During 1999 and the early months of 2000, PHOENICS has been undergoing considerable development, by way of both the introduction of new features and the refinement of old ones. Some of the highlights are described below.
The PLANT feature, which was introduced several years ago, so as to enhance PHOENICS Shareware, and was later upgraded and attached to the commercial product, is an extremely powerful model-creation tool.
There already exist a large number of PLANT-using cases in the Input-File library, as can be seen clicking here.
However, PEL, the PLANT Expression Language, does require to be mastered by its users; and entering the expressions into the Q1 file by hand is somewhat laborious and prone to error.
CHAM, in conjunction with its Moscow branch, has therefore created a stand-alone executable which facilitates the introduction of PEL expressions into a Q1 by way of a Windows-type menu, the nature of which can be seen by clicking:
This utility can be activated by way of the PHOENICS Commander Run menu.
It is supplied with a wide range of expressions from which a selection can be made via the menu, and then edited; and "help" is provided for the following items:
additional velocity
specifications
bfc-grid generation
intervention in the
calculations
inititialization
interphase transfer relations
properties
sources
time step
variables
x-,
y- and z-direction
settings
In association with Dr Michel Ferry, CHAM has attached the MIGAL multi-grid solver as an optional extra to PHOENICS. This can greatly reduce the computer-time requirements for large flow-simulation calculations. (No longer supported.)
GREX3. and its attached GX-files, have always been intended as exemplary Fortran additions which so-inclined users could either modify or copy.
In the course of time, however, their complexity and inter-twinedness had become so great that only the boldest and most experienced could dare to use them as models.
Consequently, a radical reorganization of the coding has been undertaken, between mid-December 1999 and mid-February 2000, to make it again possible for users to read, understand and profit from it.
Particular features of this work have been:
The present organization and state of the relevant open-source coding can be seen by clicking here.
The physical significance of the coding has not changed significantly; but the improved order and clarity of the open-source sector of PHOENICS-EARTH has made it possible to resume the enrichment of its contents.
For many years, PHOENICS has been supplied with a large "library" of tried and tested Q1 files, which could be used as starting points for new simulations, for demonstrations, and for instructional puprposes,
However, because of file-handling limitations which prevailed at the time at which the library was instituted, all Q1s were collected into a few very large files; and individual cases were made selectable by an ingenious (but machine-dependent) coding sequence in the Satellite.
With the passage of time, the increase of size of the library files became a serious impediment to editing and upgrading, with the result that their content remained static, and increasingly out of date.
PHOENICS 3.3, however, has a new library system, which assigns a distinct file to each case, so facilitating access, editing, loading, etc.
The current contents of the library system can be seen by clicking here,
The way is now clear for bringing into the library system a suitable selection from the large number of interesting Q1s which have been created by CHAM and by PHOENICS users generally.
Offers of cases for inclusion will be gratefully accepted; and of course the files will carry the names of their originators.
The MUSES feature, which makes extensive use of PLANT, has been further developed during the course of 1999.
It is described in the Encyclopaedia article., and extensively exploited in the creation of the SAFIR blast-furnace special-purpose program.
MUSES is likely to be especially welcomed by PHOENICS user concerned with complex chemical-engineering equipment.
HOTBOX, the version of PHOENICS which is used for the simulation of electronics-cooling problems, has been extensively improved in respect of:
HOTBOX is available for trial by way of the Remote Computing Service.
What has been said about HOTBOX can be said also about FLAIR, the special-purpose version of PHOENICS which is used for simulating the flow of air, smoke and heat inside buildings.
At the present moment the number of attached library examples is fewer, namely 28; but it will grow. They may be seen by clicking here,
Partly as a consequence of its participation in the EC-supported OSIRIS project, together with ThyssenStahl, IEHK/RWTH, RAUTARUUKI, QUANTEL and IESL-FORTH, CHAM has developed the SAFIR blast-furnace special-purpose program.
CHAM is preparing for the new age of CFD-via-the-Internet by creating an experimental Remote Computing Service, by means of which newcomers can "test-drive" PHOENICS, and experienced users can submit their jobs for running on CHAM's computers (for the moment) free-of-charge.
Like many things connected with the Internet, the reality still falls somewhat short of its promise. However, CHAM believes that all the current "teething problems" will be solved in a few years, and that "remote-computing" may become the norm.
At the present time (November, 1999), all the special-purpose programs, including SAFIR and the Virtual Wind Tunnel, are being mounted on the Simuserve servers for remote access; and preparations are being made to activate parallel PHOENICS on CHAM's PC clusters.
Four new facilities have been provided, namely:
Click above for the relevant Encyclopaedia entries.
A "time-out" facility has been provided. Click here for the Encyclopaedia entry.
The attention which has been given to radiation is best seen by inspecting the updated radiative-heat-transfer entry in the PHOENICS Encyclopaedia, and especially the "Overview" and the electronics-cooling example.
Consideration is also being given to NDO, the New Discrete Ordinates method.
This method, which will be implemented when time and resources permit, is similar in principle to the conventional discrete-ordinates method, but solves the equations in a more-economical manner.
Users who might be interested in this development, and especially those who could supply information about problems which they have solved by the conventional DO method, would be welcome as collaborators.
CHAM's main focus has been on chemistry-turbulence interactions, into which the multi-fluid model of turbulence is providing quantitative insight.
Those who are familiar with the flamelet model of turbulent combustion may prefer to start here.
