aa
SIMULATING WITH MULTI_R_DESIGNER

Simulations can be achieved to predict hydrodynamics, mass and heat transfer properties using 114 models and correlations. Among them, 16 Artificial Neural Network correlations developed at Laval University (MR_ANNs) were built from the Multi_R_Designer databank. These properties are listed in the main menu by selecting Simulation/Start new simulation. The associated models and corresponding inputs can also be visualized.

Two options are available: manual simulation or automatic simulation. An example is given below to illustrate the functionalities of the simulator.
 

Note: To fulfil the simulation process, some model inputs which are qualitative in nature were coded by the Multi_R_Designer team. These codes are summarized in the section: Coded variables. Based on the codes, assign the necessary modifications in your data file before beginning any study (see frame for the user's file). The definitions for the quantitative inputs are given in the quick guide manual.


EXAMPLE: Simulation of the frictional pressure drop () on a user's data file using all models available in the Multi_R_Designer simulator. A sample file (i.e. sample.csv) is used here to illustrate the procedures.

For the first part of the example, an automatic simulation procedure is presented when one (the frictional pressure drop) or all properties are systematically simulated. This action requires the creation of a user database (ex: bank1.data) which has the same frame as the Multi_R_Designer databank. User database offers many possibilities such as direct comparisons of user's data with those in the Multi_R_Designer databank, automatic or manual simulation with a minimum of preliminary manipulations.

For the second part, a manual simulation to predict frictional pressure drop is presented using two different file arrangements:

 
1) a user's file with a specific frame (ex: user1.csv). The user can create the frame from any .csv file (i.e. sample.csv) after he has selected the property(ies) to be simulated and their associated models.
2) a user database (i.e. bank1.data) which has the same frame as the Multi_R_Designer databank.


Finally, the user is introduced to the MEXA (Model EXpert Advisor), a tool in the simulator which provides a clear choice of the best model or correlation to predict a given property providing a range of operating conditions without any experimental data in hand.
 
Previous Page Up One Level Next Page Multi_R_Designer Tutorial Contents