1.1. BEAM4 - 3D Elastic Beam

Matrix or VectorShape Functions Integration Points
Stiffness and Mass Matrices Equation 11–15, Equation 11–16, Equation 11–17, and Equation 11–18 None
Stress Stiffness and Damping Matrices Equation 11–7 and Equation 11–8 None
Pressure Load Vector and Temperatures Equation 11–15, Equation 11–16, and Equation 11–17 None
Load TypeDistribution
Element TemperatureBilinear across cross-section, linear along length
Nodal TemperatureConstant across cross-section, linear along length
PressureLinear along length

1.1.1. Stiffness and Mass Matrices

The order of degrees of freedom (DOFs) is shown in Figure 1.1: Order of Degrees of Freedom.

Figure 1.1: Order of Degrees of Freedom

Order of Degrees of Freedom

The stiffness matrix in element coordinates is (Przemieniecki):

(1–1)

where:

A = cross-section area (input as AREA on R command)
E = Young's modulus (input as EX on MP command)
L = element length
G = shear modulus (input as GXY on MP command)
Ix = input torsional moment of inertia (input as IXX on RMORE command)
Jx = polar moment of inertia = Iy + Iz
az = a(Izy)
ay = a(Iyz)
bz = b(Izy)
fz = f(Izy)
fy = f(Iyz)
Ii = moment of inertia about direction i (input as Iii on R command)

The consistent mass matrix (LUMPM,OFF) in element coordinates LUMPM,OFF is (Yokoyama):

(1–2)

where:

Mt = (ρA+m)L(1-εin)
ρ = density (input as DENS on MP command)
m = added mass per unit length (input as ADDMAS on RMORE command)
εin = prestrain (input as ISTRN on RMORE command)
Az = A(rzy)
Ay = A(ryz)
Bz = B(rzy)
Fz = F(rzy)
Fy = F(ryz)

The mass matrix (LUMPM,ON) in element coordinates is:

(1–3)

1.1.2. Gyroscopic Damping Matrix

The element gyroscopic damping matrix is the same as for PIPE16.

1.1.3. Pressure and Temperature Load Vector

The pressure and temperature load vector are computed in a manner similar to that of BEAM3.

1.1.4. Local to Global Conversion

The element coordinates are related to the global coordinates by:

(1–4)

where:

{u} = vector of displacements in global Cartesian coordinates

[T] is defined by:

(1–5)

where:

S3 = sin (θ)
C3 = cos (θ)
X1, etc. = x coordinate of node 1, etc.
Lxy = projection of length onto X-Y plane
d = .0001 L
θ = user-selected adjustment angle (input as THETA on R command)

If a third node is given, θ is not used. Rather C3 and S3 are defined using:

{V1} = vector from origin to node 1
{V2} = vector from origin to node 2
{V3} = vector from origin to node 3
{V4} = unit vector parallel to global Z axis, unless element is almost parallel to Z axis, in which case it is parallel to the X axis.

Then,

(1–6)

(1–7)

(1–8)

(1–9)

and

(1–10)

(1–11)

The x and • refer to vector cross and dot products, respectively. Thus, the element stiffness matrix in global coordinates becomes:

(1–12)

(1–13)

(1–14)

(1–15)

( is defined in Large Strain).

1.1.5. Stress Calculations

The centroidal stress at end i is:

(1–16)

where:

Fx,i = axial force (output as FX)

The bending stresses are

(1–17)

(1–18)

where:

My,i = moment about the element y axis at end i
Mz,i = moment about the element z axis at end i
tz = thickness of beam in element z direction (input as TKZ on R command)
ty = thickness of beam in element y direction (input as TKY on R command)

The maximum and minimum stresses are:

(1–19)

(1–20)

The presumption has been made that the cross-section is a rectangle, so that the maximum and minimum stresses of the cross-section occur at the corners. If the cross-section is of some other form, such as an ellipse, the user must replace Equation 1–19 and Equation 1–20 with other more appropriate expressions.

For long members, subjected to distributed loading (such as acceleration or pressure), it is possible that the peak stresses occur not at one end or the other, but somewhere in between. If this is of concern, the user should either use more elements or compute the interior stresses outside of the program.