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

Q: What does the message "*** WARNING: You Cannot Do Structural Loads With This Database" mean?
REV 7.06

A: The message means that you imported a "Total hydrodynamic database" (you exported it using E_TOTAL command). By doing this you saved the total forces and drift forces only. This data is sufficient to assess rigid body dynamics but it is not sufficient to consider the structural response of the hull.

Q: Are second order forces included in the motions and accelerations computed in MOSES?
REV 7.06

A: It depends:

  • They are always considered in the time domain.

  • They are never considered if they are computed with RAO.

  • They are included if they are computed with SRESPONSE.
For a discussion in SRESPONSE, see How MOSES Deals With Technical Issues, Section III.G.

Q: Is there a character string limit for variables, node names, and classes?
REV 7.03

A: MOSES does have a character limit for variables, node names, and classes. For variables it is 8. For node names the * counts as a character, so there is a 7 character limit. For class names the ~ is also considered a variable, so there is a 7 character limit, as well.

Q: Is there a problem with the &APPLY command? It appears that it is using the square of the fraction I tell it to apply.
REV 7.02

A: In fact, what your file contains is: 

    &apply -fraction -load_group hullwl %prop%
    &apply -fraction -force #jyforce %prop%
Actually this is not a glitch - it is what you told MOSES to do! The first of these says to multiply the forces in load group hullwl by prop, the second one says to multiply the forces due to the load set #jyforce by prop. Thus prop for the first one and prop for the second one is prop**2. Click here to read more about &apply.

Q: Can you explain how to use &FILE?
REV 7.01

A: The &FILE command is designed to be a general interface between the MOSES user and the computer's file system.

It is not designed to replace &DEVICE -USE_FILE which, in days gone past, was the way to change the file associated with a channel; i.e. the old command 

    &DEVICE -USE_FILE G_DEVICE COW.EPS
should be replaced with 
    &CHANNEL GRA_DEV  -FILE COW.EPS
Click Here for a test that uses &FILE to open, write to, and read from a file.

Q: Can MOSES calculate the natural period of a ship?
REV 7.00

A: The mass in the equations of motion for a ship in the frequency domain depends on frequency. Thus, there is no classical natural frequency. Almost everybody talks about a natural period and when they do, they really mean where the response peaks.

Q: What formula is used in MOSES for determining radius of gyration?
REV 7.00

A: The recepies MOSES uses are those in any undergraduate dynamics book. Local plate inertia is ignored, and beams are assumed to be slender; i.e. the size perpendicular to the centerline is small in comparison to the length. Other than this, the computation is exact.

Q: We are considering to use MOSES for lowering analysis. A typical case is for instance to install a cover or a template at the seabed. We typically want to analyze 4 different phases:

  1. Lifting in air before the cover meets the surface

  2. Lowering through the surface

  3. Cover fully submerged, typically 5m below surface

  4. Cover close to seabed, a few meters above the sea bed.

I will typically model the lifting vessel and the cover as two bodies, and then connect them with sling connectors. I think that phase 1 and 3 easily can be analyzed by using the Static process approach for lifting analyses and then combine these analyses with frequency domain analyses to incorporate the vessel motions and wave forces (phase3) acting on the cover.

However; I am not sure of what the best approach would be for phase 2. I am thinking about carry out many time domain analyses of this phase, where I use different waves for each run. Then report the maximum and minimum (to check for uplift) loads in the slings for each run. After carrying out maybe 30-40 runs, I could collect the data and establish confidence intervals for the mean loads in the slings and use these results to establish limiting environmental criteria for the lifting operation. A Monte Carlo approach so to speak. Do you have any experience with these kind of analyses in Moses? Is it possible to define a "lowering velocity" for the cover when doing time domain analyses of the lifting process? As the cover is partly submerged, will Moses take buoyancy variations as the wave passes by into account? In that case, I guess I must use very small time steps in the analyses...

With respect to phase 4, I wonder whether Moses takes into account the increase of added mass as the cover gets close to the seabed.
REV 6.01

A: I normally look at the problem a bit differently. Basically, you have one body connected to another; one floating, the other being supported. The two important factors are that the length of the support line (and hence its stiffness) is being changed and the second body changes it proximity to the water and the bottom.

We recently did a study of lowering a jacket and we found that the maximum load occurred before the jacket entered the water but the maximum DAF occurred when the jacket was completely submerged. Also if the water depths are large then a resonance will occur for some length of line.

That having been said the answers to your specific questions depend on how you model the cover.

  • Morrisons equation is treated "correctly"; i.e. the current position is used in computing the added mass and the excitation. The bottom effect is ignored.

  • Diffraction is "correct" at the position used to compute the pressures. As the body is lowered a true diffraction solution does not change. There use of &DEFAULT -WAVE_RUN, which will compute the Froude Kryloff pressure "correctly", but the change in added mass is still ignored.

  • A frequency domain approach is ok before the cover can contact the water and perhaps after it is totally submerged. I would, however, use the time domain for locations from the cover being at +- the maximum wave amplitude to the water plane.

  • You can define a line change speed and simulate the lowering as a true time domain process.

  • I would probably look at three locations: cover in the air, cover just touching the still water, and cover just submerged. Now, run time domains at these positions to find which one is critical and use it to establish an environmental limit.

  • Actually, the above may not be what I would do, depending on the water depth. For deep water, you need to consider the mass and weight of the line and need to look at the response for all depths since you can have a resonance for some intermediate depth. Here, I would probably use frequency domain to find the critical depth, do a time domain at that depth, and then use the critical position to find the environmental limit.

  • The simple connectors do not include the effect of their mass. You must either add 1/3 of this mass to the cover (a standard vibration approximation) or use a rod element.

Q: I issued the command &DEFAULT -DEPTH to change the water depth after INMODEL and MOSES does not take the command. It seems that I can change the water depth only by issuing the &DEFAULT command before the INMODEL. Am I missing something here?
REV 5.05

A: Yes, you are missing something. Water depth is a property of the environment and is set on the &ENV command. The default water depth &ENV uses is set by &DEFAULT -DEPTH DEPTH, Thus to actually change the water depth, you need to issue &ENV. The reason it worked if you put the &DEFAULT before the INMODEL. is that after the INMODEL, a new environment is created which uses the defaults currently in force.

Q: I have a system composed of several "things". Should I model them as a single body with several parts, or as separate bodies?
REV 5.05

A: Unfortunately, the answer here depends on what you are interested in investigating. First, Bodies are considered to be rigid (except when using generalized coordinates). Thus, if there is to be relative motion between two "things", then they must be separate bodies. Even if there is no relative motion, there are times when you may want to use separate bodies. Suppose that you are interested in the forces which hold two things together. If these two things are parts of a single body, there is no way to "look" at the connecting forces except by doing a structural analysis. If, however, these things are modeled as two bodies connected by some system of connectors, then the connection forces are readily available. Remember, however, that these connection forces are computed assuming the bodies to be rigid, so a stress analysis will yield somewhat different results, depending on the flexibility of the bodies.

Q: How can I arrange multiple bodies in space?
REV 5.05

A: This is perhaps one of the more difficult tasks which one has to perform in MOSES. The simple answer is that you use the &INSTATE command. (For details, click here.) You need to specify the location and orientation of each body. There are several different ways of doing this, but perhaps the easiest to see is: 

&INSTATE -LOC BOD1 X1, Y1, Z1, RX1, RY1, RZ1 \
-LOC BOD2 X2, Y2, Z2, RX2, RY2, RZ2
Here, we are specifying the global location of the body origin of each body with the coordinates X1, Y1, ... Z2. To specify the orientation, we specify the Euler angles of each body, RX1, ... RZ2. Each of these are successive rotations. These are defined by first assuming the the global and body systems are aligned. One first rotates the body about the body Z axis an amount RZ, and then rotates it about the body Y axis an amount RY, and finally about the body X axis an amount RX. At the conclusion, the body is properly positioned in space.

Q: Why doesn't the conversion between metric tons and kilo-newtons change when using SI units and changing the value for the acceleration of gravity in the program?
REV 5.03

A: This confusion between mass and force units has more to do with the way SI units are typically used than with our software. In Europe, it seems common to use mass units to define a weight. As a convenience, our software allows this by accepting metric tons as a force measure. This is why changing the value of "g" has no effect on the difference between metric tons and kilo-newtons.