外文翻译(结构设计背景)

第三部分：外文翻译结构设计背景Background for Structural Design1. Practice versus TheoryWe hear much of the conflict between theory and practice. Actually, of course, there will be no conflict between good theory and good practice, although the two frequently seem at cross-purposes, particularly when both are bad. Bad theory develops from unjustifiably crude assumptions, while bad practice follows unjustifiably crude methods. When theory can be based upon correct premises and practice can be controlled by one who understands the theory involved, the two will agree. Nevertheless, there are certain considerations of practice that must be allowed to control design, particularly to facilitate construction. A few of the many problems that should influence the thinking of the designer and of the construction engineer will be discussed.2. Analytical CalculationsSince analysis precedes design, it will be useful to think over the process of analysis from the point of view of the practical designer. Analysis, to serve a useful purpose, must finally reach expression in terms of tons of steel, cubic yards of concrete, and board feet of structural timber. It is useless for the analyst or the designer to expect the construction engineer to worry about increasing the unit stress in a steel beam by a few hundred pounds per square inch above the allowable stress by the shifting of a partition. The field man knows that there are decisions he will have to make during erection that may influence the stress to a greater extent than the amount mentioned. For the same reason, he is not likely to be sympathetic when the blueprint carries a statement that a field connection is to be welded at a distance of 5 j ^ in. from a sheared edge.The accuracy of field work is seldom greater than a tolerance of in. and a sheared edge is far from a planed edge at best. The designer will cultivate the respect of the field man by avoiding such inconsistencies.With these considerations in mind, we may conclude that there is little reason for a designer to use log tables in making his usual calculations. A slide rule will provide all requisite accuracy; also, such calculations will actually command greater confidence. However, this does not justify the substitution of crude guesses for accurate analysis or for careful design calculations.Theory of ElasticityThere is no tool that has proved of greater value to the designer than the theory of elasticity. On the other hand, it is worth remembering that the significance of the word elasticity automatically rules out the effect of plastic flow or "yield". Hence, the distribution of stresses presented by this theory is the picture that would apply before any single particle had passed the yield point. As soon as any part of the structure begins to yield, the distribution of stress will change. Generally speaking, we find that plastic yielding tends to equalize stresses by a redistribution of moments, shears, and fiber stresses. The accomplished designer will be able to interpret and use the results of mathematical studies based upon the theory of elasticity, but he will not fail to readjust his ideas of structural action to allow for the influence of yielding beyond the elastic limit.3. DuctilityThis property has been mentioned as one which helps to reduce stress concentrations. For instance, according to the theory of elasticity, a small hole in a simple tension member will produce a stress concentration of three times the average unit stress in the member. Photo-elastically it hasbeen possible to measure stress concentrations around a hole of more than twice the average stress in the member. It is therefore surprising that rivet holes do not seem to reduce the ultimate static strength of a tension member (steel) by more than the influence of the reduction of effective area. The explanation must be that the steel around the rivet hole flows and thus permits a redistribution of stress so that the maximum unit stress at fracture is little greater than the average unit stress. There are innumerable similar conditions to be evaluated in structural design. All "stress raisers", such as notches, holes, threads, and cross-sectional changes, are best eliminated, but, if they are unavoidable, some reduction of their objectionable features will tie obtained from ductility.4. Cleavage or Brittle FractureA type of fracture not seen very frequently in buildings and bridges is a brittle running crack without visible yielding or plastic flow of the adjacent material. Many ships, tanks, and other steel plate structures, particularly when welded, have been destroyed by brittle fracture. When test coupons are cut from material adjacent to a brittle crack and pulled in uniaxial tension in a testing machine at room temperature, the material will usually stretch 20 percent or more in length before fracturing. Coupons removed from a weld adjacent to a brittle fracturing are likely to show even greater ductility than die parent metal. Hence, we can hardly blame either the weld or the plate itself for permitting the crack to progress catastrophically after its initiation. By checking the point of initiation, one invariably finds a stress concentration, such as a corner, a hole, or an arc strike in welding, but such concentrations of stress exist in other structures where brittle fracture does not occur.Lengthy investigations have isolated several factors that tend to produce a catastrophic brittle fracture if a small crack is initiated by a pointof high stress concentration. One significant embrittling factor is low temperature. Steels usually become brittle at a temperature well below zero degrees Fahrenheit. It is significant, however, that any temperature well below freezing will embrittle certain structural steels. The temperature below which given steel loses a significant fraction of its ductility or energy absorption before fracture, as measured by the Charpy test, is called its "transition temperature". We have learned that steels which have a transition temperature above the temperature of exposure in service are inherently subject to brittle fractured. The author observed a beam, attached to a wall column that fractured without visible ductile deformation when the wall was opened in winter for repair. The beam had served for thirty years while protected from winter temperatures by the heat of the building. Doubtless its transition temperature was above the temperature of the exposure.A second embrittling factor is triaxiality of tensile stresses. Theoretically a cube of any ductile material will lose all of its ductility and will fracture by pure cleavage if it is subjected to equal tensions of sufficient magnitude in any three perpendicular directions. Such perfection of triaxial tension is not likely to occur in a structure, but unequal tensions in three perpendicular directions are not uncommon. Any plate may be subject to biaxial tension; biaxial stress; in fact, is the usual reason for its existence. Then, as the author has shown, a third tension stress perpendicular to the plate at its mid-depth will develop from Poisson's ratio at the exact end of any tiny crack in the plate. Hence, at the end of any tiny crack-like imperfection in a plate or weld a condition of triaxial tension occurs that without doubt is an embrittling factor along with low temperature. If the imperfection develops into a visible crack, the triaxiality of tension continues to redevelop right at the end of theextending crack and thus encourages it to progress as a britde running fracture.A third embritting factor is any hidden stress that tends to build up the general tensile stress field since brittle fracture naturally does not occur under low stresses. Such hidden stresses are those due to changes in temperature of one part of a structure without equal temperature change for other integral parts, and also the residual stresses due to rolling, cooling, straightening, or forced fit during erection.A fourth embritting factor has been established by tests of prestrained material. It has been found by Mylonas and Drucker that a compressive prestrain of two or three percent across a notch followed by a relatively low tension in the same direction (less than one half of the yield tension) reduces residual ductility and may produce a brittle fracture.The four embrittling factors mentioned above are low temperature, triaxiality of tension, hidden stress fields that raise the anticipated level of stress, and loss of ductility due to prestraining in compression. Stress concentrations may be involved in the latter three factors. Each of these factors exists in some degree in every structure. In large plate structures, such as ships and tanks, the factors of embrittlement tend to combine to a dangerous degree. The designer needs to be aware of their inherent danger so that he may reduce by good design the possibility that such factors may combine to initiate a catastrophic fracture.5. The Factor of SafetySome writers have considered the factor of safety to be based upon ultimate strength, while others feel that the ratio of the elastic limit to the working stress is in reality the factor of safety. The latter point of view is certainly the more significant, but neither presents a correct picture. Theengineer is always willing to let the actual stress approach nearly the elastic limit. The range between the working stress and the elastic limit is mainly an allowance to cover unknown or partially unknown stresses.(1).Fabrication and Erection StressesIt is no secret that structural steel is handled rather roughly in the shop and in the field. Rivet holes seldom line up perfectly; hence they must be pulled into line. Welding warps and buckles the structure and leaves high residual stresses. During fabrication, bent shapes are straightened as a standard part of the fabrication process, and, of course, the elastic limit must be passed to accomplish this. The mere punching of a hole distorts the surrounding material and leaves high residual stresses. The writer is convinced that these processes will result in a structure having stresses, under the design loading, that reach the elastic limit over small areas. Such a structure would be highly unsafe if it were not constructed of a ductile material such as structural steel.(2).Knowledge of LoadsOne of the undeterminable factors in design may be the loading itself. Dead load can be estimated quite accurately, but live loading, wind, and impact, as well as traction, sway, and other inertia forces are extremely variable. Then there is the influence of temperature and the action of settling supports that often damage an otherwise well-designed structure. The engineering designer makes a sincere effort to evaluate the probable loads, but even his best judgment is unable to cope with the situation in all cases. One purpose, then, of the factor of safety is to provide some reasonable allowance for possible increased loading.(3).Knowledge of MaterialsMost design is based upon specifications that assume certain properties for the structural materials. The common specified minimum elastic limit for structural steel of one type is 33,000 lb/in2. This lower limit is controlled by mill tests. A batch of steel rolled into structural shapes has a number of coupons cut from it for testing. If we make thousands of such tests for a single batch of steel, a few will turn up that show a yield point considerably below 33,000 lb/in2. However, the chance is small that a limited number of mill tests will happen to locate the small amount of weak material. One who understands the theory of sampling is not surprised that constant strength of a product such as structural steel is not even approached. Many factors must be controlled in producing steel, each factor being permitted to vary within a limited range. These factors therefore combine to produce a variable product.6. Fabrication MethodsIt is the responsibility of the designer to understand fabrication methods and to fit each particular design to the fabrication facilities available. For instance, it is undesirable to select a beam that is longer than rolled sections stocked in local warehouses or longer than the possible situations that may need to be controlled for safe structure fabrication shop can handle properly. It is worth noting that each central warehouse provides the draftsmen in its vicinity with a list of maximum sizes of materials that are readily available. Special sizes may not he obtainable for months, even at an increased cost per pound. The designer should work with the shop man so that the resulting structure will be economical. An edge can often be finished by grinding, by milling, or, possibly, simply by burning. Knowledge of relative costs is necessary if one is to reach a proper decision.Field ErectionThe designer usually has more difficulty in cooperating with the field organization than with the shop. The reason is that field conditions are never under complete control. The weather, the soil, the kind of labor obtainable, and the vagaries of nature all combine at times to plague the field engineer so that he finds it difficult, if not impossible, to follow the exact plan presented to him. On the other hand, construction engineers are so versatile that they can usually accomplish the result desired even though some changes become necessary. The responsibility again falls upon the designer to consider the influence of all possible field conditions upon his design. Some designs must be made so that the structure can be erected by unskilled labor, while other structures may be dependent upon the services of welders and craftsmen of highly specialized qualifications. The writer knows of one bridge that was designed for transportation on the backs of camels and another that was brought to the site by airplanes. Even freight car or truck transportation introduces certain limitations that must be observed as to the over-all size or length of a given piece. possible situations that may need to be controlled for safe structural design. Standard sets of specifications are prepared under the sponsorship of the technical societies. Over a period of years such specifications have been written arid rewritten many times. The profession as a whole has used each specification and has either accepted or rejected it. Therefore, a standard set of specifications may be accepted to represent the best information available on the subject as of the date when it was written.7. Cost as a Major FactorThe previous discussion leads to the inevitable conclusion that only an economical design can be a good design. The designer will accomplish little if his structures are seldom built because of excessive cost. Therefore, the designer must balance himself between the danger of unsafe practiceon the one hand and over conservatism on the other. His best approach to the solution of this problem is to learn everything possible from the detailer, the shop man, and the construction engineer. If he knows the tolerances, clearances, and allowances introduced by the detailer, the sizes, tools and methods used by the shop, and the shapes, weights, and fits desired by the field organization, his designing is likely to be successful.In the study of costs, it is interesting to observe that certain structures commonly used in foreign countries are seldom used in the United States. There are the highest labor costs in the world, which explains the requirement of machine production in the United States. Slender structures are more likely to be found in Europe, where the high costs of material and low cost of labor make weight reduction important, a fact that is particularly evident in the field of reinforced concrete.8. SpecificationsAll structural design is controlled by specifications. Even if no limitation is placed upon the designer, he will still be very likely to depend upon a standard set of specifications for guidance. All large cities have building codes that specify not only working stresses and qualities of materials and workmanship, but such general features as window area, hallway widths, and fire provisions for a building, and similar features of other structures. The designer will follow the specifications of the local building code by necessity, but he will also usually follow the provisions of standard sets of specifications (AREA, AWS,ACI)for his own guidance. It is impossible for anyone designer to have experienced all of the al design. Standard sets of specifications are prepared under the sponsorship of the technical societies. Over a period of years such specifications have been written arid rewritten many times. The profession as a whole has used each specification and has either accepted or rejectedit. Therefore, a standard set of specifications may be accepted to represent the best information available on the subject as of the date when it was written.9. Structural FailuresThere are a great many minor structural failures, but unless there is loss of life or oilier newsworthy features about a particular failure, it never comes to the attention of anyone except the firm that repairs the damage. Frequently, the owner requests that no publicity be given to failure. Many failures are caused by improper details. It has been a habit of " handbook designers" to select members of ample size and then to connect them together inadequately. Most building failures due to wind can be traced to this weakness. Undoubtedly, this is due to the fact that member selection is often quite simple, while joint design requires a greater understanding of stress analysis.(1). SettlementCertainly the most common source of building failures is foundation settlement. The design problem involved is not to prevent settlement, which can never be done, but to obtain uniform settlement so that the structure will not be stressed thereby. For instance, if all footings of a building settle the same amount, the building will be uninjured. However, unless uniform settlement is certain, the designer should make an allowance for unequal settlement in his analysis. Hence, the ideal structure for such a location may be one that is flexible or deformable rather than rigid or ever stiff. For this reason, the simple span structure has long been pointed to as the ideal where unequal settlement is anticipated.(2).Excessive DeflectionA common error in design is to select a beam or truss properly for strength but to fail to check its load deflection. Excessive flexibility may produce cracked plaster, permit vibration amplitude to build up, or even lead to collapse. Complete collapse often results from excessive flexibility of flat roofs. The dead-load deflection produces a low spot in the roof that collects water or ice. The increased water or ice load produces further deflection which allows more water or ice to collect. The process of self- destruction is certain to continue to the point of collapse if the span is rather great, because a small added deflection produces a significant increase in water loading. The solution is either to increase stiffness to meet specifications or to camber the roof so heavily that water can never collect on it.10. ConclusionAll things considered, it is remarkable that catastrophic failures occur so seldom in structures. This fact has led many engineers to feel that absolute safety can be guaranteed by proper specifications. However, we have seen that both the loads and the strength of structural materials, members and joints are governed by the theory of probability. Therefore, although one might be able to design a structure with a probability of failure as low as one in a million, it is never possible to reduce the probability of failure for a complex structural assemblage to zero.。