<< Click to Display Table of Contents >> Navigation:  Flexcom > Data Inputs > $MODEL >
*Element

Purpose

To specify the finite element connectivity of the structural model.  

Theory

Refer to Elements for further information on this feature.

Note also that the old *NODE, *ELEMENT and *CABLE keywords have been largely superseded by the new *LINES keyword. Lines provide an automatic mesh creation facility to greatly expedite the model creation process. Using lines is a fundamentally different approach to working directly with nodes, elements and cables, although the information is ultimately handled in the same fashion internally. Since lines provide automatic mesh generation, you do not to concern yourself with explicit node and element numbering. Indeed the availability of lines makes nodes, elements and cables redundant to some degree, but they are retained for complete generality, and also to maintain downward compatibility with previous program versions. Refer to Lines for further information on this feature.

Keyword Format

Two types of line that may be mixed and/or repeated as often as necessary.

Line defining a single element:

Element, Start Node (Number or Label), End Node (Number or Label) [, v1, v2, v3, w1, w2, w3]

Line generating a number of elements from an existing element:

GEN=Master Element, No. of Elements [, Element Increment]

If the generation option is used then the master element must be defined previously. If you specify a node label rather than a node number, it must be enclosed in {} brackets. Element Increment defaults to 1. The default values of v1, v2, v3, w1, w2, w3 depend on whether the element is defined on a cable or not. Refer to Undeformed Versus Initial Positions for a detailed discussion on this topic.

Table Input

Elements - Define Directly

Elements - Generate

Notes

(a)Element numbers do not need to be continuous in a Flexcom model - as with node numbers you can use any arbitrary scheme for assigning element numbers.

(b)Articulations and spring elements are fully-fledged elements of the finite element discretisation, and you must define their connectivity, using the Elements – Define Directly or the Elements - Generate table.

(c)The specification of the components of the vectors V and W is related to the Flexcom facility for analysing a structure which is initially deformed. Basically, you specify V and W explicitly when the configuration defined by the nodal coordinates you have input does not represent a stress-free structure orientation. However, in the majority of analyses, this is not the case, and the specified nodal coordinates do represent the undeformed as well as the initial position. In this situation, the specification of V and W is optional, and should in general be omitted (with two exceptions to be discussed shortly). If you want to enter V and W values then all of the values V1, V2, V3, W1, W2, and W3 must be entered. When you do not define V and W explicitly, Flexcom calculates nominal values for these vectors using a default algorithm, based on the specified nodal coordinates. Refer to Undeformed Versus Initial Positions for a detailed discussion on this topic.

(d)The use of cables as discussed in Structures with Cables has one important ramification with regard to the orientation of the elements that comprise a cable. Where you allow the program to calculate the element orientation, as you normally will, a potential problem arises. Specifically, since the coordinates of the nodes along a cable are not known, the default algorithm of Note (c) above cannot be used. The program must, therefore, invoke an assumption or convention for this situation, and the convention used is that for all of those elements on a cable whose orientation is not explicitly defined by the user, the local axis system is defined as follows. The local x-axis is coincident with the global X-axis. The plane formed by the x and y local axes is coincident with the plane of the structure. This means that for systems defined initially in the global XY plane, the local axes are coincident with the global axes. This is of little significance in most cases, except in two situations. The first is where you want to apply a rotational boundary condition at a node that is on a cable. In this case you must ensure that the displacement term associated with the rotational constraint causes the element to be aligned with the desired orientation. The second case is where a model combines both cable elements and elements which are not on any cable. In this case you have to be careful that the undeformed configuration does not contain a bend or 'kink' where a cable and rigid element meet at a node, due to the fact that these elements have different orientations. This situation is considerably less common than the first. These issues are not discussed further here - interested readers are instead referred to Mixing Cable and Rigid Elements and Rotational Boundary Conditions and Cables.