Assignment 6 - Lu s Figueiredo
Assignment 6
1. Unit system
By opening the file neon_front_0000.rad (starter input deck), it was possible to check that the unit system was set to [Kg], [mm], [ms], as the following image shows. It was decided to stick to this system.
2. Rigid wall creation
In order to avoid issues related to initial contact between surfaces, the rigid wall was created a few milimeteres away from the front bumber, based on the node displayed in the following images:
2. Penetration check
To perform the penetration check, the tool \"penetration\" in the Tools menu was used. All the components were selected in order to check all the penetrations in the model. The result was that there were 66 penetrations in the whole model, as shown in the next image.
The problem was corrected by setting Inacti = 6. Which calculates a particular spring stiffness to be applied to the elements that are penetrating the surface.
3. COG and mass adjustment
In Hypercrash, it was determined that the mass of the front part of the car was 188.125 Kg and that its CoG coordinates are:
The CoG coordinates of the whole car model are given by:
And its mass is 1216.6 Kg
The distance between the two centers of gravity is 99.507 mm
But, in this particular case, the target weight is 700 Kg, so that\'s the weigth that should be achieved in this model.
This means that a concentrated mass of 511.875 Kg should be placed in the model so that the center of gravity of the whole model maintains its original position.
This can be accomplished using the following rule:
Where m1 is the current mass of the model, m2 is the added mass, and the center of mass is located somewhere between the two masses. To calculate the position in which the mass is going to be applied, the following calculations were performed:
This means that the concentrated mass must be placed in a node that is aligned with the nodes of both CoG\'s in the model and 136.078 mm apart from the CoG of the front part of the car.
The following image shows the relative position of the mentioned CoG\'s.
In the admas menu, I set the mass to 511.875 Kg and the type to 0 , since the mass is only applied to one node.
The next image shows the way the concentrated mass is applied to the actual model. I chose to apply the rigids to the edge that is located furher away from the wall to avoid perturbations of the results in the crash area.
4. Initial velocity
The initial velocity of 35 mph (15.6464 mm/ms) was applied to all nodes
5. Model checker
When the model checker was deployed, 5 errors occured:
All of these errors were corrected in order to run the simulation properly.
6. Main controls of the simulation
The time step was set to 0.5 ms
The simulation time was set to 80 ms
7. Cross section at node 174247 output
An output block was created so that the force applied on that section on each time step can be obtain in hyperview. The Variables table was filled with FNX, requesting the force going through the section in the x-axis direction.
8. Bumper - Rail cross section output
Similarly to the case above, FNX was also used to extract the force along the x-axis. To obtain the force coming from the Bumper deformation, a cross section was defined near node 174247 since I thought it would be the most appropriate way to correctly extract the required force.
9. Shotgun cross sectional forces output
The member where the force will be obtained and the respective cross-section are represented in the figure below. FNX will be set, just like as in the previous cases.
10. A pillar cross-section output
The next image defines the cross section used for this output
A circular cross section was used to limit the area of the infinte plane for the cross section
11. Accelerometer at the base of the B-pillar output
The accelerometer was placed in the position shown in the next image. AX was used inside the output block for this sensor because that would be the most severe component of the acceleration vector in that zone.
12. Intrusions in the dashboard at nodes 66695 and 66244
Over the course of the analysis, these nodes will be monitored in order to check for intrusions.
13. Results
a) Hyperview results
c) Force on the section definined by the node 174247
d) Bumper-rail cross section force output
e) Shotgun cross section force output
f) A-pillar cross section force output
g) Rocker accelerometer output
14. Conclusions
The two cross sections defined by the node 174247 and the bumper_rail cross section are located in the same component and are really close to each other. But, the results obtained are not consistent, since the cross section forces detected on the 174247 are almost neglible. With these results, I believe that the cross section on node 174247 was not properly defined and therefore the results obtained from there should not be considered. The only reason why a second analysis wasn\'t performed is that simillar acceptable results can be obtain in the bumper_rail cross section, which is located nearby the cross section with faulty results.
In some of the plots displayed above, the results measured for the forces assume, in a general way, positive values instead of negative ones (It would be expectable for these forces to assume mostly negative values beacuse the whole car is under compression). But this is a problem that arised from the definition of the plane normal, that, in some cross-sections, was defined backwards. And therefore, the provided results were given with the wrong sign.
Even though the results obtained for the cross section forces in the A-pillar and for the shotgun cross sections were very low, the values obtained are coherent because the A-pillar has the objective to pass the impact energy from the front of the car to the back as quickly as possible, and, therefore, it doesn\'t undergo high values of cross section force, since that force is experienced by the members on the back of the car instead. The member under the shotgun seat doesn\'t take as much energy from the impact as the side members of the chassis, for example. So the results obtained from this cross section were considered to be acceptable.
In the following image, it\'s possible to see how the addition of hole features to the parts helps them to deform in a particular way. In this case, the fender panel bends furher away from the cockpit to avoid injuries to the occupants due to the addiction of a hole in that area.
The accelerometer defined previously only measures the acceleration component in the x direction, which is considered to be the most relevant for a front impact case scenario, and given the fact that the car travels in the x direction.
There was not any intrusion in nodes 66645 or 66244. This is due to the fact that the fender bends into the dash wall but the simulation time doesn\'t allow its movement to continue into the dash wall. If the simulation time was longer or if the rigid wall was placed closer to the bumper, there would have been intrusions in the dashwall.
As it was menstioned in the previous paragraph, the fact that the rigid wall was placed so far away from the car makes most of the simulation time useless and the results may be unreliable, since better accuracy could be achieved if all the simulation time was spent in the actual crash instead on the displacement of the car form its original position to the rigid wall.