Study of contact modelling interface types, parameters, its accuracy on results and the structural mechanics of a tube with varying cross sectional area.

Study of contact modelling interface types, parameters, its accuracy on results and the structural mechanics of a tube with varying cross sectional area.

 

Objective:

The objectives of this project are

1. To mesh the given model

2. To study the effect of different interface types for contact modelling, on the accuracy of results

3. To study the correlation between geometrical change like change in cross section of the tube and its impact on the structural mechanics.

Bumper meshing:

The geometry was cleaned up, components were meshed and equivalence operation was done on elements on the intersecting portion of components BumperBeam_front, BumperBeam_front and BumperBeam_mid to maintain perfect element connectivity.

 

Contact Interfaces:

1. Contact interfaces define the behavior and mechanics of components that come into contact.

2. For example, if an object hits or collides with another object, energy is transferred through the contact surface area of the two objects.

3. If the interface between the two objects are not defined, the objects coming into any kind of contact may just pass through each other without any transfer of energy, which does not happen in real life and thus produces incorrect results.

4. Hence, whenever a scenario where one or more parts coming into contact is predicted, contact modeling or interface modelling is done to accurately simulate the physical phenomenon that occurs when parts come into contact.

5. The contact might be with the same object or component (Self-contact or self-impact) and/or different components.

6. For self-contact or self-impact, depending on the type of contact model, the edges or nodes or surfaces of the same component are selected for creating contact interface

7. For contact with other components, depending on the type of contact model, the edges or nodes or surfaces of the master and slave components are selected for creating contact interface

8. As a general rule of thumb, the component with a coarser mesh is chosen as a master component and the component with a fine mesh is chosen as slave component.

 

Model:

Since this assignment involves studying the difference in structural mechanics of the tube because of difference in geometry, the given model or part is split into two regions as mentioned below to clearly study the mechanism of deformation,stress formation,etc.

Region 1 - pshell_2mm (Dark Yellow), pshell2_2mm (Grey) – Near the wall

Region 2 - PSHELL1 (Light Gold), PSHELL2 (Light Green) – Away from the wall

 

Steps to create TH/Part:

1. Import the model in hypermesh

2. In the model browser, right click on Output Blocks and choose create

3. Click on the newly created card and in the bottom left corner, click on the small arrow near entity and click on components

4. Now select all the parts in the model.

5. In the NUM_VARIABLES option, input 2 and choose Internal energy and Kinetic energy

6. The model is then exported and run.

 

Case 1: Crash tube with type7 contact and no explicitly defined Inacti parameter.

1. In this, the supplied model has type7 contacts and all the contact parameters are default values except no explicitly defined Inacti parameter

2. Thus, the model is run as it is and the graphs and contours are recoded using hypergraph 2D and hyperview.

3. The internal energy gradually increases and at the end, remains more or less constant. The internal energy is 42011.87 Joules

4. The internal energies at end of simulation for parts in Region 1 are 7403.36 Joules and 7431.786 Joules respectively

5. The internal energies at end of simulation for parts in Region 2 are 13615.75 Joules and 13560.96 Joules respectively

6. The maximum resultant normal force is 1340.59 kN at 26.7 ms during the loading and the maximum resultant tangential force is 5.29 kN at 20.4 ms during the loading.

 

7. The contact forces contour shows that the maximum contact force is 93.85 kN

  

8. Maximum Von-mises stress is 0.696 Gpa

 

Case 2: Crash tube with type7 contact and inacti value set to 6

1. This is more or less similar to case 1 except here, the parameter inacti is changed to 6

2. When inacti=6, if initial penetrations exist, the gaps are reduced and scaled using a specific scale value to prevent locking of master and slave entities.

3. The internal energy is 42011.87 Joules

4. The internal energies at end of simulation for parts in Region 1 are 7403.36 Joules and 7431.786 Joules respectively

5. The internal energies at end of simulation for parts in Region 2 are 13615.75 Joules and 13560.96 Joules respectively

6. The maximum resultant normal force is 1340.59 kN at 26.7 ms during the loading and the maximum resultant tangential force is 5.29 kN at 20.4 ms during the loading.

7. The contact forces contour shows that the maximum contact force is 93.85 kN

 

8. Maximum Von-mises stress is 0.696 Gpa

9. As we get the same results of case 1, it can be concluded that there are no intial penetrations.

 

Case 3: Crash tube with type7, type11 contacts and inacti value set to 6

1. In this case, a type11 contact is defined for preventing edge to edge penetration

Steps to create type 11 contact

a) Open model in hypercrash

b) Double click interface card in browser

c) Create new interface – Type11

d) Define the master, slave surfaces and the parameters.

2. The model is then exported and run in hypermesh.

3. The internal energy is 42096.76 Joules

4. The Internal energy in this case is 366 Joules more than the internal energy in case 2

5. The internal energies at end of simulation for parts in Region 1 are 7585.92 Joules and 7657.9 Joules respectively

6. The internal energies at end of simulation for parts in Region 2 are 13510.58 Joules and 13624.02 Joules respectively

7. The maximum resultant normal force is 1369.65 kN at 26.6 ms during the loading and the maximum resultant tangential force is 9.63 kN at 24.6 ms during the loading.

8. The contact forces contour shows that the maximum contact force is 108.2 kN

9. The slight increase in internal energies of parts when compared to case 2 proves that there were some unaccounted edge to edge contacts in the model used in case 2

10. Hence, for this model, defining type 11 contacts is vital to get accurate results

11. Maximum Von-mises stress is 0.67 Gpa

 

Case 4: Crash tube with no notches and no boundary conditions on rigid body

1. In this case, the boundary conditions are deleted by deleting rbody_12456 in the Load Collectors option in browser of hypermesh

2. The notch is removed by either translating the elements / nodes by an appropriate distance.

3. In our case, the nodes were aligned using the align node option under node edit.

4. The model is then exported and simulation is run

5. The Internal energy is 42635.3 Joules

6. The internal energies at end of simulation for parts in Region 1 are 8725.54 Joules and 8663.47 Joules respectively

7. The internal energies at end of simulation for parts in Region 2 are 12629.81 Joules and 12616.47 Joules respectively

8. The maximum resultant normal force is 1177.91 kN at 27.4 ms during the loading and the maximum resultant tangential force is 18.36 kN at 17.4 ms during the loading.

9. The contact forces contour shows that the maximum contact force is 57.14 kN

10. Maximum Von-mises stress is 0.632 Gpa

 

Case 5: Crash tube with new notch in the middle along the whole section

1. In this case, a new notch is created in the middle part of the tube by translating the elements towards the inside of the tube.

2. The elements along the four sides/faces of the tube are selected for translation operation

3. While translating elements along the flat faces is easy by defining the axis or plane of translation, elements around the corner or curved edges are hard to move.

4. Hence, to translate the corner elements along an angle of approximately 45 degrees,

5. Three nodes are created at co-ordinates (0, 0), (1, 1), (-1, 1)

6. With the node (0, 0) as origin, two vectors are created using the remaining two co-ordinates to create vectors to define the direction of movement of elements

 

7. The model is then exported and simulation is run

8. The internal energy is 42424.85 Joules

9. The Internal energy in this case is 210.45 Joules lesser than the internal energy in case 4

10. The internal energies at end of simulation for parts in Region 1 are 7080.43 Joules and 7071.83 Joules respectively

11. The internal energies at end of simulation for parts in Region 2 are 14128.4 Joules and 14144.18 Joules respectively

12. The maximum resultant normal force is 1343.40 kN at 26.5 ms during the loading and the maximum resultant tangential force is 19.64 kN at 19.6 ms during the loading.

13. The contact forces contour shows that the maximum contact force is 80 kN

14. Maximum Von-mises stress is 0.628 Gpa

 

Case 6: Crash tube with notch in the middle using only two opposing faces

1. In this case, instead of creating a notch using all the four sides, only two opposing sides/faces are used to create the notch

 

2. The model is then exported and simulation is run

3. The internal energy is 42314 Joules

4. The internal energy in this case is 110.85 Joules lesser than case 5 and 321.3 Joules lesser than case 6

5. The internal energies at end of simulation for parts in Region 1 are 7403.15 Joules and 7418.3 Joules respectively

6. The internal energies at end of simulation for parts in Region 2 are 13757.15 Joules and 13735.39 Joules respectively

7. The maximum resultant normal force is 1235.93 kN at 27 ms during the loading and the maximum resultant tangential force is 35.71 kN at 27.4 ms during the loading.

8. The contact forces contour shows that the maximum contact force is 77.18 kN

9. Maximum Von-mises stress is 0.6715 Gpa

 

Results:

Models with different configuration of contacts, parameters and notches were simulated successfully and the observations are presented below

Table 1:

		Case 1	Case 2	Case 3
Internal energy (Joules)		42011.87	42011.87	42096.76
Internal Energy of Parts (Joules)	pshell_2mm	7403.36	7403.36	7585.92
	pshell2_2mm	7431.78	7431.78	7657.9
	PSHELL1	13615.75	13615.75	13510.58
	PSHELL2	13560.96	13560.96	13624.02
Resultant forces (kN)	Normal	1340.59	1340.59	1369.65
	Tangential	5.29	5.29	9.63
Contact forces (kN)		93.85	93.85	108.2
Max Von Mises Stress (Gpa)		0.696	0.696	0.67
Simulation run time (s)		292.28	329.85	441.69

 

Table 2:

		Case 4	Case 5	Case 6
Internal energy (Joules)		42635.3	42424.85	42314
Internal Energy of Parts (Joules)	pshell_2mm	8725.54	7080.43	7403.15
	pshell2_2mm	8663.47	7071.83	7418.3
	PSHELL1	12629.81	14128.4	13757.15
	PSHELL2	12616.47	14144.18	13735.39
Resultant forces (kN)	Normal	1177.91	1343.40	1235.93
	Tangential	18.36	19.64	35.71
Contact forces (kN)		57.14	80	77.18
Max Von Mises Stress (Gpa)		0.6321	0.6289	0.6715
Displacement (mm)		278.28	276.7	278.48
Simulation run time (s)		428.03	965.81	431.24

Learning outcome:

1. Contacts are very essential to a simulation because failure to predict the type of interaction between bodies may lead to incorrect results.

2. The type7 interface takes care of node to surface interactions but will fail to take into account edge to edge interactions

3. As the penalty method represents or portrays the contact between slave and master entities as springs that apply resistance to bodies coming in contact, there exists a non-linear relationship between stiffness of spring and the gap in case of type7 interface.

4. Initial penetrations in model may lead to locking or incorrect stiffness which eventually leads to incorrect results.

5. when Inacti=6 is set up ,the initial penetrations are removed by reducing the gap and scaling it by an appropriate scale value.

6. Type7 interface will overlook edge to edge interactions which will also lead to incorrect results.

7. Hence, both type7 and type11 interfaces are defined to cover all modes of interaction between bodies.

8. As the type11 interface restricts the edge to edge penetration and helps simulate accurate real life mechanism of contact between edges, the internal energy in case 3 is slightly more than the internal energies in first two cases.

9. The edges of the model coming into contact with other edges resulted in increased contact forces and internal energy of the model in case 3

10. Type11 interface uses very complex algorithms to calculate edge penetration and hence increases the computational load and simulation time which is evident in case 3

11. Hence, only in places where edge to edge penetration is possible, type11 interface must be defined

12. From crash perspective, the crash worthiness of a structure is calculated based on many factors like deformation, stress, energy absorption, etc.

13. Usually in crash, the lesser the normal force, lesser the harm to a person. Also, a high force applied at a small interval is more harmful than a small force applied at a long interval.

14. By comparing the displacement, resultant force and time duration of force propagation, from cases 4, 5 and 6, it can be concluded that the models in case 4 and 6 are good from a crash perspective because they are the least stiff members with more displacement and low Normal forces.

15. The stiffness of the model based on displacement in order is Case5 > Case4 > Case 6

16. The stiffness of the model based on Normal force in order is Case5 > Case6 > Case 4

17. From cases 5 and 6, it is evident that model with a notch along the whole section is stiffer than the model with notch on just the two opposite sides.

18. The notch on the four sides in case 5 resists the crumbling of the model under applied velocity.

19. Although the model in case 5 will not be good from crash perspective, this case is an example of how strategically created notches can be used to reinforce structural members.

20. The model in case 6 however offers very little resistance to crumbling and considerably low contact forces than the cases 4 and 5.

21. Notches can help to either decrease or increase the stiffness of structure depending upon the shape of notch, length of notch and number of notches.

Conclusion:

1. The given objectives were completed successfully and the observations were documented.

2. Different interface types for contact modeling and their parameters were studied in detail and contacts were modeled using those modeling techniques to study their effect on accuracy of results.

3. Models with different notch geometry and notch location were created and simulated to study the effect of notch on the results.

 

google drive link (Zip file ) - https://drive.google.com/file/d/1-9bUO5Y0P241nOX-UCeAiNbAG5yxHrSy/view?usp=sharing