Frontal Crash Analysis of Dodge Neon

DODGE NEON FRONTAL CRASH ANALYSIS

Objective:

• In this project, the frontal portion (up to B-pillar) of Dodge Neon car was crash tested to analyze the deformation and various types of forces produced during a frontal crash scenario on a rigid wall with a speed of 35mph. Along with this, the following outputs had to be requested in the analysis. And later perform the 'RADIOSS' on the model

1. Sectional force in the rails at the location of indicated node 174247.
2. The axial force received on the rails from the bumper.
3. Shotgun cross-sectional forces.
4. A pillar cross-section.
5. Acceleration curve received on the accelerometer at base of B pillar (on B pillar rocker).
6. Intrusions on the dash wall 66695, 66244.

Introduction:

• A crash test is usually done in order to analyze the vehicle beforehand and make some changes in the body, structure, or material of the BIW if required. There are different types of crash tests that are performed on a vehicle like a frontal crash, side crash, roof crush, pedestrian impact, etc. in order to ensure the safety of the passengers as well as the pedestrian.
• In this project, the frontal crash test was done on a BIW of the Dodge Neon model. The velocity of the car was set as 35mph which is one of the standard car speeds for the frontal crash according to the FMVSS norms.
• The data obtained from the crash testing is usually in terms of forces, stresses, displacements, curves, etc. which can be analyzed and the safety of the occupants can be predicted. The simulations done in the software are very much closer to the actual crashing scenario of a vehicle.
• This project was done in HyperMesh and HyperCrash for case setup and simulation, and for the post-processing, HyperView and HyperGraph 2D was used. Different types of energies, Von-Misses stresses, plastic strains, acceleration curves, and force graphs were computed.
• The vehicle has been set to crash on a planar rigid wall. Rigid walls, as the name says, are non-deformable, and fixed. Different sections were created on the parts of the vehicle body in order to analyze the forces. Along with this, accelerometers have been created at the bottom of the B-pillar on the rocker, on both sides. Mass Balancing has also been done on the body so as to get the CoG (center of gravity) just at the bottom of the driver seat or on the rail.

Procedure:

• The model was provided in .rad format. There were both, the partial model as well as the whole model. But since the node limit in the 'HYPERMESH' Student Edition is of 100k nodes, only the partial model was analyzed. The image below shows the full model.
  
  
• The partial model, which up to the B-pillar and consisted of only the essential parts, was imported in the 'HYPERCRASH' by using the ‘import solver deck’ option and was observed.
• All the Type7 interface cards (/INTER/TYPE7) which were present in the model were deleted and one single Type7 card was created for all the components. The properties of the card were set according to the recommended properties and in the slave nodes and master segments option, all the nodes of the model were selected. The image shows the Type 7 card along with its properties. This process can also be done in HyperMesh.
  
  

Penetration Check:

• The model was checked for any penetrations. If there are any penetrations in the model, the analysis will be affected. Therefore it is necessary to remove the initial penetrations or set the properties according to it.
• Penetration check was done from the ‘tools>penetration’ option, and there was 0 penetration in the model as seen below.
  
  

Rigid Wall creation:

• After the penetration check, a rigid wall was created in front of the vehicle body. The rigid wall is basically a non-yielding retaining wall which is defined by the user master node and a group of slave nodes.
• There are 4 types of rigid-wall: Infinite plane, Infinite cylinder of diameter, Sphere, and Parallelogram. Out of these, the infinite plane rigid wall was created from ‘loadcase>rigidwall>plane’ from HyperCrash, as seen below. The ‘D_search’ i.e. the search distance for the slave nodes was kept as 1200mm.
• The type of the rigid wall was set on Type2: Sliding with friction with a friction value of 0.1. A small gap was kept in between the rigid wall and the bumper for illustration purposes.
  
  

Mass Balancing:

• The initial mass of the partial model provided was approximately 188kg, while the target mass was around 700kg. The remaining mass was added manually on the body from the ‘Mass>balancing’ option in HyperCrash.
• Out of the remaining 512kg, 120kg each was added as the driver and front passenger seat. While, 250kg was added in the rear portion as the rear passenger seat, and 20kg was added as the dashboard mass.
• Thus the CoG (circled portion) came exactly on the rail under the driver’s seat where it was supposed to come in this case, as shown below. The new mass after the addition became 698kg.
  
  

Initial Velocity (/INIVEL/TRA):

• Now, some initial velocity was required to be applied to the model. This initial velocity or INIVEL was applied from the ‘loadcase>initial velocity’ option. It can also be set by right-clicking in the collector browser.
• The velocity which was supposed to be applied was 35mph according to the set standards of FMVSS. But since the unit system in the HyperCrash is kg, mm, ms, the initial velocity conversion for 35mph came out to be 15.6464 mm/ms.
• This velocity was applied along the x-axis in a translational manner. Generally, the INIVEL is applied using the global coordinate system. For node group or the ‘support’ nodes, all the nodes in the model were selected which meant that the velocity is applied to all the nodes.
  
  

Sections:

• Sections are used to identify the value of the forces passing through that portion of the component. By the use of the sections, the distribution of forces and stresses can be determined in a specific component.
• Sections had to be created in different areas of different components of the model. According to the output requests, the sections had to be made in the following areas:

1. Rails at the location of node 174247
2. Axial forces received on the rails from bumper (left and right side)
3. Shotgun cross-sectional forces (left and right)
4. A-pillar cross-sections (left and right)

• These sections were created in HyperCrash by using the ‘data history>section>create new>default’. This can also be done in HyperMesh by right-clicking in the browser area then ‘create>cross-sections’.
• The part on which the section had to be created was selected, the skew was defined using the ‘3 nodes’ option, and the rows of elements and nodes were selected having the portion of the section. This method was followed for all the sections.
• To create the section at the rails of the specified location of node 174247, that node was found by ‘node info by id’ option by specifying the node id. The ‘Iframe’ was set as 12 i.e. the CoG of the mass of selected nodes. The following image shows all the sections created.
  
  

Intrusions:

• Displacement of a specific node can be identified by opting for the intrusions parameter. In this, a node for which the displacement has to be calculated is specified. Along with it, a local skew also needs to be created in the direction opposite to it (X-direction opposite to the node). Thus the variation of the distance between the node and the origin of the local coordinate system can be computed.
• In this case, the displacements of the two nodes having id as 66695 and 66244, representing the brake pedal and clutch pedal had to be computed. This operation was done in HyperMesh under the ‘output block’ collector.
• New output block was created and the nodes were specified by right-clicking on the grnode>entity id and both the node id were specified. The image below shows the intrusions created.
  
  

Accelerometer:

• After the intrusions, accelerometers were created at the bottom of the B-pillar rocker on both sides. This can be done both in HyperMesh and HyoerCrash. In HyperCrash, this was done by right-clicking in the browser area and selecting the accelerometer option, or by selecting the same option under ‘data history’ option.
• Later, the TH of the node of the accelerometer had to be specified in order to get the curves or the graph of the accelerometers. The TH was created for both from the ‘output block’. In that, instead of nodes, the ‘accelerometers’ option was selected, and both the accelerometers were specified.

After specifying all the parameters, the model was checked for any errors. This was done in HyperCrash, under the ‘model>quality check’. There were a couple of errors related to the translational joints present in the mode. These joints were deleted straight away as they were not needed during the simulation of the model.

• Thus after clearing all the errors, and making sure all the parameters were set properly, the ‘radioss’ operation was performed in the HyperMesh from the ‘analysis’ tab. The ‘options’ menu was set as ‘-nt 4’ also called ‘multi-threading’, where 4 is the number of threads for efficiency. This number of threads must be equal to the core of the CPU, and increases by the power of 2. The radioss operation took almost 1 to 1.5 hours of time, after which all the animation files, output files, and template files were generated at the destination specified.
• After this, the engine output file was opened with the help of Notepad and the values of cycles, energy error, and mass error were observed. The observations were as follows:

Cycles: 79900

Energy error: -1.2%

Mass error: 0.0402.

The negative value of the energy error indicates that there is an energy dissipation that happened during the simulation. Also since the value is well within the limit of 15%, it is acceptable.

On the other hand, the mass error says that some negligible quantity of mass of around 40 grams was added during the simulation for the time step control or stabilization purposes. Since this value is much lower than the 2% limit, it can be accepted. Since both the energy and mass errors are acceptable, further analysis and plotting are feasible. The image below shows the energy and mass error at the end of the simulation.

• Later for viewing the animations, 'HYPERVIEW' was opened and the deformation, stress, and strain generation were in the following manner.

Displacement: The maximum amount of displacement in the model was 1252.0 mm at the 800025^th node. The following animation file shows the displacement of the model.

Von-Misses Stress: The maximum amount of Von-Misses stress generated in the model was 0.3955 KN/mm² at 147291st node.

Plastic Strain: The maximum amount of Plastic Strain generated in the model was 1.050 as seen below.

• Later in the 'HYPERGRAPH 2D', various plots of energies, forces, etc. were analyzed. Firstly, the Rigid Wall forces experienced by the rigid wall were plotted. The screen was split into two parts and the Normal Resultant force and Tangential Resultant force were plotted on them.

Rigid Wall NF and TF graphs:

• Later the different types of energies like Internal Energy, Kinetic Energy, Contact Energy, and Total Energy graphs were plotted as seen below.
  
  Internal Energy, Kinetic Energy, Contact Energy, and Total Energy:

• Due to the large mass of the model, 700kg and the high velocity of 15.6464 mm/ms, the kinetic energy in the model was very high. The initial value of the kinetic energy was around 85000J, and on the other hand, due to the large deformation of the model, the internal energy rose up to 41000J. The sum of the internal energy, kinetic energy, and contact energy equals the Total Energy value of around 85000J.
• Similarly, the section forces were plotted for the A-pillar, shotgun, bumper rail, node_174247. The following forces are from the right direction only, though they were computed for both the directions.
  
  Section Forces:
  
  
• As seen from the image above, the highest amount of forces is generated in the bumper rail as it is the first component of the vehicle body to absorb the impact. The Highest value reached up to 6.5KN.
• Similarly, the curve for the intrusion or the displacement of node no. 66695 and 66244 were plotted as shown below.
  
  Intrusion/Displacement of node 66695 and 66244
  
  
• As seen above, both the nodes which represent the accelerator pedal and the brake pedal moved by a distance of around 900mm and 850mm respectively during the crash.

Results and Conclusion:

• The most method of crash testing is the frontal crash analysis. In this project, the partial model of Dodge Neon was analyzed for frontal impact on a rigid wall. There were many different parameters specified in the case setup.
• The maximum amount of energy is being absorbed by the crumple zone in the front portion, which satisfies its purpose. Due to this, the fewer amount of forces are transferred to the passenger compartment, which in turn results in the safety of the occupants.
• The sectional forces were plotted for various sections along with the intrusions, and other forces. Thus it is very clear from this simulation that the body of the vehicle should be designed such that it will absorb all these huge amounts of forces.
• The bumper and its components like the bumper rails, shotgun, etc. play a major role in absorbing the forces. The more they absorb the forces, the higher is the safety of the occupants. To increase the absorption, notches can also be created which will initiate the deformation in the earlier stage.

Additional GDrive files of the model: Dodge Neon Frontal Crash files