Biomechanical Study Of Musculoskeletal Modelling Of Driver’s Hand

With the development of manufacturing and designing level of the vehiclesnowadays, the topic about exploring and studying the effect of vehicle operating ondriver itself has been widely put forward, which can help analyze and supervise thevehicle design from the views of ergonomics during the earlier stage of vehicle design.The musculoskeletal modeling of human system would provide the kinematicalinformation of people when operating the vehicle, and also the contact force on thejoint and the muscle force spanning along the joint. During the process of operatingand driving vehicles, for example, turning the steering wheel, changing the gear,adjusting the instruments on panel and turning on or off other instruments, the drivemay get muscle fatigue or even injury which is probably caused by the unsuitableoperating ways or forces due to the bad design of the products in vehicles. So I startedthe modeling work on finger, which is the most direct part contacting with the vehicle,and studied the possible biomechanical effect of the driver’s operating on hand byconstructing the finger musculoskeletal model to help vehicle product design andergonomic analysis. I constructed a finger tendon network model based on theprinciple of minimal total potential energy, and I found this method avoided the overstatically indeterminate problem of the traditional modelling method. The complextopological geometry of the finger tendon network was considered, and the nodes inthe network were constrained to move only along the surface of the skeleton. The pathbetween the two nodes in the network was represented by straight line connecting the nodes temporally. To improve the modeling, an automatic muscle wrapping pathsearch algorithm was put forward. This algorithm can quickly find the wrapping pathby using the mesh information on bone surface and the position of the origin andinsertion point of muscle path, where the mesh is usually obtained by medicalscanning of the bone segment. Then the muscle tendon unit length as well as themuscle moment arm can be calculated by the wrapping algorithm. The wrappingmodel was also applied to the complex topology of finger tendon network to preparefor the further biomechanical modeling of finger tendon-skeletal structure. At last Iintegrated the muscle wrapping algorithm into the finger tendon network modellingprocess, and replaced the previous straight line path of tendon component withrealistic wrapping path. The Newton-Raphson method was applied to recalculate thelength change of the tendon component by wrapping algorithm and iterationcomputation would not stop until the total potential energy of the whole finger tendonnetwork system achieves minimal value. Basically, the detailed work andachievements of the study are listed in the following:1. Comparing the simple minimal model of the finger without extensor mechanismand the full model of the finger with extensor mechanism to explore themechanism benefits and biomechanical advantages of the extensor mechanism.The mechanical model of finger musculoskeletal system during pressing wasconstructed using classical multi-body system method for calculating the muscletendon forces as well as the bone to bone contact forces. Six subjects were askedto press on the force platform using their volunteer maximal force with4differentpostures. The measurement data were then put into the two finger models. Afterobtaining the muscle tendon forces and bone to bone contact forces I used thoseresults to do the statistical analysis and concluded that the results from the twomodels are significantly different with each other.2. Constructing the model of finger tendon network based on the principle ofminimal total potential energy, which was the first to apply energy principle to thehuman biomechanical modeling area. An inverse trigonometric function wasdesigned to match the definition of the stiffness of finger tendon, which equaled0 when its length was smaller than slack length. The nodes of the finger tendonnetwork were constrained to move only on the bone surface. Therefore a simpleconstraint which limited the node moving along the line connecting the node itselfand one vertex of the triangle mesh, and the vertex was chosen due to the directionof line closest to the muscle force. To save the computing, the path between thetwo nodes in the network was represented by straight line temporally. The lengthchange and force of each tendon component in the tendon network were obtained,as well as the deformation of the topological structure. The validation was done bycomparing the length change of the tendon components with literatures, and it wasin good agreement.3. Putting forward a new muscle wrapping algorithm. By constructing the equalityequations of the composition of muscle forces along the tangent direction of thecontact mesh line, the contact points’ positions on the mesh line were obtained.Then the muscle wrapping path around the bone surface was formed by thepiecewise points on the mesh lines of the bone surface. After getting the meshinformation of the bone surface by medical CT scanning technique, and if theorigin and insertion positions were known, the wrapping algorithm canautomatically find the wrapping path quickly and precisely. According to thecomparison of the muscle tendon length and the muscle moment arm with thatfrom OpenSim, the algorithm proved to be an efficient and accurate method tocalculate the wrapping path. At the end I also applied the wrapping algorithm tothe complex finger tendon network topology to get preparation for the integrationmethod.4. Integrated the wrapping algorithm with the finger tendon network modeling basedon the principle of minimal total potential energy. The real tendon wrapping pathtook replacement of the previous straight line between two nodes. And theNewton-Raphson method was applied to calculate the equality equations. Duringthe computing process, the wrapping path was recalculated every time when theNewton-Raphson method did once iteration and the iteration would not stop untilthe total potential energy of the whole finger tendon network system achieved minimal value. It is the first time to induce the real bone geometry into themusculoskeletal modelling of human body for the purpose of revealing theanatomy characteristic, which is more precise than traditional method and fasterthan finite element method. Due to the difficulty of measuring the muscle force invivo, I only validated the results of tendon length with the literature, where thedata came from the cadaver tests.5. Taking measurement of the finger pressing on the force plate for several subjectsusing Vicon system. After learning how to use the Vicon system, I collected themeasurement data from6subjects. The data was then processed in Matlabsoftware: the gap was filled in and the data were filtered by low-pass filter. Onlythe data in the middle of the trials were used when the subject had reached asteady isometric pressing condition.Besides the finger, the other segments of the human body are also presenting anetwork state. The method based on the energy can not only be applied to the fingermodeling, it is suitable for any other segment. During the everyday operation, forexample, turning the steering wheel, changing the gear and adjusting the instrumentson panel, the muscle and joint contact forces can be obtained based on the methodproposed in this thesis, to predict the potential fatigue and injury on driver’s hand.