The digital simulation is one of the tools allowing to simulate real phenomena.

A digital simulation (known as also data-processing simulation ) indicates a process according to which one carries out a Computer program on a Ordinateur in order to simulate for example a physical phenomenon complexes (ex: fall of a body on a soft support, resistance of an oil platform to the swell, tires of a material under vibratory request, wear of a roll of the dice,…).

The scientific digital simulations rest on the implementation of ideal model often using the technique known as of the finite elements . They are thus an adaptation to the numerical means of mathematical modeling, and are used to study the operation and the properties of a system modelled like predicting its evolution of it. One also speaks about numerical Calcul.

The graphical interfaces allow the visualization of the results of calculations by synthesized images.

These data-processing simulations quickly became impossible to circumvent for the modeling of the natural systems in Physique, Chimie and Biologie, but also of the human systems in economy and social science.

Critical epistemological

The digital simulation should not obviously be confused - as the term of simulation suggests it - with the real . It is not because the computer says that will occur as that that behaves indeed like such in reality (ex: digital simulations of weather forecasting). The digital simulation is only one representation of real base on the ideal model subjacent. If the ideal model thus computerized is erroneous, the calculated results are then false and can bring to erroneous decision makings themselves.

The best way of summarizing this epistemological criticism is: " the chart is not the territory! "

Thus, the digital simulation - which tends to becoming essential tools of engineering of the complex industrial objects because of its speed of implementation (ex: simulation of crashes automobile, exploration of the aeronautical flight envelopes, simulation of atomic bombs,…) - does not have to be used without scientific catch of retreat, nor professional prudence.

The critical analysis of the results, the checking of the validity of the ideal model used, the confrontation of the results predicted with the Expérience… is as many reflexes of Engineer to be had and which form part even of the ethical then of the professional user. This so in particular not taking of ill-considered risk in the decisions of Logistic, design (ex: to see the catastrophe inherent in the investment and/or seal of the booster rocket of the American Space shuttle whose digital simulation did not detect the structural weakness) (ex: to see the financial failures in the starting of the receipts of important a Parisian amusement park because of bad digital simulations of the frequentations and the estimated expenditure of the consumers).

History

Data-processing simulation appeared at the same time as data processing for the needs for the Projet Manhattan during the Second world war, in order to model the process of detonation nuclear. The first digital simulation “civil” in theoretical physics was the Expérience of Fermi-Pasta-Ulam (1953). Since, it evolved/moved data processing parallel to.

Categories of data-processing simulation

One can distinguish three categories of simulations:

  • simulation continues, where the system appears as differential equations to be solved. It makes it possible to compensate for the analytical resolution when this one is impossible. Carried out at the beginning on analog computers, it was carried out also on Ordinateur S as well as hybrid machines , and a third type of calculators which did not have a following day, the stochastic Calculateurs.

  • the discrete simulation in which the system is subjected to a succession of events which modify it. These simulations have vocation to apply simple principles to systems of big size. Discrete simulation is divided into two main categories:
    • asynchronous or time-slicing : one each time simulates the passage of a unit of time on all the system.
    • synchronous or vent-sequencing : the arrival of the next event is calculated, and one simulates only event by event, which often allows fast simulations, although a little more complex to program.
  • simulation by agents, where simulation is segmented in various entities which interact between them. It is especially used in economic and social simulations, where each agent represents an individual or a group of individuals. By nature, its operation is asynchronous.

Methods of simulation

Examples of simulations

  • Experiment of Fermi-Pasta-Ulam in statistical physics.
  • the code MCNP developed with the National laboratory of Los Alamos is very much used for simulations in Nuclear physics.
  • Of other codes of digital simulation of accessible physical phenomena since the Net.
  • virtual Experiments (Polytechnic, Lactam, Jean-François Colonna)
  • Daisyworld of James Lovelock, in support of its theory of named self-regulation Gaïa
  • Digital simulations in mechanics (Master numerical Modeling in mechanics, University of the La Rochelle)
  • Examples of digital simulations Various possibilities of simulation

Examples of software of digital simulation

  • COMSOL MULTIPHYSICS, software finite element combining generics (mechanics of the fluids, electromagnetism, mechanics of the structures, thermics) and computing power, making it possible to study an unlimited number of interactions between various physics (coolant for example),
  • CADFLOW, software of simulation and Engineering for Plasturgie
  • Beaver, software French of simulation of the behavior of a mechanical structure,
  • Solid Dynamics, French software of simulation of the efforts in the mechanisms,
  • LMS Virtual.Lab Motion, Belgian software (Company LMS) of simulation of the efforts in the mechanisms,
  • LMS Virtual.Lab tires, software Belgian (Company LMS) of simulation of tiredness in the structures (including point and weld beads)
  • LMS software Imagine.Lab of multiphysic simulation system
  • Working Model, software of simulation of movements, collisions etc…
  • Kadviser, French software of simulation of the deductive reasoning,
  • the SAMCEF software, software Flowing Belgian of simulation of the behaviors in mechanics of the structures,
  • , software American of simulation of the fluid flows,
  • Sysnoise, Belgian software of acoustic simulation, finite elements and of borders
  • Ansys, American software of simulation by finite elements of the behaviors in mechanics of the structures, mechanics of the fluids, electromagnetism, etc
  • Systus, French software of simulation in particular used by nuclear industry,
  • ABAQUS: French software of simulation (formulations implicit and clarifies),
  • Actran, Belgian software of acoustic simulation,
  • TRUE (Temporal Reasoning Universal Development): French software (True-World company) of modeling, simulation and artificial intelligence, interfaces wysiwyg, restitution in 2D, 3D and 4D,
  • Code_Aster: Free software of Digital simulation in mechanics of the structures.
  • Moldflow, software of plastic simulation of injection
  • CFDesign, software of simulation of flow of gas and fluids

Random links:Morning Live | Werner Franz | Fauna (butterfly) | Péault | Lawrence jack