The distribution of FEAP version 8.4 is now available. Several enhancements have been incorporated in the new version. In part these include:
SUMMARY NEW COMMANDS:
Mesh Input
Solution Commands
Plot Commands
FEAP is a general purpose finite element analysis program which is designed for research and educational use. Source code of the full program is available for compilation using Windows (Intel compiler), LINUX or UNIX operating systems, and Mac OS X based Apple systems (GNU and Intel compilers).
The FEAP program includes options for defining one, two, and three dimensional meshes, defining a wide range of linear and nonlinear solution algorithms, graphics options for displaying meshes and contouring solution values, an element library for linear and nonlinear solids, thermal elements, two and three dimensional frame (rod/beam) elements, plate and shell elements, and multiple rigid body options with joint interactions. Constitutive models include linear and finite elasticity, viscoelasticity with damage, and elasto-plasticity.
The system may also be used in conjunction with mesh generation programs that have an option to output nodal coordinates and element connection arrays. In this case it may be necessary to write user functions to input the data generated from the mesh generation program.
Any problems related to difficulties in achieving a standard installation of the program should be addressed to: feap@berkeley.edu.
A small version of the system, called FEAPpv, is available free of any charge. See www.ce.berkeley.edu/feap/feappv for additional information. This program is the companion to the books: "The Finite Element Method, 6th edition, Volumes 1 and 2 (but not Vol 3)", authored by O.C. Zienkiewicz and R.L. Taylor and published by Elsevier, Oxford, 2005.
PDF files for current FEAP manuals may be obtained by downloading the following:
N.B. PDF files require Acrobat Reader to print.
A new user forum for FEAP users has been created at http://feap.berkeley.edu. This is a site to post questions, get answers, and interact with other users. We encourage users to help other users.
The file below contains a few simple test problems and the output they produce using FEAP ver 8.4. The file is in 'zip' form.
User elements may be added to the FEAP system to extend solution capabilities (See FEAP Programmer Manual). The elements given below are examples of elements which are useful in educational applications to demonstrate the behavior of finite element solutions of classical applications
The MATLAB interface is a standard part of versions 8.1 and later.
The MATLAB program permits easy solutions using many different algorithms. An interface for FEAP is provided by the user solution command routine 'umacr3.f' below. This routine permits the output to files of the non-zero values contained in the residual, tangent, mass, or damping matrices. The routine is fully operational with the 7.5 release of FEAP and except for the unsymmetric mass and damping options with release 7.4. The use is quite simple. In the solution command sequence:
A simple MATLAB program to use the tangent is
Download file from:
A direct link with MATLAB is also available. This permits FEAP to be run from MATLAB and allows for solution options available in MATLAB to be used in the analysis process. The interface is available from David Bindel at the web site listed below.
The SGI Origin machines have multiprocessor capability and solution of finite element problems using FEAP can be made much more efficient using special solution packages. The routine given below provides an interface to the SGI parallel sparse solver (users must already have the solution package from SGI for this interface to function).
An interface is available for the general sparse solver SuperLU. General information at: (http://www.cs.berkeley.edu/~demmel/SuperLU.html) Source program at (http://crd.lbl.gov/~xiaoye)
Download the source and create an archive containing the SuperLU routines for double precision real computations and, if necessary, the BLAS routines. (N.B. The SuperLU solver has been tested only in a Linux environment).
Next, download the following feap interface routines (zipped). Compile and link with the main FEAP archive and those for SuperLU created as described above. Use solution command 'splu' to activate the SuperLU solver. Subsequent commands 'tang' or 'utan' will then use SuperLU.
An interface is available for the general sparse solver UMFPACK. General information and source may be obtained from: (http://www.cise.ufl.edu/research/sparse/umfpack)
Down load the source and create an archive containing the UMFPACK archive. The interface may be obtained from:
An interface is available for the sparse solver system WSMP (Watson Sparse Matrix Package, IBM Research). General information and licensing requirements may be obtained from: (http://www.research.ibm.com/math/OpResearch/wsmp.html)
After receiving the source and license, create an archive containing the WSMP program. The FEAP interface may be obtained from:
An interface is available for the sparse solver system Pardiso. Google Pardiso for information on use. Also available with Intel compilers.
Structural Engineering, Mechanics and Materials
Department of Civil and Environmental Engineering
University of California, Berkeley