|
|
Digital Analytics News 1/25/2008 BOAT3D has received a user-interface upgrade to Version 4.1. A dedicated editor for the boat shape table now makes it simpler to work with this input dataset. On the output side, the graphic of the boat and the water surface now can appear on the screen superimposed with the numeric output values and their plots vs time. Clicking on "Go" produces a "video" of the boat in motion through the waves. 12/30/2007 A new Version 3.5 of EDTech is announced. A little "navigation bar" has been added to the program. Since EDTech has 3 primary working windows, each with a different function and different types of data files (plus several secondary windows with their own functions) we have made it possible to quickly switch windows with function keys. However, a new user may have had problems with this non-standard methodology, so for Version 3.5 we provide a "navigation bar" with buttons to select with single clicks the exact items which the user needs. Also a user may close or minimize the whole program from the navigation bar without having to shuffle working windows to get to a button. Some changes have also been made to EDTech's automatic scaling/axis numbering logic to improve appearance. 12/5/2007 "EUROKIN" is a consortium of over 10 European companies and 4 universities which was established in 1998 and "whose aim is the implementation of best practise in the area of chemical reaction kinetics" (see www.eurokin.org). In 2001 this organization published on their web site a paper describing 4 test cases for kinetic parameter estimation (www.eurokin.tudeflt.nl/publications/Paper_par-est-1.pdf). The expected orders of magnitude of the parameters were given, but the test cases were given without final results, i.e. the reader was invited to perform the parameter estimations himself as a software test. Version 5.2 DMSolver with nonlinear least-squares add-ins was used to run the 4 test cases as given, except for a few deviations from the EUROKIN directions. A report about this work is now on the Digital Analytics web site. The lead consortium author, Robert Berger, stated "I had a quick look at your report. In general the values obtained are close to the ones I obtained with several commercial packages." (Personal communication.) The goal of this work was to test Version 5.2 of the DMSolver least-squares add-ins, and I believe it also demonstrates that DMSolver is a very good software package for work in the chemical kinetics and reactor design field of technology. The user can use a very detailed model if desired without running into software limitations. Source code for the 4 test cases is given in our report. Version 2.2.0 of Free Pascal is now standard for DMSolver and the optional user-written procedure in BOAT3D, and a convenient guide program for working with the user-written procedure has now been developed. 11/11/2007 New Version 5.2 of DMSolver is announced. There are only slight changes to the Solver itself, but the add-in procedures previously available with versions 4.x have been updated and adapted to 5.+ DMSolver versions. 8/26/2007 The technical paper documenting Version 4.0 of BOAT3D has now been uploaded, and V. 4.0 now fully replaces V. 3.1 BOAT3D. 6/20/2007 New version 4.0 of BOAT3D, replacing 3.1, has now been tested with the Fridsma study data and is fully operational, with only documentation, testing with other studies, and minor cleanup of details unfinished. New Features of 4.0: There are very few user interface changes but many internal changes, justifying a designation of 4.0. The cross-sectional bottom frame lines between keel and the chines no longer have to form a straight-line vee bottom shape, i.e., there can be convexity or concavity in them. There are two new columns in the input data, one for keel and one for chine. If the user puts a 1 in both columns, a straight-line vee cross-section is obtained as with BOAT3D 3.1. If the user puts e.g. a value 1.05 in the new keel column he is requiring that the deadrise slope at the keel is to be 1.05 times the straight-line value, introducing concavity near the keel. If the user puts e.g. a value of 0.95 in the new chine column he is requiring the deadrise slope at the chine to be 5% less than the straight-line value, thus introducing concavity near the chine. Convexity is introduced similarly. The cross-sectional bottom lines are constrained to be cubic polynomials, and appropriate cubic splines are used to develop the bottom shape, which of course appears in the boat lines plot. The computations of added mass, etc., have been reorganized so that the solution computations take about 30% less cpu time. The assumptions about added mass vs. geometry, etc., are now concentrated in less program code, so it is easier for me to change them. These assumptions are slightly different than with version 3.1, and better agreement with Fridsma's experiment is obtained. Due to the ease in modification, I developed an experimental version with different hull geometry specification than 4.0 with relative ease. Small changes in the solution algorithm were made so it does not hang up when encountering minor discontinuities. 2/21/2007 Additional minor changes in EDTECH --- V 3.44. Serious bugs were fixed in the procedures to generate dashed and intermittent lines. Two new batch commands were added to copy data series within DSF databases. These were needed to support ongoing work with DMSolver nonlinear regression add-ins. The set of 4 nonlinear regression test problems presented by the EUROKIN chemical kinetics consortium (www.eurokin.org) is being used to test the DMSolver nonlinear regression routines and details are being changed in DMSolver and its add-ins. In 2007 this will become version 5.2. Normal Boat3D use does not involve programming, but the optional facility for user-written code for an added-in external force and/or moment has been reintroduced to Boat3D. This is the method original author Peter Payne used to model the use of hydrofoils, and his example foil from V2.3 has been again introduced. 10/12/2006 Minor changes in EDTECH --- Version 3.43 supercedes 3.42 to fix minor bugs or annoyances. There was a bug which introduced bad behavior and faulty text sometimes if you switched between Windows fonts and internal drawn-line fonts and then back again. Since we now use mainly always Windows fonts we did not discover this for a long time. Also, if you wanted to write data to an attached DSF with another program such as BOAT3D or TTRView, you had to completely close and exit from EDTECH before this could be done. V 3.43 now has a Detach DSF command which lets you free the DSF without other changes (then you can Attach it again after the other program has finished). Closing the Worksheet window now also detaches the DSF along with deleting Worksheet data. Also, in previous versions you could not plot a curve if there were missing values in any of the data series to be plotted. This did not happen very often, but it could be very laborious to shift data to remove missing values. V 3.43 now simply ignores any data point with either a missing X or a missing Y. DMSolver Version 5.1 has now been further tested with Flowsheet Toolkit and also with some interesting biochemistry kinetics problems. Free Pascal 2.0.4 is now used. Further work is being done on its addins and their documentation. 5/10/2006 Progress to date --- Improvements in Flash Because of other responsibilities, not as much work on Digital Analytics software has been accomplished as was planned a year ago. The Free Pascal project has come up with a new version, 2.0.2, and version 5.1 of DMSolver has been adapted to use this. While DMSolver 5.0 would solve relatively easy example problems, certain features and operational details have been added to 5.1 so that it can operate more like 4.x and handle some of the big, hard-to-converge Flowsheet Toolkit examples which caused operational problems for 5.0. 2 out of the 3 Solver addins have been successfully adapted to DMSolver 5.1 so far. A new version 2.01 of Flash Calculator is released today. Internal details were changed to give improved performance on hard-to-solve systems. The new version 5 of Flowsheet Toolkit (FT) has significant new changes besides being set up to run with DMSolver 5.x rather than the old PMSolver 4.x. The flash and the boiling/condensing heat exchanger modules have been changed to use Flash Calculator methodology. This is a more conventional methodology than used in the old FT. In the old FT flash equations including y=Kx and equations involved in calculating K were all part of the DMSolver set of equations and only the cubic equation of state was solved internal to the module. In the new FT flash equations are all internal making it more versatile, so if all-vapor or all-liquid conditions arise a solution is obtained rather than an error message. Not all minor utilities included in versions 4.x have been adapted to version 5, but some new features have been added. For example, after getting a flowsheet solution you can restart FLOWGEN, the program which you use to specify the major flowsheet modules, and use it to display the results BY MODULE. This is an extra-easy way to look at a solution compared with the typical display which does not group together all variables related to one module. Despite all these advances which get FT version 5 running, other problems and responsibilities have prevented me from doing as much testing and documentation as would be required for a truly complete system. 4/21/2005 New Chemical Engineering & Solver Versions Version 5.0 DMSolver has been further tested and is now the official version. The advantage of the new version over 4.3 is that 5.0 comes complete, and you do not need to buy and learn the large and expensive Borland Delphi software package needed previously. A new version of PMSolver is not announced; all its functions are present in DMSolver. Some seldom-used features of version 4.3 have not yet been converted over to 5.0 and will be added back in the future. In a new Version 2.0 of Flash Calculator new commands for computing bubble and dew pressures at specified temperatures have been added, and many cases which the old Flash Calculator could not handle are now possible. Two of the vapor-liquid equilibrium plot generators are also improved. Also, there is a consensus in the literature that mixtures containing hydrogen, the most volatile component of all that is commonly used in industry, cannot be modelled adequately with a conventional Peng-Robinson technique. I have added changes to the Peng-Robinson routines which improve hydrogen calculations but leave Peng-Robinson unchanged if hydrogen is not present. These hydrogen changes also apply to Distillation Guide and the Flowsheet Toolkit.
A Version
5.0 of Flowsheet Toolkit (FT)
is being developed to work with Version 5.0 DMSolver. Other changes are: The new FT is still being tested and refined. Please e-mail or call if you are interested in its status. SOME PREVIOUS NEWS EDITIONS: 2/14/2004 BOAT3D Version 3.1 See the new discussion and report about BOAT3D Version 3.1 on the BOAT3D page. 10/3/2003 BOAT3D Version 3.1 Announcement BOAT3D simulation software lets you try out and compare your planing boat designs on a digital sea before committing them to fiberglass, aluminum, or steel. And you do not have to build models or have access to expensive towing tank facilities. Version 3.0 was very much an interim version, and it has been upgraded to Version 3.1, which has many important features enabling versatile boat simulation studies. The 2-degrees-of-freedom and 3-degrees of freedom programs have been integrated together. An additional convenient mode of getting "snapshot" graphic output has been added. Peter Payne's propeller/engine simulation has been added back into the program. It had been removed to simplify version 3.0. Included in this site is an introduction to and description of BOAT3D. Simulating planing boat behavior on a PC is a very ambitious goal. To make sure BOAT3D is on the right track, I have spent time comparing BOAT3D results with data from experiments. Although a thorough literature review lists many publications on planing boats, there are only a few which present sufficiently detailed data. In this site I present a paper summarizing technical modifications which were made in BOAT3D and showing how its results compare closely with actual boat (including towing tank model) behavior.
|