Hydrostatics Questions

Q: Why do I get different Longitudinal Strength results with MOSES than I get with a different package?
REV 7.06

A: It is probably in the way you defined the weights. Consideer

     #WEIGHT *P 1000 -LDIST 0 10
this will have a weight with CG located at point *P. Now, for longitudinal strength purposes is weight will be distributed from 0 to 10 in the part system. If point *P has an x location 5 in the part system, then you get a uniform distribution. Otherwise you get a trapezoidal distribution.

Q: Why does KG_ALLOW give the same allowable KG for all angles?
REV 7.06

A: There are several things that could case this:

  • If you do not specify any criteria, then the allowable will be the BM for all cases.
  • If you have no wind (or no wind area) then you may get the same results for all angles.
  • If you have weights specified then what KG_ALLOW gives is the KG of the additional weight and this can be quite confusing.

Q: We conducted the longitudinal strength assessment of a ship with three different softwares, and the maximum bending moments, calculated by two are quite similar. But comparing these maximum bending moments to the ones computed in the MOSES show large deviations. As you said on your homepage a MOSES uses a different integration method than that described in the Principles of Naval Architecture. What kind of integration method is used?
REV 7.06

A: MOSES adds trapezoidal loads on a beam using the exact solution for the bending and shear in a pin ended beam. In other words, go to the old edition of Roark (fourth edition). Add the solution of table III.14 and table III.16. This will give you the moment, shear, and deflection at all points on a simply supported beam with a trapezoidal load applied. Now, for each load (weights and buoyancies) you simply use this formula to get the total shear, bending, and deflection.

Notice that since we begin with a pin ended beam we always get zero moment at the ends but the shear will be zero only if the body is in equilibrium under weight and buoyancy.

This algorithm is iron clad; i.e. it is always correct. If you have differences with other programs then the difference is due to the way either the buoyancy or weight is modeled (or of course, the other program could be wrong).

Q: Why does MOSES report different values for GMT and GML than other programs?
REV 7.03

A: MOSES reports most values in the body system and most other programs report them in the global system. GML and GMT can be reported with the &STATUS command.

Q: What rule or equation is being used to calculate wind force when I use the RARM command, or inside the hydrostatics menu in general?
REV 7.03

A: In general inside the hydrostatics menu the design wind rules are the same as those designated by the &env command. You can ask MOSES to report the current designation with &STATUS ENV command.

Q: I have a semi-submersible platform with mooring lines and risers. Can I simulate flooding of one of the columns in static process?
REV 7.02

A: The STATIC PROCESS and HYDROSTATIC menus are magic in the sense that they were designed to do very specific traditional things. Thus they do not have the generality of most of MOSES. In particular, these menus consider only a single body and connectors (other than a sling in Static Process) are ignored.

For your problem you will need to construct your own process:

  • Do your flooding,
  • Find equilirium, and then issue the command
  • &event_store xx
Click here to read more about &event_store. Here, xx is a number and you should assign the numbers in order as the processes evolves; i.e. start with 1 then use 2, etc.

Repeat this until you have stored events for for all situations of interest. You can then use the PRCPOST MENU to report the results.

Q: Why do I get WARNING: Equilibrium Not Found Within Tolerance when I computer righting arms with RARM?
REV 7.02

A: MOSES rolls about the designated axis and finds equilibrium at each increment. Here equilibrium is weight and buoyancy forces. In order to find equilibrium MOSES is allowed to change heave and pitch.

The residuals reported along witht he warning message are the residual arms.

This warning occurs when MOSES could not find a z and pitch combination to meet the equilibrium tolerance.

In general if you pick a roll axis of 90 degrees where the roll is done about the vessel y axis then this occurs because you are actually pitching.

Q: If I use the RARM -U_CURRENT command, will MOSES report the location of the center of pressure?
REV 7.02

A: First, MOSES does not compute the "center of pressure" as it really does not exist. Instead, MOSES computes a current force which equilibrates the wind force and then the net moment of the wind and the current force will be the heeling moment.

This is a short description of what the RARM -U_CURRENT does.

MOSES does not report all of this, but you can recapture it if you want.

  • First use &INSTATE -COND d, r, p where d, r, and p are the draft, roll, and pitch reported for a given row of output from the RARM command.
  • Next, use &ENV -wind xx 90 -current 1 90 where xx is the wind you are using with the RARM command.

  • Next, issue &STATUS FORCE.

  • Finally issue the two commands &ENV -wind xx 90 -current W/C 90 and &STATUS FORCE. Here W is the beam wind force from the pervious force status and C is the beam current force from the previous status. Now, the wind moment minus for current moment is the healing moment. Feel free to make any interpretation you wish.

Q: Why are both the displacement and waterplane areas from my MOSES model 2% below the values provided by the designer?
REV 7.01

A: This is normally a result of the surface not being closed. You wll need to review the panel definition. Whenever you are having troubles like this you need to issue the command:

Now, when you look in the output file you will find a "EXTERNAL PIECE SUMMARY" and three of the columns are "Sum Area * Normal". All three of these numbers should be zero (or very near zero) because the surface is supposed to be closed.

Q: Where is the "Center of Rotation" with the RARM command?
REV 7.00

A: There is not "center of rotation"! The ship is rotated and equilibrium is found. The vertical position is a result of the equilibrium. In the results, a draft, roll, and trim are specified. You can use these with &INSTATE -COND to set the vessel into position so you can show a picture of it.

Q: Why do I get rubbish for Longitudinal Strength results while the jacket is tipping during a launch?
REV 6.01

A: Longitudinal Strength is applicable only to a single body. To find what you watn here, do not include the jacket model for your check of longitudinal strength at tip, instead, model your tipping loads at the rocker pins using #weight. Be sure to give some nominal distance for the load distribution, using -LDIST. Hopefully, your rocker pins are located forward of the stern, so the rocker load is applied within the confines of the barge. Then, set up your equilibrium and ask for bending moments as before. Of course, the better way to do this is with the structural solver, include barge flexibility, and request the shear and bending in the barge beam for the tipping load case.

Q: What MOSES commands do I use to obtain tank capacities which shows ullage and free surface moment data? Also, how do I change the position of the sounding tube to any position of tank?
REV 6.01

A: Look at the test TANK_CAP. The .dat file shows how to define the sounding tube and the .cif file shows how to get the report you want.

Q: Is there a way to include the effects of the mooring lines in the righting arm calculation?
REV 5.10

A: Yes, but since connector forces are not included in any of the hstatic menu commands you need to manually apply a moment to the body. Use an #LSET command, and find equilibrium with &equi, not equi_h. Using &loop command to repeat this for different moment values you will end up with roll angle versus moment information that can be converted into roll angle versus righting arm. Look at the test C_RARM for a sample,

Q: Why did MOSES change from using the "LOST BUOYANCY" method when computing damaged stability to the "ADDED WEIGHT" method?
REV 5.08

A: We changed from "LOST BUOYANCY" because it was not correct! Yes, the force on the vessel was correct, but if one uses LOST BUOYANCY, the inertia of the vessel is not correct. MOSES solves dynamics problems and in many cases the effect of inertia is more important than displacement.

Now, we would not simply change one inaccuracy for another. What MOSES does is not the "ADDED WEIGHT" method as you described it. MOSES simply computes where a flooded compartment cuts the water plane and "fills" the portion below with water. An approximate velocity potential is introduced to find an approximate dynamic pressure and this pressure is integrated over the wetted surface. Thus, MOSES produces the exact hydrostatic force and an approximate dynamic force. (This is also the same procedure used for compartments filled with the -CORRECT option.)

When the force computations are viewed as MOSES performs them, there is no "free surface effect". This is simply an artifact from viewing the problem as either lost buoyancy or added weight. Rest assured, the current MOSES version is correct!

Q: I have a simple test case where I compare the results of longitudinal strength to structural results for a box shape. The structural model is composed of a centerline beam using a type of WBOX. The results do not compare very nicely, and the sign convention for shear and moment seem to be different. Why is this?
REV 5.07

A: First, the sign convention between longitudinal strength and structural elements are different. This is necessary due to the programming philosophy for longitudinal strength. The structural element sign convention can found at this Help Desk, under MOSES Sign Conventions. (For details, click here.)

Second, you are comparing an exact solution of hydrostatic bending moments to the results of a finite element structural solution. This is not to say the hydrostatic solution is any better, but simply different. The structural solution has truncation errors that vanish as the model is refined. The structural solution concentrates the hydrostatic pressure at the nodes, since MOSES will analyze any type of structure. When comparing longitudinal strength with structural results, it is not sufficient to match the plane locations with node locations. This certainly must be done, but you also need to minimize the truncation error by increasing the number of nodes. The error goes to zero as the square of the node spacing, but 50 nodes (and therefore planes) will usually provide reasonable results. Finally, you need to remember that load groups are treated differently between the hydrostatic and structural solutions. The use of the -LDIST option on #WEIGHT allows for a uniform distribution of weight over the specified distance, for the hydrostatic solution. The structural solution uses the same weight, but distributed to nodes as specified in the definition of the load group. For testing purposes you may want to use the #ELAT command.

Q: How can I retrieve the weight and buoyancy distribution along ship length from MOSES?
REV 5.07

A: There is no way in MOSES to summarize the weight and buoyancy distribution for longitudinal strength calculations. The formulation in MOSES allows for the shear information to be stored, but not the individual weight and buoyancy information. This is due to the type of integration performed to arrive at the shear curve. The integration technique ensures the shear and moment curve will close at the ends if the vessel is in equilibrium, and allows for differences in location where loads are applied. Even if the vessel is not in equilibrium, the moment curve will close. The traditional integration method specified in "Principles of Naval Architecture" does not provide this reliability, but does allow one to summarize the weight buoyancy distribution. Since we provide the shear information, which is the interesting part, we chose what we consider to be the more reliable integration method.

Q: Why does the displacement in the Curves of Form do not change when I flood a compartment ?
REV 5.06

A: You do not lose displacement because MOSES adds contents when a compartment is flooded instead of losing displacement. In the past MOSES did lose displacement, but this leads to some undesirable consequences. In particular, the flooding certainly increases the inertia of the system. Thus, we add contents.

The "added weight method" has been looked on as not being as good as lost buoyancy, because weight was not correctly added. In particular, we add the correct weight which corresponds to the free surface of the compartment being at the waterline. Thus, for equilibrium, the results are identical, but we also get the increase in inertia.

Q: How can I do righting arms from negative to positive roll angles?
REV 5.06

A: MOSES performs the righting arm calculations "about the current state"; i.e. it rotates the vessel about an axis "adding to the original state". Thus to do righting arms from say -20 to 20, you use:

&INSTATE -COND 10 -20 0
RARM 2.5 17
Your output will state at the top that the initial roll is -20 and the rolls that are reported should be added to this to get the true angle.

Q: I was looking at the hydrostatic coefficients generated from MOSES and noticed that the moment to cause trim by 0.01 degree uses BMl instead of GML in the computation. This is the same whether or not I give a value for KG. Does this approximation continue on when calculating the righting arms and using the allowable KG macros?
REV 5.06

A: This is conservative and steeped in history. Since we do not know the value of KG when the CFORM command is issued, we do not know the GML, therefore BML is the next best thing. This only occurs in the calculation of moment to change trim and is not used elsewhere in the program.

Q: When will EQUI_H and &EQUI give the same result?
REV 5.03

A: When there is no environment and no connectors in the system.