FELT(4fe) Finite Element Package FELT(4fe) NAME felt - finite element problem description file format DESCRIPTION Thefelt(4fe) file format is used by the programs of the finite element package,felt(1fe),velvet(1fe), andburlap(1fe) to describe a finite element problem. The file is human readable and consists of a friendly, intuitive syntax rather than a table of numbers. Syntactic and semantic errors are detected and reported assuring that only valid problem descriptions are used. In general white space is unimportant, arbitrary numeric expressions may be used, and case of keywords is unimportant. As per standard convention, boldface items represent keywords, italicized items represent the syntax of the grammar, and items in brackets are optional. The file syntax is shown below. [problem-description] [analysis-parameters] [object-definitions] [appearance-description]endProblem descriptionTheproblem-descriptionsection specifies the problem title and the number of nodes and elements in the problem. If this section is not specified then the problem will be unnamed and is assumed to contain zero nodes and zero elements. This section may be absent for example in defaults files which define objects but do not specify an actual problem instance. The format for aproblem-descriptionis given below.problem description[title =string] [nodes =integer] [elements =integer] [analysis = static|transient|modal|static-thermal|transient-thermal|spectral] If thetitleis missing then the problem is unnamed. Ifnodesorelementsis missing then none of that type of object are expected. The assignments can occur in any order and can even be repeated with the last assignment being used.Analysis parametersTheanalysis-parameterssection defines any parameters for a specific type of analysis. Currently, this section is used only if the analysis mode istransient. The format of theanalysis-parameterssection is given below.analysis parameters[alpha =expression] [beta =expression] [gamma =expression] [step =expression] [start =expression] [stop =expression] [Rm =expression] [Rk =expression] [nodes = [node-list]] [dofs = [dof-list]] [mass-mode = lumped|consistent] Thealpha,beta, andgammaparameters are used in numerical integration schemes (transient and transient-thermal analysis).start,stop, andstepdefine the range of time or frequency interest for transient or spectral analyses. In transient analyses,startis meaningless anddurationanddtcan be used as aliases forstopandstep, respectively.RkandRmare global Rayleigh (stiffness and mass) damping proportionality constants. Thenode-listis a comma or white space separated list of node numbers that are of interest in the analysis. Similarly, thedof-listis a list of the degrees of freedom (Tx,Ty,Tz,Rx,Ry, andRz) that are of interest. Anobject-definitionsection defines objects of a specified type. Objects include nodes, elements, materials, constraints, forces, and distributed loads. Each of these types of objects is discussed below. Multipleobject-definitionsections are allowed and the sections may occur in any order.NodesNodes are points in cartesian space to which elements are attached. A node must have a constraint and may have an optional force. A node is identified by a natural number. The syntax is as follows:nodesnode-definitionswhere anode-definitiontakes the following form:node-number[x =expression] [y =expression] [z =expression] [constraint =constraint-name] [force =force-name] [mass =expression] Thenode-numberstarts the definition. Each node must have a unique number. If a cartesian coordinate is not given then the coordinate of the previous node is used. Similarly, if no constraint is given then the constraint applied to the previous node is used. As above, the assignments can appear in any order and any number of times. As indicated above, some objects are identified by their name and some by their number. Elements and nodes have numbers while materials, forces, loads, and constraints have names.ElementsElements are linear, planar, or solid objects which are attached to nodes. Each element must have a material and may have optional loads. Furthermore, each element has a type, or definition. Like nodes, elements are identified by a unique natural number. Elements of specific type are defined with the following syntax:element-typeelementselement-definitionwhere anelement-typeis one of the following:springtrussbeambeam3dCSTPlaneStrainCSTPlaneStressiso2d_PlaneStrainiso2d_PlaneStressquad_PlaneStrainquad_PlaneStresstimoshenkohtkbrickctgrodand anelement-definitionhas the following form:element-numbernodes = [node-list][material =material-name] [load =load-name-list] Theelement-numberstarts the definition. Each element must have a unique number. If no material is given then the material applied to the previous element is used. Theload-name-listis a list of up to three loads to apply to the element. Thenode-listis a comma or white space separated list of node numbers. Each type of element requires a certain of nodes and in some cases a special "null node" which is numbered zero may be used to indicate a gap or filler in the list.MaterialsElements are made of a type of material. Each material has a name and certain physical properties not all of which may be used by any one element. The syntax for defining materials is as follows:material propertiesmaterial-definitionswherematerial-definitionhas the following form:material-name[color =string] # color forvelvet[E =expression] # Young's modulus [Ix =expression] # moment of inertia about x-x axis [Iy =expression] # moment of inertia about y-y axis [Iz =expression] # moment of inertia about z-z axis [A =expression] # cross-sectional area [J =expression] # polar moment of inertia [G =expression] # bulk (shear) modulus [t =expression] # thickness [rho =expression] # density [nu =expression] # Poisson's ratio [kappa =expression] # shear force correction [Rk =expression] # Rayleigh damping coefficient (K) [Rm =expression] # Rayleigh damping coefficient (M) [Kx =expression] # thermal conductivity in the x-direction [Ky =expression] # thermal conductivity in the y-direction [Ky =expression] # thermal conductivity in the z-direction [c =expression] # heat capacitance Thematerial-namestarts the definition. If an attribute of a material is not specified then that attribute is zero. The assignments may occur in any order. Thecolorspecifies the color to use in drawing the material withinvelvet, and is ignored by other applications.ConstraintsConstraints are applied to nodes to indicate about which axes a node can move. The syntax for defining a constraint is as follows:constraintsconstraint-definitionswhereconstraint-definitionhas the following form:constraint-name[color =string] # color forvelvet[tx = c|u|expression] # boundary translation along x axis [ty = c|u|expression] # boundary translation along y axis [tz = c|u|expression] # boundary translation along z axis [rx = c|u|expression|h] # boundary rotation about x axis [ry = c|u|expression|h] # boundary rotation about y axis [rz = c|u|expression|h] # boundary rotation about z axis [itx =expression] # initial displacement along x axis [ity =expression] # initial displacement along y axis [itz =expression] # initial displacement along z axis [irx =expression] # initial rotation about x axis [iry =expression] # initial rotation about y axis [irz =expression] # initial rotation about z axis [vx =expression] # initial velocity along x axis [vy =expression] # initial velocity along y axis [vz =expression] # initial velocity along z axis [ax =expression] # initial accel. along x axis [ay =expression] # initial accel. along y axis [az =expression] # initial accel. along z axis Theconstraint-namestarts the definition. A value ofcfor a boundary condition indicates that the axis is constrained; a value ofuindicates that the axis is unconstrained. An expression indicates a displacement (non-zero) boundary condition and may contain thetvariable for time varying boundary conditions in transient analysis problems. The initial dislacement, velocity and acceleration specifications are only used in transient problems. A value ofhfor a rotational boundary condition indicates a hinge. By default, all axes are unconstrained. Thecolorspecifies the color to use in drawing the constraint withinvelvet, and is ignored by other applications.ForcesForces, or point loads, may be applied to nodes. The syntax for a force definition is as follows:forcesforce-definitionswhere aforce-definitionhas the following form:force-name[color =string] # color forvelvet[Fx =expression] # force along x axis [Fy =expression] # force along y axis [Fz =expression] # force along z axis [Mx =expression] # moment about x axis [My =expression] # moment about y axis [Mz =expression] # moment about z axis [Sfx =expression] # frequency-domain spectra of force along x axis [Sfy =expression] # frequency-domain spectra of force along y axis [Sfz =expression] # frequency-domain spectra of force along z axis [Smx =expression] # frequency-domain spectra of moment about x axis [Smy =expression] # frequency-domain spectra of moment about y axis [Smz =expression] # frequency-domain spectra of moment about z axis Theforce-namestarts the definition. If the force or moment is not specified then it is assumed to be zero. Theexpressionsfor forces may be time-varying. Time-varying expressions include the single variabletto represent the current time in the solution of a dynamic problem or consist of a list of discrete (time, value) pairs. Frequency varying expressions for spectra can also usewto represent the independent variable (radial frequency). Thecolorspecifies the color to use in drawing the force withinvelvet, and is ignored by other applications.LoadsDistributed loads, or loads for short, are applied to elements. The syntax for a defining a distributed load is as follows:distributed loadsload-definitionswhere aload-definitionhas the following form:load-name[color =string] # color forvelvet[direction =dir] # direction [values =pair-list] # local nodes and magnitudes Theload-namestarts the definition. Thediris one ofLocalX,LocalY,LocalZ(local coordinate system),GlobalX,GlobalY,GlobalZ(global coordinate system),parallel, orperpendicular. Thepair-listis a sequence ofpairs. Apairis a node number and an expression enclosed in parentheses. The node number refers to the position within the element rather than referring to an actual node. Thecolorspecifies the color to use in drawing the load withinvelvet, and is ignored by other applications.Appearance DescriptionTheappearance-descriptionsection is used byvelvetto describe the appearance of a problem. This section is currently not used byfelt. The appearance includes the state of the drawing area and any tool figures. This section consists of two subsections, thecanvas-configurationsection and thefigure-listsection. Thecanvas-configurationsection has the following syntax.canvas configurationcanvas-parameterswhere acanvas-parameterhas the following form: [node-numbers =boolean] # node numbering [element-numbers =boolean] # element numbering [snap =boolean] # snap grid status [grid =boolean] # visible grid status [snap-size =expression] # snap grid size [grid-size =expression] # visible grid size [node-color =color-name] # node color [element-color =color-name] # element color [label-font =font-name] # labeling font [tool-color =color-name] # tool figure color [tool-font =font-name] # text figure font [x-min =expression] # x-axis minimum [x-max =expression] # x-axis maximum [y-min =expression] # y-axis minimum [y-max =expression] # y-axis maximum [x-pos =expression] # x position of drawing area [y-pos =expression] # y position of drawing area [width =expression] # width of viewport window [height =expression] # height of viewport window [scale =expression] # scale of drawing area Abooleanis eithertrueorfalse. Acolor-nameis the name of a valid X11 color. Similarly, afont-nameis the name of a valid X11 font. The last five parameters are probably not very meaningful to the user. Thefigure-listsection has the following syntax.figure listfigure-definitionswhere afigure-definitionhas the following form:figure-type[x =expression] # x coordinate [y =expression] # y coordinate [width =expression] # width [height =expression] # height [start =expression] # starting angle [length =expression] # arc length [text =name] # text string [color =name] # color [font =name] # text font [points = [point-list]] # line points Thefigure-typestarts the definition and is one ofrectangle,polyline,text, orarc. Note that not all properties have meaning for all figures. Any unneeded property is ignored. If a color or font property is not given then the previous property is used. Thepoint-listis a list of (x-coordinate, y-coordinate) pairs.ExpressionsAnexpressioncan be either constant or time-varying. As discussed above, time-varying expressions contain the variabletor consist of a list of discrete (time, value) pairs. If a time-varying expression is given where a constant expression is expected, the expression is evaluated at time zero. Anexpressionhas one of the following forms, where all operators have the precedences and associativities given to them in the C programming language.expression?expression:expression# in-line conditionalexpression||expression# logical orexpression&&expression# logical andexpression|expression# integer inclusive orexpression^expression# integer exclusive orexpression&expression# integer andexpression==expression# equalityexpression!=expression# inequalityexpression<expression# less thanexpression>expression# greater thanexpression<=expression# less than or equalexpression>=expression# greater than or equalexpression<<expression# integer shift leftexpression>>expression# integer shift rightexpression+expression# additionexpression-expression# subtractionexpression*expression# multiplicationexpression/expression# divisionexpression%expression# integer remainder-expression# arithmetic negation!expression# logical negation~expression# integer bitwise negation(expression)# enforce precedencesin (expression)# sinecos (expression)# cosinetan (expression)# tangentpow (expression,expression)# power (exponentiation)exp (expression)# exponentiallog (expression)# natural logarithmlog10 (expression)# base-10 logarithmsqrt (expression)# square roothypot (expression,expression)# Euclidean distancefloor (expression)# floorceil (expression)# ceilingfmod (expression,expression)# floating point remainderfabs (expression)# absolute valuenumber# literal valuet# current time Finally, a discretely valued expression has the following syntax, where the optional+indicates that the list represents one cycle of an infinite waveform.(expression','expression)... [+] AUTHOR Thefeltfile format was developed by Jason I. Gobat (jgobat@mit.edu) and Darren C. Atkinson (atkinson@ucsd.edu). SEE ALSO corduroy(1fe), felt(1fe), velvet(1fe), xfelt(1fe), corduroy(4fe). Version 3.00 8/4/95 FELT(4fe)