Page 44 - Using ANSYS for Finite Element Analysis Dynamic, Probabilistic, Design and Heat Transfer Analysis
P. 44

composite materials   •   31
                          You may also write your own subroutines to define the element and
                      layer  coordinate  systems (USERAN and USANLY); see the Guide to
                      ANSYS User Programmable Features for details.
                          Layer thickness—If the layer thickness is constant, you only need
                      to specify TK(I), the thickness at node I. Otherwise, the thicknesses at the
                      four corner nodes must be input. Dropped layers may be represented with
                      zero thickness.
                          Number of integration points per layer—This allows you to deter-
                      mine in how much detail the program should compute the results. For very
                      thin layers, when used with many other layers, one point would be appropri-
                      ate. But for laminates with few layers, more would be needed. The default
                      is three points. This feature applies only to sections defined through the sec-
                      tion commands. Currently, the graphical user interface (GUI) only allows
                      layer real constant input of up to 100 layers. If more layers are needed for
                      SHELL99 or SOLID46, the R and RMORE commands must be used.



                      2.2.2.2  Defining the Constitutive Matrices

                      This is an alternative to specifying the individual layer properties and is avail-
                      able as an option KEYOPT (2) for SOLID46 and SHELL99. The matrices,
                      which represent the force–moment and strain–curvature relationships for the
                      element, must be calculated outside the ANSYS program as outlined in the
                      ANSYS, Inc. Theory Reference. They can be included as part of the solution
                      printout with KEYOPT(10). The main advantages of the matrix approach are:


                        •  It allows you to incorporate  an aggregate  composite  material
                           behavior.
                        •  A thermal load vector may be supplied.
                        •  The matrices may represent an unlimited number of layers.

                          The  terms  of  the  matrices  are  defined  as  real  constants.  Mass
                      effects are incorporated by specifying an average density (real constants
                      AVDENS) for the element. If the matrix approach is used, detailed results
                      in each layer cannot be obtained since individual layer information is not
                      input.



                      2.2.2.3  Sandwich and Multiple-Layered Structures

                      Sandwich structures have two thin faceplates and a thick, but relatively
                      weak, core. Figure 2.11 illustrates sandwich construction.
   39   40   41   42   43   44   45   46   47   48   49