Practical Considerations in Modeling;

Interpreting Results
Introduction
In this section we will discuss some modeling considerations and guidelines,
including mesh size, natural subdivisions, and the use of symmetry and associated boundary conditions,
We will also introduce the concept of static condensation, which enables us
to apply the basis of the CST stiffness matrix to a quadrilateral element.

Finite Element Modeling

Finite element modeling is partly an art guided by visualizing physical interactions taking place within a body, In modeling the user is confronted with the difficult task of understanding physical behavior taking place and understanding the
physical behavior of various elements available for use, Matching the appropriate
finite element to the physical behavior being modeling is one of many decisions
that must be made by the modeler. Understanding the boundary conditions can
be one of the most difficult task a modeler must face in construction a useable
finite element model.
Aspect Ratio and Element Shape
The aspect ratio is define as the ratio of the longest dimension to the shortest
dimension of a quadrilateral element, In general, as the aspect ratio increases,
the inaccuracy of the finite element solution increases,
Consider the five different finite element model shown in the figure below. A
plot of the resulting error in the displacement at point A of the beam verse aspect
ratio is given. In addition, the numerical answers are given in the following table.

CIVL 7117 Finite Elements Methods in Structural Mechanics

Page 286

CIVL 7117 Finite Elements Methods in Structural Mechanics

Page 287

Case

Aspect
Ratio

Number of
Nodes

Number of
Elements

Point A

Point B

% Error

1.1

84

60

-1.093

-0.346

5.2

1.5

85

64

-1.078

-0.339

6.4

3.6

77

60

-1.014

-0.238

11.9

6.0

81

64

-0.886

-0.280

23.0

24.0

85

64

-0.500

-0.158

56.0

Exact Solution

-1.152

-0.360

In general, elements that yield the best results are compact and regular in
shape will: (1) aspect ratios near one; and (2) corner angles of quadrilaterals
near 90.
CIVL 7117 Finite Elements Methods in Structural Mechanics

Page 288

Use of Symmetry
The use of symmetry will often expedite the modeling of a problem. Symmetry allows us to consider a reduced problem instead of the actual problem. This
will allow us to use a finer discretization of element with less computational cost.
Natural Subdivisions at Discontinuities
There are a variety of natural subdivisions for finite element discretizations.
For example, natural locations of nodes occur at concentrated loads or discontinuities in loading, other types of boundary conditions, and abrupt changes in geometry of materials.
Sizing of Elements and Mesh Refinement
A discretization depends on the geometry of the structure, the loading, and the
boundary conditions. For example, areas of high, rapidity changing stresses require a finer mesh then regions where the stress is constant.

CIVL 7117 Finite Elements Methods in Structural Mechanics

Page 289

Here the use of symmetry is applied to a soil mass subjected to a foundation

loading (66 nodes and 50 elements). Note that at the place of symmetry the displacements in the direction perpendicular to the plane muse be zero. This is
modeled by rollers at nodes 2 - 6.

CIVL 7117 Finite Elements Methods in Structural Mechanics

Page 290

The figure above illustrates the use of triangular elements for transitions from
smaller quadrilaterals to larger quadrilaterals. The transitions are required since
CST elements do not have immediate nodes along their edges. If an element had
an intermediate node, the resulting equations would be inconsistent with the energy formulation for the CST equations.

CIVL 7117 Finite Elements Methods in Structural Mechanics

Page 291

CIVL 7117 Finite Elements Methods in Structural Mechanics

Page 292

CIVL 7117 Finite Elements Methods in Structural Mechanics

Page 293

Infinite Medium
A typical example of infinite medium is a soil foundation problem. The guideline for the finite element model is that enough material must be included such
that the displacements are nodes and stresses within the elements becomes
negligibly small at locations far from the foundation load. The level of discretization can be determined by a trail-and-error procedure in which the horizontal and
vertical distances from the load are varied and the resulting effects on the displacements and stresses are observed. For a homogeneous soil mass, experience has shown the influence of a footing becomes insignificant if the horizontal
distance of the model is taken as approximately four and six times the width of
the footing and the vertical distance is taken as approximately four to ten times
the width of the footing.
Checking the Model
The discretized finite element model should be checked carefully before results are computed. Ideally, a model should be checked by an analyst not involved in the preparation of the model, who is then more likely to be objective.
Preprocessors with their detailed graphical display capabilities now make it comparatively easy to find errors, particularly with a misplaced node or missing element or a misplaced load or boundary condition. Preprocessors include the ability to color, shrink, rotate, and section a model mesh.
Checking the Results and Typical Postprocessor Results
An analyst should probability spend as mush time processing, checking, and
analyzing results as spent in data preparation.

CIVL 7117 Finite Elements Methods in Structural Mechanics

Page 294

Plate with a hole in the center.

Deformed shape of plate superimposed over an unreformed shape

CIVL 7117 Finite Elements Methods in Structural Mechanics

Page 295

The wrench in this example is modeled by 307 constraint strain triangular elements (plane stress assumption). Below is a plot of the deformed shape of the
wrench over the original mesh.

CIVL 7117 Finite Elements Methods in Structural Mechanics

Page 296

On the right is a plot of the stress in the wrench and the plot on the left is a
plot of the defected shape. This analysis was preformed using WinFElt finite
element structural analysis system.

CIVL 7117 Finite Elements Methods in Structural Mechanics

Page 297

Equilibrium and Compatibility of Finite Element Results

An approximate solution for a stress analysis problem using the finite element
method based on assumed displacement fields does not generally satisfy all the
requirements for equilibrium and compatibility that an exact theory-of-elasticity
solution satisfies. However, remember that relatively few exact solutions exist.
Hence, the finite element method is a very practical one for obtaining reasonable,
but approximate, numerical solutions.
We now describe some of the approximations generally inherent with finite
element solutions.
1.

Equilibrium of nodal forces and moments is satisfied. This is true because the global equation F = Kd is a nodal equilibrium equation whose
solution for d is such that the sums of all forces and moments applied to
each node are zero. Equilibrium of the whole structure is also satisfied
because the structure reactions are included in the global forces, and
hence, in the nodal equilibrium equations.

2.

Equilibrium within an element is not always satisfied. However, for the

constant-strain bar and the constant-strain triangle, element equilibrium
is satisfied. Also the cubic displacement function is shown to satisfy the
basic beam equilibrium differential equation, and hence, to satisfy element force and moment equilibrium.

3.

Equilibrium is not usually satisfied between elements. A differential element including parts of two adjacent finite elements is usually not in
equilibrium (see the figure below). For line elements, such as used for
truss and frame analysis, interelement equilibrium is satisfied. However,
for two- and three-dimensional elements, interelement equilibrium is not
usually satisfied. Also, the coarseness of the mesh causes this lack of
interelement equilibrium to be even more pronounced.

CIVL 7117 Finite Elements Methods in Structural Mechanics

Page 298

4.

Compatibility is satisfied within an element as long as the element displacement field is continuous; hence, individual elements do not tear
apart.

5.

In the formulation of the element equations, compatibility is invoked at

the nodes. Hence, elements remain connected at their common nodes.
Similarly, the structure remains connected to its support nodes because
boundary conditions are invoked at these nodes.

6.

Compatibility may or may not be satisfied along interelement boundaries. For line elements such as bars and beams, interelement boundaries are merely nodes. The constant-strain triangle remain straight sided
when deformed and therefore, interelement compatibility exists for these
elements. Incompatible elements, those that allow gaps or overlaps between elements, can be acceptable and even desirable.

CIVL 7117 Finite Elements Methods in Structural Mechanics

Page 299

Convergence of Solution
When the mesh size is reduced - that is the number of elements is increased we are ensured of monotonic convergence of the solution when compatible and
complete displacement functions are sued.

Case

Number of Elements

Number of Nodes

Aspect
Ratio

Point A

12

21

-0.740

24

39

-0.980

32

45

-0.875

64

85

1.5

-1.078

80

105

1.2

-1.100

Exact Solution
CIVL 7117 Finite Elements Methods in Structural Mechanics

-1.152
Page 300

Interpretation of Stresses
In the stiffness or displacement formulation of the finite element method, used
in this course, the primary quantities determined are the interelement model displacements of the assemblage. Secondary quantities, such as stress and strain,
are computed based on these nodal displacements. In the case of the bar and
constant-strain triangles, stresses are constant over the element. For these elements, it is common practice to assign the stress to the centroid of the element
with acceptable results.
An alternative procedure sometimes is to use an average (possibly weighted)
value of the stresses evaluated at each node of the element. This averaging
method is often based interpolating the element nodal values using the element
shape functions. The averaging method is called smoothing. While the results
from smoothing may be pleasing to the eye, they may not indicate potential problems with the model and the results. You should always view the unsmoothed
contour plots as well. Highly discontinuous contours between elements in a region of an unsmoothed plot indicate modeling problems and typically require additional refinement of the element mesh in the suspect region.

Static Condensation
Lets consider the concept of static condensation and used it to develop the
stiffness matrix of a quadrilateral element. Consider a general quadrilateral element as shown below.

CIVL 7117 Finite Elements Methods in Structural Mechanics

Page 301

An imaginary node 5 is temporary introduced at the intersection of the diagonals

of the quadrilateral to create four triangles. We can superimposed the stiffness
matrices of the four triangles to create the stiffness matrix of the quadrilateral
element, where the internal imaginary node 5 degrees of freedom are said to be
condensed out so that they never inter the final equations.
Lets start by partitioning the equilibrium equations:

K11 K12 de Fe
K
d = F
K
22 i
21
i
where di is the vector of displacements corresponding to the imaginary internal
node, Fi is the vector of loads at the internal node, and de and Fe are the actual
displacements and loads, respectively. Rewriting the above equations we gives:

K11de + K12d i = Fa
K 21de + K 22d i = Fi
Solving for di gives:

d i = K 22 1K 21de + K 22 1Fi
Substituting the above equation, we obtain the condensed equilibrium equation:

kc dc = Fc
where

kc = k11 k12k 22 1k 21
Fc = Fe k12k22 1Fi
where kc and Fc are called the condensed stiffness matrix and the condensed
load vector, respectively. An advantage of the four-CST quadrilaterals is that the
solution becomes less dependent on the skew of the subdivision mesh. The
skew means a directional stiffness bias that is built into a model trough certain
discretization patterns.

CIVL 7117 Finite Elements Methods in Structural Mechanics

Page 302

The stiffness matrix of a typical triangular element, call it element 1, labeled with
nodes 1, 2, and 5 is given as:
(1)
k11
(1)
k (1) = k21
(1)
k51

(1)
k12
(1)
k22
(1)
k52

(1)

k15
(1)
k25

(1)
k55

where kij (1) is a 2 x 2 matrix. The assembled stiffness matrix for the quadrilateral
is:

CIVL 7117 Finite Elements Methods in Structural Mechanics

Page 303

Example Problem
Consider the quadrilateral with internal node 5 and dimensions as shown below. Apply the static condensation technique.

Using the CST stiffness matrix for plain strain, we get:

1

1.5 1.0 0.1 0.2 1.6 1.2

1.0 3.0 0.2 2.6 0.8 5.6

E 0.1 0.2 1.5 1.0 1.6 1.2

[k (1) ] = [k (3) ] =

4.16 0.2 2.6 1.0 3.0 0.8 5.6

1.6 0.8 1.6 0.8 3.2 0.0

1.2 5.6 1.2 5.6 0.0 11.2

4

1.5 1.0 0.1 0.2 1.6 1.2

1.0 3.0 0.2 2.6 0.8 5.6

0.1
0.2
1.5
1.0
1.6
1.2
E
[k (2) ] = [k (4) ] =

4.16 0.2 2.6 1.0 3.0 0.8 5.6

1.6 0.8 1.6 0.8 3.2 0.0

1.2 5.6 1.2 5.6 0.0 11.2

CIVL 7117 Finite Elements Methods in Structural Mechanics

Page 304

The resulting assembled matrix before static condensation is:

After partitioning, the condensed stiffness matrix is:

CIVL 7117 Finite Elements Methods in Structural Mechanics

Page 305

Flowchart for the Solution of Place Stress/Strain Problems

The following flowchart is typical for a finite element process used for the
analysis of plane stress and plane strain problems.

CIVL 7117 Finite Elements Methods in Structural Mechanics

Page 306

Example Problem

Consider the thin plate subjected to the surface traction shown in the figure
below (as discussed in Chapter 6). Use WinFElt to model this problem. Assume
plane stress conditions.

Assume plane stress conditions. Let E = 30 x 106 psi, = 0.30, and t = 1 in. Determine the nodal displacements and the element stresses.
Lets discretize the plate into two elements as shown below:

The general WinFElt input file format is:

problem description
title = " "
type = static : transient : static-substitution : modal : spectral :
static-thermal : transient-thermal
nodes
= nnn
elements = nnn

CIVL 7117 Finite Elements Methods in Structural Mechanics

Page 307

analysis parameters
start
= xxx
stop
= xxx
step
= xxx
beta
= 0.25
gamme
= 0.5
alpha
= 0.0
nodes
= [nnn, nnn, ...]
dofs
= [Tx, Ty, Tz, Rx, Ry, Rz]
mass-mode = lumped : consistent
nodes
x_name_x elements
material properties
forces
distributed loads
constraints
end

For this problem the WinFElt file is:

problem description
title = "example_p291"
nodes
= 4
elements = 2
analysis = static
analysis parameters
dofs
= [Tx, Ty]
nodes
1
x
2
x
3
x
4
x

= 0
= 0
= 20
= 20

y
y
y
y

= 0
= 10
= 10
= 0

z
z
z
z

=
=
=
=

0
0
0
0

constraint = pin
constraint = pin
constraint = free force = x_force
force = x_force

CSTPlaneStress elements
1
nodes = [1,3,2]
material = plate
2
nodes = [1,4,3]
material = plate
material properties
plate
t = 1.0
E = 30e6
forces
x_force

nu = 0.30

Fx = 5000.0

distributed loads
CIVL 7117 Finite Elements Methods in Structural Mechanics

Page 308

constraints
pin
Tx = c
free Tx = u

Ty = c
Ty = u

end

In this input file the keywords: title, nodes, elements, and analysis
define the type of problem. Under analysis parameters section the keyword
dofs define the degrees of freedom for each node.
Next the nodes are defined using the format:
nnn

x = xxx

y = xxx

z = xxx

force = x_name_x

constraint = x_name_x

where nnn is the node number, x, y, and z are the coordinates of the node,
force is the name of the point force, and constraint is the name of the cons
taint definition.
Next the elements are defined using the following format:
nnn

nodes = [nnn, nnn, ...]

material = x_name_x

load = x_name_x

where nnn is the element number, the nodes parameter defines the nodes that
form he element, material defines the element material properties, and load
define any distributed loads on the element.
Next the element material properties are defined using the following format:
x_name_x

A = xxx
Ix = xxx
Rk = xxx
Kx = xxx

t = xxx
E = xxx
G = xxx
nu = xxx
Iy = xxx
Iz = xxx
J = xxx
Rm = xxx
Ky = xxx
Kz = xxx
c = xxx

kappa = xxx

where x_name_x is the name of the material and the remaining keyword define
the section properties.
Next the nodal forces are defined using the following format:
x_name_x

Fx = xxx
Mx = xxx

Fy = xxx
My = xxx

Fz = xxx
Mz = xxx

where x_name_x is the name of the nodal force and the remaining keyword define the magnitude and direction of the concentrated force.
Next the elemental distributed loads are defined using the following format:
x_name_x

direction = x_name_x

values = (1, xxx) (2, xxx)

where x_name_x is the name of the element distributed load the remaining keyword define the magnitude, shape, and direction of the distributed load.
CIVL 7117 Finite Elements Methods in Structural Mechanics

Page 309

Next the constraint conditions are defined using the following format:
x_name_x

Tx = c:u

Ty = c:u

Tz = c:u

Rx = c:u

Ry = c:u

Rz = c:u

where x_name_x is the name of the constraint condition, the keyword u and c
state whether the degree of freedom is unknown or constrained.
The input file is terminated with an end statement.
After opening the input file in WinFElt, click on the solve button and an new window will open with displaying the results:
** example_p219 **
Nodal Displacements
----------------------------------------------------------------------------Node #
DOF 1
DOF 2
DOF 3
DOF 4
DOF 5
DOF 6
----------------------------------------------------------------------------1
0
0
0
0
0
0
2
0
0
0
0
0
0
3
0.00060958 4.1633e-06
0
0
0
0
4
0.0006637 0.00010408
0
0
0
0
Element Stresses
------------------------------------------------------------------------------1:
1004.8
301.44
2.4019
1004.8
301.43
0.19566
2:
995.2
-1.201
-2.4019
995.2
-1.2068
-0.13812
Reaction Forces
----------------------------------Node #
DOF
Reaction Force
----------------------------------1
Tx
-5000
1
Ty
-3002.4
2
Tx
-5000
2
Ty
3002.4

CIVL 7117 Finite Elements Methods in Structural Mechanics

Page 310

The results are similar to those presented in the textbook on page 298. In addition, WinFElt has some graphical visualization capabilities. By selecting the appropriate options on the Controls menu, a color contour of the displacements can
be plotted.

If more elements are utilized it may be possible to visualize the smoothed stress
contours.

Note that the stress contours are not correct and most likely due to the fact the
CST elements do not provide continuous inter-element stress values.

CIVL 7117 Finite Elements Methods in Structural Mechanics

Page 311

WinFElt help file for basic keywords

:Problem description
The problem description section is used to define the problem title and the number of nodes
and elements in the problem. The problem description section is the only section which you
cannot repeat within a given input file.
:nodes = integer
:elements = integer
These numbers will be used for error checking so the specifications given here must match the
actual number of nodes and elements given in the definition sections. Note that the definitions
for nodes and elements do not have to be given in numerical order, as long as nodes 1 ... m
and elements 1 ... n (where m is the number of nodes and n is the number of elements) all get
defined in one of the element and node definition sections in the file.
:analysis = problem type
defines the type of problem that you wish to solve. Currently it can either be static, transient,
static-substitution, modal, static-thermal, transient-thermal, or spectral. If you do not specify
anything, static analysis will be assumed.
:Nodes
The nodes section(s) must define all of the nodes given in the problem.
:x = constant expression
:y = constant expression
:z = constant expression
Each node must be located with an x, y, and z coordinate using x=, y=, and z= assignments.
Coordinates are taken as 0.0 if they are not otherwise defined. If a coordinate is left unspecified for a given node, it takes the value for that coordinate from the previous node.
:constraint = string
A node must have a constraint assigned to it by a constraint=statement. The constraint name
specified by string must be a valid constraint defined in the constraints section of the input file.
The default constraint leaves the node completely free in all six degrees of freedom. Like coordinates, if a constraint is left unspecified, the node will be assigned the same constraint as
the previous node.
:force = string

CIVL 7117 Finite Elements Methods in Structural Mechanics

Page 312

Forces (applied point loads and moments) on a node are optional and are applied using the
force= statement; if a force is not specified there will be no force applied to that node.
:mass = constant expression

specifies an optional lumped mass at this node.

:Elements
An element definition section begins with the keywords xxxxx elements where xxxxx is the
symbolic name of a type of element. All elements under this section heading will be taken to
be the given element type. There could be multiple sections that defined beam elements, but
each must begin with the keywords beam elements. If there were truss elements in the same
problem, they would have to be defined in sections which began with the keywords truss elements. Currently available types are spring, truss, beam, beam3d, timoshenko, CSTPlaneStrain, CSTPlaneStress, iso2d_PlaneStrain, iso2d_PlaneStress, quad_PlaneStrain,
quad_PlaneStress, htk, brick, rod, and ctg. A FElt problem is not limited to one element type;
the routines for assembling the global stiffness matrix takes care of getting the right parts of
the right element stiffness matrices into the global matrix.
:nodes = [n1, n2, ...]
Each element must have a list of nodes to which it is attached. The node list is defined with the
nodes=[ ... ] statement. The length of the list inside the square brackets varies with element
type, but must always contain the full number of nodes which the element type definition requires.
:material = string
A material property must be assigned to every element with a material= statement. If the material is never specified, an element will take the same material property as the previous element. If nothing ever gets assigned to an element, the default material property will have zeros
for all of its characteristics.
:load = string
An element can have up to three optional distributed loads. Each load is assigned with a separate load= assignment. The named load in each assignment must be a valid object defined in
the distributed loads section.
Each element type may treat a distributed load differently so you should be careful that the
names given for the loads on a given element match the names of distributed loads which are
defined in a manner conformant with what that element type is expecting.
:Material properties
The material properties section(s) is quite simple. Each material has a name followed by a list
of characteristics, such as
steel E = 30e6 A = 1.0 ...
CIVL 7117 Finite Elements Methods in Structural Mechanics

Page 313

Note that no element types require a material property with every characteristic defined. In
fact, most only use three or four characteristics. Density (rho) is always necessary for the materials in a transient, modal, or spectral analysis problem if you actually want your elements to
have any inertia. Different element types can certainly use the same material as long as that
material definition contains the right characteristics for each type of element that uses it.
:E = constant expression

Specifies the material's Young's (elastic) modulus.

:A = constant expression

Specifies the material's cross-sectional area.

:t = constant expression

Specifies the material's thickness.

:rho = constant expression

Specifies the material's density.

:nu = constant expression

Specifies the material's Poisson's ratio.

:G = constant expression

Specifies the material's bulk or shear modulus.

:J = constant expression

Specifies the material's torsional stiffness.

:Ix = constant expression

Specifies the material's I_xx moment of inertia.

:Iy = constant expression

Specifies the material's I_yy moment of inertia.

:Iz = constant expression

Specifies the material's I_zz moment of inertia.

:kappa = constant expression

Specifies the material's shear force correction factor.

:Rk = constant expression

Specifies the material's Rayleigh stiffness

damping coefficient.

:Rm = constant expression

Specifies the material's Rayleigh mass damping coefficient.

:Kx = constant expression

Specifies the material's thermal conductivity along

the x-direction.

:Ky = constant expression

Specifies the material's thermal conductivity along

the y-direction.

:Kz = constant expression

Specifies the material's thermal conductivity along

the z-direction.

:c = constant expression

Specifies the material's heat capacitance.

CIVL 7117 Finite Elements Methods in Structural Mechanics

Page 314

:Constraints
:Tx = constraint specification
:Ty = constraint specification
:Tz = constraint specification
:Rx = constraint specification
:Ry = constraint specification
:Rz = constraint specification
The constraints section(s) must define all of the named constraints within the problem. Each
constraint is defined by a name followed by a list of DOF specifications of the form Tx=? Ty=?
Tz=? Rx=? Ry=? Rz=? where the ? can either be c for constrained, u for unconstrained or a
valid, possibly time-dependent, expression for cases where a displacement boundary condition
is required (e.g., settlement of support, or time-varying temperature along a boundary in a
transient thermal analysis problem). Note that specifying Tx=c is equivalent to Tx=0.0. The T
and R refer to translation and rotation, respectively and the subscripted axis letter indicates
that the specification refers to translation along, or rotation about, that axis. An additional
specification of h or hinged is allowed for the rotational DOFs for cases where it is necessary
to model a nodal (momentless) hinge. Note that a hinge specification currently has meaning
only on beam, beam3d and timoshenko elements.
If a specification is never made for a DOF, then the problem is assumed to be unconstrained in
that degree of freedom. Getting the constraints right is an important part of getting a reasonable solution out of a finite element problem so you should be aware of what DOF are active in
a given problem (this will depend on which types of elements are being used ... even in a 2-d
problem, the global stiffness matrix will take displacement in the z direction into account if
there are 3-d elements in use, consequently, those displacements should be constrained).
:ITx = constant expression
:ITy = constant expression
:ITz = constant expression
:IRx = constant expression
:IRy = constant expression
:IRz = constant expression
In order to account for initial conditions in transient analysis problems, a constraint specification may also include the initial (at time t=0) displacements, and velocity and acceleration in
the translational DOFs. Initial displacements are given with ITx=, IRz=, etc.
:Vx = constant expression
:Vy = constant expression
:Vz = constant expression
:Ax = constant expression
:Ay = constant expression
:Az = constant expression
You can specify initial accelerations with Ax=, Ay=, and Az=. Initial velocities are given by Vx=,
Vy=, and Vz=. Unspecified velocities and displacements will be taken as 0.0. If there are no
initial accelerations specified (i.e., none of the nodes have a constraint with an acceleration
CIVL 7117 Finite Elements Methods in Structural Mechanics

Page 315

assigned) then the initial acceleration vector will be solved for by the mathematical routines
based on the initial force and velocity vectors. If any of the nodes have a constraint which has
an acceleration component defined (even if that component is assigned to 0.0) then the
mathematical routines will not solve for an initial acceleration vector; they will build one based
on the constraint information, assigning 0.0 to any component that was not specified. What
this means is that you cannot specify the initial acceleration for only a few nodes and expect
the mathematical routines to simply solve for the rest of them. If you specify any of the initial
accelerations then you are effectively specifying all of them.
:Forces
:Fx = variable expression
:Fy = variable expression
:Fz = variable expression
:Mx = variable expression
:My = variable expression
:Mz = variable expression
The forces section(s) defines all of the point loads used in the problem and actually looks a lot
like the constraints section. The magnitudes of the forces in the six directions are specified by
Fx=? Fy=? Fz=? Mx=? My=? Mz=?. If the value for a given force component is not given it is
assumed to be zero. The directions for both forces and the constraints as defined above
should be given in the global coordinate system (right-hand Cartesian).
If you are doing transient analysis, the force definitions can be more complicated than a simple
numerical assignment or expression. FElt allows you to specify a transient force as either a
series of time-magnitude pairs (a discrete function in time) or as an actual continuous function
of time. This latter fact means that you can define a force as Fx=sin(t) rather than having to
discretize the sine function. These continuous forcing functions are a special case of expressions as discussed above.
:Sfx = variable expression
:Sfy = variable expression
:Sfz = variable expression
:Smx = variable expression
:Smy = variable expression
:Smz = variable expression
For spectral analysis problems, you can explicitly specify input spectra using Sfx=, Smy=, etc.
if you want to to compute the actual power spectrum of the output. These spectra can be analytic functions of w or discrete frequency, power pairs.
:Distributed loads
The distributed loads section(s) must contain a definition for each distributed load that was assigned in the element definition section(s). A valid definition for a distributed load is a symbolic
name followed by the keywords direction=xxx and values=(n,x) (m,y) ... The direction assignment must be set to one of parallel, perpendicular, LocalX, LocalY, LocalZ, GlobalX, GlobalY,
GlobalZ, radial, axial. The valid directions for a given element type, and what those directions
CIVL 7117 Finite Elements Methods in Structural Mechanics

Page 316

refer to for that element type are described in the individual element descriptions in chapter~\ref{elements. The values assignment is used to assign a list of load pairs to the named
load. A load pair is given in the form (n, x) where n is the local node number to which the
magnitude given by x applies. Generally, two pairs will be required after the values= token.
The imaginary line between the two magnitudes at the two nodes defines an arbitrarily sloping
linearly distributed load. This allows you to specify many common load shapes: a constant distributed load of magnitude y, including cases of self-weight, a load which slopes from zero at
one node to x at a second node, or a linear superposition of these two cases in which the load
has magnitude y at node 1 and magnitude $x+y$ at node 2. The definition of this load would
be:
distributed loads load_case_1

direction=perpendicular

values=(1,-2000) (2,-6000)

:Analysis parameters
The analysis parameters section is required only if you are doing some type of transient, modal, or spectral analysis (e.g., analysis=transient, analysis=spectral in the problem description
section). For modal analysis it is simply used to set the type of element mass matrices that will
be formed, but for transient and spectral analyses it contains information that further defines
the problem and the parameters for the numerical integration in time.
:start = constant expression
the start of the frequency range of interest in spectral analysis
:stop = constant expression
the end of the time (transient analysis) or frequency (spectral analysis) range of interest.
:duration = constant expression
an alias for stop= in transient analysis problems.
:step = constant expression
the time or frequency step to be used between the start and stop points
:dt = constant expression
an alias for step= in transient analysis problems.
:beta = constant expression
:gamma = constant expression
:alpha = constant expression
integration parameters in the structural and thermal dynamic integration schemes
:mass-mode = for the types of element mass matrices that should be formed, either lumped
CIVL 7117 Finite Elements Methods in Structural Mechanics

Page 317

or consistent (note again that this is the only assignment that is required in this section for a
modal analysis problem);
:nodes =[integer, integer, ... ]
defines a list of nodes which you are interested in seeing output for. The list of nodes should
just be a comma delimited list of valid global node numbers.
:dofs = [ ... ]
defines the list of local DOF that you are interested in for the nodes that you are interested in.
The list of DOF should be a list of symbolic DOF names (Tx, Ty, Tz, Rx, Ry, Rz). You will get
solution output for each of these DOF at each of the nodes that you specified in the node list.
:Rk = constant expression
:Rm = constant expression
global Rayleigh damping parameters with Rk= and Rm=. If either of these parameters is nonzero then the global damping matrix will be formed using these two parameters and the global
mass and stiffness matrices as opposed to being formed from elemental Rayleigh parameters
and elemental mass and stiffness matrices.

CIVL 7117 Finite Elements Methods in Structural Mechanics

Page 318

Problems

18.

Do problems 7.1, 7.3, and 7.4 on pages 332 - 343 in your textbook A First
Course in the Finite Element Method by D. Logan.

19.

Work problems 7.7, 7.10, and 7.15 on pages 332 - 343 in your textbook A
First Course in the Finite Element Method by D. Logan using WinFElt.

CIVL 7117 Finite Elements Methods in Structural Mechanics

Page 319