Makers of MOSES - The Final Answer in Marine Simulation
Rev. 5.08 Bug Fixes
REV NO: REV 508.001
DATE: August 20, 1998  Discovered By: Roger Burger 
Description Of Problem

A &ptmod vessel sometimes does not produce pictures with connectors shown.

REV NO: REV 508.002
DATE: August 28, 1998  Discovered By: Larilla Templeton
Description Of Problem

If one asks for tank capacities for an exterior compartment, MOSES will "hang".

REV NO: REV 508.003
DATE: August 26, 1998  Discovered By: Larilla Templeton
Description Of Problem

The &surface menu will abort when combining blocks on a PC with an arithmetic exception.

REV NO: REV 508.004
DATE: August 28, 1998  Discovered By: Bahrain
Description Of Problem

The color (and ratio if annotated) for a joint is the value for the last brace and not the maximum.

REV NO: REV 508.005
DATE: September 01, 1998  Discovered By: Scott Holland
Description Of Problem

The new multi-server ethernet security does not work properly.

REV NO: REV 508.006
DATE: September 01, 1998  Discovered By: Paul Dixon
Description Of Problem

The maximum amount of memory allowable on a PC inadvertently was reduced from 512 gig to 96 megs.

REV NO: REV 508.007
DATE: September 04, 1998  Discovered By: Roger Burger
Description Of Problem

Fix 508.004 corrected the joint problem, but it broke the plotting of beam unity ratios.

REV NO: REV 508.008
DATE: September 05, 1998  Discovered By: Peter Lai
Description Of Problem

The report of Panel Exciting Forces has bad values for the yaw moment if one has more than 1 heading.

REV NO: REV 508.009
DATE: September 07, 1998  Discovered By: Franck Langevin 
Description Of Problem

If one uses -BLENG 101, where 101 is an element name, MOSES does not interpret it correctly. 

REV NO: REV 508.010
DATE: September 07, 1998  Discovered By: Curtis Owens
Description Of Problem

Longitudinal Strength is not quite right for a model with plate elements.

REV NO: REV 508.011
DATE: September 07, 1998  Discovered By: R. Nachlinger
Description Of Problem

In some strange cases, loops do not work properly. 

REV NO: REV 508.012
DATE: September 09, 1998  Discovered By: R. Nachlinger
Description Of Problem

If one has the following:     &if .false. &then       &set o = dog       %dog     &endif He will get an error that the variable dog has not been defined, but it is not used.

REV NO: REV 508.013
DATE: September 17, 1998  Discovered By: Gerogina Moldanado
Description Of Problem

If one issues an RARM command immediately after a CFORM, then the reported weight and KG will be incorrect.

REV NO: REV 508.014
DATE: September 17, 1998  Discovered By: Roger Burger 
Description Of Problem

On some linux machines, plots are sometimes interrupted and execution stopped with an X error.

REV NO: REV 508.015
DATE: September 19, 1998  Discovered By: Larilla Templeton
Description Of Problem

In some cases, the RENAME command in the &SURFACE Menu will not remove points that it should. This results in an error message Panel Has duplicate vertex.

REV NO: REV 508.016
DATE: September 20, 1998  Discovered By: R. Nachlinger 
Description Of Problem

The previous fix, REV 508.015 was not quite correct.

REV NO: REV 508.017
DATE: September 21, 1998  Discovered By: Georgina Moldonado 
Description Of Problem

If one has many beams of a class which enter Appendix b5b of the AISC code, then the results will not be valid.

REV NO: REV 508.018
DATE: September 22, 1998  Discovered By: Peter Lai
Description Of Problem

If one models a diffraction mesh as a piece with -perm 0 and another piece to account for the buoyancy with -diftype none, then a floating point exception may occur on a pc. On other machines, the viscous damping will be incorrect.

This fix solves the pc problem and will produce viscous damping based on the diffraction mesh on other machines.

REV NO: REV 508.019
DATE: September 23, 1998  Discovered By: Curtis Owens 
Description Of Problem

On big problems (over 8000 elements in a part), a PC will act strange and give incorrect answers.

REV NO: REV 508.020
DATE: September 25, 1998  Discovered By: R. Nachlinger
Description Of Problem

Fix Rev 508.018 causes problems with pieces that are not in the water.

REV NO: REV 508.021
DATE: September 26, 1998  Discovered By: Larilla Templeton
Description Of Problem

If one has a long set of data to describe an emitted compartment in the &SURFACE Menu, it will be truncated.

REV NO: REV 508.022
DATE: September 28, 1998  Discovered By: Curtis Owens 
Description Of Problem

If one omits a name for a piece on &DESCRIBE PIECE more than once, MOSES will add the data to the first piece with omitted name.

REV NO: REV 508.023
DATE: September 29, 1998  Discovered By: Curtis Owens 
Description Of Problem

Wind force applied to panels out of the water is not correctly applied to the structural model.

REV NO: REV 508.024
DATE: September 29, 1998  Discovered By: Curtis Owens
Description Of Problem

Disk space is wasted with cases -time if one has many headings and frequencies. In particular, this fix shrinks a pst data base file for 3.1 gig to 1 gig for 750 load cases and 90 time combinations.

REV NO: REV 508.025
DATE: October 05, 1998  Discovered By:Steven Tang 
Description Of Problem

Wrong minmax arguments for tnkl

REV NO: REV 508.026
DATE: October 05, 1998  Discovered By: Marcelo Canga 
Description Of Problem
Bending and Shear diagrams using the hydrostatic menu, using the moment-report command were incorrect. Some changes were done in hst_prop.f during rev5.08. To fix the problem, hst_prop.f from rev5.07 was copied back to this release.
REV NO: REV 508.027
DATE: October 16, 1998  Discovered By: Guillaume Le Marechal 
Description Of Problem
The rao command in the freq_resp menu sometimes produces a floating point overflow
REV NO: REV 508.028
DATE: October 21, 1998  Discovered By: Guillaume Le Marechal 
Description Of Problem
The &compartment command with the -simple option was computing the wrong free surface correction when it was used with specific gravities other than the default.
REV NO: REV 508.029
DATE: October 24, 1998 Discovered By: Guillaume Le Marechal
Description Of Problem

For very big problems (17000 nodes @ 6 dof per node), one can receive an error message "Too Many Global Stiffness Blocks".

REV NO: REV 508.030
DATE: October 24, 1998 Discovered By: Curtis Owens
Description Of Problem

For some cases the heading interpolation will not give good results. This normally occurs when one has an asymmetric model and uses a 0 degree heading.

REV NO: REV 508.031
DATE: October 26, 1998 Discovered By: Curtis Owens
Description Of Problem

For very large problems (band width greater than 3200 nodes), the profile minimizer fails to work properly. Thus, for these problems, MOSES takes quite a bit longer to factor the stiffness matrix than necessary.

REV NO: REV 508.032
DATE: October 30, 1998 Discovered By: Greg Hook
Description Of Problem

The documented command, "SPECTRUM", is not a valid command, but SPECTRAL is.

REV NO: REV 508.033
DATE: October 30, 1998 Discovered By: Georgina Maldonado
Description Of Problem

If one has a body with multiple parts and the last one does not have a mesh, then &PLTMODEL MESH will yield an error message that there is nothing to plot.

REV NO: REV 508.034
DATE: November 02, 1998 Discovered By: Curtis Owens
Description Of Problem

The stresses computed for shear Z is incorrect.

REV NO: REV 508.035
DATE: November 02, 1998 Discovered By: R. Nachlinger
Description Of Problem

The previous fix for large problems did not really fix the problem. The real problem was that for the problem with 14800 nodes and 18000 elements, product of the band width times the number of degrees of freedom is greater than the largest positive 32 bit integer. This fix also changes the files used in structural solving to minimize the problem of having a single file larger than 2 gigabytes (a limit on most operating systems).

Another fix may be forthcoming with regards to the profile minimizer. At the moment, the particular problem of interest generates stiffness matrix files of 2,050,000,000 bytes, so anything larges will fail the 2 gig limit.

REV NO: REV 508.036
DATE: November 05, 1998 Discovered By: Curtis Owens
Description Of Problem

As mentioned before, the profile minimizer did not do a good job on very large problems (14800 nodes, 25300 elements). This is a total rewrite of the minimization process. It not only makes the profile minimizer work properly, but reduces the time for profile minimization on this problem from 950 seconds to 20 seconds a pass.

REV NO: REV 508.037
DATE: November 08, 1998 Discovered By: R. Nachlinger
Description Of Problem

The last fix did not work on a Silicon Graphics Machine.

REV NO: REV 508.038
DATE: November 11, 1998 Discovered By: Curtis Owens
Description Of Problem

If one has a zero radius brace, one can get hung in an infinite loop when computing offsets.

REV NO: REV 508.039
DATE: November 12, 1998 Discovered By: Curtis Owens
Description Of Problem

This fix solves several minor problems with the STRUDL converter.

REV NO: REV 508.040
DATE: November 13, 1998 Discovered By: R. Nachlinger
Description Of Problem

The previous fix broke the SACS, DAMS, and STRUCAD conversion if one has reference nodes.

REV NO: REV 508.041
DATE: November 18, 1998 Discovered By: Tim Nolte
Description Of Problem

On some machines, it is possible to get hung in an infinite loop when computing local offsets.

REV NO: REV 508.042
DATE: November 18, 1998 Discovered By: Curtis Owens
Description Of Problem

It is not possible to have more than 4096 individual selectors in either the -selec or -except lists of a selection criteria.

REV NO: REV 508.043
DATE: November 20, 1998 Discovered By: C. Owens
Description Of Problem

The STRUDL converter is extremely slow on large problems.

REV NO: REV 508.044
DATE: November 24, 1998 Discovered By: Curtis Owens
Description Of Problem

Some really large problems (17500 nodes 25000 elements) may cause a segment violation on a linux machine when building the degree of freedom table.

REV NO: REV 508.045
DATE: December 01, 1998 Discovered By: Curtis Owens
Description Of Problem

Linux cannot have a file bigger than 2 gig. This is a recoding of the basic database file system to "get around" this limitation.

REV NO: REV 508.046
DATE: December 11, 1998 Discovered By: Yvan Leipold
Description Of Problem

Generalized degrees of freedom do not work for a system composed of more than 1 body.

REV NO: REV 508.047
DATE: December 19, 1998 Discovered By: Carl Webb
Description Of Problem

Sometimes a &connector @ -l_horizontal command will fail to find a solution. This occurs when one has a short, stiff top segment and a lot of much more flexible second segment.

REV NO: REV 508.048
DATE: January 04, 1999 Discovered By: Guillamue Le Marichal
Description Of Problem

Longitudinal strength has problems with vertical beams. It gives warnings about zero length loads and the results are not good.

REV NO: REV 508.049
DATE: January 23, 1999 Discovered By: R.Nachlinger
Description Of Problem

The building of spectral post-processing cases is extremely inefficient. The answers are correct, but the time for member and joint checks can be up to 8 times faster with this fix.

REV NO: REV 508.050
DATE: January 27, 1999 Discovered By: Abam
Description Of Problem

The .acct macro does not work in the cus file as advertised.

REV NO: REV 508.051
DATE: January 28, 1999 Discovered By: Guillaume Le Marechal
Description Of Problem

Under a seriously convoluted set of circumstances involving redefining environments and computing RAOs in the FREQ_RESP menu, the raos may be wrong due to improper interpolation.

REV NO: REV 508.052
DATE: January 29, 1999 Discovered By: Georgina Maldonado
Description Of Problem

For some strange joints, a Windows machine will abort with a negative number raised to a non integer power.

REV NO: REV 508.053
DATE: February 10, 1999 Discovered By:Roger Burger
Description Of Problem

The "projected area" reported with compartment summaries is twice what it should be.

REV NO: REV 508.054
DATE: February 12, 1999 Discovered By: Paul Dixon
Description Of Problem

If one has non-convex panels, then the mesh will not be properly refined.

REV NO: REV 508.055
DATE: February 23, 1999 Discovered By: John Windus
Description Of Problem

In the process post-processing menu, the last column obtained via DRAFT is the same as the next to last.