Load Groups - Ultramarine.com Load Groups

The load group is a generalization of a nodal load. Here, a collection of load attributes is associated together as a group. For simulations, the load due to each of the attributes is computed and applied to the proper body. A load group also has a list of points associated with it. This list is given a name (normally the name of the load group) and is called a load map. When a stress analysis is performed, the total load on the group will be distributed to the nodes associated with the points by a least squares technique. Thus, if there is only one node associated with the group, it acts as a nodal load.

There are two primary differences between a load group and a nodal load. First, load groups allow for the association of not only loads, but also load attributes, with nodes. Secondly, the load attributes of a load group can be distributed to more than one node. In addition to the obvious advantages of using attributes instead of loads, the load group allows one to define gross properties to entire bodies when he is not interested in the structural details.

The load map may be either defined with a set of selectors as described below, or MOSES will define it for you. In particular, if one has not defined a map explicitly, the loads will be mapped to the points where load attributes have been defined. Whenever a body or part is defined, MOSES will automatically define a load group with the same name as the body or part.

Load groups are defined in much the same manner as bodies and parts. In other words, one issues:


     &DESCRIBE LOAD_GROUP, LG_NAME, :NODE_SEL(1), .. :NODE_SEL(n), -OPTIONS

where the options are:


     -DAMP_FUSE, YES/NO

     -AMASS_FUSE, YES/NO

and all load group attributes which follow belong to the specified load group. Here, LG_NAME is the name which one wishes to give to the following load group, and :NODE_SEL(i) are a set of selectors defining the points to which the loads will be mapped. The force from a load group is automatically distributed to the specified nodes as if the nodes were connected to the point of application by a rigid structure. Note, however, the rigid structure is not required, since the force is distributed by a least squares technique, as shown in Figure 21. The options -DAMP_FUSE and -AMASS_FUSE control the use of matrices defined with #AMASS and #DRAG commands in the frequency domain. By default, they are not used. If YES/NO is set to YES, the matrices will be included.

One can obtain the values of some of the present attributes of a load group with the string function:


     &LOADG(:LSEL, -OPTION)

the valid options are: -PERCENT, -WEIGHT, -RADII, and -CG. The string returned here is the name of the load group followed by the applicable values for each load group which matches the selector :LSEL. This value is multiplied by the load group multiplier and margin. The CG and radii of gyration are returned in the part system.

One can define mass, added mass, viscous damping, linear damping, wind, and buoyancy attributes for a load group. Here, the linear damping is not strictly associated with wave radiation, since it is a constant. Both the added mass and viscous drag are computed according to Morison's Equation. Most of the load attributes defined are applied at a point previously defined by the user. If the point reference is omitted, then the loads will be applied at the part origin.

Many option apply to all load attributes described, In particular, all load group attributes are assigned a Category of the default "extra" category by default. If one wishes to alter this, he may use the -CATEGORY option where indicated. Thus, one can have load attributes associated with different categories within the same load group. Also, MOSES associates a multiplier with the load group as a whole. This multiplier may be changed as can the multipliers for categories with the &APPLY command. This allows one to be able to "turn off" an entire load group as well as alter the force computations for some of the attributes.

Each of the attributes allows one to define a NUMBER, defined by an option, -NUM_APPLIED, which multiplies the results of an attribute before it is applied. In other words, the force, mass, drag, and added mass are first computed based on the properties defined. Then, each of these quantities are multiplied by the number specified before they are applied to the body.

The forces which will be applied are determined by a set of multipliers defined by the -WIND, -DRAG, and -AMASS options. These multipliers are similar to shape coefficients in that a force is computed and then the multiplier is applied. If any of these options are omitted, the corresponding multiplier is set to one. The forces which result from these commands are computed according to Morison's Equation. Wind only acts on the area defined by #AREA if its center is above the water surface. Similarly, water loads are only attracted when the center of area is below the water surface. -WAVE_PM is used to define WAVMUL which is a multiplier for wave particle velocity and acceleration. If WAVMUL is greater than zero, it is used to factor the wave particle velocity and acceleration before it is added to the other velocities and accelerations to compute a force. The default value of WAVMUL is zero, in which case wave velocity and acceleration will not be considered for the load attribute.

The buoyancy due to these commands is defined by the -BUOY_THICK option. Here, BTHICK is a thickness (inches or mm) which, when multiplied by the submerged area, will yield the buoyancy force.

Weight can also be defined with the other commands by using one or both of the options -TOT_WEIGHT or -MULT_WEIGHT. Here, WT is a weight, in bforce, which will be applied at the point of application. With the -MULT_WEIGHT option, WMULT is a weight/area (bforce/blength**2) which is multiplied by the area defined by either a #PLATE or a #AREA command to yield a weight applied at the centroid. Both options can be used on the same command.

For any of the load attributes that follow, the options:



     -TEXTURE, NAME_TEX, X_SCALE, Y_SCALE


     -COLOR, COLOR(1), FRAC(1), ... COLOR(n), FRAC(n)

can be used to define the color and texture of the attribute. These will be used when one asks for a picture with -COLOR MODELED. Here, NAME_COL is any color which has been previously defined. See the section on Colors for a discussion on defining colors. The NAME_TEX value for -TEXTURE is the name of a file in either /X/data/textures or /X/data/local/textures (here MOSES is store in /X). The X_SCALE and Y_SCALE are scale factors which will be applied to the texture. The NAME_TEX of NONE will yield a null default texture.

Perhaps the most popular of the load group class commands is the one which associates a weight with the group. The form of this command is:


     #WEIGHT, *PT, WT, RX, RY, RZ, -OPTIONS

and the available options are:


     -LDIST, X1, X2

     -NUM_APPLIED, NUMBER

     -CATEGORY, CAT_NAME

This command instructs MOSES that a weight of WT bforce is attached to the part at the location specified by the point *PT. This weight has radii of gyration RX, RY, and RZ (feet or meters) about the point *PT. The -LDIST option defines the longitudinal distance over which the weight will be applied when computing traditional longitudinal strength. Here, X1 and X2 are the beginning and ending longitudinal coordinates (feet or meters) of the interval over which the weight will be applied.

To define an applied force within a load group, one uses the command:


     #LSET, *PT, FX, FY, FZ, MX, MY, MZ

This command defines an applied generalized force with the magnitude of the components given by FX, FY, etc. (bforce and bforce-blength). The force is applied at the point defined by *PT, and is a member of the load set #LSET. As with all user defined load sets, one must "activate" the load set with an &APPLY command before it will actually be applied.

To instruct MOSES to build an added mass matrix for the load group, one should issue:


     #AMASS, *PT, DISP, CX, CY, CZ, RX, RY, RZ -OPTIONS

and the available options are:


     -CATEGORY, CAT_NAME

Here, the added mass will be DISP/G, where G is the gravitational constant. CX, CY and CZ are the added mass coefficients, and RX, RY, and RZ are added radii of gyration, taken about the point specified by *PT.

Similarly, one can define a constant, linear drag matrix with the command:


     #DRAG, *PT, DISP, D(X), D(Y), D(Z), R(X), R(Y), R(Z) -OPTIONS

and the available options are:


     -CATEGORY, CAT_NAME

Here, the force that will be produced is:

     F(X) = DISP/G * D(X) * V(X)
     F(Y) = DISP/G * D(Y) * V(Y)
     F(Z) = DISP/G * D(Z) * V(Z)
     M(X) = DISP/G * R(X)**2 * OMEG(X)
     M(Y) = DISP/G * R(Y)**2 * OMEG(Y)
     M(Z) = DISP/G * R(Z)**2 * OMEG(Z)

Where G is the gravitational constant, DISP, D(i) and R(i) are the quantities specified, V(i) is a velocity at the point *PT, and OMEG(i) is an angular velocity in rad/sec.

Point buoyancies can be associated with the load group by:


     #BUOY, *PT, DISP, -OPTIONS

and the available options are:


     -CATEGORY, CAT_NAME

     -NUM_APPLIED, NUMBER


     -TOT_WEIGHT, WT

Here, DISP (bforce) is the magnitude of the buoyancy which will be applied. Again, this is applied at the location specified by the point *PT, and is applied only when this point is below the water surface. Weight can also be specified with this command by use of the -TOT_WEIGHT option. Here, WT is a weight (bforce) which will be applied at the same point as the buoyancy.

The following two commands, #AREA and #PLATE, are somewhat similar in that they both define an area which attracts water and wind forces. The format of the first of these is:


     #AREA, *PT, AX, AY, AZ, -OPTIONS

and the available options are:


     -CATEGORY, CAT_NAME

     -NUM_APPLIED, NUMBER

     -WIND, WINMUL

     -DRAG, DRGMUL

     -AMASS, AMSMUL

     -WAVE_PM, WAVMUL

     -BUOY_THICK, BTHICK

     -TOT_WEIGHT, WT

     -MULT_WEIGHT, WMULT

Here, AX, AY, and AZ (ft**2 or m**2) define the components of an area concentrated at the point specified by *PT.

In contrast to #AREA, a #PLATE command defines a distributed area, and load attraction does not depend on the location of the center of area. The form of this command is:


     #PLATE, *PNT(1), *PNT(2), ............, -OPTIONS

and the available options are:


     -CATEGORY, CAT_NAME

     -NUM_APPLIED, NUMBER

     -WIND, WINMUL

     -DRAG, DRGMUL

     -AMASS, AMSMUL

     -WAVE_PM, WAVMUL

     -BUOY_THICK, BTHICK

     -TOT_WEIGHT, WT

     -MULT_WEIGHT, WMULT

Here, one specifies up to four vertices of a polygon by points *PNT(i). These vertices must be input in the order of one tracing the outline of the plate. MOSES will compute the portions which are submerged and those which are above the water surface, applying the proper forces in each regime. The level of detail used in the force calculation is defined by the -MAXAREA and -MAXREFINE and the method of computing drag are defined with options of the &PARAMETER command. The options for #PLATE function in the same manner as they do for #AREA, with one exception. When the submerged portion of a plate is computed, an aspect ratio is also computed. The added mass for the plate is that computed for a rectangular plate according to DNV Classification Notes 30.5.

Occasionally, it is convenient to describe a load attribute as a tubular, but without actually adding a tubular element to the model. This can be performed with the command:


     #TUBE, OD, T,  *PT1, *PT2 -OPTIONS

where the available options are:


     -CATEGORY, CAT_NAME

     -NUM_APPLIED, NUMBER

     -WIND, WINMUL

     -DRAG, DRGMUL

     -AMASS, AMSMUL

     -WAVE_PM, WAVMUL

     -TOT_WEIGHT, WT

     -MULT_WEIGHT, WMULT

     -BUOY_DIA, BOD

Here, OD (inches or mm) is the diameter of the tube used to calculate environmental forces and T (inches or mm) it the thickness. The tube is positioned between the points *PT1 and *PT2. The first eight options behave the same as for the #AREA command, and the -BUOY_DIA option defines the diameter used to calculate buoyant forces for the attribute. The weight computed for this element is as described about for #AREA except that another weight is computed using the thickness, the OD, the current default density, and the length of the tube.

Wind and current forces can be input to MOSES and then associated with a load group. This is achieved with the command:


     #TABLE, T_NAME, *PNT, -OPTIONS

Here *PNT is the point of application of the force and the available options are:


     -WAVE_PM, WAVMUL

     -CATEGORY, CAT_NAME

When computing current forces, the wave particle velocity is computed a the minimum depth of the location of *PNT or the water surface. The wind speed is always computed at standard anemometer height so the wind is always allied regardless of the *PNT location.

For this command, T_NAME is the name of a table of user defined force coefficients for wind and current. This table is defined using the &DATA menu:


     &DATA A_TABLE, T_NAME, FLAG

where T_NAME is the name of the force table. In this menu, the commands available are:


     ANGLE, ANG1

     WIND_ARE, WAX, WAY, WAZ, WAMX, WAMY, WAMZ

     CURR_ARE, CAX, CAY, CAZ, CAMX, CAMY, CAMZ

The ANGLE, WIND_ARE, and CURR_ARE commands are repeated for each angle, for up to 36 angles. If FLAG is specified as REFLECT, the angles specified should range from 0 to 180 degrees. If this is not specified, the angles specified should be from 0 to 360 degrees. These angles are either wind angles relative to the body system, or relative body/current velocity angles. The body velocity used is that computed at *PNT, specified in body coordinates as X, Y and Z. The force acting at *PNT is calculated by multiplying the specified coefficients by the square of the relative velocity. These force coefficients also create damping in the frequency domain. For the WIND_ARE and CURR_ARE commands, X, Y and Z are force coefficients, while MX, MY and MZ are moment coefficients. The prefix WA is for wind area, and CA is for current area. Remember to exit this menu using END_&DATA.

If one wishes, he can associate a Tanaka Damping load attribute with a load group. This is accomplished with the command:


     #TANAKA  WETSUF -OPTIONS

and the available options are:


     -ROLL, SECTION, FRACTION, BLOCK, DEPKEEL, KG, BEAM, BILRAD

     -PITCH, SECTION, FRACTION, BLOCK, DEPKEEL, KG, LENGTH, BILRAD

     -PERIOD, T(1), T(2), ...., T(n)

     -ANGLE, AN(1), ...., AN(n)

The two options -ROLL and -PITCH are instructions to compute a set of Tanaka data and add it to what exists, and one can have up to 20 different occurrences of these options. The degree of freedom which will be effected by the damping is defined by the option name. Here, WETSURF is the total wetted surface of the body (ft**2 or m**2), and one can think of each pair of -ROLL and -PITCH options as defining the damping for a part of the body which has a fraction, FRACTION, of the wetted surface. All of the other variables for the option apply to the piece being defined. SECTION defines the type of section, and must be either BOW, MIDBODY or STERN. BLOCK is the block coefficient, DEPKEEL is the distance from the waterline to the keel (feet or meters), KG is the vertical center of gravity above the keel (feet or meters), BEAM is the breadth of the body (feet or meters), LENGTH is the length of the body (feet or meters), and BILRAD is the bilge radius (feet or meters). No pitch damping is produced by default. If you define no additional data on the #TANAKA command, you will be put into a new submenu menu.

Once in the submenu, the equivalent linear damping coefficients are defined with the commands


     R_TANAKA PER, VDM(1), ......, VDM(N)

or


     P_TANAKA PER, VDM(1), ......, VDM(N)

Here, PER is one of the periods specified with the -PERIOD option, and VDM(i) are the coefficients corresponding to the angles specified with -ANGLE. Notice that VDM(1) corresponds to ANGL(1), VDM(2) with ANGL(2), etc. The units for VDM are bforce-sec-feet or meters. R_TANAKA refers to roll damping coefficients, while P_TANAKA is for pitch. This menu is "exactly" the same as the one you have before with I_TANAKA.

The final command which can be used for defining load group attributes is different from the others. Here, instead of defining specific load attributes, one defines the wind and current load attributes of an entire vessel. The particular form of this command is:


     #TANKER, SIZE, TLEN, TDEP, TBEAM, AEX, AEY, LCP, -OPTIONS

and the available options are:


     -CATEGORY, CAT_NAME

     -CBOW

     -YAW_FACTOR, YF

     -WAVE_PM, WAVMUL

This command causes wind and current forces to be computed for a tanker of the specified size according to the data presented in Prediction of Wind and Current Loads on VLCCs by the Oil Companies International Forum (OCIMF). NOTICE, these curves assume that the vessel is defined so that the X axis goes from bow to stern and that the keel of the vessel is at the origin of the part. These forces will be added to any other forces computed for the load group. The option -CBOW should be used if the tanker has a bulbous bow, and the option -YAW_FACTOR controls the yaw viscous damping during a time domain simulation. The OCIMF formulae were derived for computing the wind and current force on a stationary tanker. Instead of simply using the current velocity in the formulae, MOSES uses the relative current tanker velocity, so that viscous damping is obtained as well as applied force. This approach works quite nicely for the basic forces and yaw moment, but produces zero moment for a tanker which has zero velocity about the center of pressure regardless of the yaw angular velocity. To overcome this problem, an additional term has been added which depends on the lateral coefficient and the yaw angular velocity. The -YAW_FACTOR option specifies a multiplier for this extra term. A value of zero for YF means that the term will not be used while a value of one means that it will be used with no modification.

Here, SIZE is the size of the tanker in thousands of deadweight tons (e.g. a SIZE of 100 would denote a 100,000 DWT tanker). If all of the other dimensions are omitted, they will be interpolated from an internal database. The other data are: length, depth, beam, extra frontal area, extra lateral area, and distance from the origin to the longitudinal center of the tanker. Here the units are feet or meters for distances, and ft**2 or m**2 for areas. MOSES uses the major dimensions to compute the wind and current areas, and the extra areas are for wind only. There is an internal database of default extra areas for AEX and AEY. These values are replaced when a non-zero value is used for AEX or AEY. In particular, the lateral wind area is given by TDEP minus the draft times TLEN plus AEY. Likewise the frontal wind area is (TDEP - DRAFT) * TBEAM + AEX.

In some cases, using #TANKER may cause forces in a direction opposite from one expects. This is a function of the OCIMF data, and not a problem with the software. There are lift forces involved with the ship shaped hull forms this data represents, sometimes causing this behavior. The OCIMF results are based on extensive wind tunnel and tank tests on typical tankers. We have incorporated digitized versions of these curves into MOSES, therefore, what is derived is essentially what was measured. The question now becomes why are these forces in a direction opposite to what one may expect - the answer is probably lift (but this depends on your expectations!), and is what permits us to sail and fly. The hull will behave like an aerofoil where the flow does not separate immediately at the bow (and particularly if it is cylindrical as far as wind loads are concerned). As a consequence, the longitudinal force components may be "negative", i.e. up current or upwind, for some directions. Rest assured, however, the resultant force is always downweather so we can't sail or fly for nothing - there is a net drag!