Assignment 7
Objective: To conduct a side crash analysis using RADIOSS. Performing it to extract the forces experienced at the rails and intrusions at the hinge pillar, b pillar, and fuel tank region.
Case Setup and Execution:
Unit check: Open 0000.rad file to check the units. It shows that the units used here are [ Kg, mm, ms ]. Thus the units followed here are and in the 300k model is [kg, mm, ms].
Interface Creation: Now we need to create the interface. The interface here used is TYPE 7 as it is a universal contact. Go to Solver > right-click and select Inter > select the card image as Type 7 > For slave and master nodes select all the 98 components > set the other values as mentioned in the image below.
Rigid Wall Setup: Rigid wall can be created in the model viewer panel. Right-click on the panel select the rigid wall option > ensure the geometry type to be a cylinder, then in the engineering data set the base node (Xm, Ym, Zm) as 2150, 1009, 2000 > now set the normal as 0,0,1 as the plane is perpendicular to the x-axis and the opposed the motion of the frontal biw model. Set the slide as 2 (this is sliding with friction) also here I set the Dsearch as 1000.0 and the Coulomb friction must be 0.1 as asked in the question. The diameter of the cylindrical rigid wall used for this analysis is 273 mm, this number is obtained through web research.
Initial velocity: The provided velocity is 35m/hr which is 15.646 mm/ms. Go to the solver browser > right click > create > under engine keywords select INIVEL > select the type as TRA (as the type of motion is translation) > ( as it is only in y direction) Set Vy = 15.646 > in Grnod_id select all the nodes.
Time step and run-time: In the solver, browser go to the engine keyword, under DT select DT/Noda > Select eng_dt_noda_file > change Tmin to 0.0005 as time step is given as 0.5 microseconds. Whereas to set the runtime, under the engine keywords select RUN > then in the eng_run file set Tstop to 80.0. For the frequency, I set it as 0.5 by editing the Anim/DT engine file.
Mass addition and Positioning of CG: To admass and change the center of gravity to match it with the full-scale model we use the HYPERCRASH application. To access hypermesh go to Applications> click hypercrash > set the units as KN, kg, mm & ms > then drag the edited frontal crash file. Then on the other hand import the 300k model but set the units as N, T, mm, s.
In the full-scale model go to the Mass tab and select balancing. A new tab opens which shows the entire mass of the full-scale body. The mass of the full-scale model is 1216.555 Kg with its Cg coordinates to be (X = 3124.1 mm Y = 5.77mm). Now the mass in the frontal crash should be added in such a manner that the CG of the frontal crash BIW is approximately the same as the full-scale model at X and Y coordinates.
Now pull up the side crash model in the hypercrash window and double click on the Added Mass and then click create new. Select the Type=0,1,2,3,... as necessary to position the CG closer to the full-scale model. I added masses using mostly type 1 card to position the CG.
The closest CG I could obtain was (2940.5mm, 5.22mm). With the mass of the model to be 710.46 Kg.
Now when this is done go to the model in hypermesh and create sents for the nodes used to add mass. Then in the solver browser, go to ADMAS > right-click to create > then create the similar masses at the same position as in the HYPERCRASH window. This method is chosen as it is easier to view the change and add mass in the hypercrash window.
Model checker and penetration checker: To check for penetrations in the hypermesh window go-to tools > then click penetration check > under selections select all the type 7 contact group > then click run. No penetrations were observed in my model. If there are any penetrations then click click depenetrate or use the translate feature.
To check model go-to tools > model checker > Select RADIOSS block > then run it. Several errors and a warning will appear with a short description of it. Read the descriptions and performs actions accordingly to omit those errors. If there are any free nodes or an empty feature then we can delete it.
Sections & Intrusion: Two sections have to be created to extract the forces experienced by those parts in that section throughout the crash process. To do so in the model browser we right-click > create > then select cross-section > select the parts that have to be included > then select the base nodes (we can use the provided node id here) and the N1, N2 direction. Section to be created is for B-pillar, and hinge pillar.
To calculate the intrusions we need to create springs. Three intrusions are recorded thus we have to create 3 springs in the model. In the browser we go to mesh > create >1D elements > select springs > insert the node id. The properties in the given image are assigned to the created springs.
Output blocks: In the model browser > right click on the output blocks > select create > then in entity id enter the type of entity for which the TH has to be extracted. Use this to create an output block for sections, nodes, and springs. An output block to record the velocity at node 337773 must be created.
Results and Discussion
Section Forces: The following plots show the variation of resultant and normal force through the sections created at the B pillar and hinge pillar region:
Total resultant force plot:
Resultant Normal Force plot:
In both the plots, we notice that the hinge pillar experiences more force at the section. Even though the B-pillar is closer to the area of impact the force experienced by it is seen to be relatively low in comparison to the hinge pillar. The variation of force through both the section is seen to be similar however the intensities are different. The maximum force experienced by the hinge pillar is 6kn whereas the Bpillar experiences 3kN of force which is half of what the hinge pillar experiences.
Intrusions: To calculate the intrusion the resultant elongation in the created spring is recorded for the entire crash. Here its seen that the maximum intrusion is observed in the fuel tank region, which is around 875 mm for a crash duration of 80ms. Followed by B-pillar which is around 740mm at the end of the 80th ms. The least intrusion is observed at the hinge pillar which is almost 500mm. The increment in the intrusion is quadratic and if we observe a little more in detail then it can be noticed that the rate of increase in the intrusions at the fuel tank and B-pillar begin to decrease at the end of 80ms whereas for the hinge pillar the intrusions keep increasing.
The intrusions in the fuel tank region can be decreased by adding a rail from the fuel tank region to the other side of the vehicle to obstruct the deformation undergone due to crash.
Peak Velocity: The following plot is the velocity plot at node 337773 to record its peak velocity during the crash. Due to the crash, the velocity rapidly drops in the first 10 ms of the simulation.
Conclusion
The side crash was correctly performed in hypermesh using RADIOSS. There was no mass error observed in the simulation but an energy error of -1.2% was recorded which is completely under the error tolerance. The required outputs were appropriately recorded and plotted using hypergraph 2D. In the outputs, it was noticed that the hinge pillar experiences more force at the section and the intrusions at the fuel tank region are the most.
Google Drive: https://drive.google.com/drive/folders/1QQOe0qPYXolInyMV4hyAhhyO8a1TJbAL?usp=sharing