SEED Guides Selecting Structural Configuration < >

The design of any solid object for whatever purpose must include some reassurance that its structural behaviour is adequate. The structural ability to resist distortion under load depends on two sets of properties, those inherent in the materials, such as elastic and plastic behaviour, and those resulting from the spatial configuration of the material, such as section properties or geometric arrangement. This Design Procedural Guide assumes some familiarity with the concepts of structural stiffness, statically determinant conditions, structural redundancy, and the behaviour of materials usually used for structural purposes.

The distinguishing property of a static structure is one, or several, fixed supports or foundations, regarded as rigidly interlinked or whose relative stiffness is known, providing sufficient reaction stiffness to resist any wholesale movement of the structure. Each support may be capable of providing a reaction in one or more of the six degrees of freedom (i.e. one linear and one rotational freedom in each of three non-coplanar directions, usually assumed to be, but not necessarily, orthogonal). Support stiffness is usually regarded as effectively infinite or with some pre-determined value. (Structures able to move with significant variation of acceleration, in at least one degree of freedom, will be dealt with in a companion guide. Situations where sudden changes of loading or repeated loading are likely to occur leading to vibration or fatigue will be covered in the units dealing with particular cases such as MPT 6.2, Component Design - Shaft with Fluctuating Loads.)

Newton's Third Law of Motion can be briefly stated as "action and reaction are equal and opposite". In static structural terms that implies that the total effect of a loading imposed on a body and the resultant reactions from its supports must balance, BUT since all forces and reactions on a body do not act at the same point, there must be some "mechanism" of force distribution to interconnect points of application and reaction. The structure of an object provides that mechanism and can be thought of as a maze of paths for the "flow" of forces between where they are applied and where they are reacted. Sometimes forces or reactions, or both, are spread throughout some or all of the body (such as the forces resulting from weight or magnetic fields) or distributed over part of its surface as pressure or shear (that is perpendicular or tangential to the surface). This analogy of flows is a powerful aid to the designers imagination in making design proposals. The task of designing structure is to ensure that the object being designed can withstand all of the loads likely to be imposed whatever other task the object may be intended to perform. Newton's Law also applies to any part of a structure isolated by a series of imaginary cuts, so the forces acting across all of the cut faces must be in balance with any external forces and reactions acting on that part. This is not only an aid to analysis, but can give a qualitative appreciation of the necessary form of the structural design. In many cases, the load pattern itself may be considered as a two or even one dimensional problem, even so, all structures must be robust in all directions, particularly to resist buckling failures which will happen in whichever is the weakest plane.

Issues to be considered

Loading
The pattern of loading the body must resist is usually specified as a series of normal service, and extreme load cases, including support conditions and definitions of failure, such as:

• permanent deformation (usually defined as local strain beyond an elastic limit, but sometimes in other ways),
• excessive deflection (either from a single loading or building up from successive loadings as small increments of permanent deflection, known as shake down),
• buckling (local or more general compressive instability),
• dynamic instability (unacceptable vibration or oscillation),
• fatigue (progressive incremental damage leading to failure after repeated loading),
• breakage (ultimate failure leading to partial or complete collapse).

A major distinction is made between loadings which occur as part of normal service, and those which only occur in extreme circumstances. Extreme loads may be 'proof loads', which must be sustained without damage and after which the structure may continue in normal service, or g ultimate loads', under which the structure is expected simply to behave in a predictable and safe way, but is not expected to sustain further use without repair. It is the need to sustain these unusual and extreme cases which often determines the majority of the structural design. Load cases rarely match actual service conditions. They are intended to provide a set of worst cases so giving confidence in the design. Care should be taken in specifying such cases to avoid dominating a design by a load which is very unlikely to occur or in adding extreme load cases together in impossible ways. However an extreme case might happen in a sensible combination with a normal service loading. Each load case is usually given a "factor of safety", or sometimes one for each type of possible failure, which reflects the degree of confidence which must be placed in the design and a recognition of the consequences of failure. Combining such factors with conservative analytical assumptions and material property values may produce unnecessarily sturdy designs.

Shape

Generally speaking, the simpler the shape, and the shorter and more direct the connections between loads and reactions, the more efficient the structure can be. The shape of the structure should therefore, if possible, complement the pattern of loading. As the match becomes less perfect, the compensating stresses grow, and eventually dominate the design. Where the shape is almost totally constrained by other considerations (perhaps aesthetic requirements, access holes, ducting, or spaces needed by other devices), the stress patterns may become very complex, and an efficient structural shape may only be achievable by interaction with a computer model, or through a large test programme. It may be possible to move load application points, redistribute forces, or modify the shape of other devices, to give a more convenient structural configuration without compromising its purpose. These possibilities should be examined in the early stages to give the structural designer the maximum freedom from constraint.

Structural action

Will it be a truss, a frame, shear panels, some hybrid, a shaped solid, or something more complex? Sometimes the use will dictate the type, other applications give a range of choice. Bridges may be cantilevers, beams, trusses, suspension bridges, arches, or combinations of these, with single or multiple spans and wide or multi-level decks. Sometimes the industry, the user environment, or non-structural requirements (such as aesthetics, or weather protection for users) determine the basic type. Once determined, it becomes the underlying basis for the details of the design. The basic differences between these types of action depend on the way in which they deal with load paths. A truss uses the inherent stability of triangulation to divide and recombine load paths in a balanced way between application points and supports. It consists of an assembly of (usually) straight members each connected only at its ends by pinned joints and carrying only direct loads between those ends. Load and support points are at the nodes unless a local spanning structure, which may also be one of the truss members, transfers the loads to them. A frame uses the moment carrying capacity of its members and joints to carry a more complex combination of bending, shear and direct loading along the various paths available. A panelled structure transfers shear between coplanar members in closed polygons (usually rectangular). This allows the members to take on the directly loaded role normal in trusses without the need for triangulation. Local stiffening may be needed at load and support points to distribute forces into the panels and also periodically to avoid panel buckling. A shaped solid does not have separately definable members but carries continuously varying stress patterns as the load paths spread to fill the available material.

Static Determinancy

The static determinancy of a structure must cover the combination of the structure itself and its support constraints. There must be sufficient support, in all six degrees of freedom, to balance all the imposed loads otherwise the structure is not static. If there is potential for the structure to be, or to nearly be, statically determinate the structural design is usually simplified. If the structure is not statically determinate, load will be distributed around the structure in proportion to the relative stiffness of the paths available. This distribution can be very sensitive to small changes in design or errors in construction thus making the design inherently less reliable.

The basic procedure for considering each of these issues in turn is outlined in the flow chart below. The rest of this Guide considers these stages in more detail.

Flow Chart 1 - DESIGN PROCEDURE FLOW CHART CHART 1