Assignment 6

Frontal Crash Analysis of BIW

Objective:

To perform Frontal crash analysis on BIW and obtain output requests at the following sections:

• Sectional force in the rails at location of indicated node 174247.
• Axial force received on the rails from bumper.
• Shotgun cross sectional forces.
• A pillar cross section.
• Acceleration curve received on the accelerometer at base of B pillar (on B pillar rocker).
• Intrusions on the dash wall 66695,66244

Case setup and execution:

The given reduced element model is imported into HyperCrash and if we check the connection quality,

Menu -> Quality -> check connectivity of tree selection

HyperCrash shows some connectivity issues in the model. Some of the issues are because of the deletion of elements in the reduced element model. Those connections are deleted.

A rigid wall with friction 0.1 is created in front and very close to the bumper and Dsearch is provided a value of 300. It tells the solver to use all the nodes within that proximity to use as slave nodes for the rigid wall and so it won’t cross the rigid wall.

Menu -> Loadcase -> Rigid Wall

The full 300K element model is opened in another HyperCrash session and both the models are compared for mass distribution and Center of gravity. Mass is added on the reduced model to match the Full scale model in order to capture the exact crash scenario and it is made sure that both model’s center of gravity coincides (The Center of gravity would fall below the driver’s seat in most of the cases).

Menu -> Loadcase -> Added mass

Menu -> Mass -> Balancing (to view total mass and C.O.G)

The rigid body node of front suspension towers are constrained in the z direction to make it more realistic. A highly stiff suspension springs at front won’t absorb energy (won’t compress or expand) and rather pass it to the rear.

Loadcase -> Boundary condition

An initial velocity of 35mph is given to the model in x direction

Menu -> Loadcase -> Initial Velocity (35mph = 15.6464m/s) and select entire model nodes

Sections are created at various locations asked in the question to analyze the forces passing through them

Data History -> Section -> Section

1 Accelerometer is placed on both B pillars at the base.

Data History -> Accelerometer -> Accelerometer

Now sections and accelerometers area created, we have to record time data in these blocks

Time history -> create new -> TH of section

Time history -> create new -> TH of accelerometer

The question asked to plot displacement of 2 nodes near pedals, so we need to record time data for nodes.

Time history -> create new -> TH of nodes (and enter the node IDs mentioned in the question) and movable skew is created below the driver’s seat to analyze the intrusion of the node with respect to the driver. It’s always preferable to create a skew at a node which will be free from deformation to capture the exact deformation of the nodes in pedal area.

A global contact interface is applied with recommended properties on the whole model. Type 11 contact is applied to specific areas of concern where edge to edge contact is expected. After creating contact interfaces, it is made sure there is no penetrations and intersections in the model. After running simulation, on viewing the animation file, few more parts were showing edge to edge intersection and penetration. So, more Type 11 contacts are added to resolve the issue.

We also have to include interface to analyze contact forces

Time history -> create new -> TH of interface

Engine file is edited as shown:

Run: 80ms

DT_ELTYPE_KEYWORD_IFLA : Time step - 0.5 to 0.1 microseconds (the model already has Time step defined for each type of elements)

ANIM DT: Tfreq = 2ms

ANIM_ELTYPE_RESTYPE: ELEM variables = EPSP, ENEM, VONM

ANIM_MASS: On

MON: On

PARITH: On

Print: -100; 55

Tfile: DT_HIS = 0.5

Checks and Debugs:

Now the model checker option in HyperCrash is used to check for errors and warnings in the model.

Menu -> Quality -> Model Checker -> Run

The errors and warnings are corrected, and the model is now ready for running Radioss simulation.

Starter output file (checked Final mass, COG, no incompatible kinematic condition, Normal Termination), Engine Output file are checked, and no issues found.

Mass error is 1.71% and is within 3% allowable limit

Energy error is -2.1% and is within 5% allowable limit

(Hourglass energy + Contact energy)/Total Energy is 2.6% (found by obtaining max value of Hourglass energy, contact energy, and Total energy plots) and is within 15% allowable limit

There is no sudden drop-in Time step (viewed in Engine output file)

VonMisces animation looks smooth

Hence, we can say simulation is good.

Results and conclusion:

• Sectional force in the rails at location of indicated node 174247 and mirror section

The graph shows the forces passing through these rail sections. The max force is 12.265kN for Left section and around 9.39kN for right section.

• Axial force received on the rails from bumper - Left and right side

The graph shows the forces received by the rail from bumper. The max force is around 4.913kN for Left section and around 5.774kN for right section.

• Shotgun cross sectional forces - Left and right side

The graph shows the forces passing through these shotgun sections. The max force is around 23.4026kN for Left section and around 20.727kN for right section.

• A pillar cross section - Left and right side

The graph shows the forces passing through these A pillar cross sections. The max force is around 16.828kN for Left section and around 12.918kN for right section.

• Acceleration curve received on the accelerometer at base of B pillar (on B pillar rocker) - Left and right side

The graph is an unfiltered acceleration data from accelerometer kept at base of B pillars

Saefilt95 is used to filter the obtained noisy acceleration curve to filter around 60Hz. Mirror padding and Fwd-Back direction option is used in the filter.

• Intrusions on the dash wall 66695,66244

The graph shows the intrusions of these nodes (present in the dashboard) in the x direction (into the driver’s compartment).

Mov2 skew is used here. Moving skew defines a local coordinate system, defined by three nodes. At each time, the actual orientation of the skew is recalculated according to the actual position of these nodes.

Mov and Mov2 skew are similar skews differing only in the way axes are defined. Fix skew fixes the local coordinate system with respect to the nodes and doesn’t change orientation.

Blue and red lines are displacements with respect to a skew system present in the cross member (under driver’s seat). Since this skew undergoes deformations in all directions, the curve looks wavery.

Green and pink lines are displacements with respect to a skew system present in the far end of the floor. Since this is a stable point (attached to a rigid body and located at far end from barrier), the curve looks smooth.

Negative value indicates the relative distance between the nodes and the system is getting smaller than the initial value.

Since analysis didn’t include a dummy to measure any injury caused by this intrusion, it is not possible to conclude if this is good or bad. In an analysis with dummy, if the displacement needs to be limited, some design modification (can make it thicker, change material) can be made in the rails, shotgun to absorb more energy.

66244 peak intrusion – 834.982mm

66695 peak intrusion – 794.631mm

The graph shows resultant displacement of the nodes with respect to 2 skew systems. First 2 – skew system under driver’s seat. Last 2 – skew system on floor far from barrier.

Blue and green lines coincide

Red and pink lines coincide