Assignment 6-Frontal Crash Simulation Challenge

AIM :

       To simulate the frontal crash using the body in white components, make a model as per the given requirement and extract the required results in the post-processing.

OBJECTIVES :

       Check the unit system.

       Create an appropriate interface, recommended parameter & friction force is 0.2 between the components.

       Check the penetration & Intersection between the components.

       Create the rigid wall, 0.1 as a friction force between rigid wall & car.

       Adding the extra mass to attain the 700kg as a target mass.

       Apply the initial velocity to the car. 

       Keep the time step minimum of 0.1micro sec & maximum of 0.5micro sec.

       Run the model up to 80 milliseconds.

       Finding the sectional force in the rails, shotgun & A-pillar component.

       Finding the axial force received on the rail from the bumper.

       Calculate the acceleration of the car.

       Calculate the intrusions on the dashboard.

       Check the model to ensure good quality.

       Plotting all the required results. 

INTRODUCTION :

       A Frontal crash is one of the most common crash tests that result in serious injury or death. A Frontal crash test helps us to evaluate how well a vehicle might protect the passenger when the vehicle collides head to head on another vehicle or any rigid barrier.

       Automotive companies are using two methods to conduct crash tests such as physical crash tests & Virtual crash tests.

Physical Crash test:

      In a physical crash test, we are using the real vehicle to conduct the frontal crash to collide on the reinforced concrete wall. An automotive vehicle collides on the wall or barrier at some speed, then calculate the deformation of the vehicle & How airbags are opened when a vehicle collides on the barrier.

       An "anthropomorphic test device" is commonly used as a "Crash test dummy" in the automotive testing vehicle, it is used to calculate what are impact forces a passenger suppose to receive.

       For each crash test, have to use the new vehicle. It will lead to an increase in the cost & time.

Virtual Crash test:

       In a Virtual crash test, by using powerful computers we can perform the same crash analysis and get the proper results whatever we want. We have to set up the model as same as in real crash scenarios & get the results.

       We can perform an N number of tests with some changes as per our requirement. Eventually, we save a lot of money & time.

       But, the accuracy of the results is slightly changed then the physical test.

 

PROCEDURE FOR OBTAINING THE OBJECTIVES :

1) Checking the Unit system :

      Open the "Begin_card", check the unit system which is followed by Milligrams, millimetres & seconds unit system. And also we can check the unit system in the starter file.

     

2) Apply interface to components:

      Create a Type 7 (Nodes to the surface) contact, set the parameters as per the recommended crash application & then provide 0.2 as the coefficient of friction.

Igap : Determines how the size of the gap is calculated.

An "Igap = 2" is a set variable gap to take into account the true distance between parts.

GAPmin : Minimum gap for activation of the interface.

A "GAPmin = 0.5" is a specify the minimum thickness of the model to avoid numerical issues. Typically, it takes half of the thinnest part.

Inacti : Action to takes if initial penetration exists.

An "Inacti = 6" is to remove initial penetrations where possible. Elsewhere, reduce to less than 30% of the defined gap value and adjust the gap with the "Inacti" parameter.

Istf : Affects how the stiffness of the interface is calculated.

An "Istf = 4" is to set stiffness of interface based on the softer of master and slaves.

Iform : Friction formulation. 

A "Iform = 2" is sliding forces are computed using the stiffness of the interface, which usually results in a bigger time step.

Stmin : Minimum stiffness to use in the interface.

A "Stmin" is a specify a minimum stiffness in the contact to avoid too soft contact

Idel : Behaviour of slave and master segment if an element fails.

An "Idel" is removed slave nodes from contact because of element deletion.

 

Have changed the contact card image as per the requirement as below.

3) Checking the Penetration & Intersections :

     Open the "Penetration & Intersections" checking tab, chosen all groups as a selection, keep "0" as Minimum penetration depth and "1" as a thickness multiplier.

     There's no intersections & penetration are presented in between the components as marked below.

     So, we can move to the next process with this model.

4) Creating the rigid walls :

      Create a new "Rigid wall" in the model browser, first & foremost thing is to select the centre node on the front bumper, then it showing the node coordinates. We have to tweak the coordinate number to fix the wall.

       A "Normal " parameter represents the coordinate of the normal vector.

       To give the friction between the rigid wall & components, we must switch to "Sliding with friction" and give the 0.1 as a coefficient of friction.

 

     

     After creating the rigid wall in front of the car, it looks as below.

     

5) Adding the extra mass to attain target mass :

     A given model total mass is 188 kg only. But, actually full-scale 300k nodes model mass is 700 kg.

     So, we have to provide the remaining mass on a given model.

    

     Create a new Add mass card, set the ground id.

     Chosen "0" as a Mass type which means, a mass is added to each node of the node group. So, we have to provide the mass value is "Remaining mass is divided by the number of nodes selected on the model"

     Mass = 512/3517

             = 0.145 kg

    

     Nodes are selected from the end of the floor panel as below.

    

    After adding the mass to the model, it reaches the target mass of 700 kg.

    

6) Apply the Initial velocity :

    Create an "INIVEL" card for applying the boundary condition on the model.

    Create a new ground id, select all components as entity id.

     Provide the 15.64 as velocity in x translational direction & the remaining two directions are constrained.

    

      An initial velocity is applied to global coordinate systems, that's why we didn't consider skew id or local coordinate system.

     After gave the velocity on x translational direction, a number is showing from the model.

    

7) Set the time step :

DT_ELTYPE_KEYWORD_IFLAG_SUPPORT:

     In this option, we can create different types of time step control options such as INTER, NODA, BRICK, QUAD, SHELL, TRUSS, BEAM, SPRING & other few.

     After, we have to select the type of time step control such as CST, CST1, CST2, AMS, STOP, DEL.

     In this case, we have created INTER & NODA as a type of time step option

ENG_DT_INTER:

     This method is used for controlling the time step from the element.

     A Type of time step option is "INTER" and the time step control type is "DEL"

     Keep the 0.67 as a scale factor on the time step.

     Keep the 0.0005 ms as a minimum time step.

     If the model time step is reduced below the minimum time step, this algorithm will add some mass to the element so which will lead to maintaining a time step which more than the minimum time step.

    

ENG_DT_NODA:

      This method is used for controlling the time step from the Node.

     A Type of time step option is "NODA" and the time step control type is "CST"

     Keep the 0.67 as a scale factor on the time step.

     Keep the 0.0001 ms as a minimum time step.

     As same, If the model time step is reduced below the minimum time step, this algorithm will add some mass to the node so which will lead to maintaining a time step which more than the minimum time step.

    

8) Creating the sectional cross-sections:

     To find the force which passes through rails, shotgun, A-pillar. we must create a sectional cross-sectional in there which location we want to look at it.

     First of all, create a "Moving frame" on the location where we want to measure.

     To create a moving frame, we have to create a new system and properly select the three nodes.

    

    Next, followingly create a new section, similarly, properly select the nodes. Properly assign the frame id and provide the delta T value as 0.1 & Alpha value as 1.67.

    Create a new ground shell id, select the three or four rows of elements properly on where we want to calculate the sectional force.

    Make sure to switch with "12" as an Iframe which keeps the coordinate at the location of the centre of gravity.

    

     Here below is how I select the elements for calculating the force passing through this sectional cross-sectional area.

     

    Create "Cross-sections" in the model browser, all data are stored by default once we created the section in the solver browser.

    

     Finally, create a new set & name it a cross-section. Assign the created cross-section into entity id. 

    

    Similarly, do the same process for all locations where we want.

9) Create the accelerometer node:

    Create a new accelerometer card in the solver browser

   

    Select the one node on the B-pillar rocker as a node id. Provide a value of 1.67 as a cutoff frequency.

    Calculate the acceleration to the global coordinate system.

    Similarly, do it on another side too.

   

   Create an accelerometer card in the model browser, provide the cutoff frequency as 1.67.

   

    Create a new card under the Output blocks & name it an accelerometer.

    Assign both accelerometer cards as an entity id.

   

10) Create Intrusions nodes:

    For calculating the intrusion of the dashboard wall while the car has collided on the rigid wall.

    Create a two new skew on the cross rail. Create an intrusion node as suggested and calculate the dash wall intrusion along the x-axis.

    

    Create a node time history, appropriately select the node and assign the newly created skew to that.

    Similarly, do it on another node as well.

    

11) Check the quality of the model :

     In this model, there's no major error that will not cause the analysis.

     So, we move this model to the next process which is solving.

    

 

ANALYSIS OF THE MODEL :

     Select the "Radioss" sub-panel under the "Analysis" panel, save the file in the separate folder, and click the radioss button to solve the problem by solver itself.

     After calculating the model, a solver is generated different files such as the Starter listing file, Engine listing file, Animation files, Time history file and then restart file.

    

      As we know, a time step is calculated in each & every cycle. If the model time step value is less than 0.001, a solver will take a default time step value which 0.001. Because we enabled "CST" which means the solver will use default constant time step if model time step value falls below minimum nodal time step value. To keep the time step as constant, it will add some mass on the node for which control the time step.

      In the Zeroth cycle, a modal time step is less than the nodal minimum time step. A Control method has to be used constant default time step which is 0.001 and a mass has been added on the node. If the mass added percentage is lies between 1 to 3, that means those will be acceptable.

      An error is positive which means there's some energy is generated in the body.

      

    

      As we can notice, a solver has mentioned constant time step up to end with the help of constant time step control. Eventually, a mass percentage was also added to the node.

      In the few cycles, errors are negative which means some energy has been dissipated. But, it comes under the acceptable range.

      If the time step value decreasing more, a mass added percentage also be increasing more.

      

POST-PROCESSING :

      Post-processing is a process of extracting useful data from the solved model.

Hyperview :

      After completion of solving, switch to "Hyperview" from the "HyperMesh" section.

      Select the appropriate file in an "h3d" format and open the same file in the Hyperview interface.

Contour plot of Displacement:

      A maximum displacement reaches up to 1251 mm after 80 milliseconds ran.

       

Contour plot of Von Mises stress:

       Stresses are distributed in the entire model. More stresses are generated on the front rail, shotgun, bumper & other front components.

       A generated values are a low number because the initial velocity is less.

       As know, balancing masses are added to the end of the floor panel. If the mass is higher, the inertia is also higher. As we can see from the simulation, a mass added area have tried to move forward when a car has collided on the wall. Because it obeys the second law of motion.

      

 

Hyperview :

      Switch to the "HyperGraph 2d" interface instead of the "Hyper View" interface for plotting the required data.

      Import the appropriate time history file.

Energies plot:

      Kinetic energy is decreased over a period of time. Because a velocity is decreased over the period of time due to the car collided on the rigid wall. Kinetic energy is directly proportional to velocity. It has started at 85000 and ended up with 45000.

      Internal energy is increased over a period of time. Because it has some work done in the car.

      There's no hourglass energy is generated since we used the QEPH shell element formulation.

      The total energy is a combination of Internal energy & kinetic energy.     

     

Dash wall Intrusion at 66695 node:

      There's no intrusion happened up to 25 ms, after that intrusion starts quadratic in nature, reaches 225 mm over the period of time.

      At the end of the simulation, 225 mm intrusion has happened. That's not a recommended one, because passenger's legs get injured due to that intrusions.

     

Dash wall Intrusion at 66244 node:

      This is a different node, it has a different intrusion. On that node location, an intrusion starts from the early, increased quadratically in nature over the period of time.

      A maximum intrusion is 830 mm which causes severe leg injury.

     

Acceleration on the left side:

     Maximum acceleration is 0.20 mm/ms^2 at the beginning, it is captured from the left accelerometer node.

     Apart from that, 0.15 mm'ms^2 acceleration is captured a few times.

     

Acceleration on the right side:

       From the right side accelerometer point of view, 0.5 mm/ms^2 is the peak acceleration value.

      

Cross-sectional forces :

      A total resultant force is the resultant of x, y & z components.

A-Pillar left side :

       On the left side A-pillar, the sectional peak force is around 1 kN.

      

A-Pillar right side :

      On the right side A-pillar, the sectional peak force is around 1.8 kN. This force is higher than the left side.

      

Axial force on the bumper left side :

       On the left side bumper, a peak axial force is 10 KN

      

Axial force on the bumper right side:

      On the right side bumper, a peak axial force is 7.5 KN. It is slightly lesser than the left side.

      

Sectional force on rail left side:

      On the left side-rail, a peak sectional force is around 4.5 KN.

      

Sectional force on rail right side:

       On the right side rail, a peak sectional force is around 16 KN. It is a higher value than the left side.

      

Sectional force on Shotgun left side:

      On the left side shotgun, a peak sectional force is almost 18 KN.

      

Sectional force on Shotgun right side:

      On the right side shotgun, a peak sectional force is almost 20 KN. It is slightly higher than the left side.

      

 

CONCLUSION :

       Successfully ran the frontal crash simulation as per the given requirements on the given BIW components.

       Calculated all the outputs requests such as sectional forces, axial forces, acceleration & Intrusion of the dash wall.

       There are some negative errors are generated due to the solver has added some amount of mass on the node for controlling the time step. To control the negative error, we should reduce the nodal minimum time step, but the computation time will be increased. No need to reduce the nodal time step in this case, because those errors come within an acceptable range.