Loads, Strength, and Structural Safety
A. Loads
Loads that act on structures are usually classified as dead loads or live loads. Dead loads are fixed in location and constant in magnitude throughout the life of the structure. Usually the self-weight of a structure is the most important part of the dead load. This can be calculated closely, based on the dimensions of the structure and the unit weight of the material. Concrete density varies from about 90 to 120 pcf ( 14 to 19 kN/m3) for lightweight concrete, and is about 145 pcf (23 kN/m3) for normal concrete. In calculating the dead load of structural concrete, usually a 5 pcf ( 1 kN/m3) increment is included with the weight of the concrete to account for the presence of the reinforcement.
Live loads are loads such as occupancy, snow, wind, or traffic loads, or seismic forces. They may be either fully or partially in place, or not present at all. They may also change in location.
Although it is the responsibility of the engineer to calculate dead loads, live loads are usually specified by local, regional, or national codes and specifications. Typical sources are the publications of the American National Standards Institute, the American Association of the State Highway and Transportation Officials and, for wend loads, the recommendations of the ASCE Task Committee on Wind Force.
Specified live loads usually include some allowance for overload, and may include dynamic effects, explicitly or implicitly. Live loads can be controlled to some extent by measures such as posting of maximum loads for floors or bridges, but there can be no certainty that such loads will not be exceeded. It is often important to distinguish between the specified load, and what is termed the characteristic load, that is, the load that actually is in effect under normal conditions of service, which may be significantly less. In estimating the long-term deflection of a structure, for example, it is the characteristic load that is important, not the specified load.
The sum of the calculated dead load and the specified live load is called the service load, because this is the maximum load which may reasonably be expected to act during the service life of the structure. The factored load, or failure load which a structure must just be capable of resisting is a multiple of the service load.
Loads are random processes; more precisely, they are stochastic processes. However, in order to match the requirements of the methods of calculation actually used in most structural specification (allowable stresses and semi-probabilistic methods), each load is also characterized by the parameters representative of the different computational methods .
The loads can be classified with respect to their effect on the structure (static or dynamic) or with respect to their variation of intensity. Loads can also be classified with respect to some particular aspect, such as limited or not limited, having long or short duration, dependent or not on human activities etc.
Loads consist of: ( 1 ) concentrated and distributed forces ( direct actions ), (2) imposed deformations ( indirect actions ).
A load is assumed as a single load if it is not related to any other load or imposed deformation acting on the structure. In practice more than one single load acts on the structure, although it is convenient to consider each load separately.
Loads are random processes; more precisely, they are stochastic processes. However, in order to match the requirements of the: methods of calculation actually used in most structural specifications ('allowable stresses and semi-probabilistic methods), each load is also characterized by the parameters representative of the different computational methods',
The loads can be classified with respect to their effect on the structure (static or dynamic ) or with respect to their variation of intensity. Loads can also be classified with respect to some particular aspect, such as limited or not limited, having long or short duration, dependent or not on human activities etc.
B. Strength
The strength of a structure depends on the strength of the materials from which it is made.The properties of steel that have been described so far are applicable only if the ambient temperature stays within reasonable proximity of 70。F, say, from 30 to ll0。F. The properties of steel do not change appreciably for temperatures up to approximately 300。F to 400。F, although the stress-strain curve shows increasing nonlinearity when the temperature exceeds 250。F. It is therefore fortunate that most structures will never experience heat levels that go past these points, and even in those that do,the high temperatures are normally of very short duration and appear only over asmall portion of the structure. A typical example is what may take place in a structure during a fire: The temperatures may reach high levels, but only for a very short
time, and normally only in highly localized spots. Exceptions do appear, as has been evidenced by some of the more spectacular fire-related collapses (McCormick Place,in Chicago, Illinois, for example), but these are fortunately few and far between.Realistic conditions are better represented by what took place during the full-scale fire test that was conducted in a parking garage in Scranton, Pennsylvania,in 1972: Damage was localized, and most of it was easily repaired, for example, by cleaning blackened areas and replacing damaged tiles.Nevertheless, it is important to know how heat affects the material, and, if necessary, to take heat into account in the design process.
The relationships between the temperature and the primary strength and stiffness characteristics of steel are shown in Fig. 5.6. For all practical purposes, Fy, Fu, and E show decreasing values as the temperature increases, although the rate of decrease becomes significant only after the temperature has reached approximately 1,000*F. Fu actually exhibits a slight increase between 250* and 600*F, which is due to the phenomenon of strain aging. The yield and ultimate stresses have dropped to about one-half of their room temperature value at 1,100--1,200'F; at this level E has reached about 60 percent of its original value. The level of E is actually more important for the structure, since deflections are directly proportional to the modulus of elasticity. The phenomenon of creep will also come into play if the loaded structure is subjected to increased temperatures for an extended period.
Tempering of steel is normally done in conjunction with quenching, such as when the high-strength quenched and tempered steels are produced. The rapid cooling of the steel due to quenching will produce the hard, fine-grained structure called martensite. Although very strong, martensite is also very brittle, and the tempering is done to reshape the crystalling structure only to the extent that ductility and toughness are increased, but strength is maintained.
Any amount of heat input and subsequent cooling will produce a certain level of built- in or Y ii'~ stresses, due to the restraints of the material and the structure to the contractions that 3must take place. This occurs very prominently in welded joints; it will also occur throughout any structure or part of it that has been heated. If the heat has been applied unevenly, and the contractions are not restrained, a certain amount of distortion is bound to appear. This may make structural fitup difficult, but appropriate application of heat and controlled cooling may relieve such problems.
The designer that is concerned about welding and other residual stresses in joints with high degrees of restraint may choose to redesign the connection geometry to avoid problems. Reference details the solution to such problems as they pertain to lamellar tearing, but the recommendations are excellent advice for the design and fabrication of welded connections in general. It is also possible to use stress relieving in some form, as was done for a number of the beam-column-diagonal connections in the exterior frames of the John Hancock Center in Chicago, Illinois. However, stress relieving is usually an expensive procedure that is limited by the size of the heating ovens and the structural pieces that are to be treated. In most cases it is probably preferable to use other means of reducing the alleged problems of built-in stresses.
Local quenching effects may appear as a fire is extinguished in a structure, and water from the fire hoses hits heated steel. However, it is rare for local quenching effects to occur over anything but a minor area. The structural effect is therefore minimal, but heat treatment can be applied to remove any problem spots, if the owner of the structure is leery of doing nothing. Naturally, if the members have been deformed badly they may require replacement anyway. However, the material in itself does not usually suffer irreparable damage due to a fire.
C. Structural Safety and Reliability
Safety requires that the strength of a structure be adequate for all loads that may conceivably act on it. If strength could be predicted accurately and if loads were known with equal certainty, then safety could be assured by providing strength just barely in excess of the requirements of the loads. But there are many sources of uncertainty in the estimation of loads as well as in analysis, design, and construction. These uncertainties require a safety margin.
In recent years engineers have come to realize that the matter of structural safety is probabilistic in nature, and the safety provisions of many current specifications reflect this view.
Separate consideration is given to loads and strength. Load factors, larger than unity, ate applied to the calculated dead loads and estimated or specified service live loads, to obtain factored loads that the member must just be capable of sustaining at incipient failure. Load factors pertaining to different types of loads vary, depending on the degree of uncertainty associated with loads of various types, and with the likelihood of simultaneous occurrence of different loads.
More structural reliability theory is concerned with the rational treatment of uncertainties in structural engineering and with the methods for assessing the safety and serviceability of civil engineering and other structures. It is a subject which has grown rapidly during the last decade and has evolved from being a topic for academic research to a set of well- developed or developing methodologies with a wide range of practical applications.
Uncertainties exist in most areas of civil and structural engineering and rational design decisions cannot be made without modelling them and taking them into account. Many structural engineers are shielded from having to think about such problems, at least when designing simple structures, because of the prescriptive and essentially- deterministic nature of most codes of practice. This is an undesirable situation. Most loads and other structural design parameters are rarely known with certainty and should be regarded as random variables or stochastic processes, even if in design calculations they are eventually treated as deterministic. Some problems such as the analysis of load combinations cannot even be formulated without recourse to probabilistic reasoning.
Until fairly recently there has been a tendency for structural engineering to be dominated by deterministic thinking, characterized in design calculations by the use of specified minimum material properties, specified load intensities and by prescribed procedures for computing Stresses and deflections. This deterministic approach has almost certainly been reinforced by the very large extent to which structural engineering design is codified and the lack of feedback about the actual performance of structures. For example, actual stresses are rarely known, deflections are rarely observed or monitored, and since most structures do not collapse the real reserves of strengths are generally not known. In contrast, in the field of hydraulic systems, much more is known about the actual performance of, say, pipe networks, weirs, spillways etc., as their performance in service can be relatively easily observed or determined.
Most structural design is undertaken in accordance with codes of practice, which in many countries have legal status, meaning that compliance with the code automatically ensures compliance with the relevant clauses of the building laws. Structural codes typically and properly have a deterministic format and describe what are considered to be the minimum standards for design, construction and workmanship for each type of structure. Most codes can be seen to be evolutionary in nature, with changes being introduced or major revisions made at intervals of 3-10 years to allow for: new types of structural form, the effects of improved understanding of structural behaviour, the effects of changes in manufacturing tolerances or quality control procedures, a better knowledge of loads, etc.
The lack of information about the actual behaviour of structures combined with the use of cedes embodying relatively high safety factors can lead to the view, still held by some engineers as well as by some members of the general public, that absolute safety can be achieved. Absolute safety is of course unobtainable; and such a goal is also undesirable, since absolute safety could be achieved only by deploying infinite resources.
It is now widely recognized, however, that some risk of unacceptable structural performance must be tolerated. The main object of structural design is therefore to ensure, at an acceptable level of probability, that each structure will not become unfit for its intended purpose at any time during its specified design life. Most structures, however, have multiple performance requirements, commonly expressed in terms of a set of serviceability and ultimate limit states, most of which are not independent; and thus the problem is much more complex than the specification of just a single probability.
There is a need for all structural engineers to develop an understanding of structural reliability theory and for this to be applied in design and construction, either indirectly through codes or by direct application in the case of special structures having large failure consequences, the aim in both cases being to achieve economy together with all appropriate degree of safety. The subject is now sufficiently well developed for it to be included as a formal part of the training of all civil and structural engineers, both at undergraduate and post-graduate levels. Courses on structural safety have been given at some universities for a number of years.
