1. Three sets of test results on cadaver knee have been reported in literature. The legforms, dummies and mathematical models do not validate under all these three test conditions. A computational model such as THUMS can be expected to validate under multiple boundary conditions. This work involves validation of the FE model of knee obtained from THUMS against all these three test conditions.

The mesh in THUMS has been modified to obtain the FE model corresponding to that of a pedestrian. Validation under the test conditions used by Kajzer et al (1999) and shearing test conditions used by Kajzer et al (1997) has been achieved. The FE model also validates against the four point bending test condition used by Kerrigan et al (2003). The work also focused on the injury process of knee. The injuries in simulations for all test conditions agreed well with the respective autopsy results. It was also found that scaled FE models of knee do not validate against the test conditions.

Super imposed view of knee

Figure 1: Knee shear and bending test set up (Kajzer et al., 1997, 1999) modeled for simulations

Figure 2 : Bending test set up (Kajzer et al., 1999)

2. Finite Element simulation of a lower extremity model is used to (1) determine which of the muscle parameters maximum force capacity (Fmax), initial activation levels (Na) and maximum muscle contraction velocity (Vmax) affect ligament strains the most and (2) to identify which muscles affect the knee response the most in low speed, just below the knee, lateral impact. Simulations have been performed with Fmax, Na and Vmax varying from their reference values. Sensitivity of ligament strains to variation in muscle parameters has been studied. It is observed that knee response is more sensitive to Fmax and Na than Vmax. Amongst the muscles varied, reduction in the Fmax and the Na in the hamstring and the gastrocnemius muscles affects the knee ligament strains the most. The hamstring parameters significantly affects the ACL, the PCL as well as the MCL strains whereas, change in the gastrocnemius parameters affects only the MCL strain.

                                                            

   Figure 3 : Simulation set up used in the present study (Soni A et al., 2009)

    

3. A full body pedestrian model with active muscles has been developed. FE simulations have been conducted using the full body model and front structures of a car. Two preimpact conditions, that of a symmetrically standing pedestrian representing a cadaver and an unaware pedestrian have been simulated. Stretch based reflexive action was included in the simulations for an unaware pedestrian. Results show that due to muscle contraction (1) peak strains in all the knee ligaments reduce (2) VonMises stresses in tibia and fibula increase to their ultimate stress limits and are predicted to fail and (3) knee joint effective stiffness increases by 67% in lateral bending.

                                                       

Figure 4 : Simulation set up used in the present study showing the PMALE and FE model of the car front (Soni A et al., 2008)

4. Unlike car occupants, pedestrian crashes occur in a variety of postures (like stationary, walking, running or jogging etc). Muscle contraction, required to maintain the initial posture, modifies the load at the knee joint in rapid loading conditions. This study investigates the effect of muscle active forces on lower extremity injuries for various impact locations and impact angles for a freely standing pedestrian. Three different pre-impact conditions of a freely standing pedestrian, representing a cadaver, an unaware and an aware braced pedestrian, have been simulated for each impact orientation. Stretch based reflexive action was included in the simulations for an unaware pedestrian. It is concluded that strain in knee ligaments is dependant on impact locations and angles and the MCL is the most vulnerable ligament. Further, due to muscle effects, except when the impact is on the knee, peak strain values in all the ligaments are lower for an unaware pedestrian than either for a cadaver or for a fully braced pedestrian.

Figure 5 : Simulation set up to study the effects of muscle forces in a freely standing pedestrian for five impact locations(left) and four impact angles at below knee level (right)   (Chawla A et al.,2007)

5. Three variables, viz. height of impact, pedestrian offset with respect to car centre and impact speed, have been considered here. Full scale car pedestrian FE simulations have been performed using the full body pedestrian model with active lower extremities (PMALE) and front structures of a car model. Two pre-impact conditions of a symmetrically standing pedestrian, representing a cadaver and an unaware pedestrian have been simulated. It is concluded that (1) with muscle contraction risk of ligament failure decreases whereas risk of bone fracture increases (2) ligament and bone strains are dependent on the impact location (3) chances of ligament injuries are higher when the impact occurs near the outer corner of the car (4) risk of bone fracture increases with speed (5) bone fracture reduces the risk of ligament failure.

    

Figure 6 :  Simulation setup to study the effects of muscle forces in a symmetrically standing pedestrian for (a) impact at four heights (b) four standing positions in front of car and (c) five impact speeds

CLICK HERE TO VIEW THE ANIMATION