Head Impact against Rigidwall and Bonnet

HEAD IMPACT SIMULATION OVER A RIGIDWALL AND CAR HOOD

INTRODUCTION:

A subpart of a Crash analysis performed in major industries is pedestrian safety and protection. In this Vehicle safety engineering field, the engineers are required to design the automobile especially the front half i.e the bonnet and car hood in such a way that under circumstances which involves accident of collision with pedestrians the vehicle has to impart less damage to the civilian as much as possible.

While it is proven from studies that during collisions with pedestrians the major injury is imparted on the head region with minor injuries on the torso and thigh region. Crash safety engineers consider the head which is more prone to injury that could lead to concussions or even death to be their main priority during crash analysis.

The factor that is used by engineers to determine the level of damage imparted onto the head region is determined using the HIC factor which stands for Head Injury Criterion / Coefficient.

`HIC = max([1/(t1-t2)inta(t).dt]^2.5.(t2-t1))`

where `a(t)` -> average deceleration over a certain period of time.

The exponent 2.5 is widely preferred for head injury.

OBJECTIVE:

The objective of this project is to perform a head impact simulation of a headform meshed geometry against a rigid wall and a hood of a car and comparing the resultant stress and strain developed on either body including the HIC36 factor of the simulation.

The simulation

CASE 1: SIMPLE HEAD MODEL AGAINST RIGIDWALL

Since creating an entire keyword file for head impact simulation with the headform and hood can be intimidating, at first we will try to replicate the same simulation using a simple sphere shaped shell element against a rigid wall.

1. Create a *SECTION_SHELL and a *MAT_24_PIECEWISE_LINEAR_PLASTICITY with a shell element formulation 2 and a shell thickness of 5 mm.
2. A RIGIDWALL_PLANAR is created from the CreEnt -> Rigidwall -> Planar Option normal to z-axis and about a distance of 10 mm from the bottom node of the shell element.
3. SET_NODES are created containing every node of the headshell part and an INITIAL_VELOCITY card containing the velocity rate of 40kmph which is about 11.11 mm/ms and along a direction of -ve Z-axis towards the rigid wall. NOTE: Since our head part consists of a sphere, rotation about an axis will not be used. However, in the case of real headform geometry, the pedestrian head will be aligned so as the front part (face) is set for an impact angle of about 35 to 75 degrees.
4. During impact, if huge deformation occurs it could potentially lead to contact of elements with the adjacent elements in the same shell component. Hence, to impart a self-contact among each element of the part a CONTACT_AUTOMATIC_SINGLE_SURFACE card is used. The only parameter that needs to be initialized in this keyword is the slave set ID, which in our case will be the entire shell part.
5. A *CONTROL_TERMINATION keyword is specified to initialize the end time of the simulation which is about 5 ms.
6. In addition, the CONTROL_ENERGY card is also initialized to compute the hourglass and stonewall (rigid wall) energies during the analysis.
7. Finally, the DATABASE keyword is created specifying the output parameters that need to be computed during the simulation including the animation d3plot with a time interval of 0.1 ms and ASCII_Outputs like GLSTAT- to calculate energies, NODOUT- for nodal displacement/acceleration and HIC value and the RWFORCE - rigid wall force. The required nodes from which the results need to be extracted are specified in a SET_NODES and reference in a *DATABASE_HISTORY_NODE_SET card.
8. After submitting the simulation, the post-processing is done. While the ASCII_Output parameters will be clearly analyzed for the real head model, the Von-mises stress experienced by the bodies during impact is shown below,

CASE 2. HEADFORM AGAINST A RIGIDWALL

Unlike the previous case, where the head part was a complete sphere, this model involves translation and rotation of the part hence creating an impact angle of 45 degrees with the rigid wall and the bonnet. This can be done using the *INCLUDE_TRANSFORM and DEFINE_TRANSFORMATION keywords in a separate keyword file (.k or .dyn) and referenced to the headform model using the above-mentioned keywords such that any changes made don't directly affect the headform model which can act as a basis for all pedestrian safety analysis.

1. DEFINE_TRANSFORMATION

This keyword has to be initialized before the INCLUDE_TRANSFORM to prevent errors. This keyword lets us perform certain transformations like SCALE, ROTATE, etc to the part file reference in the INCLUDE keyword. For our model, we will rotate the model about the x-axis at an angle of 45 degrees. This can be achieved by specifying the a1-a3 parameters which define the axis of rotation and a4-a6 the point through which the vector passes through and finally a7 where the angle of rotation in degrees can be specified.

2. INCLUDE_TRANSFORM

This keyword is a modified form of the regular *INCLUDE_INCLUDE keyword file, that it helps us make changes/ transform the imported model prior to creating it on the viewport. Certain offsets of IDs can also be changed however since our model consists of very few parts we will not use the offset command.

3. PART - SECTION & MAT

The headform part which has to replicate the behavior of a real pedestrian head is modeled using a SECTION_SOLID. Since this headform part constitutes of different sub-parts, each sub-parts are modeled using different materials with the Skull, Accel block, Backplate modeled using a MAT_20_RIGID material, and the Skin with a MAT_77_O-OGDEN_RUBBER which is a type of hyperelastic material.

The rest of the keywords - *CONTACT_AUTOMATIC_SINGLE_SURFACE, *CONTROL_TERMINATION & ENERGY, DATABASE_ASCII_Output & BINARY_D3plot & HISTORY_NODE_SET, *RIGIDWALL_PLANAR are created with the same parametric values as used for the Simple head model.

POST-PROCESSING

The Von-mises stress experienced by the pedestrian head model is shown below,

The ASCII_Outputs referenced in the DATABASE card namely the GLSTAT and NODOUT results are shown below,

From the above graph, we can clearly understand the energies developed during the impact. From the d3plot we can visualize that the impact occurs at a time step of 1 ms and from hereon the stress experienced by the headform gradually increases which leads to a gradual increase in the internal energy developed on the model and simultaneous drop in kinetic energy of the model.

Similarly, we can see a gradual increase in the hourglass energies developed which changes non-linearly after the point of impact. However, the stonewall energy experienced in the rigid wall is relatively low, the reason being it has more stiffness and the impact of headform doesn't cause stress occurrence at the wall. This is a valid result since the same would occur in real life if the pedestrian head were to collide with the bonnet of a car i.e the head would acquire more damage than the bonnet.

The resultant nodal acceleration is shown below,

The HIC coefficient plotted along with the resultant acceleration for a selected node in the headform is shown below,

This is a relatively low HIC value considering the impact was calculated from a relatively low distance of about 10 -15mm from the rigid wall. The above HIC value represents a very minor head injury which according to the AIS (Abbreviated Injury Scale), there is a high probability of migraine and small concussions but with very little probability of severe skull fractures.

However, in real cases, the damage could be significantly high considering the acceleration factor as it increases at a sudden rate when calculated from a significant distance of the head from the bonnet of the car.

This modification will be made while simulating the same with a bonnet in the next case

CASE 3: HEADFORM AGAINST A BONNET

This case involves the same simulation but the impact occurs on a bonnet similar to real-life pedestrian accidents. From the results mainly the HIC36 value obtained from the previous simulation, few minor changes have been made to the part so as to replicate a real pedestrian accident. This is achieved by taking into consideration of the civilian distance from the bonnet and the velocity rate at which it occurs while maintaining the velocity at 40 kmph.

The regular keyword changes are,

1. PART - SECTION & MAT (HOOD)

The hood in order to replicate that of a real-life scenario is not modeled as a rigid wall since the stiffness of a rigid wall is significantly high to experience a considerable deformation under high rate impact circumstances. Hence the hood is modeled using an Aluminium material using the MAT_ELASTIC keyword.

2. CONTACT_AUTOMATIC_SURFACE_TO_SURFACE

Unlike previous cases involving rigid walls, where the contact between any part model and the created rigid walls are pre-existing hence, while considering the contact between any other models excluding a rigid wall has to be initialized. For successful contact in an impact test, the models are compiled onto a SET_SEGM since specifying segments during contact are more appreciable than parts/node sets.

Thus the create SET_SEGM are referenced in the AUTOMATIC_SURFACE_TO_SURFACE card with the Slave ID being that of the Head model and the Master ID relating to the segments of the BONNET.

The rest of the keywords are created much similar to previous cases.

POST-PROCESSING

The Von-mises stress generated during impact is shown below,

During impact, the elements that came in contact with the bonnet experienced a stress a high as 9.079 `N / (mm)^2` and after impact from the below figure we can clearly see that the head experiences permanent deformation with the stress of about 4.013 `N /(mm)^2` over the entire from (face) region.

The HIC36 value of the simulation shown below,

It is obvious that we would experience a high HIC value considering the modified parameters and the high stress and permanent deformation experienced by the part. In this case, we get a HIC value of 3570 which is a relatively high value comparing the previous case.

This high HIC value could mean, the dummy (pedestrian) could experience severe damage to the head with a high possibility of high concussion along with multiple skull fracture, neurological damage, and possibly even death.

CONCLUSION:

Hence from the above-gathered simulation data and results, we can conclude that depending on the velocity rate, distance, cross-sectional area, material, and various other factors could alter the impact results causing either low injuries like concussions or severe injuries like skull fractures for the pedestrians involved.

This simulation helps us understand how the bonnet design of a car impacts the injuries experienced by the pedestrian and hence new engineering designs could be imparted in ways to prevent such severity of damage to the pedestrian in case of real road accidents.