Week - 5 - Modelling Spotwelds

Objective:

To perform a simulation between brackets (connected using spotweld) and a rigid wall and to derive the requested outputs.

Introduction:

• Spotwelds are physical connections between two (or multiple) layers of parallel sheets of metal. The only thing needed for creating a spotweld connection is its coordinate. The weld coordinate does not have to be on a FE node position, but the layers of parts to be connected should be reasonable parallel.
• In Ls-Dyna Spotweld can be modelled using beam element and *CONTACT_TIED_SHELL, *CONSTRAINED_SPOTWELD, *CONSTRAINED_GENERALIZED, using solid element and *CONTACT_SPOTWELD.
• In this assignment the simulation was performed to simulate a crash test of the bracket parts connected using spotweld made of beam and solid element accordingly and the results obtained during the post-process are then compared.

Procedure:

• The given model is imported onto the workpage of the LS-Pre-post using File>Open>LS DYNA Keyword file.

Assigning material and section to the brackets.

• Materials can be classified as elastic materials, elastic plastic materials, hyper elastic materials, viscoelastic materials etc, in which metals and plastics can be modelled using elastic plastic material and is assigned using *MAT_024 (PIECEWISE_LINEAR_PLASTICITY) card from the keyword manager. If failure mode is highly complex or the material is a polymer, composite, thermoplastic, then *MAT_187 card can be used.
• Here the material for the brackets is given as steel and the propeties such as mass density, youngs modulus, poissons ratio, tangent modulus etc were assigned to the material card. Tangent modulus ( Slope of stress strain curve) is taken as 2% of the Youngs modulus.
• A section id with shell card is created for the part with (ELFORM=2) using Keyword manager>All>Section>Shell and a thickness of 1.2mm is assigined.
• Now these created material id and section id are assigned to the brackets using the part option under keyword manager.

Modelling spotweld using beam element.

• To model spotweld, beam elements were created using Mesh>Element generation>Beam and thereby picking up the nodes where the beam to be created. Here four spotwelds were used in a symmetric manner and are occupied in same part id.

• When the beam element gets created Ls-Dyna defaults creates a mat and section id, where it can be deleted or used accordingly.
• *SECTION_BEAM were created with ELFORM=9, also thickness value and the cross section (Tubular) type were also assigned.

• *100_SPOTWELD card has been used to assign the matrial for the spotweld and the parameters such as mass density, youngs modulus, poissons ratio, tangent modulus, EFAIL, NRR, NRS,NRT, MRR, MSS, MTT etc were assigned accordingly. Here its important to mention the TFAIL inorder to perform a time based failure.
• Spotweld failure can be performed as time based failure and force based failure. Time based failure can be used if the property of the weld is unknown (correlation with some experimental data is needed).

• The section and material card created for the beam element are now been assigned the part id under keyword manager.

Creating a Rigid wall.

• The given model does not contain any rigid wall needed for the analysis. A new rigid wall is been created using Create Entity> Rigid Wall.
• Here a planar wall is created by defining a vector with one point and a normal. The rigid wall is then offseted to a distance of 10mm from the bracket.
• It should be noted that the normal created by the rigid wall and the velocity of the brackets should be in an opposite direction. If its been in the same direction the bracket gets penetrated through the wall resulting a wrong simulation.
• Since the generated rigid wall and direction of velocity were in same direction, The normal is made opposite to each other by using the rotate command and specifying the angle and axis its to be aligned.

Assigning velocity to the parts.

• Before assigning the velocity to the the bracekts, a node set containing all the nodes of the brackets has to be created. This can be done using create entity option or by set option under keyword manager.
• The distance between the bracket and Rigid Wall is around 10mm and with a initial velocity of 13.88mm/ms the impact should happen at t=0.7ms.
• The velocity is assigned from the *INITIAL _VELOCITY card under keyword manager and a velocity of 13.88mm/ms is assigened in X direction.

Giving contact definition.

• Contact in LS DYNA is defined using contact surfaces and occurs when one surface of the outer surface of a model penetrates another segment.
• Here an *AUTOMATIC_SINGLE_SURFACE card from keyword manager is been used and Slave segment set ID (SSID) is assigned to the part set containing both the parts of bracket and leaving the MSID blank.
• Another contact card of *TIED_SHELL_EDGE_TO_SURFACE were used to define the contact between the spotweld and the brackets. Whenever modelling a spotweld, the master set id will always be the parts that is being connected using the spotwelds and slave set will include all the spotweld definitions.

Creating Set data of nodes and elements.

• A set of nodes an elements are created at the middle part of the bracket in order to find out the cross sectional force (SECFORCE), Acceleration of the node etc at the middle part along its lenght. Other method to do this process is to create a cross section plain in some coordinate along the bracket using *DATABAE_CROSS_SECTION_PLANE. Here first method has been followed.
• These sets can be created using Create entity> Set Data.

Assigning termination time.

• *CONTROL_TERMINATION_CARD from keyword manager is used to define the simulation time using ENDTIME otpion.
• Since the brackets hits the rigid wall with a velocity of 13.88mm/ms and the impact happens at 0.7ms. therefore the ENDTIME is assigned as 1ms (Additional time after the impact is added to get the crash, bouncing, buckling effect etc).

Database.

• The oputput data needed for the analysis is generated using *DATABASE card.
• DATABASE_BINARY_D3PLOT under keyword manager ensures d3plot files are written in which it stores the stress strain contours of the deformed system.
• ASCII under keyword manager can generate different output such as force generated in a contact, total energy, material engery etc. Here output for ELOUT, GLSTAT, MATSUM, RCFORCE, NODOUT, SLEOUT,SWFORCE are been requested.
• DT (Time intervel between the output) is an essential parameter which basically denotes the time intervel after which the state of solution need to be written and the value is kept as 0.01 for this analysis.
• *DATA_BASE_CROSS_SETION card can be used to generate the section force by assigning the NSID and SSID accordingly.
• *DATABASE_HISTORY_NODE card was also created to get the acceleration plot of a node in the middle of a crash box.
• *DATABASE_EXTENT_BINARY card was provided by specifying STRFLG which write strain tensor data to d3plot and elout.

Model check up and running the simulation.

• The model is checked for any errors and if any error occurs it should be corrected further.
• The .k file is been saved and runned using LS-DYNA manager.
• While running the simulation, It can be seen that spotweld failed at a time as per the data provided.

• If Normal termination is showed after running the file, its then ready to do post processing.

Modelling spotweld using solid element.

• To model spotweld using soild element, a hexa element type were created using Element tools>Element editing and picking up the four nodes where the solid element to be created. Here also four solid elements were used as that of the one modelled using beam elements and occupied in the same part id.

• *SECTION_SOILD was used to create a section card with an ELFORM of 1.
• The same material card with all the properties that was used for the above model was assigned to this model too.
• The section and material card created for the solid element are now been assigned the part id under keyword manager.
• Apart from the *AUTOMATIC_SINGLE_SURFACE contact, *CONTACT_SPOTWELD card is also defined and thereby assingning the master set and slave set accordingly.

All the other procedure remains same as that of the model discussed above.

Results:

• Modelling of spotweld using beam and soild element were performed and simulated crash test for the modelled parts on the rigid wall sucessfully.
• Results for the required outputs were also collected and made a comparision between the simulation performed using beam and soild element.

Animation of the final simulation.

Spotweld forces.

Spotweld Axial force.

• The above plots shows the axial force for the beam and solid element respectively. When the part gets crashed to the rigid wall axial force gets to a maximum value and reaches zero when the spotweld fails. Here the axial force is shown maximum for the solid element when compared to beam element, this is beacuse the solid element (Hexa) are more stiffer compared to beam element.

Spotweld shear force.

• The maximum magnitude of the shear force generated for the solid element is found to be much higher than that of beam element. Since the model is symmetric the curve line for the respected spotwelds are seen to be drawn close to each other.

Stress-strain plots.

Effective stress(v-m).

Spotweld modelled with beam element. Spotweld modelled with solid element.

• Von Mises is a theoretical measure of stress used to estimate yield failure criteria in ductile materials and is also popular in fatigue strength calculations (where it is signed positive or negative according to the dominant Principal stress), whilst Principal stress is a more "real" and directly measurable stress.
• Here the stress plot remains same for both the cases and it can be seen that the stress rises to 376 MPa which means after the impact the brackets yields and undergo plastic deformation.

Effective plastic strain.

• From the plot it can be seen that the strain curve attains a peak value and become constant. Its due to the plastic deformation occuring in the brackets when its crashes the rigid wall.
• The stain value for the element-101639 is found out to be maximum for solid element compared to beam element.

Cross-sectional force generated for the top bracket.

• The above plots shows the magnitude of secforce in the top bracket with respect to time.
• The magnitude of the force along X-direction is larger compared to other two Y and Z direction, Since the motion of the brackets are along the X- Axis.
• The resultant force for the spotweld modelled with solid element is found to be maximum compared to beam element.

Acceleration plot of a node in the middle of top bracket.

• The above plot shows the acceleration of Node-501440 along X,Y,Z direction and also the resultant acceleration with respect to time.
• The parts intially moves with a constant velocity and when it crashes the rigid wall its starts deaccelerating. The magnitude of the acceleration in Y and Z axis is small and the acceleration along X- axis for spotweld modelled using beam element is found to be `166(mm)/(ms^2)` and that of solid element is found to be `284(mm)/(ms^2)`which has higher magnitude.

Energy plots.

• The above plot shows various energy plots for the parts modelled with spotweld using beam and solid element respectively in which the plots of solid element shows higher energy values.
• The plot for the Kinetic energy is constant for some time since the brackets moves with a constant velocity and when it crashes it deaccelerates and kinetic energy of the brackets decreases which is shown by a decline curve in the plot.
• Internal energy of the crash box shows a steep increase after the crash happens. Here the kinetic energy has been stored as internal energy following the law of conservation of energy theory.
• Total energy remains constant.
• Hourglassing is a state of strain, which is free of energy (ZEM: Zero Energy Mode) and can emerge in case of one-point-integrated solid- (hexahedrons) and shell elements. Hourglass modes are mostly caused by concentrated loads and contact (contact force at several nodes ). Hourglassing can be reduced by refining the mesh, Using a fully integrated element type etc. but takes more computation time when fully integrated element type is used.
• Here the hourglass energy is very small and is kept with in the range.
• Sliding energy (GLSTAT - global) says about the numerical validity of the contacts. If friction is zero, than the GLSTAT value should be 1% or less of the internal energy. Since there is no sliding occuring, the value of sliding energy is equal to zero.

Conclusion

• The .k file for the crash simluation has been made from the scratch and perfomed the simulation with all the required outputs.

............................................................................................................................................................................................

File: https://drive.google.com/drive/folders/17zcjsexViLA2Bj9TrdkNkIljPaNfN3BW?usp=sharing