Rev 4.27 is a minor release, with several new features. Primarily, the numerical algorithms have been improved to yield better, more realistic results. First, the scheme for checking the number of users has been changed. In the past, if jobs aborted, one had to "reset" to purge the accounting database of the aborted job. This is no longer necessary! In fact, the procedure reset and the saved help file are no longer supplied.
Several new figures have been added to the reference manual. Also, the figures now are automatically produced when the manual is generated. In most cases, they will be emitted in the file root.gdv in the format specified for graphics output.
Another new feature is the concept of a border color for graphics. The need for this concept arose from the use of color printers. In the past, black and white were simply reversed from the screen to the color printer. This simply did not produce good looking color plots of structural unity ratios. With the new concept, results are aesthetically similar between the two devices.
All of the options of &ENV which have not been documented for several years have been deleted. This may cause some old jobs to no longer run. Also, the manner in which diffraction is computed has been changed. In the past, all elements within a "repeated group" interacted. Now, all elements within a given body interact. As a result, the REPEAT command which allowed for repeating of hydrodynamic data is no longer allowed.
A technical problem with "water springs" has been fixed. In the past, new water springs were generated each time the STRUCT menu was entered. This resulted in warnings that elements were being redefined. In some cases, errors would result. Now, the springs are simply redefined so that no messages appear.
The manner in which forces are computed for a #TANKER load group has been changed. In the past, only the current velocity was used to compute the force. Now, the relative vessel-current velocity is used. In addition, one now has several new ways to control the computation of the forces.
The primary changes have to do with time domain simulations. Several new options have been added to the TDOM command: -FACT_CONV, -NO_CAPSIZE, and -NEWMARK. The first of these allows one to specify a fraction of the force due to the convolution which will be applied, and it replaces the old option -NO_CONV. If a factor of zero is specified, then no convolution force will be applied. The -NO_CAPSIZE option turns off the check for capsizing and is useful for simulations where one knows the a body will rotate through large angles. Finally, the -NEWMARK option allows one to use a Newmark integration method to solve the equations of motion instead of the predictor/corrector method. In preliminary testing, the Newmark scheme has appeared to be quite a bit superior to the old method in that it retained stability for larger time steps and induced much less numerical damping.
In the past, several approximations were made when the diffraction hydrodynamic forces where applied to a body in the time domain. In particular, it was assumed that the body did not move. For most cases, this was a reasonable assumption, but for single point moorings it was not. These forces are now computed accounting for the current vessel heading and location. Finally, a new term was added to the equations of motion to account for the apparent current acceleration in the vessel frame of reference. This is a quite small term, but it is necessary if one is to predict the swinging nature of a tanker on a hawser due to wind and current correctly.
Finally, we need to mention a change that occurred with Rev 13.522. This version fixed a problem with the phase angle of the heave RAO. With the test cases we ran, this change did not affect the results significantly. However, for some cases where the transverse center of gravity of the cargo in question is a substantial distance from the centerline of the vessel, the user may notice a difference in the results.