Frontal crash analysis of a car ( BIW )
Objective :
Frontal crash analysis - To import a given model and analyze its behaviour in various aspects by feeding required inputs , and requesting required outputs using tools in hyperworks , through which the analysis of the model during frontal crash could be determined
Procedure :
- Open hypermesh and choose RADIOSS interface
- Then import the 0000.rad file
- Check for correct unit system --> Kg mm ms ( As per requirement )
- Open model in browser tab and select cards --> Begin card
- Check for the units of mass , length , time
- After importing , check for the penetrations in the model , since penetration leads to simulation errors and timestep issues
- In this process, hypercrash was more convenient in checking for penetrations issues , thus hypercrash application was opened
- Go to Applications in menu bar and select Hypercrash
- Once Hypercrash was opened , contact interface from browser tab selected
- Select all the interfaces created and click the check penetration for selected objects option
- The quality window appears showing all the penetrations and intersctions of the model
- There where no penetrations found in this case
- Then, hypermesh window is again opened to assign the simulation run time
- In the browser tab open cards under which select ENG_RUN
- To assign the simulation run time in ms , enter required values in Tstop ( In this case 80ms assigned )
- Then , to assign Timestep open solver browser
- In solver browser , right click --> create --> engine keywords --> DT --> DT
- In which assign the required timestep in Tmin ( In this case 0.5 was assigned )
- After which initial velocity to be assigned for which solver browser is opened
- Right click --> create --> boundary conditions --> INIVEL
- In which as per the direction of global axis and required velocity to be mentioned accordingly
- In this case as per global axis , x-direction is mentioned with required velocity ( To be mentioned in mm/ms )
Requirement velocity 35mph
1 mm/ms = 2.23 mph
35mph = 15.6 mm/ms - After which Grnod_id to be selected , Right click on Grnod_id and select create
- Under Entity IDs select all the nodes of the model
- Click proceed and initial velocity is assigned as per requirement
- After which rigid wall is created
- To create rigid wall , in the model browser , right click --> create --> rigid wall
- The Engineering data , Normals are entered with reference to the 300K model
- Silde value and Dsearch values are entered
- Rigid wall has been created
- Next interfaces are to be created for the model
- In the model browser , right click --> create --> contact
- In the contact window , enter ther recommended parameters like Type of contact , Istf , Igap , Idel , Stmin , Fric , Gapmin , Inacti , Iform
- Then for selecting Masters and slaves right click the Surf_id(M) and Grnod_id(S) respectively and under Entity IDs select the components for master and slaves
- As per requirement , the given model's mass has to reach a target mass of 700kg from its initial mass by adding mass to it and bring the CG ( Centre of Gravity ) to required position ( near centre cross beam )
- To know the initial mass and CG of the model
Tools --> Mass details --> Mass , Cog and Inertia - Select mass tab to view total mass ( Initially )
- Using the COG tab and the node ID , initial COG of the model is found by creating a node
- To add mass to the components , in the model browser , right click --> solver mass
- Under which select the required mass type , mass and grnd_ID( Right click --> create and under Entity_IDs select the Nodes for which mass to be added )
- After adding masses over the components , check the total mass again so as to reach the target mass
- Also check the COG of the model to reach the required position
- Thus the target mass and COG position reached
- Then as per output requests , ( Over TH )
Intrusions on the dash wall to be determined after simulation - As per requirement node ID 66695 and 66244 intrusions are to be determined using TH plots
To request TH for these intrusion nodes position , from solver browser , right click --> create --> TH --> node - Select the nodes from Entity IDs
- Accelerometer to be mounted over base of B pillar ( B pillar rocker ) to analyze acceleration curve
- In Solver browser , right click --> create --> TH --> ACCEL
- On the accelerometer window , right click on Entity ID's and the in node_ID select the node over the rocker ( similarly other side )
- Thus to create acceleration curve in output TH graph, accelerometer was created over B pillar ( on rocker )
- Similarly as per requirements , sectional forces were to be studied using the outputs from TH data
- For this the position with node ID's were specified
- Over these positions Sectional area were to be selected
- Using the given node IDs , the nodes and the position to create sections were determined ( this was done using find options and node ID entered )
- All the required positions were identified
- Similarly on the appropriate positions on the other side of the model have to be specified
- In each position , section selection were made ( Similar procedure for all position )
- To create section open solver browser right click --> create --> SECT --> SECT
- Enter N1, N2 , N3 to specify directions
- Right click on grnod_id to enter Entity_ID ( where 2 rows of nodes selected )
- Right click on grshel_id to enter Entity_ID ( where elements covering the 2 rows of nodes were selected to specify sectional force )
- Assign deltaT , alpha , Iframe as per requirement
- Similarly as per requirement , 10 sectional positions were created under SECT in solver
- These were to be plotted in TH during analysis , thus these cross sections were to be defined under TH card
- Open TH card in solver and under node select cross section and right click to select all the cross sections created under Entity IDs
- All the inputs were specified as per requirements
- Before simulation run , model checker performed to check the model for better result during simulation run and to avoid errors
- To check , Tools --> Model checker --> Radioss Block
- Model checker tab appears
- Right click on Error and select Run to determine the errors and the Errors tab opened to view the errors
- If no errors , proceed for the simulation run else solve the errors and then run the simulation
SIMULATION RUN :
- After which Simulation RUN could be performed using Analysis panel in main menu
- Choose RADIOSS option and select the required parameters , input file to save the generated files
- Click RADIOSS to run the simulation and files will be generated in input file location
- After simulation run , a dialog box shows the total number of cycles and simulation time
- After which open the 0001.out file that was generated during simulation run and check the energy error & mass error
- Energy error to very maximum level of +5% is acceptable model , -5% represents a good model and better to be close to 0%
At ideal conditions -15% - Using page window layout , split the window for convenienece
- Select Hyperview option in toolbar and click load model to upload .h3d file from the generated files during simulation run
- Click Apply
- The simulation runs for a period of 80ms for 17 frames
Frame - 8
Frame - 15 - Observe the animation RUN and analyse the changes occuring in the elements and their reaction for a period of 80ms
- Similarly analyse using contour ( select contour icon on toolbar )
- Select result type as Displacement or Vonmises , averaging method as simple
Also use maximum and minimum measurement plotiing using measures icon in the toolbar - Click Apply
- Contour frame - 8
Contour frame - 15 - Next open Hypergraph , import T01 file using data file option to plot graphs
- Plot ,
Y Type - Global variables
Y request - Hourglass energy , Total Translational energy , Internal energy , Kinetic energy , Contact energy , Mass
Y component - MAG - Then select - for plotting acceleration curve
Y Type - Accelerometer 1 , Accelerometer 2
Y request - 1 accelerometer 1 , 2 accelerometer 2
Y component - ACC - Resultant acceleration - Then select - for plotting intrusions due to displacement during deformation
Y Type - Intrusion 1 , Intrusion 2
Y request - 66695 , 66244 ( As per requirement )
Y component - Resultant displacement - Then select - for plotting sectional forces over the 10 cross sectional areas
Y Type - Cross section
Y request - cross section 1 - cross section 10
Y component - Total resultant forces - These graphs define the plotted details and variations over a period of 80ms
- Thus the model have been fed with inputs and required outputs were drawn after simulation , and analysed for the changes in its nature over a defined period of 80ms
Observations over outputs after simulation :
Animation :
The car model is allowed to crash in the front region , where the deformation starts from the bumper and travel through the interior parts of the car
Initially the stress were maximum at the bumper and slowly passes through the inner components
With reference to the contour plots, the maximum stress was observed near the dash board region on frame 15 for a total frame of 17 at a value of 0.399
The stresses reach its maximum value when the component absorption of forces over the region were high
Sectional forces :
These are forces that are analyzed over a particular region to determine amount forces passing through it
As per requirement , 5 sections on each side of the car model were to be created and analyzed for sectional forces ( Totally 10 sections )
- A pillar cross-sections
By observing this , the forces across both the A pillar were not similar
This was perhaps due to difference in selection of nodes and elements over those cross-sections
Or the transmission of forces through the region would differ due to material , property nature
If there were softer material in the deformation , maximum of the forces would be absorbed by that material and less forces would be transmitted
Similarly in case of harder material present , then maximum forces would have been transferred across it to the next component
Both the cross-sections in A pillar , seems to raise the forces , over a period of time ( As time increases )
Since A pillar deformation starts increasing over a period of time , increase in this stresses were observed ( Since A pillar is located behind few components that first gets deformed and then comes to A pillar ) - Rail component
By observing this , the stresses over both the rails were almost similar shows both the rails have undergone similar deformation
Since the Rails were obviously present in the front of the car model , as the crash starts to deform , the stresses starts initially at the bumper and then passes through the rails
So initially the stresses were high on the raiils and then the stresses passes though the rails to other components
Then over a period of time , again the stresses seems to increase in the rails due to friction with other components coming in contact ( End of the simulation 70ms - frame 15 ) - Shot gun
By observing this, there seems to be difference in shotgun stresses , might be due to sectional selection and also due to deformation pattern
Absorbing more forces during deformation results in increase in its stresses
This absorption of stresses differ due to nature of material and property , also the forces received by that component/part from the previous components
Thus there would be differences in the stresses of every component
Since the shotgun positioned after the rails , the forces received by the shotgun were after the rails
In TH , for a time of 80ms , it could be seen that the forces initailly were low and then increases ( This is due to , once the rails receive the forces initailly , it then passes remaining forces )
Thus for a period of 80ms , the forces received by the shotgun were at the period of 20ms to 45ms approx. - Hinge pillar
Similar to other components , there were differences observed in the hinge pillars due to several reasons like explained before ( depends on absorption of forces with nature of material and property )
Initailly the forces received by the hinge pillars were low due to its positioning , gradually increases as time increases
Forces seems to be received less compared to bumper rails and shot gun , since the hinge pillar located behind those components and obviously the stress would have been absorbed by them and remaining reaches the hinge - Inner pillar
After bumper rail, these were the components to receive more forces and results in higher stresses
Due to position of this component right next to rails and the area of cross section was much more than shot gun , hinge pillar and A pillar , ( the stress concentration were seems to be more )
Inner pillar , initially at low stresses due to time . Since initially the rails were only components receiving forces and as time passes the inner pillar receives the forces
Overall view on sectional forces :
These forces were plotted to analyze the amount of forces received and passing through them over a period of time during deformation.
The amount for forces transferred and received depends on the position of the components
If the component was present at very beginning or front of the model were deformation or collision starts , thats the point where the stresses were maximum ( Bumper rail )
And gradually as time passes, deformation increases and forces get transferred to other components
These forces gets transferred only after that particular component has received enough stresses and the remaining stresses gets passed to next component
Also the stress depends on the area of cross-section ( Higher the area more the total stress observed in that particular area )
And finally when the stresses pass through component by component , the stresses received by the component present at the end of the sequence , expected to be minimum
Intrusion of dash :
- To determine the intrusion of a particular position ( as per requirement ) TH were plotted with time against displacement
- There were 2 nodes ( 66695 & 66244 ) given to find its intrusion , which were plotted for displacements over a period of simulation time 80ms
- And found to have displacements almost similar for a period of 80ms
- There seems to be very gradual increase in the displacement due to deformation
- Displacement of about 900mm for a period of 80ms was observed
Acceleration curve :
- To determine the acceleraion over a particular position , accelerometer mounted and plotted to TH to determine acceleration over a period of time
- Acceleration generally is change in velocity with time
- By observing the graph , it shows that the acceleration has increased at the time period of 50ms to its maximum position
- This was due to the velocity at that time period would have been maximum
Output file 0001.out observation :
Time step :
Time step was observed to be constant throughout the simulation for a period of 80ms ( No fluctuation or change in timestep )
Energy error :
As far as the energy error , it was well maintained to its limit
Initially starts at 0.0% and increases to 0.1% --> 0.8% ( shows absorption of energy )
Then decreases to 0.0% --> -1.8% ( Shows dissipation of energy )
But overall the energy error was maintained near to 0% which was a good sign
Mass error :
Generally this error occurs when timestep decreases and to manage or to increase the timestep the masses are added to the model which in turn reflects in the mass error
In this case the mass error increases gradually with increase in time due to additional masses added to balance the COG and attain the target mass of 700kg
Rectify :
Avoid adding mass through nodes over a particular region , mass to be distributed evenly throughout during component preprocessing
Normal termination :
Finally after checking for outputs in the engine output file , to check whether the simulation have been considered good
At the end of the simulation , " NORMAL TERMINATION " would be printed in the engine output file
Thus the output file observation have been done using 0001.out file
Result and Conclusion :
Thus for a given model , frontal crash simulation have been carried out using Hyperworks ( Hypermesh ) , under recommended inputs and required outputs
Analysis for the simulation have been studied
Model and files :
.h3d - https://drive.google.com/file/d/1tACn02UDFI4QnLGCwWibtYl1kkA1iS5X/view?usp=sharing
T01 - https://drive.google.com/file/d/1-m1v8XDD3u-EYpLIpsDVl8fl42SDmh0S/view?usp=sharing
Model - https://drive.google.com/file/d/1T-wTNHeI1KZjPXDxQQ9ya18cXaHvDeE8/view?usp=sharing