FR2839636A1 - Animal body biomechanical modelling system for crash test dummy validation measures frequency response to on forehead impulse and derives two mode modal model - Google Patents

Abstract

An animal body biomechanical modelling system measures the acceleration response at the rear (2) of a volunteers head to a an impulse (4) excitation on the forehead (1), calculates the frequency response and produces a two or more mode modal analysis (5) based analytic model over the range up to 20 Hz for use in crash test simulations and model validation.

Description

The invention relates to a method for biomechanical modeling of a part

  of the body of an animal, an application of such a model to the simulation of the behavior of said body part in a crash test, a physical or numerical manikin whose biomechanical behavior is s agency to be similar to that of such a model (the model or manikin being then referred to as "biofidele") as well as an application of such a model to the validation and calibration of a part of a physical or numerical manikin. o In various fields, including the improvement of the safety of motor vehicles, attempts have been made to develop models for reproducing the behavior of an animal, particularly a human, in a variety of ways. accident situations. One of the goals of this modeling is to allow car manufacturers to optimize the safety of occupants in such situations. In particular, rear shock simulation is used to optimize

  the safety provided by the seats and the headrests of a motor vehicle.

  To do this, we use classically physical or digital models that vent agencies to have a similar behavior to that of o animal considered. Thus, in pla, cant this type of dummy in the vehicle during a crash test type test and measuring the stresses it undergoes, one can foresee the possible traumatisms that can be caused on the part of the

body of modeled animal.

  Problems that arise are those of the modeling parameter eVou of the calibration of manikins, so as to obtain modeled cinematics which

  are comparable to those actually observed in the event of an accident.

  To be as realistic as possible, the setting should be done with data, such as stiffness, damping and parameters.

  inertial, which are specific to the part of the body considered.

  The current methods for evaluating these data are based on the temporal response measured on the living in all the necessary conditions, and the models based on these methods are used to reproduce the part of the

  body in all its static and dynamic complexity.

  But, for reasons of ethics and limits of physical tolerance, one can not

  s not submit volunteers to tests that could cause injuries.

  Therefore, we can use cadavers to set the models.

  But, the models based on these models have an almost cadaverical behavior that is not very representative of that of the living, mainly

  o during the simulation of violent shocks.

  Indeed, the parameters useful in the models of the prior art are largely a function of the action of the muscles, ligaments, as well as the interaction between all the organs forming the part of the body under consideration.

  s according to the different possible anthropometric characteristics.

  As a result, the parameters investigated in the context of temporal analysis are largely inaccessible, which often leads to the use in the models of hypothetical, average or empirical parameters which, in

  in most cases, only access to global behaviors.

  Therefore, the behavior of the models based on these models is not entirely satisfactory, in particular, they do not allow to predict

  certain phenomena observed for example in rapid cinematography.

  To remedy these drawbacks, the invention proposes a method for biomechanical modeling of a body part of an animal that is performed in vivo in the frequency domain by means of a modal analysis, so as to develop a biofidel model. of minimal complexity capable of taking

  o Count the influence of muscular action.

  For this purpose, and according to a first aspect, the invention proposes a method for biomechanical modeling of a part of the body of an animal, said method comprising steps providing for: - subjecting the part of the body to dynamic impulse demand or vibratory; - measure the response induced by said solicitation; - calculate the frequency transfer function between solicitation and response; - perform a modal analysis of the biological structure of the body part; analytically model the biomechanical behavior of the part of the

  body on the basis of said modal analysis.

  According to a second aspect, the invention proposes an application of such a model to the simulation of the behavior of said body part in a type test.

crash test.

  According to a third aspect, the invention proposes a physical or numerical manikin whose biomechanical behavior is designed to be analogous.

  to that of the analytical model obtained a method as defined above.

  According to a fourth aspect, the invention proposes an application of the method defined above, to the validation and / or calibration of a portion of a physical or numerical manikin, in which the biomechanical behavior of the manikin is compared to that of an analytical model obtained by setting the process described above on said part of the manikin. Other objects and advantages of the invention will appear during the course of

  description which follows, with reference to the accompanying drawings in which:

  FIG. 1 is a schematic view of an experimental device for the implementation of the method of biomechanical modeling of the head-neck assembly of a volunteer; FIG. 2 is a Bode diagram of the transfer function obtained by implementing the method with the device of FIG. 1; FIG. 3 represents a damping spring mass system with two degrees of freedom; FIG. 4 is the Bode diagram of the transfer function calculated for a system represented in FIG. 3; Figure 5 schematically shows the correspondence between the head-neck assembly and the inverse dual-pendulum model used; FIG. 6 represents the variation of the modulus of the transfer function computed at point S of the model represented in FIG. 5; FIG. 7 represents the imaginary part of the dynamic stiffnesses calculated respectively at the ON, OH and S points of the model of FIG. 5; FIGS. 8a and 8b respectively represent the modal deformations

  correspond to the two modes of resonance identified in Figure 7.

  According to the invention, a modal frequency analysis is used so as to know the data that can be used in a behavior model of a

  body part of an animal, including the frequencies of resonance,

  resonance, modal stiffness, modal damping, modal deformities. For this purpose, the part to be modeled is excited at a low level of energy in

  a frequency band adapted, impulse or vibratory.

  Thus, it is possible to realize the modeling in vivo without risking to generate

  so the slightest lesion to the volunteers on whom the data is collected.

  Moreover, the modal analysis is particularly adapted in the case of species because it allows to determine the properties inherent to the part of the body considered by studying its response to a known or measurable solicitation that will excite its own modes of vibration that can be mechanisms

of lesions.

  According to the invention, it is possible, for example, to apply a pulse to the part of the body to be modeled and to measure the acceleration induced by the said pulse. By an adequate mathematical treatment, it is then possible to obtain the function of

  frequency transfer between the pulse and acceleration.

  Under the effect of the impulse, the part of the body is excited over a whole range o of frequencies which depends directly on the nature of the stress. Depending on the nature of the part to be modeled and in particular its quasi static stiffness, we will choose a frequency band and therefore a solicitation

  adapted to perform the desired analysis.

  In relation with FIG. 1, an experimental device is described

  to obtain the transfer function of the whole head-neck of a human.

  In this case, the frequency band of analysis can be typically between 0 and 20 Hz, the energy of frequencies higher than 20 Hz being in the

  Practical little useful for the researched modeling.

  To obtain this frequency range, the shock must be rather soft type with a low transmitted energy, that is to say, unlikely to generate

  injuries to the volunteer on whom the data is collected.

  However, this methodology is transposable by the person skilled in the art to others

  body parts of an animal to model.

  To obtain the transfer function, the volunteer is equipped on the front with a force sensor 1 and on the rear part of the head of an accelerometer 2. A lever arm 3 is associated at one end with a fixed structure The other extremity which is movable in rotation being provided with an impactor 4. When the lever arm 3 is actuated, the impactor 4 is arranged to apply a pulse on the force sensor 1. In the analysis considered, the shock to be rather soft type, the impactor 4 eVou the force sensor 1 can be

  equipped with a rubber tip that plays the role of low-pass filter.

  The device further comprises an acquisition station 5 input signals, output, said station for calculating the transfer function between these

two signals.

  In the specific example, during the acquisition, the volunteer is seated in a chair. The lever arm 3 is then adjusted, in mass and in length, so as to apply the desired pulse to the force sensor 1 whose volunteer is equipped, said pulse being applied from front to back, substantially in the

  sagittal plane and at the forehead.

  It should be noted that, due to head rotation after impact, the normal component of force and acceleration does not carry the total values of applied force and acceleration. However the maximum deflection of the head never exceeds a degree, which allows to neglect this effect. This is why, in the context of the analysis presented, the recorded force and acceleration are assimilated to a force and a linear acceleration according to X. FIG. 2 is a Bode diagram of the transfer function obtained with the device. presented above, said diagram showing, as a function of frequency, respectively the module or apparent mass and the phase of the

transfer function.

  From the analysis of this diagram, it appears that: - at low frequencies, the apparent mass drops regularly until

  present a minimum towards 1.4 Hz, which clearly evokes a resonance.

  Likewise, the phase crott following the progression of the apparent mass, from -180 to -32 passing to -90 at the frequency of 1.4 Hz; whereas the phase has a minimum of 3.1 Hz, the apparent mass

  present a maximum at this frequency. This behavior evokes an anti-

  resonance of the head-neck assembly; at around 5 Hz, the module curve again presents a minimum for a phase of -16, corresponding to a point of inflection and therefore to a second resonance; - above 7 Hz, the two curves stabilize at a value of 3.8 kg

  o for the apparent mass and -2.8 for the phase.

  The modal analysis thus shows a behavior stiffness at the low frequencies, a mass behavior with the high frequencies as well as two pie of resonances respectively with 1.4 Hz and 5 Hz and an anti-resonance with 3.1 s Hz. The answer of [ head-to-neck assembly due to impulse force demand thus gives a typical transfer function of a model a

  localized parameters has two degrees of freedom.

  The biomechanical model of this set can therefore be represented by a damper spring mass system with a two degree of freedom as represented in FIG. 3. Indeed, such a system presents the modeled Bode diagram represented in FIG. 4, which is much similar to

  s experimental Bode diagram (Figure 2).

  Thanks to the invention, we have thus obtained a biomechanical modeling of the head-neck assembly which is particularly simple, said modeling can thus easily be used to simulate the behavior of the whole assembly.

so a crash test.

  For this purpose, we chose a model with double inverse pendulum whose segments

  rigid suppositions represent respectively the neck and the head (see figure 5).

  As defined in FIG. 5, the pendulum has the center ON and OH, whose inertial parameters JN and JH and the mass mN and mH (at the centers of gravity GN and GH) as well as the parameters of the spring shock absorber systems (kN, cN) and (kH,

  CH) implants at each center of rotation, define the model.

  In general, the module of the transfer function in terms of mass

  Apparently calculated at the S end of this system is shown in Figure 6.

  But the determination of the system implies that of the following parameters: o - Mass of the party: Me; - mass of the neck: Mcou; - inertia of the head along the Y axis: JH; - inertia of the neck along the Y axis: JN; distance between the center of gravity of the head and CO: aH; distance between the center of gravity of the neck and C7: aN; distance between CO and the position of the sensors: LH;

  - neck length between C7 and CO: LN.

  With regard to the geometric parameters of the neck and the fete, they are easily obtained for example by radiographic, and one places 0NOH = LN,

OHS = LH, ONGN = aN, OHGH = aH.

  Inertias, masses as well as the positions of the centers of gravity are more difficult to obtain directly on the living, but wind deductible of studies

known anthropometrics.

  Starting from these parameters, it is then possible to optimize, for example by a test of the% 2, the transfer function modeled with respect to the transfer function.

  experimental so as to deduce the values of (kN, cN) and (kH, ch).

  In a particular example, for a male volunteer, having the following parameters: - Me = 4.7 kg; MCOU = 1.7 kg; - JH = 0.0209 kg. m2; - JN = 0.0038 kg.m2; - aH = 0.059 m; - aN = 0.065 m; s - LH = 0.08m; LN = 0.13M; The following stiffness and damping parameters are obtained: - 14.765 <kN <18.240 (in Nm / rad); to - 9.684 <kH <14.900 (in Nm / rad); 0.394 <cN <0.506 (in Nms / rad);

- 0 <cH <0.057 (in Nms / rad).

  And the resonance values respectively at 1.4 Hz and 5.9 Hz, as well as anti

resonance at 4.5 Hz.

  We therefore note that the damping at the level of the head-to-head connection is very

  falble, see nonexistent, unlike the amortization between the neck and the torso.

  The stiffness winds are quite close to each other, that of the first pivot seeming to be

show slightly higher.

  To validate the model obtained, tests were conducted on several staff to assess the influence of the anthropomorphic characteristics of

  the individual on the transfer function.

  These tests showed a great similarity in the transfer functions, especially with regard to the frequencies of resonance and anti resonance. This shows that the frequency behavior is substantially independent of the morphology of the individual, which makes the model particularly interesting for use in simulating

  crash test type because it is representative of a large population.

  The tests mentioned above were carried out on volunteers whose eyes were firm during the shock in order to limit the influence of the muscles in the measurements. However, open-eye tests and contracted neck muscles were also conducted. The transfer functions obtained allowed a very fine analysis of the influence of muscular contraction on the dynamics of the system. Modal analysis also makes it possible to characterize the different modes of vibration of a structure subjected to a given excitation. In the case of a living structure such as the head-neck assembly, we can deduce the modal deformities respectively at the point ON, OH and S carry the imaginary part of it.

  o the transfer function at each of these points (Figure 7).

  Thanks to these modal deformities one can better understand the mechanisms of lesions and their conditions of appearance. In particular, thanks to the knowledge of the modal deformities, one can optimize the systems of

  protection of the occupant of a vehicle, including seats and / or

  head, so as to filter out the dangerous frequencies thus identified.

  Concerning the head-neck assembly, it is deduced that the resonance at 1.4 Hz corresponds to a neck extension (FIG. 8a) whereas the resonance at around 5 Hz

  o corresponds to an S-shaped deformity of the cervical spine (Figure 8b).

  The method according to the invention thus makes it possible to identify the modes of vibration of a biological structure, to deduce stiffnesses and dampings which can be applied to models of minimum complexity so that

  s realistically simulate the behavior of the structure.

  In addition, this methodology is also particularly suitable for validation and calibration of physical or digital manikins. For the former, the experimental modal analysis technique described above is used. For the latter, a numerical modal analysis is carried out in order to extract the eigenvalues expressing the resonance frequencies and the associated eigenvectors which express the modal deformations. So, we

  obtains as many eVou calibration and objective calibration parameters.

Claims (9)

  A method of biomechanically modeling a body part of an animal, said method comprising the steps of: - subjecting the body part to dynamic impulse or vibratory loading; measure the response induced by said solicitation; - calculate the frequency transfer function between solicitation and response; o - perform a modal analysis of the biological structure of the body part; - to analytically model the biomechanical behavior of the part of the   body on the basis of said modal analysis.   s 2. Method according to claim 1, characterized in that the body part   is a head-and-neck set of a human.   3. Method according to claim 2, characterized in that the solicitation is applied on the front of a volunteer, the answer being measured on the part back of the party.   4. Method according to claim 2 or 3, characterized in that the applied bias is such that the frequency band of the modal analysis is between 0 and 20 Hz.   5. Method according to any one of claims 1 to 4, characterized in that   that only the first two modes of frequency response are used for modeling.   6. Method according to any one of claims 1 to 5, characterized in that   that the measured response corresponds to the acceleration induced by the solicitation. 7. Application of a biomechanical model of a body part of an animal obtained in accordance with the method according to any one of   1 to 6, to the simulation of the behavior of said part of   body in a crash test type test.   8. Physical or numerical manikin whose biomechanical behavior is designed to be analogous to that of the analytical model obtained by the   Method according to one of Claims 1 to 6.   o 9. Application of the method according to any one of claims 1 to 6, to the   validation eVou to the calibration of a part of a physical or numerical manikin, in which the biomechanical behavior of the manikin is compared to that of an analytical model obtained by the implementation of the   method according to any one of claims 1 to 6 on said part of