• health monitoring, health management, and novel control and sensing technologies, including MEMS, smart materials, new sensors, and actuators

Each of these areas will require a better understanding of the processes and phenomena involved and more reliable prediction of interactions among the elements of the process or device.


Probabilistic Methods

The design of aircraft structure considers interactions among many complex processes, including material selection, fabrication, assembly, operations, and maintenance. The primary material parameters that affect a structural design are strength and stiffness, which are traditionally characterized by deterministically derived design property values, or “allowables.” Allowables are statistically reduced values based on experimental data and on assumed statistical models or distributions. Historically, structural design criteria are established by reducing design limits and ultimate conditions by a “safety factor” of 1.5, based on the ratio of ultimate strength to yield strength of common structural metals. In addition to engineering simplicity, the most important reason for using deterministic approaches is that design criteria expressed in terms of a margin of safety are more readily accepted by regulators and customers.

In the deterministic approach, a structure is designed to operate with the simultaneous occurrence of poorest allowable material quality and the most severe operating environment, level of damage, and service load conditions. That is, uncertainties are handled using conservative safety factors, a safe approach when most of the data describing the uncertainties are incomplete. As a result, current structural designs are very conservative and very heavy.

Probabilistic structural design/analysis is based on the principles of structural mechanics and uses conventional structural analysis tools, such as closedform solution of mathematical equations and finite element methods, to solve structural problems involving the variables. In the probabilistic approach, the actual strength or stiffness distributions are developed using analysis models that account for the probability of material defects, dimensional tolerances in structural component fabrication and assembly, variations in operational loads, and the probability of in-service and maintenance damage. Using probabilistic models, the safety and reliability of structures can be assessed over their entire lifetimes. Probabilistic structural design/analysis has been used to solve a variety of engineering problems, including spacecraft engines, durability analyses, and risk assessment of existing structures.

The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement