/INTER/TYPE2

Format

Read this input, if Spot_flag = 20, 21, or 22:

Read this input, if Spot_flag = 25, 27 or 28:

Definitions

Comments

1. Interface TYPE2 is a kinematic condition; no other kinematic condition should be set on any node of the secondary surface, except when Spot_flag =25, 27 or 28.
2. The d_search is computed as (see Tied Interface (TYPE2) in the Radioss Theory Manual):
   (1)

   $$d_{\mathit{search}}=\frac{1}{n}\cdot {\displaystyle \sum _{i=1}^n{d}_i}$$
   with,

   $n$

   Being the number of main segments

   $d_i$

   Tthe total length of all the main side segments

3. Main nodes of an interface TYPE2 may be secondary nodes of another interface TYPE2 only if the hierarchy level of the first interface is lower than the hierarchy level of the second interface. Hierarchy levels are only available with Spot_flag =2. It does not work if Spot_flag =0 or Spot_flag =1.

   A possible workaround is using Spot_flag=2, which corresponds to the default formulation (Spot_flag=0); except that it is not compatible with /DT/NODA/CST.

4. Spot_flag =2 is equivalent to formulation 0; except that it is not compatible with nodal time step /DT/NODA/CST.
5. Spot_flag =4 is recommended to connect SPH particles to a surface (refer to Smooth Particle Hydrodynamics (SPH)).
6. Spot_flag = 20, 21 or 22 can include falure and be used to model a glue connection. It is not compatible with nodel time step /DT/NODA/CST. The stress is computed for each secondary node according to the "equivalent" surface around the node.
   In this case, the force in secondary node will be scaled by reduced force coefficient Fac_N (Fac_T), which is computed as:(2)

   $$\mathit{Fac}\_N=\text{min}\left\{\sqrt{\frac{{\sigma }_{N\_\text{max}}{}^2}{\text{max}\left[{\left({\sigma }_N\left(t\right)\right)}^2,{10}^{-20}\right]}},1\right\}$$

   (3)

   $$\mathit{Fac}\_T=\text{min}\left\{\sqrt{\frac{{\sigma }_{T\_\text{max}}{}^2}{\text{max}\left[{\left({\sigma }_T\left(t\right)\right)}^2,{10}^{-20}\right]}},1\right\}$$

   The reduced force is compared to the max value:

   if ${\sigma }_N<{\sigma }_{N\_\max }$ , then Fac_N =1, which means the force will not be reduced.

   if ${\sigma }_N>{\sigma }_{N\_max}$ , then $Fac\_N=\sqrt{\frac{{\sigma }_{N\_\max }^2}{\max \left[{\left({\sigma }_N\left(t\right)\right)}^2,{10}^{-20}\right]}}$ , which means the force will be reduced.

   Here the max value will be defined by the user with:(4)

   $${\sigma }_{N_{\max }}=\mathrm{Fscale}\left(\dot{\sigma }\right)\cdot {\mathrm{f}}_{sn}\left(\frac{\text{Δ}{X}_N}{Fscal{e}_{dist}}\right)$$

   (5)

   $${\sigma }_{T_{\max }}=\mathrm{Fscale}\left(\dot{\sigma }\right)\cdot {\mathrm{f}}_{st}\left(\frac{\text{Δ}{X}_T}{Fscal{e}_{dist}}\right)$$

   (6)

   $$\mathrm{Fscale}\left(\dot{\sigma }\right)=Fscal{e}_{stress}\cdot {\mathrm{f}}_{sr}\left(\frac{\dot{\sigma }}{Fscal{e}_{str\_rate}}\right)$$

   While, ${\mathrm{f}}_{sn}$ , ${\mathrm{f}}_{st}$ and ${\mathrm{f}}_{sr}$ are functions of fct_ID_sn, fct_ID_st and fct_ID_sr.

   Once the rupture criterion (defined by Rupt) is reached, the contact will be deleted.

   Here:

   • ${\sigma }_{N\_\max }$ is the maximum normal stress value defined by fct_ID_sn
   • ${\sigma }_N\left(t\right)$ is the normal stress
   • $\sigma$ _{T_max} is the maximum tangential stress value defined by fct_ID_st
   • ${\sigma }_T\left(t\right)$ is the tangential stress
   • Fscale_stress is the input constant stress factor
   • fct_ID_sr is the input variable coefficient
   • fct_ID_sn and fct_ID_s are the input stress-displacement functions
   • I_sym permits to choose between symmetric or asymmetric rupture (traction/compression). The initial direction from main surface to the secondary node defines the positive side (traction). If the distance is zero (secondary node lies on the main surface), the rupture will be symmetric, even with I_sym =1.

   This failure option (Spot_flag = 20, 21 or 22) can not be used in implicit.

7. Spot_flag =30: Secondary mass/inertia/stiffness distribution to the main node is based on the Kirschoff model: bi-cubic form functions are used instead of linear (standard formulation). It allows a softer contact behavior since the element shape curvature is taken into account in the force/moment transmission.
   Warning: This formulation is not compatible with solid elements, as it requires rotational DOF.
8. If flag I_del2 =2, then when a 4-node shell, a 3-node shell or a solid element is deleted, it is also removed from the main side of the interface (the kinematic condition is suppressed on relative secondary nodes).
9. The options I_del2 =1 and I_del2 =2 act if the main element is deleted using explicit deletion in Radioss Engine (using the keyword /DEL in Radioss Engine Input (/DEL/SHELL, /DEL/BRICK, ...)).
10. If I_filtr is set to 1, the normal and tangential stresses are filtered with an alpha filter, as:
    (7)

    $${\sigma }_N\left(t\right)=Alpha\cdot {\sigma }_N\left(t\right)+\left(1-Alpha\right){\sigma }_N\left(t-dt\right)$$

    (8)

    $${\sigma }_T\left(t\right)=Alpha\cdot {\sigma }_T\left(t\right)+\left(1-Alpha\right)\cdot {\sigma }_T\left(t-dt\right)$$
11. Spot_flag =25 (penalty formulation) will keep the penalty formulation during the whole run. The secondary node (of this contact) could also be the secondary node of another kinematic option, like rigid body.

    The penalty stiffness is constant, calculated by default as the mean nodal stiffness of main and secondary side. The stiffness factor, Stfac, may be used to modify it, if needed. The penalty stiffness will be multiplied by Stfac.

    A critical viscous damping coefficient (Visc) allows damping to be applied to the interface stiffness.

12. If Ignore = 1, 2, or 3, the secondary nodes without a main segment found during the Starter, are deleted from the interface.
13. If Ignore ≠ 1000, d_search is used.
    If Ignore = 2 or 3 and d_search = 0, d_search is computed, for each secondary node as:(9)

    $${\delta }_1=0.6\left(thickness\_secondary\_node+thickness\_main\_segment\right)$$

    (10)

    $${\delta }_2=0.05\left(main\_segment\_diagonal\right)$$

    $d_{search}=\max \left({\delta }_1,{\delta }_2\right)$

    For shells:

    • thickness_secondary_node = shell thickness of secondary
    • thickness_main_segment = shell thickness of main
    For solids:

    • thickness_secondary_node = 0
    If Ignore = 2:

    • thickness_main_segment = $\frac{Element\_volume}{Segment\_area}$
    If Ignore = 3:

    • thickness_main_segment = 0
    If Ignore = 2 or = 3:

    • Thickness is retained in the following order: first from /PART definition, from /SHELL or /SH3N definition, then from /PROP definition.
14. The contact is compatible with 2D-plane and -axisymmetrical simulations only for Spot_flag=0 and in case of connecting to solid elements with Spot_flag=0, then moments are not transferred.
15. If flag I_del2 =1, then when all 4-node shells, all 3-node shells and all solid elements belonging to a main segment are deleted, this segment is also removed from the main side of the interface (the kinematic condition is suppressed on relative secondary nodes).
16. Spot_flag = 25, 27 or 28: Interface penalty stiffness is computed from both main segment stiffness K_m and secondary node stiffness K_s, depending on I_stf flag:
    • I_stf = 1: $K_n=Stfac\cdot K_m$
    • I_stf = 2 (default): $K_n=Stfac\cdot \frac{K_m+K_s}{2}$
    • I_stf = 3: $K_n=Stfac\cdot \max \left(K_m,K_s\right)$
    • I_stf = 4: $K_n=Stfac\cdot \min \left(K_m,K_s\right)$
    • I_stf = 5: $K_n=Stfac\cdot \frac{K_m\cdot K_s}{K_m+K_s}$
17. If I_the >1, the material of the secondary side and main need to be a thermal material, using finite element formulation for heat transfer (/HEAT/MAT).

    Thermal conduction is computed when the secondary node falls into contact.

    The heat exchange is computed from main to secondary and from secondary to main:(11)

    $${\varphi }_{cond}=K_{the}\left(T_s^{}-T_m^{}\right)$$
18. By default, I_proj =1 is used to avoid having the wrong mass distribution when the secondary node is projected outside of the main element. The mass and inertia are distributed on the closest edge based on the projection of the secondary node on this edge. Use I_proj =2, to obtain the same results as Radioss version 14.0 or older.

    
    
    Figure 1.

19. When using the penalty formulation Spot_flag=25, moments cannot be transmitted from the secondary nodes to a main segment. Therefore, it is not recommended to use it for any connection where the secondary nodes have rotational degrees of freedom. This would include: shell to shell, spring to shell, shell to solid where the shell is secondary and solid is main. Due to this limitation and the lower robustness compared to kinematic formulations, it is recommended to use the mixed kinematic and penalty formulation, Spot_flag =27 and 28.
20. Spot_flag =27 and 28 are a mixed kinematic and penalty formulation tied contact. By default, the kinematic formulation is used. Any secondary nodes with incompatible kinematic conditions are automatically switched to the penalty formulation. Incompatible kinematic conditions with rigid bodies, imposed displacements, imposed velocities, imposed accelerations, other tied contact secondary nodes, or boundary conditions will cause the switch to penalty formulation. A WARNING message is printed in the Starter output file when secondary nodes are switched to penalty formulation.

    The penalty formulation stiffness is constant and calculated using I_stf and Stfac. A critical viscous damping coefficient (Visc) allows damping to be applied to the interface stiffness. The penalty formulation can transfer moments from the secondary nodes to the main segment.

21. Unlike Spot_flag =1, Spot_flag =28 does not add any mass at time=0 when the main surface of the tied contact is a shell element. If the main surface is a solid element there could be some mass added. No mass is added when Spot_flag =27 is used.