Nonlinear-code
所属分类:数学计算
开发工具:Fortran
文件大小:239KB
下载次数:41
上传日期:2012-02-06 08:42:07
上 传 者:
pl122022
说明: 这是fotran的非线性有限元程序,包括1D,2D程序
(fortan NFEM )
文件列表:
Nonlinear code\1D nonlinear\license.txt (1566, 2011-06-01)
Nonlinear code\1D nonlinear\MA_Crisfield_Chapter1_page_3.m (4405, 2011-06-01)
Nonlinear code\nonlin2\acce.f (1625, 1996-10-25)
Nonlinear code\nonlin2\arcl.f (3717, 1996-10-25)
Nonlinear code\nonlin2\asmb.f (3413, 1996-10-25)
Nonlinear code\nonlin2\b13e.f (4024, 1996-10-25)
Nonlinear code\nonlin2\b13s.f (3351, 1996-10-25)
Nonlinear code\nonlin2\bcon.f (1873, 1996-10-25)
Nonlinear code\nonlin2\brac.f (2548, 1996-10-25)
Nonlinear code\nonlin2\brsw.f (2130, 1996-10-25)
Nonlinear code\nonlin2\conv.f (1094, 1996-10-25)
Nonlinear code\nonlin2\crot.f (1181, 1996-10-25)
Nonlinear code\nonlin2\eign.f (4664, 1996-10-25)
Nonlinear code\nonlin2\elem.f (3211, 1996-10-25)
Nonlinear code\nonlin2\erro.f (411, 1996-10-25)
Nonlinear code\nonlin2\exel.f (1226, 1996-10-25)
Nonlinear code\nonlin2\extd.f (6167, 1996-10-25)
Nonlinear code\nonlin2\forc.f (3629, 1996-10-25)
Nonlinear code\nonlin2\i2212.inp (1206, 1996-10-24)
Nonlinear code\nonlin2\i2221.inp (985, 1996-10-24)
Nonlinear code\nonlin2\i2222.inp (968, 1996-10-24)
Nonlinear code\nonlin2\i2231.inp (1675, 1996-10-24)
Nonlinear code\nonlin2\i2232.inp (1678, 1996-10-24)
Nonlinear code\nonlin2\i2233a.inp (1674, 1996-10-24)
Nonlinear code\nonlin2\i2233b.inp (1658, 1996-10-24)
Nonlinear code\nonlin2\i2234a.inp (1658, 1996-10-24)
Nonlinear code\nonlin2\i2234b.inp (1658, 1996-10-24)
Nonlinear code\nonlin2\i2234c.inp (1658, 1996-10-24)
Nonlinear code\nonlin2\i2235a.inp (1689, 1996-10-24)
Nonlinear code\nonlin2\i2235b.inp (1690, 1996-10-24)
Nonlinear code\nonlin2\i2235c.inp (1729, 1996-10-24)
Nonlinear code\nonlin2\i2235d.inp (1689, 1996-10-24)
Nonlinear code\nonlin2\i224.inp (1014, 1996-10-24)
Nonlinear code\nonlin2\i225a.inp (6156, 1996-10-24)
Nonlinear code\nonlin2\i225b.inp (6158, 1996-10-24)
Nonlinear code\nonlin2\i225c.inp (6141, 1996-10-24)
Nonlinear code\nonlin2\imamn.f (1925, 1996-10-25)
Nonlinear code\nonlin2\inck.f (2592, 1996-10-25)
Nonlinear code\nonlin2\inck1.f (3571, 1996-10-25)
Nonlinear code\nonlin2\init.f (2267, 1996-10-28)
... ...
Input Manual
Each numerical data set is headed by a line of key words
explaining the data listed below. Although the selection of
these key words can be different, it is compulsory to have a
line of characters before every numerical data block.
Under the top line ANALYSIS, lies the very first numerical
data either 0 or 1, which will decide whether to perform a
simple check on the input data. 0=No; 1=Yes. The second data
block defines the global information of the model:
NV= Total No. of Degrees of Freedom;
NE= No. of elements in the structure
NBCON= No. of constrained nodes
NLOAD= No. of nodes where loadings are applied
NMATE= No. of elements which have different EA's
NANIT= No. of elements which have different initial
internal forces
NDIM= The dimension of the problem (two or three).
The next data card gives:
ITEYL= Type of strain measure(1=Engg; 2=Green ; 3/4=Log
with/without volume change),
POSS= Poisson's ratio
EA= Young's modulus times the area for all elements
except those listed in NMATE
ANIT= Initial element internal force for all elements
except those listed in NANIT
If NMATE and /or NANIT are not zeros, the next card(s) will
have:
if NMATE is not 0, Different EA's (Elem. No;EA.....)
if NANIT is not 0, Different ANIT's(Elem. No;ANIT......)
The data under the NODAL COOR. card provides the coordinates
x and y for the 2D case and x,y,z for the 3D case for each node in the
structure. For large structures this can be tedious. So instead
one can use other pre-processors to generate the nodal coordinates
and have them copied here. Whichever way one wishes to input the
coordinates, one should have NN*NDIM data in this block,
starting from the x and y(or x,y and z) coordinates of the
first node to the last. Here NN is the number of nodes.
The ELEM CONNECTION card tells the matrix assembler ASMB
what are the global nodal numbers for each of the two
element nodes. So the first pair of integers are the global
nodal numbers for the first element; the second pair for the
second element, and so on. This again can be produced using
a standard pre-processor and one can copy the nodal connections.
The external loadings are input from card LOADINGS, which
consists of NLOAD*(NDIM+1) groups of data. In each group the
first data is the nodal number where loading is applied. The
next NDIM real numbers are the actual nodal forces in NDIM
directions (x,y,z)
The boundary conditions are defined in the data section
BOUNDARY CONDITION. All nodes are assumed free initially
i.e. IBC(1-NV)=0. '0' here means free . These are modified
by the relevent constraints imposed on NBCON nodes. In each
of the NBCON groups of data, the first number denotes the
global nodal number, while the following two (for 2D) or
three (for 3D) integers designate the type of boundary
conditions, 0=free; 1=fixed; -1=displacement controled.
The output file is normally quite big, as it contains a lot
of information generated in the solution process. To have a
quick evaluation of the results, one can write the load
factor plus the displacement(s) at a number (no more than
10) of key degrees of freedom. This can be easily done by
first specifying the number of output variables and what
their global D.O.F. are under the header OUTPUT VARIABLES.
The information is written in a column-wise format, ready as
input for most of the standard graphic packages.
The effects of linear springs can be simulated through the
option EARTHED SPRINGS. To input this information, one first
gives the number of each springs (up to NV). Then one provides
for each spring, its stiffness and the D.O.F. attached to it.
The card FACI,.... begins the solution control data part.
FACI= load increment size
NINC= total number of increments
IWRIT= 0:brief output
1:detailed output
IAUTO= 0:fixed increments and no step halving with
convergence failure
1:automatic incrementation
IARC= 0:load contol
1:MAC cylindrical arc-length control
2:Riks orthogonal plane arc-length control
3:Ramm updated orthogonal arc-length control
4:Fried orthogonal trajectory arc-length control
5:Rheinboldt specific displacement control
6:Powell and Simon incremental work control
ILOAD= determines the sign of the load increment by;
1:current stiffness parameter CSTIF
2:A negative pivot
IACC= 0:do not use secant Newton to accelerate
iteration
1:employ secant Newton to accelerate iteration
IRES= 0:present analysis is not a re-start
1:present analysis is a re-start
IBRAC= 0:do not bracket the singular point
1:bi-section bracketing
2:'0.618' section bracketing
3:interpolation based on the determinent of K
4:interpolation based on the product of the
smallest and biggest pivots of K
5:interpolation based on the smallest pivot of K
6:direct computation of singular point from the state
given by the one indirect branching method(IBRAC=5 above)
-6:direct computation of singular point from underformed
state
7:semi-direct bracketing with bi-section
-7:semi-direct bracketing with interpolation on PMIN
ICRIT= The critical point No.to be bracketed (eg. the first or
second from the undeformed state)
IBRSW= 0:no branching
1:branching with the 1st eigenmode as
predictor
2:branching with an asymmetric predictor
using 2nd order terms
3:multiple bifurcation with lowest eigenmode, the amplitudes
of which could be supplied by the user(see section 22.3.2)
4:by Fujii's line search
5:by Seydel's approximation to the Ist eigenmode
6:Riks's orthogonal to the primary path
7:adaptive force=0.01*MAX(Diag(Kt)*DT(IROW) with Rheinboldt's
arc-length control
The controled displacement component IROW can be supplied by the user
or it can be related to the smallest pivot.
IROW= the row No. Of Kt,which has to be replace by a unit vector
produce Fujii's search direction or Seydel' predictor
for branching
ICVCK= 1: resd. force norm scaled by external force norm
2: resd.force norm scaled by reaction norm
3: iterative disp. norm scaled by total disp. norm.
BETOK= convergence tolerance
ITERTY= 1: full Newton-Raphson iteration
2: modified Newton-Raphson iteration
NITMAX= maximum number of iterations allowed for each
increment
NLSMX= number of line searches desired(no more than eight)
EPSI= converging tolerance for bracketing singular point
SHIF= the amount of shift required when performingan eigen
analysis on a near singular Kt
If IAUTO is not 0, input the following data for automatic
incrementation;
IDES= desired number of iterations
FACMX= the maximum load increment allowed
FACMN= the minimum load increment allowed
ISWCH= 0: never switch to arc-length control
1-5: switch to the desired type of arc-length control if
the current stiffness parameter gets smaller than a preset value
IPRED= 0:use Euler predictor
1:use 2nd order predictor
ICORT= 0:use Newton corrector
1:explicit 3rd order corrector
2:implicit 3rd order corrector
If ISWCH=1, input:
CSTIFS= the value of the current stiffiness parameter at which one
switches from load to arc-length control
If IARC is not 0, input:
DLDES= the desired incremental arc-length
DLMAX= the maximum incremental arc-length
DLMIN= the minimum incremental arc-length
If IACC is not 0, input:
R1C= the lower cut-out for not applying acceleration
R2C= the upper cut-out for not applying acceleration
If NLSMX is not 0,input:
PERMLS= line search tolerance factor
AMPMX= maximum amplification factor
ETMXA= maximum step length
ETMNA= minimum step length
近期下载者:
相关文件:
收藏者: