Modelling physiological features of Human body behavior in car crash simulations

Transcription

1 Modelling physiological features of Human body behavior in car crash simulations Michel BEHR, Yves GODIO, Maxime LLARI, Christian BRUNET Laboratoire de Biomécanique Appliquée (Marseille, France) Presented by Yves GODIO ABSTRACT Human numerical models are widely used to investigate injury mechanisms involved in car crash configurations. One limitation of these models is linked to the time dependency of biological tissues mechanical properties, as a result of various physiological modifications. To answer this limitation, we present one possible approach to evaluate the influence of internal body pressures (mainly resulting from breathing) on the result of a frontal impact, by modelling main muscles responsible for respiration. A finite element model of a full human body in driving position was derived from the Radioss Humos model. Modifications first included integration and interfacing of a user-controlled respiratory muscular system (intercostals, scalene, sterno-cleido-mastoidius and diaphragm muscles). A volumetric measurement procedure for lungs and both thoracic and abdominal chambers was also implemented using the Radioss Monitored Volume tool. A frontal impact at an initial speed of 50km/h was simulated, for two different stages of respiration (at rest and after a 2l inspiration). Internal abdominal pressure (IAP) in both cases was compared. The result of the IAP influence on the diaphragm cupolas maximal strain during both simulations was recorded and compared in order to determine a qualitative risk indicator. The Humos model, initially dedicated to road safety research, could be turned to a rather promising 3D model of respiration with only minor modifications. Our approach also underlined the need to implement in human models physiological behaviors such as muscle contraction when investigating injury mechanisms and risks. Laboratoire de Biomécanique Appliquée ( Faculté de Médecine nord Bd Dramard Marseille, France Tel: +33 (0) michel.behr@inrets.fr 1

2 Introduction General context: Passive road safety Realistic experimental protocols Injury mechanisms validation Accidentology Input data for numerical simulations Injury mechanisms validation Experimentation Protocol set up Input data for simulations Models validation Numerical simulation Towards virtual traumatology Our approach is coupling real accident data and experimental tests with PMHS to develop biofidel Human body numerical models Can physiological events (such as muscular contraction) alter injury mechanisms? 2

3 Introduction Steps of investigation of current study: 1. Contracting muscle modelling 2. Development of a respiration module in a FE human body model 3. Respiration and frontal impact simulation 3

4 Problem definition Contracting muscle modelling The muscle bundle: Muscle body architecture : 3D bricks merged with 1D action lines (series of small-sized contraction units) Each contraction unit is composed of a passive component (relaxed muscle) and an active component (bracing) in parallel a p a a p a p 4 a p p a: active element p: passive element

5 Problem definition Contracting muscle modelling Contraction formulation (classical muscle mechanics) δ CP Passive Component Non linear elastic behavior (in traction) f p ( δ ) = α( β. δ )² CA δ CP. δ r c A 0 Ta t Active Component The generated force depends on instantaneous muscle length and strain rate f 1 Optimal length f p δ 5

6 Problem definition Modification of the Radioss HUMOS model 1. Monitored volumes definition: Whole Abdomen (PIA) stomach (gastric PIA) lungs (air transfer) 2. implementation of active muscles Diaphragm Intercostals Scalenes Sterno Cleido Mastoidius Abdominal wall HUMOS model Initially defined at rest (Functional Residual Capacity - FRC) 6

7 Problem definition Modification of the Radioss HUMOS model scalenes SCM Muscle contraction : 2 approaches 1D contractile cells for Diaphragm, scalenes, SCM and Intercostals Bulk Modulus over-estimation for abdominal wall (*3, according to Gennisson et al, 2005) Intercostals Muscles Number of springs Diaphragm (2D contractile) 200 Intercostals (1D inferior and superior) Internals 236 Externals 236 Scalenes (1D) 28 Sterno Cleido Mastoidius (1D) muscular activations Diaphragm (Behr et al, 2006) 7

8 Problem definition Contraction parameters Model of respiratory muscle contraction unit (CU) F max = PCSA * F s where F is maximal voluntary contraction force PCSA : Physiological Cross Section Area of considered contraction unit F s # 25 N/cm² (Herzog,1999) RESULTS: Pretension function Diaphragm Fmax = 38 N per CU Sterno-Cleido-Mastoidius and Scalenes Fmax = 100 N Intercostal internal and external sup. Fmax = 42.5 N per CU Force (N) Fmax *Fmax 1 2*Fmax Intercostal internal and external inf. Fmax = 34 N per CU Relative elongation (%) 8

9 Analysis Contracting muscle validation Verifying the contraction unit behavior for a whole muscle bundle Traction/compression cycles (light blue curve on figures), up to 50% of deformation. Boundary conditions : proximal extremity fixed, imposed velocity (1ms -1 ) on other extremity Passive force (N) Resulting force (on tendons) without contraction Theoretical Simulation Deformation (%) Active force (N) 1000 Resulting force (on tendons) with maximal contraction Theoretical Simulation Deformation (%) % Physiological window 40% 20% 400 0% 400 0% % % time (ms) time (ms) 9

10 Analysis Validation of a respiratory cycle by contracting muscles dv=2litres dh=4.0cm At rest t=0ms Full inspiration t=300ms From rest (FRC) to Full inspiration reached in 300ms led to : - 2 litres lungs volume increase (physiological Total Lung Capacity is 3.5litres) - ~15 rib angle variation in sagittal plane for rib 1 (in agreement with Ratnovsky et al, 2005) - 4cm vertical deflection of the diaphragm cupola (3.1cm according to Blaney et al, 1997) 10

11 Analysis Frontal impact simulation 2 Frontal impact conditions : 50 Km / h At rest and at full inspiration (~2litres) Respiratory cycle from 0 to 300ms After 300 ms, deceleration of the sled 11

12 Analysis Relative evolution of the Intra-Abdominal Pressure (IAP) Intra Abdominal Pressure Intra Abdominal Pressure 25 impact impact At Functional Residual Capacity At Total Lung Capacity dp (kpa) Sled Velocity inspiration Velocity (m/s) dp (kpa) At Functional Residual Capacity At Total Lung Capacity Sled Velocity Velocity (m/s) time (ms) time (ms) The same frontal impact at two stages of respiration led to : - 4kPa IAP offset at full inspiration (coherent with Hodges et al results) - during impact, IAP increase is up to 30%higher after 2l inspiration at full inspiration, the maximal strain level (injury criteria) of the diaphragm cupola reaches 8% (against 1% at rest) 12

13 Discussion, Conclusion Internal body pressures: the diaphragm contraction induces an initial pressure offset and increases the maximal strain level during a frontal impact (injury risk possibly increased) Possible improvements : tuning of the respiration module evaluation of realistic IAP injury criteria (experimentally) Physiological events, such as muscular contraction, should be considered in biomechanics of impact (at least up to a reasonable level of impact energy, still to be determined) 13

14 Acknowledgements & References We would like to acknowledge Professor Yves Jammes expertise for the evaluation of the respiration module J.L. Gennisson et al., Human muscle hardness assessment during incremental isometric contraction using transient elastography, J Biomech 38 (2005) W. Herzog, Biomechanics of the musculo-skeletal system, 2nd édition, Wiley ed. Ratnovsky et al., Anatomical model of the trunk for analysis of respiratory muscles mechanics, Resp Physiol 148 (2005), Blaney et al., Sonographic measurement of diaphragmatic displacement during tidal breathing manoeuvres, Physiother Theor Pract 13 (1997), Hodges et al., Intra abdominal pressure increases stiffness of the lumbar spine, J Biomech 38 (2005), Behr et al., 3D reconstruction of the Diaphragm for virtual traumatology, Surgical Radiologic Anatomy 28 (2006),