Drop test over different ground surface
DROP TEST ON DIFFERENT MATERIALS:
INTRODUCTION:
Drop test is a type of impact simulation which takes place at a fraction of seconds. This type of simulation is performed either to determine the weak points of a cellular phone or in this report, what effect arises when the desired object falls on different objects which exhibit different material and sectional properties in its geometry.
OBJECTIVE:
To perform impact simulation of a solid object, in this case a cellular phone, against a Shell and Solid section which acts as a ground with different material properties can be accomplished using the explicit FEA solver LS-DYNA.
A drop test simulation of a mobile phone onto a ground object made of both Shell and Solid Section elements and with different material properties is performed to determine the effect of stress and strain generated during the impact between the two bodies in both the cases.
KEYWORD SETUP:
The main solid object (cellphone) is designed to be about 150.4 mm in length and the object is dropped on the ground from a pre-determined height of 20mm.
The unit system chosen is Kg/mm/ms
1. *PART KEYWORD - *SECTION & *MAT
Both the section and material properties of the main object can be assigned within the *PART keyword.
The main object (i.e cellphone) is modeled using a SECTION_SOLID ELEMENT with an Element formulation (ELFORM) of 1 which represents a “Constant shell and solid element”.
In case of MATERIAL, since most of the mobile phones used are made of aluminum, in this simulation we utilize the material MAT 024 (PIECEWISE-LINEAR-PLASTICITY) which can exhibit similar properties when its attributes are as shown below,
CASE 1: SOLID RIGID SECTION
In the case of the ground part, for the first case, the section is as similar to that of the cellphone i.e it is modeled using SECTION_SOLID with EFORM of 1.
The material chosen for modeling the ground part is a MAT 20 RIGID material whose properties are as shown below,
Note that in this case, a boundary condition at the node sets is not required as each section element is constrained on all DOF in space using the above-mentioned properties.
CMO = 1, Restricts specified DOFs in the global coordinates
CON1 = 7, Constraints translation in all x, y and z directions
CON2 = 7, Constraints rotation in all x, y and z directions
CASE 2: ELASTIC SHELL SECTION
The SECTION_SHELL property for the ground property can be assigned either from *SECTION or the *PART keyword with an element formulation, ELFORM of 2, Belytschko-Tsay element formulation.
For the material attribute, the ELASTIC MAT 001 material is chosen with the following values,
2. *INITIAL_VELOCITY
Since the main object and the ground are placed relatively close i.e a distance of 20mm approx. An initial velocity of 5 mm/ms is applied along the negative Z-axis.
However unlike most drop tests, the orientation of the drop will not be specified since this project aims at only finding the effects caused by different ground plane on the object itself.
3. *BOUNDARY_SPC_SET (SHELL GROUND)
This keyword is applied only at the end nodes in the ground Shell section elements because as stated before, the Solid ground section is made up of rigid material that constraints every element nodes on all DOFs in space.
Hence, in this case, the nodes are constrained in all DOF using the *BOUNDARY_SPC_SET keyword for the highlighted node-set shown below,
4. *CONTACT
This keyword is used to specify the type of contact experienced between the two bodies upon impact. Hence, for both cases, Shell and solid ground, we select an AUTOMATIC_SURFACE_TO_SURFACE contact keyword where the SEGM_SET for both the main object and ground object is selected with the Slave Segment (SSID) being the main (cellphone) object and the Master Segment (MSID) being the ground section.
In the above figure, we can see how each element in both the main and ground object constitute a segment. Hence each segment in the ground is collectively saved as a set, gndSEGM and similarly the same goes for the main object SEGM_SET, phSEGM.
5.*CONTROL_TERMINATION
This keyword is used to specify the total time duration for the simulation, which in our case is 5 ms.
6. *DATABASE
This keyword controls the output parameters and the necessary parameters that need to be computed from the analysis.
6.1. BINARY_D3PLOT:
This is the most important output file required. The BINARY_D3PLOT contains the animation file for the simulation. The only required input parameter that needs to be specified for this keyword file is the frequency or time interval. This parameter is used to represent the time interval/ frequency when the state of the analysis must be stored and displayed during the animation. Hence in our case we specify a DT (Time interval) value of 0.5.
This means that for every 0.5 ms the state and result of the analysis will be saved onto the d3plot which can be later viewed in LS-PrePost for the simulation.
6.2. ASCII_Output:
This keyword is used to extract the required data for the user from the analysis. For both the cases, the following Output files are requested with a time interval, DT=1 ms
ELOUT – Element Output data
GLSTAT – Global Statistics
MATSUM – Material Energies
NODOUT – Nodal data.
POST-PROCESSING:
Hence the results can be visualized in LS-PrePost using the D3PLOT file that we chose from *DATABASE.
The results for both individual cases are,
CASE 1: SOLID RIGID GROUND SECTION
We can plot the Von-mises stress each regular time intervals as,
FringeComponent->Stress->Von mises stress
From the above fig, we can observe how effective von-mises stress propagates at different states/time intervals. However, the solid ground doesn't experience stress at the point of impact reason being it has high stiffness and other parameters like the distance and velocity of the moving object being relatively slow.
NOTE: The above images represent the stress components at two-interval of time. Hence the behavior of the objects during the impact cannot be clearly understood with data extracted from two states. Hence the results can be validated accurately by increasing the frequency in the D3PLOT.
Similarly, the ASCII_Outputs can be extracted in the Post-processing window in LS-PrePost.The nodout Output files contain the nodal displacement/velocity of the selected node sets. This can be accomplished by creating, in this case, the bottom nodes of the cellphone into a NODE_SET and this can be referenced in the HISTORY_NODE_SET sub-parameter in the *DATABASE keyword. The end result is shown below,
The above figure represents the z-displacement of the selected NODE_SET. Similarly, other parameters for the selected set like displacement, velocity, acceleration can also be plotted. Below is a similar data plot for the resultant velocity of the selected node-set.
From the above graph, we can clearly visualize that upon till impact the bottom node-set move at a constant velocity of 5mm/ms as specified. During impact, the resultant velocity becomes zeros as the objects come in contact, and later the velocity for each node depending on its position from the contact region spikes up due to the reaction force caused by the ground.
CASE 2: ELASTIC SHELL GROUND SECTION
Similarly, we can observe the impact effect in case of an Elastic Shell element as shown below,
Hence, from the above figure, we can observe the von-mises stress component experience by the Ground shell elements and no stress experienced by the moving object reason being the stiffness of the ground section being low compared to that of solid ground and the main object. The ground section can also be visualized to be deforming being modeled as an elastic element that imposes less reaction force against the moving object which causes its deformation.
Similarly, the ASCII_Outputs can be extracted in the Post-processing window in LS-PrePost. The glstat data extracted is plotted below,
Another ASCII_Output parameter which we can utilize to compare the difference between both the cases is, nodout which is plotted below.
The above figure represents the resultant velocity of the bottom node set of the main object. It can be observed that while each node maintains the specified velocity upon till impact, unlike the previous case, here we can observe that after the time interval of the impact, say after 4ms, the velocity of the node-set gradually decreases. While in the previous case the velocity gradually increased after the impact caused due to reaction forces, in this case, since the ground shell element is incapable of applying reaction forces onto the phone due to its low stiffness and its own deformation, it is true that the velocity would decrease gradually.
CONCLUSION:
Hence in regards to the results obtained from above, we can conclude that in case 1, which involves a solid ground section with a rigid material, stress is induced only on the phone and not on the ground with strain experienced on neither of them. Whereas in case 2 involving a shell section with an elastic material property, the ground deforms due to low stiffness with a low or negligible strain.
The above-simulated problem and data gathered can be in fact used to model real-world applications involving an accidental drop of cellular phones on either a solid ground like graphite or a membrane-like cloth or plate which could easily mimic the second case.