Structural Classes Structural Classes

There are quite a few different section types which one may use in defining a class for a structural beam or plate:

     ~CLASS SHAPE_NAME                        PT PW -OPTIONS

     ~CLASS TUBE    A B C D               -OPTIONS

     ~CLASS CONE    A B C           PT PW -OPTIONS

     ~CLASS BOX     A B C D         PT PW -OPTIONS

     ~CLASS WBOX    A B C D E       PT PW -OPTIONS

     ~CLASS PRI     A B             PT PW -OPTIONS

     ~CLASS IBEAM   A B C D         PT PW -OPTIONS


     ~CLASS TEE     A B C D         PT PW -OPTIONS


     ~CLASS ANGLE   A B C D         PT PW -OPTIONS



     ~CLASS PLATE   A B C D               -OPTIONS

Here, the letters A through H which follow the section type are dimensions (inches or mm) which describe the size of the section and are defined in Figures 5 and 6 at the end of this section. The remaining two pieces of data, PT and PW define the thickness and width (inches or mm) of any attached plate one wishes to include in the section. Any attached plate is always attached at the side of the section in the beam +Z direction. The area of the plate is not included in the axial area of the section, but the inertia and resulting change in neutral axis is considered. Notice that a PLATE can be corrugated if one inputs the three additional lengths which define the corrugation. If only one number is input, the plate is flat.

All of these sections except the LLEG are connected to the nodes at the neutral axis unless instructed otherwise with the option:


Here, WHERE must be either TOP or BOTTOM. When this option is used, the node will be attached at the center of the "top" or "bottom" of the section. With symmetric section this is quite clear, but with non symmetric ones it is a bit harder to describe. An angle, for example, will be connected at the center of the flange if BOTTOM is used, but the bottom of the web when TOP is used. A LLEG section is special in that by default, the node is connected to the center of the tubular portion.

The shapes are as show below. The option:


can be used on shapes which are not symmetric about the neutral axis to reflect the shape about it. This has the same effect as specifying

     -CA 180

as an option on all of the elements which use this class.

In addition to the standard options discussed above, there are several ones specific to beam classes, the first group of these are:



which define fatigue properties for the elements. For a discussion of the -SCF and -SN options, see the sections on associating SCFs and SN curves with fatigue points.

The next group contains the section options:


     -POINTS, Y(1), Z(1), AY(1), AZ(1), ..... \
     Y(n), Z(n), AY(n), AZ(n)

     -P_FY, FY(1), FY(2), .... FY(n)

     -M_P, Zy, Zz

     -P_N, Pn

     -ETA, ETA

     -F_TYPE, TYPE

If one wishes to override the section properties computed from the dimensions, then he should use the -SECTION option. Here, AREA is the cross-sectional area (inches**2 or mm**2), IY and IZ are the moments of inertia (inches**4 or mm**4), J is the polar moment of inertia (inches**4 or mm**4), and ALPHAY and ALPHAZ are the shear area multipliers. The ALPHAs are numbers which transform average shear stresses into the true shear stresses at the neutral axis, i.e.


If any of these values are less than or equal to zero, the program computed value will be used.

Normally, MOSES determines "critical" points at which to compute stresses. Sometimes, one wishes to define these points himself. Stress points are defined with the -POINTS option. Here, Y(n) and Z(n) are the beam system Y and Z coordinates of the point and AY(n) and AZ(n) are the values of alpha for that point. (These are the same alphas discussed above except that instead of computing the stresses at the neutral axis do it at the nth point.) The option -P_FY allows one to define the yield stress at the "critical points". If one uses this option, he really should define his own points so that he is certain of the physical location of each point. The options -M_P and -P_N define the plastic moments and the nominal axial strength of the section. NOTICE If you use the -SECTION option to change the properties, you probably will also need to use the -M_P and -P_N options as well. The option -ETA defines the exponent "eta" in the interaction formulae of the AISC LRFD code check. Finally, the option -F_TYPE defines the fabrication type of the section. Here, TYPE must be either FABRICATED or COLD_FORGED.

The next class of Structural Class options are:

     -LEN, L




In general, MOSES allows for an element to have different properties along its length. To define such an element, one should have a Class definition command for each segment of the element. When more than one segment is defined, the first set of properties are associated with the beginning or "A" end of the element, and the last set of properties associated with the "B" end of the element. The lengths of the segments are defined by the -LEN option with L (feet or meters) being the length. For BEAM classes, the length of every segment but one should be defined. MOSES will then compute the length of this segment so that the total length of the element is correct. Also, one can define the segment lengths with the -PERL option. Here, PERL is the percentage of the total length of the element which will be attributed to a given segment.

When one uses a shape type of CONE, MOSES will automatically divide the beam into a number of prismatic segments based on the number specified with the -REFINE option. Thus, if -REFINE is not specified, then the beam will consist of a single tube with a diameter which gives the correct volume. The thickness of the approximate cylinder is

     T = Ti * ( R1 + R2 ) / ( 2 * Ra )

where Ti is the thickness input, R1 and R2 are the radii at the ends and Ra is the diameter of the approximate cylinder. This is an approximation of the correct thickness for small values of (R1-R2)/L. Here, the first dimension given is the outside diameter (inches or mm) at the "beginning node" of the element, the second dimension is the thickness (inches or mm), and the third dimension is the outside diameter (inches or mm) at the "end node" of the element. The options which alter the load attributes are not honored for cones, and one cannot use a cone section as the one with zero specified length if when defining beams composed of different shape types if the section is refined.

MOSES has an ability to automatically redesign a class of members so that all members within that class have favorable code checks. All resizing is performed on a subset of the shape table (more will be said about this in the next section). The subset considered during resizing is defined by a single selector, and two limits define with the -RDES option. The program will consider only those shapes which match the selector. For tubes, only sections which satisfy the d/t and kl/r limits specified will be considered. MOSES will consider all shapes selected to produce a shape which will yield a minimum cost and which satisfies the code check criteria.

MOSES allows one to define "stiffeners" for structural elements. Both longitudinal and transverse stiffeners can be defined, and they add both stiffness and weight to a model. In addition to adding weight and stiffness, stiffeners are used in checking some codes. In general, stiffeners are associated with a previously defined class. The weight added to elements by stiffeners can be eliminated with the use of the -ST_USEW option of either &DEFAULT or on the element definition command, or by specifying -WTPLEN to be zero when defining the stiffener class. The weight is computed by the weight per length of the stiffener times its length times the number of stiffeners. If one does not specify a class for the stiffeners, then the stiffeners are "magic" in that they are weightless and automatically pass any checks on their properties.

While these are conceptually simple, one can easily become confused by the details. While the form of the options used to define stiffeners is the same for all elements, the details differ. Thus, let us begin by considering stiffeners on generalized plates. Here, the options used to define stiffeners are:



The option which begins with -T defines transverse stiffeners and that beginning with -L defines longitudinal ones. Here, longitudinal stiffeners are parallel to the element X axis and transverse ones are perpendicular to the X axis. For both of these, options, ~STIF_CLASS is the class name which will be used to define the stiffener, and have been defined previously. WHERE defines the "vertical position" of the stiffener. WHERE may be either +Z, or -Z. If +Z is used, the stiffeners will be connected to the "top" size of the plate and for -Z, the bottom side. If WHERE is omitted, INTERNAL or +Z will be used.

Longitudinal stiffeners on beams are similar to longitudinal stiffeners on plates, and are defined with the -L_STIFF option discussed above and


Also, here WHERE can also have the additional values of INTERNAL of EXTERNAL which make sense for closed sections. In reality, +Z and EXTERNAL are the same as are -Z and INTERNAL. The difference between -L_STIFF and -LN_STIFF is that the first defines the location of the stiffener by a spacing (here SPACE is in inches or mm) and the second by the number of stiffeners. Obviously,

     SPACE  = DISTANCE / ( NUMBER + 1 )

Where DISTANCE is the distance over which the stiffeners are applied. For tubes, the distance is the circumference (inner or outer depending on WHERE). All other shapes are composed of rectangles. Here, DISTANCE and NUMBER are used for each rectangle of the section and DISTANCE is the longer dimension. Thus, for a "PRI" section, the DISTANCE is the greater of A and B. Fractional stiffeners are used and they are "smeared" over the full width. Thus, adding stiffeners increases both of the inertias of the section.

Transverse stiffeners on beams are the most complicated and are defined with either of the two options:



Here, the values have the same meaning as those for longitudinal stiffeners on beams. The new value, LENGTH, will be discussed in a minute. One of the problems with transverse stiffeners is that they are used for two purposes: stiffeners against hydrostatic collapse and to stiffen joints. Since transverse stiffeners on beams are normally used on tubes, we will call them rings here. To avoid confusion, rings will be used to stiffen joints only if the class has more than one segment and the stiffeners are in the segment closest to the joint}.

Transverse stiffeners suffer from the same problems as does buckling lengths. If an element is not fully supported on both ends, then the longitudinal stiffener spacing may be longer than the element length! This is where the value LENGTH comes in. It provides a "DISTANCE" to be used in the conversion from spacing to number of stiffeners. You can specify three things for LENGTH: a length in feet or meters, the token LENGTH, or the token BLENGTH. If LENGTH, is specified, the length of the element is used, if BLENGTH is specified, the minimum of the two buckling lengths is used, a number input will be used directly, and if this parameter is omitted, the segment length will be used.

Let us consider two examples. First, suppose that we have an element which is actually part of a logical beam, and suppose the logical beam has a single hydrostatic ring. If all elements of the logical beam had the same properties, they could be defined with the class:


This class will place one ring in the center of the logical beam and have a stiffener spacing of half the minimum buckling length. Each element would have the same stiffener spacing, but a stiffer weight of the weight of the ring times the element length divided by the buckling length. As another example, consider a beam for which we need a joint ring at one end. The class


We must use a segment to define the joint ring, hence the first segment. The hydrostatic ring defined for the second segment will be the middle of the second segment and the stiffener spacing here is half the length of the second segment.