Rev. 5.08 Bug Fixes

Rev. 5.08 Bug Fixes

Printable Version

Release Information

Downloads

Bug Fixes for Minor Version

Information for Current Release

Information for Next Release

Update Previous Release

Update for MOSES Beta

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 resuced 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 dupilcate 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 dmaping 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 tnk_pfl

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 assymetric 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 forthcomming 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 strenght 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 advertized.

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.

Ultramarine, Inc. | Suite 325 | 3100 S. Gessner | Houston, Texas 77063 | (713)975-8146

Site Map

Copyright(C) 1996-2008 Ultramarine,Inc.