SIMULATION OF A SIDE CRUSH OF A DODGE NEON BIW

TITLE : SIMULATION OF A SIDE CRUSH OF A DODGE NEON BIW

AIM : Simulate the frontal crush of the dodge neon on a rigid wall and to post proccess the forces generated or transfer through specific cross-sectional areas of the car using Hypermesh (Radioss) and Hypercrush.

A crash test is a form of destructive testing usually performed in order to ensure safe design standards in crashworthiness and crash compatibility for various modes of transportation (see automobile safety) or related systems and components.

• Frontal-impact tests: which is what most people initially think of when asked about a crash test. Vehicles usually impact a solid concrete wall at a specified speed, but these can also be vehicle impacting vehicle tests. SUVs have been singled out in these tests for a while, due to the high ride-height that they often have.

• Moderate Overlap tests: in which only part of the front of the car impacts with a barrier (vehicle). These are important, as impact forces (approximately) remain the same as with a frontal impact test, but a smaller fraction of the car is required to absorb all of the force. These tests are often realized by cars turning into oncoming traffic. This type of testing is done by the U.S.A. Insurance Institute for Highway Safety (IIHS), EuroNCAP, Australasian New Car Assessment Program (ANCAP) and ASEAN NCAP.
  Small Overlap tests: this is where only a small portion of the car's structure strikes an object such as a pole or a tree, or if a car were to clip another car. This is the most demanding test because it loads the most force onto the structure of the car at any given speed. These are usually conducted at 15-20% of the front vehicle structure.

• Side-impact tests: these forms of accidents have a very significant likelihood of fatality, as cars do not have a significant crumple zone to absorb the impact forces before an occupant is injured.

• Pole-impact tests: A difficult test which places a large amount of force on a small proportion on the side of the vehicle.
  Roll-over tests: which tests a car's ability (specifically the pillars holding the roof) to support itself in a dynamic impact. More recently, dynamic rollover tests have been proposed in lieu of static crush testing (video).[1]
  Roadside hardware crash tests: are used to ensure crash barriers and crash cushions will protect vehicle occupants from roadside hazards, and also to ensure that guard rails, sign posts, light poles and similar appurtenances do not pose an undue hazard to vehicle occupants.

• Old versus new: Often an old and big car against a small and new car,[2][3] or two different generations of the same car model. These tests are performed to show the advancements in crash-worthiness.[citation needed]
  Computer model: Because of the cost of full-scale crash tests, engineers often run many simulated crash tests using computer models to refine their vehicle or barrier designs before conducting live tests.

• Sled testing: A cost-effective way of testing components such as airbags and seat belts is conducting sled crash testing. The two most common types of sled systems are reverse-firing sleds which are fired from a standstill, and decelerating sleds which are accelerated from a starting point and stopped in the crash area with a hydraulic ram. It can also be used to evaluate the whiplash protection of a vehicle's seat.

Full Car Model

Reduced Frontal car

UNIT SYSTEM 

The unit system which was used in this model was kg ms mm

PROCEDURE

• Create an a contact interface to model how the parts interact when they come into contact with each other
• Check for element penetrations
• Check for connectivity of all parts
• Creation of rigid wall
• Mass Balancing
• Applying initial velocity
• Creation of cross-section for force analysis
• Creating time history output request
• Creation of  animation output request
• Applying custom time step
• Model Check Quality
• Post processing results of animation (Plastic strain , displacement, Vonmises stress)
• Time step, enegy error and mass error debugging
• Plotting time history data of sectional forces, accelerometer , specific energy and Intrusion displacement
• Conclusion

CONTACT INTERFACE

A self contact is created for the car using the Type 7. All parts are selected as its a self contact interface interaction Properties used can be seen below.

PENETRATION AND CONNECTIVITY CHECK

it is very important to check for element which have penetrated into the gap because this can affect the simulations results and drops to negative. The simulation becomes unstable. Below is an image of the penetration check in hypercrush. As it can be seen there where no penetration found in the model.Also checked below is the connectivity of all parts and ther was no unconnected parts.

Isolating these parts, 1D elements are created on these parts to establish all neccesory connections. An example is shown below of connections on the rails.

RIGID WALL CREATION

Rigid wall is a body with very high stiffness and is made to be static and to receive the side crush of the dodge. The cordinates of the rigid wall is defined first by picking a node on the door to specify the tail cordinates. The head cordinates is created by copying that of the tail and only giving a little gap to the normal direction. This gap is defined as small as possible in order to reduce CPU computing time during simulation.The search distance of possible slave nodes that could hit the rigis wall is be selected together with a search tolerance. 0.1 sliding frequency is selected for the wall.

MASS BALANCING

iT IS IMPORTANT TO CONTROL THE DISTRIBUTION OF MASS IN THE MODEL TO REFLECT THE ACTUAL POSITION OF THE C.O.G OF THE FULL MODEL OF THE CAR. The full model of the car has its centre of gravity somewhere closer to the front drivers seat. Masses are added at the front and back links to push the COG to the middle part of the car.This masses must also be added to reach a desired weight of 700kg. The current weight before balancing was 166kg. 500kg was added to the front links and 34kg is added to the side links. The centre of gravity can be seen in place as a dot in the car model below.

INITIAL VELOCITY

Intial velocity kinematic condition of 35mph is added to propel the model horixontally in the x direction to the rigid wall to simulate the crush , with a friction of 0.1

CREATION OF CROSS-SECTION

Cross-sections are created for specific parts in order to study the force distribution in these parts. Parts are listed below

1. Cross_Member_1
2. Cross_MemBer_2

TIME STEP 

Custom time step is specified with the following parameter

• Constant element timestep is used with cst turned on to switch to small strain when large element deformation happens.
• A factor of 0.67 is specified NODA /CST with INTER/CST factor 0.67 and time step o.ooo1
• Addition of mass to elements which produces low timestep
• Simulation time of 80ms
• Time step of 0.0005ms

CREATING TH OUTPUT REQUEST

In order to post proccese the results of the simulation , the required data must be rquested. It can be seen below the requested uotput file ad the variables selected is left at default (DEF). All the sections are requested together. Intrusions at certain locations with a reference skew

1. B_Pillar_Intrusions
2. Fuel_Tank_Intrusions
3. Hinge_Pillar_Intrusions
4. Inner_Door_Intrusions_Velocity

CREATION OF ANIMATION OUTPUT REQUEST

In order to post proccese the results of the simulation , the required data must be rquested. The animation of vonmisses ,hourglass and specific energy, Time step and Mass variation is request by creating with ANIM_ELTYPE_RESTYPE.

MODEL QUALITY CHECK

It is important to check for possible errors that may occur during simulation and model checker shows all the errors together with warnings. Below it can be seen that no errors was seen during this check.

TIMESTEP AND ENERGY ERROR DEBUGGING

The time step through the simulation was as expected with the minimum time step not less than 0.0001ms and the energy error was well below 15%. Overall the mass change with a percentage of 0.000% . There was no change in mass during the simulation.

POST PROCCESING ANIMATION OUTPUT

  ANIMATION

ANIMATION_SIDE

ANIMATION_FRONT

ENERGY TIME DATA

The flow of energy through the simulation was fairly as expected . The kinetic energy decreasing with significant increase in internal energy with time.Because energy absorption is the goal ,the whole of K.E which starts at the highest value should get dissipated and transferred to the internal energy. The more internal energy the more the deformation of the components. The copntact energy which is the energy spent in moving penetrations out of the gap zone is almost constant with a slight increase. This is good as its supposed to be less than 0 to 5% of the total energy. The  total energy was slightly decreasing due to the loss of energy and this is expected from the negative energy error recorded. This is due to the timestep control of using CST applying small strain switching element deformation. Hourglass is well stabilized as its seen constant.

MASS PLOT

Overall the mass change with a percentage of 0.000% . There was no change in mass during the simulation.

 TIME-STEP PLOT

Has it can be seen from the diagram above , there are some fluctuations in the time steps. There was a decrease early on at 10ms but the was increase as the simultions went on. At the end of the simulations the timestep decreased. This decrease is acceptable because the timestep was well above the critical time step value. 

The time step was kept above the critcal value of 0.0001ms.

Energy error of -2.0 was recorded this indicates some energy being dissipated

Simulation Time = 9522.63 s

No of cycles = 340756

CROSS-SECTIONAL FORCES

CROSS_MEMBER_1

From the cross_member_1 where the driver's front seat is loacated, this crush resulted in a normal(Y) force of 10KN at 60ms from the crush.

CROSS_MEMBER_2

 From the cross_member_2  the crush resulted in a normal(Y) force of 10KN at 18ms from the crush. This is due to the fact that the second crossmember was closer to the rigid wall and started deformation before the member_1.

Fuel_Tank_Intrusions

From 0 the intrusions into the fuel tank are was 270mm. The negative sign is due to the direction of the refernce skew created.

Hinge_PIllar_Intrusions

From 0 the intrusions into the fuel tank are was 224mm. The negative sign is due to the direction of the refernce skew created.

B_Pillar_Intrusions

From 0 the intrusions into the fuel tank are was 550mm. The negative sign is due to the direction of the refernce skew created.

Inner_Door_Intrusion_Velocity

The peak velocity of the door was 15mm/ms at the early stages but upon impact began to decrease due to the rigid bodies presence. The velocity reduced by 87% during crush.

CONCLUSION

The side crush was performed successfully with mass ,timestep an energy value all being balanced with errors. The B_pillar received the most forces upon impact and can be fatal for any occupant therefore the stiffness of that should be increased by improving the material used. The intrusion by the hinge and fuel need to be reduced by adding more components such as beams, absorbtion pads and cross rods to absorb more energy before they reach these critical areas.