Modeling - Rigid End Offsets
Modeling - Rigid End Offsets
Objectives:
1) Model the rigid portion of a structure
2) Use rigid end offsets to account for neutral axes shifts
3) Use rigid members to model offsets
Up to now, we have used the centerline dimensions of the real structure to create our
numerical models. The figure below shows an example of a real and the numerical
model of a structure.
This implies some assumptions about the structure that we have represented in our model.
The main assumption is that the depth of the member is small enough as not to affect the
results. In reality, our beam or column members stop at the joint. The structural joint
then connects the beams to the columns. The figure below shows the connection of the
beams and columns to joint elements.
In steel structures where 36-inch members and reinforced joints are used, the centerline
dimensions may not be acceptable. In concrete structures, the extreme rigidity of the
joint may cause the use of centerline dimensions to be inaccurate. The way to model this
effect is through the use of rigid end offsets when the flexibility of the joint can be
ignored.
Rigid end offsets allow you to model the fact that the ends of a member are not
considered to be part of the flexible portion. A model using rigid end offsets can be
represented schematically as shown below.
Notice how a rigid member replaces the portion of the member that extends from the
center of the joint to the end of the flexible portion of the member. This effect can be
handled in two ways:
The use of extra nodes and members is an easy to use method that works in all programs.
The use of transformation matrices must be done at the program level and hence must be
an option in the program. We will look at the modeling method using extra nodes and
members first.
The most straightforward method of including rigid end offsets is to convert the model
with the rigid members into an analysis model. This means replacing the rigid sections
with small rigid beams. The model would look like:
Notice that each rigid end adds additional nodes and members. The DOF required for the
rigid end model are identical to the original structure, X and Y translation and Z rotation.
The choice of the properties for the rigid beam segments needs to again follow the rule of
thumb of 1000 times stiffer than the connected element. In this case, that means
choosing a moment of inertia 1000 times that of the connected element.
In the case of concrete or reinforced steel joints, this model will give good results.
However, the connection is still assumed to be rigid and the stiffness of the joint is
neglected. It is possible to add flexibility to the joint by using two nodes and springs
between them. This is identical to using a beam with a zero torsional moment of inertia
to model a hinge. By making the torsional stiffness represent the flexibility of the joint,
joint deformation can be included.
This same procedure of rigid members can be used to include the effect of a member
connected to another member but having different neutral axes. The most common use is
to include the effect of a pre-stressing tendon in a concrete beam. The concrete beam is
modeled using a standard beam element. The pre-stressing element is modeled using a
Note that the tendon is located at a distance d from the neutral axis of the beam. This
effect can be modeled using a rigid end offset for the truss element. The model shown
below can be used.
Beam Member
Tendon
Rigid Members Eccentricity
Truss Member
Notice how rigid arms connect the beam to the truss. This has the effect of causing the
truss to stiffen the rotational stiffness of the beam as well as its axial stiffness. This
model can be created using either of the two methods above. If the truss element will
allow initial forces, such as pre-tensioning or thermal effects, then pre or post-tensioning
effects can be analyzed. The truss force at a distance from the neutral axis causes the
required additional moment.
The stiffening effect of the truss can be seen by looking at the deflected shape of the
combined structure shown below
Using the above figure we can create the required SSTAN input file. The file uses the
centerline nodal locations and includes the rigid ends for the beams. Notice that the rigid
end are given in the global coordinate system. Hence, the sign of the offset matches the
global coordinate directions. For example, the tops of the columns have a negative Y
offset singe they are downward from the joint. Another point to notice is that the
distributed load is applied to the beam but the beam is 18" shorter (due to the rigid ends.
Therefore, 18" worth of the distributed load is neglected. The required input file is:
After running SSTAN, we can plot the structure to verify the input. Looking at the plot
we can see the rigid end offsets shown for each member. The maximum moment is
found under the beam results. The output for the beam is:
Note that the maximum positive moment is also given for loaded members.