2-D analysis in LS-DYNA ------------------------ 1. Nodes must lie in the x-y plane, i.e., the z-coordinate is zero. For axisymmetric problems, the y-axis is the axis of symmetry and all nodes must have x-coordinate >= 0. Furthermore, the element normal for 3- or 4-node elements should be in positive z, else automatic shuffling of the connectivity may invalidate orthotropic material directions, as in the case of AOPT=0. 2. Four-noded, 2-D continuum elements are designated as shells for convenience of input. Use shell formulation 13 for plane strain continuum elements, formulation 14 for high-pressure axisymmetric continuum elements (high pressure problems such as those involving high explosive), or formulation 15 for general purpose axisymmetric continuum elements. Set NIP=4 in *section_shell to obtain a fully-integrated (2x2) formulation (applies to element formulations 13 and 15 only). The default is single point integration and is thus subject to hourglass modes. See Section 3.9 of the 2006 LS-DYNA Theory Manual (downloadable from www.lstc.com). For plane strain and axisymmetric shells, use beam formulations 7 and 8, respectively (see Remark 1 below). TT1 and TT2 in *section_beam are ignored for these beam formulations. TS1 and TS2 are the shell thicknesses at node1 and node2, resp. QR/IRID in *section_beam gives the number of through thickness integration points. Examples: http://ftp.lstc.com/anonymous/outgoing/support/FAQ_kw/beam.type8.circpl.pin.k http://ftp.lstc.com/anonymous/outgoing/support/FAQ_kw/contact_2d_beam7.k 3. Contact or coupling is done via *contact_2d_automatic_OPTION. 4. R-adaptivity is available for 2D continuum elements. Compare http://ftp.lstc.com/anonymous/outgoing/support/FAQ_kw/pen-noadapt.k to http://ftp.lstc.com/anonymous/outgoing/support/FAQ_kw/pen-adapt.k 5. See *part_adaptive_failure to allow an adapted part to split into two pieces. 6. From a material standpoint, a 2D continuum element is classified as a solid. From an element standpoint, we refer to these as shells only because the input syntax for shell elements (3 or 3 or 3 or 4 nodes) fits the bill. 7. See TYPE=2 in *control_bulk_viscosity (applies to 2D plane strain and axisymmetric elements). 8. See remark under *section_shell regarding units for shell formulation 14 and 15. Regarding ASCII output: "The attached examples all use the *case command to compare ASCII force output (bndout, secforc, nodfor, spcforc, rcforc, ncforc) from shell type 14 (case14.* output) and shell type 15 (case15.*). Shell type 15 consistently reports forces per unit radian. Shell type 14 sometimes reports forces per unit radian (rcforc, ncforc), sometimes reports forces per unit length (nodfor, spcforc), and sometimes, as in the case of secforc and bndout, you can't always say which is being output... force per radian or some other unit. This, obviously, is very confusing for shell 14 users. My suggestion is to modify shell 14 so that ALL force output is reported as force per radian, just as shell 15 does. In the meantime, I'll add a warning in the User's Manual concerning inconsistency of force output for shell 14." 6/9/16 Ticket#2016060810000186 Bugzilla 12319 9. Added MPP support for *CONTACT_2D_AUTOMATIC_SINGLE_SURFACE and *CONTACT_2D_AUTOMATIC_SURFACE_TO_SURFACE in 971 R7.0.0. 10. To define a contact force transduer, use *CONTACT_2D_FORCE_TRANSDUCER.