Crash box against a Rigid wall simulation

CRASH BOX SIMULATION:

INTRODUCTION:

A Crash Box is a structure that is capable of absorbing high energy during impact thereby reducing the damage that could occur to the object or component it is integrated with.

The most common use of a crash box is for its application in automobiles. The Bumper of a car attached in the front and in the back act as a crash box, by minimizing the impact energy that could've been transferred to the car, by absorbing it.

OBJECTIVE:

The objective of this project is to perform an impact simulation of a Crash Box in two different cases with different thickness 1.2mm and 1.5mm onto a rigid wall and to study the various effects like stress, strain, energies, contact forces, rigid wall forces, acceleration, etc.

While crash boxes used in real life are complicated in design as its main purpose is absorbing energy, in this project, we will be considering a simple design consisting of shell elements and modeled using MAT 024 (i.e Steel material) and the rigid wall will be modeled using the LS-DYNA keyword *RIGIDWALL_PLANAR.

NOTE: This study does not replicate the actual effects of a modern crash box used in automobiles but merely the effect of a simplified version of a crash box onto a rigid wall.

SETUP KEYWORD:

1. *PART - *SECTION & *MAT

Section and Material of the crash box are modeled using a *SECTION_SHELL with an element formulation ELFORM of 2 (Belytschko-Tsay), a shell thickness of 1.2mm and 1.5mm (in different analysis) and MAT 024 PIECEWISE_LINEAR_PLASTICITY respectively and with keyword properties, as shown below,

2. *RIGIDWALL_PLANAR

This keyword is used to create a RIGIDWALL positioned at a distance of 20mm away from the front-end of the crash box.

This rigidwall can be created as follows, CreEnt -> Rigidwall -> Cre -> Planar -> N1-N2

The above-selected option lets us define a rigid wall by specifying two nodes that determine the tail and head of the normal vector of the wall.

3. *INITIAL_VELOCITY

This keyword is used to specify the velocity the crash box is moving at. We assume in this case, an initial velocity of 50Kmph, which is about 13.88 in the gm/mm/ms unit system.

The velocity keyword is used at a set of nodes, representing the entire number of nodes available in the part. This node set can be created using, CreEnt -> Set Data -> *SET_NODE

NOTE: The same can be achieved using the *BOUNDARY_PRESCRIBED_MOTION_SET keyword where more parameters like velocity, acceleration, and displacement of the crash box can be defined. This keyword can be made use of when the velocity of the system is unknown but the end result i.e amount of displacement required to travel is pre-determined.

4. *CONTACT

During impact or crash, in this case, there is a possibility that elements pertaining to the object itself could come into contact with each other. Hence the keyword *CONTACT_AUTOMATIC_SINGLE_SURFACE is defined.

If the user fails to specify this keyword, penetrations of each element onto others in the object could occur resulting in error simulation.

While defining this keyword, the only required parameter is the Slave ID (SSID). The Master ID is not-defined as our contact doesn't involve contact between two different objects. Hence to specify the SSID, we select the part while the SSTYP constitutes option 3.

5. *CONTROL

This keyword is used to add certain parameters that must be met or taken into consideration during the analysis.

5.1. *CONTROL_TERMINATION

This keyword specifies the duration of the simulation. In this case, we assume the impact occurs within a time period of, say 2 ms.

5.2. *CONTROL_ENERGY

This keyword is used to specify the various energies that need to be considered and computed during the simulation.

For this analysis, since we have a rigid wall present, we can compute the Stonewall energy (Rigidwall energy, RWEN). Similarly, we also require the hourglass energy (HGEN).

6. *DATABASE

This keyword specifies the output parameters that need to be extracted during the simulation.

6.1. ASCII_Option

Here the required output parameters that need to be computed are specified and written in a separate file. In our case, the following files are written for every time interval of,

GLSTAT - Global Statistics involving KE, IE, hourglass energy, etc.

NODOUT - Nodal properties like displacement, acceleration, etc.

RCFORCE - Interface force

RWFORCE - Rigidwall force (since our model consists of a rigid wall)

SLEOUT - Sliding interface energies

SECFORCE - Cross-sectional forces

6.2. BINARY_D3PLOT

This is the most important keyword where the output of the animation file is written (.d3plot).

For our simulation, we specify a time interval of 0.1, which represents that for each 0.1 ms the output data of the analysis is written onto the animation file.

6.3. CROSS_SECTION_SET

Since we specified, the Cross-sectional energy like X, Y, Z forces that need to be computed. This keyword lets us specify the SET to which the required forces need to be computed.

6.4. HISTORY_NODE_SET

Here, the *SET_NODE containing the required nodes for which the Nodal output parameters that need to be computed are defined.

NOTE: The above simulation is performed again with a shell thickness of 1.5 mm, which can be specified in the keyword *SECTION_SHELL

POST-PROCESSING:

On performing the analysis with the above-mentioned parameters for the two cases, the following results are obtained.

We can observe that the crash box bounces off the rigid wall, the reason being the crash box is traveling at a very low velocity and having a higher stiffness which prevents the crash box from buckling.

The Von-mises stress experienced at the crash box before and after the impact is shown below,

Similarly, other results obtained for both the cases (1.2 and 1.5) at the state of impact are shown below,

Stress at X, (1.2mm and 1.5mm)

From the above result, we can observe that most of the stress concentration is at the first half side of the crash box with a maximum of 151.5Pa in 1.2 mm and 152.0Pa in 1.5 mm. This is a less significant difference since the thickness is only about 0.3 mm.

Stress at Y,(1.2mm and 1.5mm)

From the above result, we can observe that the Stress in the Y direction (i.e in the lateral direction) is about, 161.6Pa in case 1 and 163.3Pa in the latter.

Mean Strain, (1.2mm and 1.5mm)

We can observe that the strain in both X and Y direction is zero. This is because the travel velocity of the system is relatively low and the dimension of the crash box is quite large with more stiffness to experience even a significant amount of mean strain in the system.

However, the Effective plastic strain which measures the elements that are subjected to yielding during the impact is shown below,

Plastic Strain, (1.2mm and 1.5mm)

The effective plastic strain increases whenever the material is subjected to yielding, i.e., when the state of stress is on the yield surface.

The above figure gives a clear representation of the elements in the crash box that are actively yielding.

GLSTAT - Kinetic, Internal and Total energy, (1.2mm and 1.5mm)

We can observe the Energies experience in the system. We can observe that after the state if impact the Kinetic energy gradually decreases while the Internal energy increases caused due to the development of stress within the system.

GLSTAT - Hourglass and Sliding energy, (1.2mm and 1.5mm)

From the above graph, it is obvious that the Sliding energy is not present in the analysis as the crash box bounces off the rigid wall. There is a considerable amount of hourglass energy experienced by the system after the impact. This hourglass effect can be visualized at the end nodes that come in contact with the rigid wall during impact, as they are the ones subjected to deformations.

Rigid wall force (RWFORCE), (1.2mm and 1.5mm)

As expected, during impact there will be an external force experienced by the rigid wall by the crash box. This force experienced by the rigid wall is about 14e4 N in case of a 1.2 mm crash box and 17e4 N in case of the crash box with a thickness of 1.5 mm.

Sliding Interface energies (SLEOUT), (1.2mm and 1.5mm)

As stated before since there is no Sliding effect like the sliding force is experienced by the contact regions, it is possible that there would be no Sliding interface energies either.

Acceleration, (1.2mm and 1.5mm)

From the above fig, we can observe that during impact the acceleration of the system along its travel direction (i.e X) varies as there occurs a change in velocity. This acceleration is negative since the velocity drop during impact and regains velocity as it bounces off the rigid wall which can be visualized above as the acceleration reverts back with less deviation.

This acceleration of a node along the length of the crash box (X-direction) reaches a low limit of about -1000 `m / s^2` in case 1 and about -1100 `m / s^2` in the case of the crash box with 1.7mm shell thickness.

Cross-sectional forces (SECFORCE), (1.2mm and 1.5mm)

We can observe that the during impact the specified cross-sectional set along the length of the crash box (X-axis) is about 450 N in case 1 and about 600 N in the second case.

Resultant Cross-sectional force, (1.2mm and 1.5mm)

Finally, the resultant cross-sectional force experienced by the crash box irrespective of a single direction is about 480 N in case 1 and about 600 N in the second case.

CONCLUSION:

Hence from the above results obtained we can visualize and understand how the crash box at different conditions alter its effects by changing a single parameter, the thickness of the crash box.

Moreover, we can observe as a whole, that the crash box bounces off the rigid wall with small values of stress experienced and without and buckling or folding effects. However, a different result like, buckling and folding effect could occur if *PRESCRIBED_MOTION is applied onto the crash box which specifies the distance of travel of the system. Hence the system would have no other way than to buckle and fold.