Week - 5 - Modelling Spotwelds

COMPARISON OF SPOTWELD MODELLED AS BEAM AND HEXA ELEMENT USING LS-DYNA

AIM: To model spot weld as beam and hexa elements and create a complete simulation file for crash analysis from the given FE model of assembly of parts and produce the following deliverables.

1. Input .k file and output files (d3plot, glstat, sleout, rcforc, secforc, swforc)
2. Animation of the final simulation
3. Compare axial force, shear force, resultant force and spotweld length plots of spotweld modelled as beam and hexa element.
4. Plot of all energies (total, internal, kinetic, hourglass, sliding)
5. Cross-sectional force generated in the middle of the given assembly of parts (along its length)
6. Acceleration plot of a node in the middle of the given assembly of parts (along its length)
7. Maximum directional stress and strain along the length of the given assembly of parts (X strain, Y strain etc)

INTRODUCTION:

Spot welding is one of the oldest welding methods for joining metal plates using strong current and is widely used in all industries such as automobile, shipbuilding and aerospace sectors. In this simulation the spot weld is modelled as beam and hexa element and results are compared for both the cases.

Following are the different assumptions and information required:

1. The given assembly of parts should crash into a rigid wall created using *RIGIDWALL_ keyword.
2. Material for the given assembly of parts is steel.
3. Thickness of the given assembly of parts is 1.5 mm.
4. Initial velocity for the given assembly of parts is around 50 kmph.
5. The unit system used is gm-mm-ms.
6. The spotweld is modelled as case (i) beam element, (ii) hexa element.
7. The simulation is carried out for both the cases and results are compared and interpreted.

PROCEDURE:

The given LS-Dyna keyword file is opened in LS-PrePost using option File>Open>LS-Dyna Keyword File as shown in the fig. 1.

Fig.1. Imported FE model.

SPOTWELD MODELLING:

Mesh>EleGen>Beam>Two_Node_Sets.

Fig.2.1 Beam element.

Mesh>EleEdit>Create>Hexa.

Fig.2.2 Hexa element.

The spotwelds are modelled as case (i) beam element (ii) hexa element. In the first case six beam elements are created as shown in fig. 2.1. In the second case using element edit option the hexa elements are created in the position of beam elements as shown in fig. 2.2.

PART DEFINITION:

The part is defined by assigning the section properties and material properties. In this model there are three parts i.e. top bracket, bottom bracket and spotweld.

Section properties:

Keyword manager>SECTION>SHELL.

Fig.3.1. Section shell.

The section properties of brackets are assigned as shell element with 1.5 mm thickness and ELFORM=16, Fully integrated shell element (very fast).

Keyword manager>SECTION>BEAM.

Fig.3.2. Section beam.

The section properties of spot weld in the first case are assigned as beam element with 3 mm thickness and ELFORM=9, spotweld beam, see *MAT_SPOTWELD (Type 100)

Keyword manager>SECTION>SOLID.

Fig.3.3. Section solid.

The section properties of spot weld in the second case are assigned as solid element and ELFORM=1, constant stress solid element (default),

Material properties:

The brackets and spot welds are assigned with steel material with following parameters as shown in table.

Steel material properties	Value
Mass Density [gm/mm3]	7.85e-3
Young’s Modulus [MPa]	210e3
Poisson's ratio	0.3
Yield stress [MPa]	250
Tangent modulus [MPa]	210e2

Keyword manager>MAT>024-PIECEWISE_LINEAR_PLASTICITY.

Fig.4.1. Steel material

MAT24 (Piece wise linear plasticity) material card is used to assign the steel material properties to the assembly of parts. The MAT24 represent Piecewise linear isotropic plasticity. With this material model it is possible to consider the effect of the strain rate.

Keyword manager>MAT>100-SPOTWELD.

Fig.4.2. Spot weld material.

This material model applies to beam element type 9 and to solid element type 1. The failure models apply to both beam and solid elements.

Part definition:

Keyword manager>PART

Fig.5. Part definition of assembly of parts and spot weld.

BOUNDARY CONDITIONS:

Rigid wall:

Create Entity>Rigidwall>cre>Planar>1n+NL

Fig.6. Rigid wall.

The rigid wall is created at a distance of 10 mm away from the brackets.

Initial Velocity:

Keyword manager>Initial>Velocity

Fig.7. Initial velocity.

The brackets are assigned with initial velocity of V_x= 13.88 mm/ms is applied along x direction to crash against the rigid wall.

CONTACT CONDITION:

Keyword manager>CONTACT>AUTOMATIC_SINGLE_SURFACE

Fig.8.1. Contact between Rigid wall and brackets.

The AUTOMATIC_SINGLE_SURFACE contact type is assigned for the contact between rigid wall and brackets. It is quite helpful to apply this contact method in the crash models because all the elements are included in one single set and LS-DYNA considers also when a part comes into contact with itself. The FS and FD that are static and dynamic friction coefficient with a value of 0.08 is entered in the contact card.

Keyword manager>CONTACT>TIED_SHELL_EDGE_TO_SURFACE

Fig.8.2. Contact between Spotweld beam element and brackets.

The TIED_SHELL_EDGE_TO_SURFACE contact type is assigned for the contact between spotweld beam elements and brackets.

Keyword manager>CONTACT>SPOTWELD

Fig.8.3. Contact between Spotweld solid element and brackets.

The SPOTWELD contact type is assigned for the contact between spotweld solid elements and brackets.

CONTROL FUNCTION:

Keyword manager>CONTROL>ENERGY

Fig.9. Control energy.

The control energy function is enabled for computing the hourglass energy, stonewall energy and sliding energy.

Keyword manager>CONTROL>TERMINATION

Fig.10. Control termination.

The control termination function is enabled to specify the end time of simulation. The termination time is set for 5 ms to capture the effect of brackets striking the rigid wall.

DATABASE OPTION:

Keyword manager>DATABASE>BINARY_D3PLOT

Fig.11. Database binary_D3plot.

The time step value of 0.1 ms is given for the BINARY_D3PLOT and 0.01 ms in the DATABASE_ASCII option for GLSAT, MATSUM, NODOUT, RCFORC, RWFORC, SECFORC, SWFORC and SLEOUT.

Keyword manager>DATABASE>EXTENT_BINARY

Fig.12. Database binary extent.

DATABASE_EXTENT_BINARY card with STRFLG =1, is used to compute the elastic strain in the model.

Keyword manager>DATABASE>HISTORY_NODE

Fig.13. Database history node.

DATABASE_HISTORY_NODE card is used to compute the acceleration of a 501385 node in the model.

Keyword manager>DATABASE>CROSS_SECTION_PLANE

Fig.14. Data base cross section plane.

DATABASE_CROSS_SECTION_PLANE card is used to compute the sectional force of the model.

The keyword file created is checked for errors using the option keyword manager>model check. The keyword file is saved using ‘.k’ extension and is made to run in the solver by getting normal termination message for both the cases i.e. (i) Spot weld modelled as beam element, (ii) Spot weld modelled as hexa element

RESULTS:

The D3plot output file is opened in LS-PrePost using option File>open>LS-Dyna binary plot.

1. The animation of Von-Mises stress contour is as shown below.

case (i) case (ii)

2. The animation of Effective plastic strain contour is as shown below

case (i) case (ii)

3. Plot of Axial force in spotweld

Fig.15.1 Plot of axial force in spotweld modelled as beam element.

Fig.15.2 Plot of axial force in spotweld modelled as hexa element

In the first case the axial force transferred to the spot weld gradually fluctuates with a maximum value of 1.65 kN. In the second case the axial force transferred to the spot weld rapidly fluctuates with a maximum value of 0.848 kN comparatively less than the first case, since the spotweld is modelled as solid hexa element which is rigid compared to beam element.

4. Plot of Shear force in spotwelds

Fig.16.1 Plot of shear force in spotweld modelled as beam element

Fig.16.2 Plot of shear force in spotweld modelled as hexa element

In the first case the shear force generated during collision increases suddenly to value of 1.06 kN and thereafter decreases gradually. In the second case the shear force generated during collision increases suddenly to value of 1.9 kN and thereafter fluctuates till rebounding.

5. Plot of Resultant force in spotwelds

Fig.17.1 Plot of resultant force in spotweld modelled as beam element

Fig.17.2 Plot of resultant force in spotweld modelled as hexa element

In the first case the resultant force generated in the spot weld gradually increases during collision and fluctuates with maximum value of 1.79 kN. In the second case the resultant force generated in the spot weld increases suddenly during collision and fluctuates with a maximum value of 1.91 kN.

6. Plot of Length of spotwelds

Case (i)

Case (ii)

Fig.18 Plot of spotweld length

From the graph for the first case the length of the spot weld is gradually increasing till a time of 2.48 ms, since the spotweld is modelled as 1D beam element. In the second case the length of spot weld remains constant since it is modelled as solid hexa element and is able to withstand the force generated during the collision.

7. Global energy plots

The energy plots comprising of kinetic energy, internal energy, total energy, hourglass energy, sliding energy and stonewall energy were plotted for both the cases of spotweld. The graph represents an energy balance of a dynamic test on an assembly of parts and it can be noted that as the kinetic energy decrease, the internal energy increases as expected from the theory. The other important things to consider are that the total energy has to remain constant and the sliding interface must remain low.

Case (i)

Case (ii)

Fig.19. Global energy plots

From the energy plot graph, it is observed that for the first case the kinetic energy is reduced during the time of collision to a value of 73.9x10³ N-mm and the internal energy is increased to a value of 68x10³ N-mm. For the second case the kinetic energy is reduced during the time of collision to a value of 72.7x10³ N-mm and the internal energy is increased to a value of 67.9x10³ N-mm.

8. Sectional force plots

Case (i)

Case (ii)

Fig.20. Plot of sectional resultant force.

In the first case a resultant sectional force of 5.7 KN is developed during the collision with rigid wall and increases to a maximum of 9.81 KN at time 1.67 ms and decreases while the crash box is rebounced.

In the second case a resultant sectional force of 10.4 KN is developed during the collision with rigid wall and increases to a maximum of 10.8 KN at time 1.69 ms and decreases while the crash box is rebounced.

9. Acceleration plot of a node

The resultant acceleration is plotted for a node 501385 in the assembly of parts.

Case (i)

Case (ii)

Fig.21. Plot of resultant acceleration of a node in the model.

In the first case a resultant acceleration of 106 mm/ms² is developed during the collision with rigid wall and increases to a maximum of 733 mm/ms² at time 3.54 ms and decreases while the brackets are rebounced.

In the second case a resultant acceleration of 358 mm/ms² is developed during the collision with rigid wall and increases to a maximum of 1060 mm/ms² at time 4.78 ms and decreases while the brackets are rebounced.

10. Maximum directional stress and strain along the length of the assembly of parts.

Case	Maximum Stress, MPa			Maximum Strain
	X-direction	Y-direction	Z-direction	X-direction	Y-direction	Z-direction
1	535.9 in Element 100988	509.3 in Element 101737	466 in Element 104313	0.0131 in Element 104316	0.01 in Element 104318	0.0137 in Element 102392
2	457.8 in Element 102388	322.8 in Element 107725	342 in Element 100921	0.0065 in Element 101601	0.0073 in Element 100953	0.0102 in Element 101660

11. Plot of v-m stress vs effective plastic strain

Case (i)

Case (ii)

Fig.22 Plot of v-m stress vs Effective plastic strain.

The v-m stress vs effective plastic strain is plotted for the element 101712. From the graph, the maximum stress generated in element 101712 for the first case is 608 MPa and for the second case is 310 MPa which is greater than the yield stress value hence the thickness of the model has to be increased.

CONCLUSION:

1. The keywords necessary to build the spot weld modelling analysis was created from scratch.
2. The model was solved successfully and the results are compared for both the cases i.e. (i)spotweld modelled as beam element (ii) Spotweld modelled as hexa element.
3. The force transferred in the first case is better than the second case.
4. The maximum stress and strain induced in the first case is more compared to second case.
5. The deformation of the spotweld is more in the first case compared to second case.
6. The first case is preferred for ease modelling with less computional time but for better results second case is preferred.
7. Learned to build a solver deck for modelling spotweld as beam and hexa element for crash analysis in LS-Dyna and extract the outputs to interpret and compare the results.

Google Drive Link:https://drive.google.com/drive/folders/1ET8ZxjSHF-9sCIeARe2N_D43oo5CffG-?usp=sharing