Kt
Knowledge tools

Et
Example tools

Mt
Model tools

Rt
Risk tools

St
Structural tools

Ht
Historical timeline

Generic Rt Models

Rt contains a number of built-in models for hazards, responses, and damage, as well as economic, environmental, and social impacts. The background and documentation of these models are provided in Mt. Rt also contains several generic models that promote powerful ad hoc modelling. They are described below:

Algebraic Expression

A unique feature in Rt is that it is possible to implement user-defined models in a variety of ways, without any need for software compilation. One possibility is to write an arbitrary mathematical expression in the "Algebraic Expression" model. The expression can contain any of the basic mathematical functions. Parameters are input to this model by using parameter names in the expression. The output is given to an automatically generated Generic Response object.

Scripts

An even more powerful approach for implementing user-defined models is the Script model. By utilizing the ECMA Script Language it is possible to implement complex algorithms as JavaScript models in Rt, without any recompilation. The language specification for ECMA is available here, and many other references on Java Script programming are available online. The input to the Script model is a parameter list that names the parameters that are give to the script. A path to the location of the script file and an indication of whether the implementation includes DDM sensitivities are also given. This model does NOT automatically generate any response objects. Instead, Generic Response objects must be created, in which the name of the script model is given. The script must compute the value of the variables that have exactly the same name as these Generic Response objects. Parameters from Rt, like random variables, are employed in the script by utilizing the exact parameter name in the script. The naming convention to return "DDM sensitivities" from a script model is d_response_d_input as shown in the example that is posted in Et.

Root-finding

This model solves a general nonlinear single-variable equation and gives the root to an automatically generated Generic Response object. The input is:
  • Expression: The nonlinear equation, with x as the unknown variable, for example: x^2 – sin(x)
  • Algorithm: The solver algorithm, which can be: Newton, Secant, Steffenson, Bisection, False Position, and Brent Dekker
  • Maximum Iterations: Maximum number of iterations
  • Tolerance: The acceptable error in the result
  • Initial Root: The starting point for the search, used in the derivative-based algorithms Newton, Secant, and Steffenson
  • Lower Bound: The lower bound of the search interval, used in the bracketing algorithms Bisection, False Position, and Brent Dekker
  • Upper Bound: The upper bound of the search interval, used in the bracketing algorithms

Random Variable with Random Parameters

This model is employed when the distribution parameters of a random variable are themselves random variables. The distribution type of that random variable is given as input, together with a list of parameters that serve as its parameters, in the order prescribed for the standard distribution types in Rt. The input also includes a standard normal random variable that is transformed inside the model to represent the actual random variable. The output is a Generic Response object with the realization of the random variable.

External Software

Specific model are implemented in Rt to link directly with certain external software applications, which are listed below. Before utilizing these models it is necessary that you have the external application installed on your computer.
  • OpenSees
    • Input
      • Parameter List: List of the parameters that are mapped into the OpenSees model
      • Executable File: Absolute path to the opensees.exe file, for example C:\Program Files\OpenSees\opensees.exe
      • Input File: Absolute path to the location of the OpenSees inputfile, for example C:\mystructure.tcl This file contains $x entries, where “x” is the name of an Rt parameter, for example: uniaxialMaterial Concrete01 1 -$fprime -0.002 0.0 -0.006
    • Output
      • This model does NOT automatically create any response objects. Instead the analyst creates File Response objects. In those, specify the name of the OpenSees model and the name of the file where OpenSees stores the response
  • EMME
    • Input
      • Parameter List: List of the parameters that are mapped into the EMME model
      • Input Scenario File: This is a so-called template inputfile to EMME that contains Rt parameter names
      • Output Scenario File: This is the final inputfile to EMME, in which the parameter names have been replaced by their value
    • Output
      • This model does NOT automatically create any response objects. Instead the analyst creates File Response objects. In those, specify the name of the EMME model and the name of the file where EMME stores the response