Active human body models for virtual occupant response, step 2

Purpose and goal

For the project the overall goal was to develop a mathematical model (finite element) human body model that can predict human motion in a heavy braking vehicle (autonomous braking) prior to crash. The model will be used by the industrial partners Autoliv and Volvo Cars to develop integrated safety systems that can be activated prior to crash. Such systems will improve the protection of the occupants in a car collision and reduce the number of killed and injured car occupants. The project has primarily addressed FFI´s global objective ´Technology is developed with the potential to reduce the number of traffic fatalities by one third as the parliamentary milestone for 2020´ and FFI´s goal to ´Contribute to increase the research and innovation capacity in Sweden and therefore ensure the competitiveness of the automotive industry

Results and expected effects

Within this project, Jonas Östh completed a doctoral degree and Jóna Marín-Ólafsdóttir completed a licentiate degree at Chalmers University of Technology. In total 9 publications in peer review journals. The Beta version of the SAFER A-HBM was further developed to enable simulation of bracing in autonomous and driver voluntary braking; and to simulate postural control for events with a lateral component. To enable this, a number of further developments were identified and implemented. The main results were: 1.To simulate driver interaction with the steering wheel, muscles with active postural control were implemented for the upper extremities. The active muscle models for the spine and upper extremities were merged into the SAFER A-HBM. To simulate driver braking, lower extremity muscles were implemented in the SAFER A-HBM. 2. To provide validation data, a set of volunteer experiments were carried out. 20 volunteers (11 male, 9 female) were subjected to autonomous and driver braking as drivers and passengers, with two seat belt systems in a Volvo V60 driving on rural roads in the Gothenburg area. Prior to testing, the volunteers were instrumented with surface electromyography (EMG) and the maximum voluntary contractions (MVC) were measured. During braking data collected were: vehicle accelerations, occupant kinematics from video data, surface EMG normalized to the MVC, steering column forces, seat indentation, foot-well and braking forces, seat belt forces and payout. The driver response to autonomous braking was significantly different from driver braking. Shoulder belt pre-tension before braking induced muscle activity in the upper extremities, this was most prominent for females. 3.The SAFER A-HBM was validated with regard to the volunteer data for autonomous braking. 4.Anticipatory postural responses were implemented with feed forward control of the active muscle elements and the response was compared to the volunteer data for driver braking. 5.To define a modeling methodology of lateral postural control, muscle recruitment strategies were studied by analyzing the cervical muscle activity of volunteers subjected to perturbations in eight different directions. It was found that the activation patterns varied with direction. Anterior muscles (SCM and STH) were most active during forward (0°) and forward oblique (±45°) perturbations whereas posterior muscles, aside from SPL, were most active during rearward (180°) and rearward oblique (±135°) perturbations. A combination of anterior and posterior muscles was active during lateral (±90°) perturbations. 6.A strategy to implement lateral posture control with closed loop control was defined. A first version of the lateral control tested for the cervical muscle, indicate that multiple controllers are needed at each location.

Approach and implementation

The project was a cooperation between academia and industry. It has contained both mechanical volunteer testing and development of mathematical models, beside development of methodology for active muscle responses. The project included three industrial partners and one academic partner, involving senior engineers and researcher as well as two Ph.D. students. One of the Ph.D. students was mainly involved in planning and carrying out the mechanical volunteer tests with support from the industrial partners Volvo Cars and Autoliv. For the tests, Volvo Cars supplied the vehicle used for testing and supported the installation of the test equipment in the vehicle. Autoliv supplied the restraint systems used in the tests and also supported the installation. In total, 20 Volunteer tests were carried out and analysed. The results were used in the development and validation of the mathematical model. The other Ph.D. student was mainly involved in analysing volunteer tests carried out at ´University of Vancouver´. Both students were involved in the development and validation of the mathematical human body model.