Engineering design is an exploratory creative process, a process of analysis, synthesis, decision-making, evaluation and optimization according to certain goals. In this process, a series of design parameters and design variables such as load, displacement, stress, strength, stiffness and life are defined to describe many engineering things and provide a large number of numerical calculations for engineering design; in addition, mechanical principles The design of the plan, the formulation of the mechanical structure and structural modeling, the selection of materials, parameters and coefficients, and the determination of process procedures all require engineering and technical personnel to make inferences, reasoning and decisions based on experience and intelligence. When analyzing and designing a mechanical system, it is necessary to establish a mathematical model for the complex mechanical system that can be used for analysis and design. At this time, the complex mechanical system must be simplified or idealized, so that the simplified mathematical model and When compared with actual mechanical systems, there are often errors or uncertainties. If certain errors and uncertainties are ignored during the modeling of a mechanical system and are not included in the expected data, it is impossible for the designer to trust the analysis based on this model. and designs can be applied to actual mechanical systems.
The uncertainties encountered in engineering practice mainly include uncertainty, unfixedness, unreliability, unpredictability, vague meaning, variability, incompleteness, unknown yet boundedness and irregularity, etc. For example, we call the difference between the true value and the approximate value of a physical quantity an error. In actual measurements and calculations, it is rarely or impossible to know the true value of a physical quantity. Usually only the approximate value of the physical quantity and the error limit are known. Therefore, the error is a kind of uncertainty. In mechanical design, determined structural parameters and determined mathematical models are generally used for analysis and calculation. In engineering practice, there are usually errors and uncertainties related to loads, material performance parameters, geometric dimensions, calculation models, initial conditions, boundary conditions, and joints of structural components. Although they may be small in most cases, the combination of these errors and uncertainties may cause large deviations or unpredictability in the response of the structure.
There are various errors and uncertainties in machinery. Some parameters have manufacturing errors, installation errors or uncertainties; some parameters have calculation errors and measurement errors; some mechanical systems have different characteristics when they are in different working conditions. Parameter values; some parameters have a certain parameter variation range; some parameters cannot be accurately measured or given at present.
Almost all design variables carry a certain degree of uncertainty, and in engineering analysis and design, they can be divided into three categories according to their sources:
(1) Physical uncertainty In engineering design, there are many physical quantities, such as load, material properties, geometric dimensions, etc., that are dispersed. This type of uncertainty that is directly related to physical quantities is usually called uncertainty of physical quantities.
(2) Statistical uncertainty The analysis of the distribution characteristics of variables and the determination of distribution functions are inseparable from statistics and judgment. However, any statistics and judgment are based on samples of variables, and the capacity of any sample cannot be infinite. The distribution pattern of any variable is determined through statistics and inference. Therefore, the statistical method of any parameter contains uncertainty. This uncertainty caused by statistics is called statistical uncertainty.
(3) Uncertainty of the model In engineering design and analysis, a model of the relationship between input quantities and output quantities must be established. The models are usually established based on mechanical principles or experience. Different models may be established for the same practical problem, and the models may be Certainty can also be uncertainty. However, in mechanical design, mechanical parts or systems are very complex, and ordinary mathematical models cannot fully reflect the actual structural conditions. Various assumptions made on the model and simplification of complex boundary conditions will cause the model to contain uncertainties. The uncertainty of the model will have a great impact on the results of mechanical design analysis and should be paid enough attention to.
