|
<< Click to Display Table of Contents >> Monte Carlo Simulation |
  
|
|
<< Click to Display Table of Contents >> Monte Carlo Simulation |
|
Because of the complexity of calculations, a Monte Carlo simulation technique was used to translate the uncertainty about the input into the uncertainty about the output. The magnitude of this uncertainty is established in terms of the probability distributions of the decision variables namely
•% shattered concrete (%SH)
•% pumping (%PU)
•% faulting (%FA)
•International Roughness Index (IRI)
In each simulation run, a generator of random numbers is used repeatedly, to create a value for each individual input variable, according to the triangular distribution specified for this variable on the Control Page. The random values of axle loads are generated using an empirical distribution given on the Axle Loads Page. Values thus generated are then used in formulas that comprise the design procedure. The output items thus obtained from the procedure are recorded by the computer. As the simulation progresses the average values of the decision variables generated so far are displayed by the meters on the Control Page.
At the end of simulation, there is, for example, a record of 100 000 values of each applicable decision variable. From these, probability distributions are plotted and shown on the Distribution Page.
In principle, the Monte Carlo simulation method pre-tries a pavement design. During each of the above-mentioned number of simulations, one possible design-outcome scenario is created. According to this scenario, the design will result in certain adverse consequences. These are the life cost, and pavement afflictions. The main pavement afflictions are shattered concrete, pumping, faulting, excessive crack spacing and excessive roughness. At the end of a simulation run, many thousands of outcome scenarios are thus available for analysis.
The program then summarizes these outcome scenarios in terms of probability distributions, averages, standard deviations, and confidence intervals of the decision variables. The figures thus obtained are compared both with objective criteria and subjective expectations. The decision to accept, modify, or reject the examined design configuration is based on these comparisons. Making this decision is the designer's right and responsibility.
The principal objective of the cncPAVE program is to make the designer foresee the consequences of his decision. Since the evaluation of consequences is very fast, the program is particularly suited to be used dynamically, to perform a succession of trial-and-error experiments directed by the user, and thus quickly arrive at a logical and rational design solution.