10.1080/ 15389580309856

[8]. EEVC WG17 adult headform FE model

www.arup.com/dynayna [9]. Thomas Frank, Daimler Chrysler AG, Artur Kurz,

LASSO Ingerieurgesellschaft mbH, Martin Pitzer, PENG GmbH, Michael Sollner, Porsche AG, ―Development and Validation of Numerical Pedestrian Impactor Models‖,4th European LS-DYNA Users Conference. Occupant

[10]. Sunan Huanga, Jikuang Yanga, ―Optimization of a

reversible hood for protecting a pedestrian‘s head during car collisions‖ paper No. 42 (2010) 1136–1143http://www.sciencedirect.com/science/article/pii/S000145750900342 X

[11]. Directive of European parliament relating to the

protection of pedestrians, amending directive 70/156/EEC Brussels, 2003

[12]. J.S. Ruan, T. Khalil, A.I. King, Dynamic response of

the human head to impact by three-dimensional finite element analysis, J Biomech Eng 116 (1) (1994) 44-50 http://www.ncbi.nlm.nih.gov/pubmed/8189713

[13]. Zhang, L., Hardy, W., Omori, K., Yang, K.H.,King, A.I.

2001. ―Recent advances in brain injury research: a new model and new experimental data‖, Bioengineering Center, Wayne State University.

[14]. A.M. Nahum, R. Smith, C.C. Ward, ―Intracranial

pressure dynamics during head impact‖, Proceedings of the 21st Stapp Car Crash Conference, Vol. SAE Paper No. 770922, 1977. http://www.mendeley.com/research/intracranial-pressure-dynamics-during-head-impact/

[15]. Al-Bsharat, W.N. Hardy, K.H. Yang, T.B. Khalil, A.I.

King, S. Tashman,Brain/skull relative displacement magnitude due to blunt head impact: new experimental data and model, Proceedings of the 43rd Stapp Car Crash Conference, Vol. Paper No. 99S-86, 1999.

[16]. L. Zhang, K.H. Yang, A.I. King, A proposed injury

threshold for mild traumatic brain injury, J. Biomech. Engrg.-Trans., ASME 126 (2004) 226–236. [45] C. Zhou, T.B. Khalil, A.I. King, A new model comparing impact response of the homogeneous and inhomogeneous human brain, in: Proceedings of the 39th Stapp Car Crash Conference, 1995. SAE paper no. 952714.

[17]. L. Zhang, K.H. Yang, A.I. King, Comparison of brain

responses between frontal and lateral impacts by finite element modeling, J Neurotrauma 18 (1)(2001) 21-30.

[18]. C. Zhou, T. B. Khalil, A. I. King. 1995. ―A new model

comparing impact responses of the homogeneous and inhomogeneous human brain‖. SAE n°952714.

[19]. Yasuhiro Dokko, Jim Manavis, Peter Blumburgs, Jack

McLean, Liyn Zhang, King H.Yang, Albert I.King ―VALIDATION OF THE HUMAN HEAD FE MODEL AGAINST PEDESTRIAN ACCIDENT AND ITSTENTATIVE APPLICATION TO THE EXAMINATION OF THE EXISTING TOLERANCE CURVE‖, Paper No. 322

[20]. Biomedical Engineering, Advanced Human Modeling

Laboratory, Advanced VirtualHuman for Safety Improvement. FE modeling flyer-07 - College of Engineering - Wayne State.

[21]. Horgan, T. J. & Gilchrist, M. D., 2003, ―The creation of

three-dimensional finite element models for simulating head impact biomechanics‖, International Journal of Crashworthiness,inpress.

[22]. 2008 International Workshop on Modelling,

Simulation and Optimization, Lei Gang, Bao

http://www.who.int/choice/publications/RTIDistribution%20Article.pdf

http://www.who.int/choice/publications/RTIDistribution%20Article.pdf

http://www.sciencedirect.com/science/article/pii/0734743X94800336



http://www.autoliv.com/wps/wcm/connect/5f997a004ce4f3f8ac68eef594aebdee/Pedestrian.pdf

http://www.autoliv.com/wps/wcm/connect/5f997a004ce4f3f8ac68eef594aebdee/Pedestrian.pdf

http://www.tandfonline.com/doi/pdf/10.1080/

http://www.arup.com/dynayna

http://www.sciencedirect.com/science/article/pii/S000145750900342

http://www.sciencedirect.com/science/article/pii/S000145750900342

http://www.ncbi.nlm.nih.gov/pubmed/8189713

http://www.mendeley.com/research/intracranial-pressure-dynamics-during-head-impact/

http://www.mendeley.com/research/intracranial-pressure-dynamics-during-head-impact/

http://www.google.com.my/url?sa=t&rct=j&q=(wsuhim)%20wayne%20state%20university%20fe%20head%20injury%20model&source=web&cd=20&ved=0CGsQFjAJOAo&url=http%3A%2F%2Fengineering.wayne.edu%2Fpdf%2Fhumanmodelinglab.pdf&ei=RhZeT6bWD8nhrAfi7e3ODA&usg=AFQjCNF__rG3MThKyQqmrP2hvbrIjbuanw

http://www.google.com.my/url?sa=t&rct=j&q=(wsuhim)%20wayne%20state%20university%20fe%20head%20injury%20model&source=web&cd=20&ved=0CGsQFjAJOAo&url=http%3A%2F%2Fengineering.wayne.edu%2Fpdf%2Fhumanmodelinglab.pdf&ei=RhZeT6bWD8nhrAfi7e3ODA&usg=AFQjCNF__rG3MThKyQqmrP2hvbrIjbuanw


30


Chunyan, Liu Xianjue. ―Finite Element Analysis of Head-Ground Collision‖

[23]. Voo K, Kumaresan S, Printar FA et al. Finite element

models of the human head[J]. Med Biol Eng Comput, 1996, 34(5): 375-381.

[24]. Masami Iwamoto, Yuko Nakahira, Atsutaka Tamura,

Hideyuki Kimpara, Isao Watanabe, Kazuo Miki. ―DEVELOPMENT OF ADVANCED HUMAN MODELS IN THUMS‖ Keynote 7, 6th European LS-DYNA Users‘ Conference

[25]. Galford, J.E., and McElhaney, J.H., "A viscoelastic

study of scalp, brain and dura". Journal of Biomechanics 3, 1970, pp. 211-222.

[26]. Daniel Baumgartner, Daniel Marjoux, Remy Willinger,

Emma Carter, Clive Neal-Sturgess, LuisGuerra, Luis Martinez, Roger Hardy: ―PEDESTRIAN SAFETY ENHANCEMENT USING NUMERICAL METHODS‖ paper2007.

[27]. Chan M., Nahum A.M., 1980, Intracranial pressure: a

brain injury criterion, Proc. of the 24th Stapp Car Crash Conf., SAE Paper n°801304, Ward C.C.

[28]. Willinger R., D. Baumgartner, 2003, ―Human head

tolerance limits to specific injury mechanisms‖, Int. J. of Crashworthiness, vol. 8, n°6, pp. 605-617.

[29]. Trosseille, X., Tarriere, C., Lavaste, F., Guillon, F., and

Domont, A., "Development of a FEM. of the human head according to a specific test protocol", Proc. of 36th Stapp Car Crash Conference, SAE Paper No.922527, 1992.

[30]. Giovanni Belingardi,Giorgio Chiandussi,Ivan

Gaviglio,Politecnico di Torino, Dipartimento di Meccanica. Italy: ―DEVELOPMENT AND VALIDATION OF A NEW FINITE ELEMENT MODEL OF HUMAN HEAD‖. Paper Number 05-0441

[31]. Tamer El Sayed , Alejandro Mota , Fernando

Fraternali , Michael Ortiz,‖ Biomechanics of traumatic brain injury‖ , 2008http://aero.caltech.edu/~ortiz/Pubs/2008/ElsayedMotaFraternaliOrtiz2008_Biomechanics.pdf

[32]. T. El Sayed, Constitutive models for polymers and

soft biological tissues, PhD. Thesis, California Institute of Technology, 2007.

[33]. http://www.med.har-vard.edu/AANLIB/ [34]. P. Thoutireddy, J.F. Molinari, E.A. Repetto, M. Ortiz,

―Tetrahedral compositeFinite elements‖, Int. J. Numer. Methods Engrg. 53 (2002) 1337–1351.http://aero.caltech.edu/~ortiz/Pubs/2002/ThoutireddyEtAl2002.pdf

[35]. L. Zhang, K.H. Yang, A.I. King, Comparison of brain responses between frontal and lateral impact by finite element modeling, J. Neurotraum. 18 (2001) 21–30.

[36]. S. Kleiven, Finite element modeling of the human

head, Ph.D. thesis, Royal Institute of Technology, Stockholm, Sweden, 2002.

[37]. S. Kleiven, H. von Holst, Consequence of head size

following trauma to the human head, J. Biomech. 35 (2002) 153–160.

[38]. K.K. Mendis, R.L. Stalnaker, S.H. Advani, A

constitutive relationship for large deformation finite element modeling of brain tissue, J. Biomech. Engrg. 117 (1995) 279–285.

[39]. T J Horgan, M D Gilchrist, ―Influence of FE model

variability in predicting brain motion and intracranial pressure changes in head impact simulations‖ paper No.0299,2004

[40]. Anderson R.W.G. ; ‗A Study on the Biomechanics of

Axonal Injury‘,2000 [41]. http://digital.library.adelaide.edu.au/dspace/bitstream/

2440/37911/6/01front.pdfPatran r2a, 2002, MSC/Patran User‘s Manual, MSC.

[42]. SHUGAR, T A and KATONA, M G. Development of a

finite element head injury model. ASCE J. Eng. Mech. Div., EM3(E101/173):223–239, 1975.

[43]. Lorenson, B., VTK open source software,

<http://public.kitware.com/VTK/>. [44]. TURQUIER, F, KANG, H S, TROSSEILLE, X,

WILLINGER, R, LAVASTE, F, TARRIERE, C and DOMONT, A. Validation study of a 3d finite element head model against experimental data.Proc. of the 40th Stapp Car Crash Conf., SAE paper no.962431, 1996.

[45]. DIMASI, F, MARCUS, J H and EPPINGER, R H. 3-d

anatomic brain for relating cortical strains to automobile crash loading. In Proc. 13th Int . Tech. Conference. on Experimental Safety Vehicles, Paper No. 91–S8–O–11, 1991.

[46]. RUAN, J S. ―Impact Biomechanics of Head Injury by

Mathematical Modeling‖. PhD thesis, Wayne State University,1994.

[47]. ZHOU, C, KHALIL, T B and KING, A. Viscoelastic

response of the human brain to sagittal and lateral rotational acceleration by finite element analysis. In Proc. of the 1996 International IRCOBI Conf. on the Biomechanics of Impacts, pages 35–48.

[48]. WILLINGER, R, TALEB, L and KOPP, ―C M. Modal

and temporal analysis of head mathematical models‖. J. Neurotrauma, Vol. 12:743–754, 1995.

http://aero.caltech.edu/~ortiz/Pubs/2008/ElsayedMotaFraternaliOrtiz2008_Biomechanics.pdf



http://www.med.har-vard.edu/AANLIB/

http://aero.caltech.edu/~ortiz/Pubs/2002/ThoutireddyEtAl2002.pdf



http://digital.library.adelaide.edu.au/dspace/bitstream/2440/37911/6/01front.pdf

http://digital.library.adelaide.edu.au/dspace/bitstream/2440/37911/6/01front.pdf


31


[49]. Gennarelli T.A., et al.; ‗Axonal Injury in the Optic nerve: A Model Simulating Diffuse Axonal Injury in the Brain‘, J. Neurosug., 71, 244-253, 1989

[50]. Thibault L.E., et al.; ‗The Strain Dependent

Pathophysiological Consequences of Inertial Loading on Central Nervous System Tissue‘, International Conference on the Biomechanics of Impacts, Bron,1990

[51]. Ueno K., et al. ; ‗Development of Tissue Level Brain

Injury Criteria by Finite Element Analysis‘, Journal of Neurotrauma, Vol.12, No.4, 1995

[52]. Eppinger R.H., et al. ; ‗SIMon Theoretical Manual‘,

SIMon Version 1.0 Beta distributed by the USDOT/NHTSA, 2001

[53]. Wikipedia, the free encyclopedia. Coup contrecoup

injury - Wikipedia, the free encyclopedia en.wikipedia.org/wiki/Coup_contrecoup_injury

[54]. Willinger, R., Kang, H.S., Diaw, B.M. 1997.

―Développement et validation d‘un model méchanique de la tête humaine‖.

[55]. Gennarelli T.A., et al.; ‗Axonal Injury in the Optic

nerve: A Model Simulating Diffuse Axonal Injury in the Brain‘, J. Neurosug., 71, 244-253, 1989

[56]. Claessens, M., Sauren, F., Wismans, J.

1997.―Modeling of the human head under impact conditions: a parametric study‖. SAE 973338.

[57]. K. Weinberg, A. Mota, M. Ortiz, A variational

constitutive model porous metal plasticity, Comput. Mech. 37 (2006) 142–152.

[58]. K. Weinberg, M. Ortiz, Shock wave induced damage

in kidney tissue, Comput. Mater. Sci. 32 (2005) 588–593.

[59]. T. El Sayed, A. Mota, F. Fraternali, M. Ortiz, A

variational constitutive model for soft biological tissues, J. Biomech. 41 (7) (2008) 1458–1466.

[60]. Gao Chunping, ―FINITE ELEMENT MODELLING OF

THE HUMAN BRAIN AND APPLICATION IN NEUROSURGICAL‖, PHD Thesis, Department of Mechanical Engineering National University of Singapore.

[61]. TROSSEILLE, X, TARRIERE, C, LAVASTE, F,

GUILLON, F and DOMONT, A. ‗Development of a F.E.M. of the human head according to a specific test protocol‘, Proc. 36th Stapp Car Crash Conference, SAE Paper No. 922527, Society of Automotive Engineers, Warrendale, PA, 1992.

[62]. HORGAN, T J and GILCHRIST, M D. ‗The creation of

three dimensional finite element models for simulating head impact biomechanics‘, International Journal of Crashworthiness, 2003 8 (4) 353–366.

[63]. ABAQUS 5.8. ABAQUS/Standard User‘s Manual.

HKS, 2000.

http://en.wikipedia.org/wiki/Coup_contrecoup_injury

http://en.wikipedia.org/wiki/Coup_contrecoup_injury

INTERNATIONAL JOURNAL OF SCIENTIFIC & … · This research aimed at a review the types of FE head...

Documents

Transcript of INTERNATIONAL JOURNAL OF SCIENTIFIC & … · This research aimed at a review the types of FE head...