Converting Models Questions

Marine Engineering Specialists -- Bentley Systems has acquired Ultramarine's MOSES Software [ Press Release ]

Q: I converted a SACS model and when I looked at the result I found a class ~HO|1. Where does this come from?
REV 7.05

A: MOSES converts the ring stiffeners in the SACS model. To do this, there must be a class for the ring. The class ~HO|1 is the class for the ring defined in SACS grup HO.

Q: Why do I get the message:

      *** WARNING: Load Attribute Has No match *J1739   *J1759 ?

REV 7.00

A: The problem is that you do not have unique node names and class names for each cargo model. You should use the -JPREFIX and -CPREFIX options for &convert to add a prefix to the node and class names so that there is no confusion. Click here to see what the Reference Manual has to say.

An even easier way is to use the sample file we provide in:

   \ultra\hdesk\tools\cnv\cnv_ck.cif

Copy this file to your working directory, and alter as needed, and run moses with cnv_ck as the root. The will convert the models and give you a summary of their weights. The modified version will, perhaps, look something like:

     $
     $****************************************       variables
     $
     &set mod_type   = sacs    sacs
     &set sac_files    = ban.sac zel.sac
     &set dat_files    = ban.dat zel.dat
     &set prefix         = w       z
     $
     $****************************************       now convert and sum
     $
     &insert &info(pgmpat)/hdesk/tools/cnv/doit.cif

What the above says is that will convert two models. Both models are SACS models and the SACS model files are ban.sac and zel.sac. The converted MOSES models will be placed in the files ban.dat and zel.dat. Finally, the prefix w will be added to nodes and classes for the ban model and the prefix z will be added to the zel model. If you have another model, simply add another column to each of the 4 variables.

Notice above I used "w" for the prefix for the ban model because "b" is usually reserved for the barge prefix. In fact, all of the barges in our library use b as a prefix.

Q: Why did the MOSES SACS converter did not properly convert my SACS loads?
REV 7.00

A: You did not use &convert -loads load. Most of the time, loads in SACS are treated as weights for conversion purposes, and treating these as actual applied loads has little value. The only reason we have the -loads option is for comparison purposes. If you use &convert -loads load in MOSES Rev 6.00 and older, be sure to add "&set fact = 1" to your input before the jntlod macro is used.

Q: I have a structural model of a topside converted from SACS to MOSES format with ZCG which is too low. Is there a simple way to correct this?
REV 6.01

A: This is a common problem. Since the equipment weights for the topsides are applied at the deck levels, the vertical CG is always off. Since you want to perform a structural analysis, simply overriding the VCG using &weight is not accurate, the loads are not applied where they belong. Historically, some engineers have solved this by correcting the SACS model. They add fake structural members that join at a common point above the various deck levels, then load this common point as desired. This is rather cumbersome.

The other option is to correct the model in MOSES, using load groups. You have to remove some of the original equipment loads, and replace these with load groups that place the load at the proper location, yet distribute the loads to appropriate structural nodes. A macro is provided below to make this easy, but use at your own risk. You should also review the section the MOSES Reference Manual on load groups. You need to be using Rev 6.01 and above for this macro to work.

 
        ========== Macro linp.mac ==================
 
        ¯o linp prefix enumb wt x y z nodes
           &local nlist  n
           &local prefix = *%(prefix)0000
           &loop n ( %nodes )
              &set nlist = %nlist &string(overlay  %prefix %n)
           &endloop
           &describe load %enumb %nlist
              #weight &number(abs %wt) -cen %x %y %z -cat equipment
        &endmacro
 
        ========== Sample file using linp.mac =========
 
        $       name      wgt    x       y    z   ----- nodes -------
        $
        linp q q_sub    14.07 -31.66  -4.68 23.67 3706 3707 3708 3709
        linp q q_celler  9.77 -14.26  -6.90 42.93 1209 1207 1228 1230
        linp q q_z-7475 34.50 -36.50  37.00 32.00 1231 1230 1233 1232
        linp q q_z-7440 10.45 -38.44  -3.46 32.11 1257 1256 1254 1255
        linp q q_z-7441 10.45 -38.44 -16.46 32.11 1212 1209 1256 1257
        linp q q_z-7460  7.26  49.00  33.83 32.19 1228 1260 1237 1235
        linp q q_z-7486  7.70  32.08  40.75 33.33 1228 1260 1237 1235
        linp q q_J-1095  7.37  16.58  53.71 34.63 1283 1285 1240 1242
        linp q q_z-7465 16.17  47.25  46.25 34.77 1228 1260 1237 1235
        linp q q_z-7495 42.77 -11.50  -9.00 35.00 1209 1208 1220 1253 1254 1256
        linp q q_z-7496 42.77  11.50  -9.00 35.00 1208 1207 1263 1265 1220
        linp q q_z-2905 71.61   4.75  40.50 35.15 1230 1229 1228 1235 1234 1233
        linp q q_z-7435 82.39 -22.54 -38.25 38.20 1200 1201 1202 1208 1209 1212
        linp q q_J-1020 29.83 -13.00  15.75 40.50 1253 1220 1229 1230
        linp q q_J-1021 29.83  13.00  15.75 40.50 1220 1265 1228 1229
        linp q q_z-7445 35.20 -41.11  15.00 42.00 1255 1517 1254 1230 1231

Include this information in your MOSES deck model.

Q: I get strange error messages when trying to convert a SACS model. The files I used worked before, only the SACS model is different. Why is this?
REV 5.10

A: Your SACS model has more commands after the first END command in the file. While these may be SACS plotting commands, MOSES is trying to interpret these as real MOSES commands. Place an &fini command after the first END command, and the remainder of the file will not be read.

Q: I converted a SACS model that included AREA and WEIGHT cards. When I read the resulting model into MOSES, I get errors, why is this?
REV 5.10

A: We convert the AREA card from page 2-4-9 of Volume I of the SACS manual. You are using the AREA card from page 13 of the LAUNCH/FLOATATION manual, which is rarely used. We cannot convert both formats. The WEIGHT command also comes from LAUNCH/FLOATATION, and is not converted.

Q: Why when I convert a SACS file where the units are metric do I get incorrect Young's modulus?
REV 5.09

A: The MOSES to SACS converter does not support the SACS ME option for units. This is because it has never made sense to us to use kilograms as a force unit, and very few users ever utilize this option. Either change your SACS model input to reflect the MN option, or edit your resulting MOSES model to change the stress units.

Q: I am converting a SACS model file into MOSES to perform a launch analysis. What will happen to the pre-defined Ky, Kz, Ly & Lz values that have been defined in the SACS data file ? Do I have to redefine them or will MOSES automatically redefine them accordingly?
REV 5.06

A: The buckling length factors are converted properly from SACS, no further action is required on your part. In MOSES, there is a direction cosine matrix created for each element that defines the relationship between the element coordinate system and the model input coordinate system. This allows the jacket model to be rotated in any manner, without requiring a subsequent change to the buckling length factors. Refer to "MOSES Sign Conventions" under the Help Desk. (For details, click here.)