Background Information

Weight Classification Systems

On a vessel, in accordance with Archimedes Principle the weight of all the components, crew, effects, and other loads is equal to the weight of the volume of water that the vessel displaces.  In order to be able track weight growth, as well as assist in cost estimating, and predicting the estimated weights for new designs, all the individual components on a ship are categorized into groups based on the function of that component.  In the US for modern naval vessels, a system called the Extended Ship Work Breakdown Structure (ESWBS) is currently used. ESWBS is an extension of the previously used Ship Work Breakdown Structure (SWBS) and for what I am considering here they are pretty much the same.

Prior to the mid 1960s (or so) though, a different system, called the Bureau of Ships Consolidated Index (BSCI) was used.  Although this system is in general similar to the SWBS/ESWBS system, there are some differences between how some specific items are categorized.

Additionally, overseas there are other, nationally based systems, including a system used in the UK based on their Naval Engineering Standards (NES).  As with the BSCI system this system is in general similar to SWBS/ESWBS, however here are some differences between how some weights are categorized.

For the most part I believe that most of the data I have collected is based on either the SWBS/ESWBS or the BSCI system (depending on the date that the vessels were designed/built) however, I am pretty certain that a couple of the data points I have collected are in the UK NES system.  At some point in time I hope to go back and clarify which data points are based on which system, but for now I believe that there is currently a mixture of the different systems represented.

Overall, all three systems tend to break ship weights into the following categories:

  • Group 100 - Structural weights
  • Group 200 - Propulsion system weights
  • Group 300 - Electrical system weights
  • Group 400 - Command, Communications, Computers, Controls, Intelligence, Surveillance, & Radars (C4ISR)
  • Group 500 - Auxiliary systems
  • Group 600 - Outfit & Furnishings
  • Group 700 - Weapon systems
  • Group F - Loads (which is identified as Group 800 in the NES system - I believe)
  • Group M - Margins

In general, groups 100 to 700 add up to give the vessel's basic light ship weight, though for early stage designs a number of margins (from Group M) are often added to address the uncertainty in early stage design numbers. Also, if ballast is required it is often added in as well.

Light ship weight plus loads (such as crew & their effects, fuel, stores, and munitions, etc) add up to give the vessel's full load displacement.

Finally, since it is realized that over the life of a vessel it will probably grow in weight, due to added systems, ship alterations, and other various additions, typically a service life allowance is also included to reflect the allowable expected growth of the vessel.  As such, the full load displacement of the vessel when new (at delivery) plus its service life allowance would be equal to the vessel's full load displacement at the end of its service life.  I believe that for non-US vessels some of the terminology will be different (ie that may not call all the weights, margins, and allowances exactly the same) but I believe, in general, the overall concepts are similar.

Resistance & Powering

Resistance

Resistance Components - As a ship moves through the water it creates a drag force, or resistance.  In general this resistance consists of several major parts, and some additional minor components.  Specifically, there is;
  • the frictional resistance generated by the contact of the hull's surface with the water,
  • a wavemaking resistance generated by the hull as it pushes the water aside as it passes,
  • a wind resistance caused by the above water portion of the vessel as it interacts with the air,
  • the resistance of the appendages on the hull below the water, including such things as rudders, propeller shafts and brackets, as well as potentially other items like bilge keels, sonar domes, stabilizers, bow bulbs, and stern flaps,
  • added resistance in waves, and
  • other items such as wave breaking resistance, spray resistance, and an added resistance caused by the control surfaces (such as the rudders) as the ship makes minor steering c orrections in an effort to maintain its course and added resistance due to hull roughness and fouling.
For simplicity, it is not uncommon to combine some of these components together.  Specifically, wavemaking, wave breaking, and spray resistance (where applicable) are often rolled into a single category called "Residuary" resistance. In order to estimate the resistance of a new design it is necessary to make a reasonable estimate of the components of resistance.

Resistance Estimating Techniques - If data exists on a similar ship it is sometimes possible to use that data for estimating the resistance of a new design.  However, if the new design is significantly different from the existing design this may not be fully adequate.  Additionally, over the years work has been ongoing in trying to develop mathematical means of estimating a ship's resistance by modeling the ship's hull in a computer and using Computational Fluid Dynamics (CFD) methods.  However, this is somewhat time consuming and not well suited to early stage estimation, where full details of a ship's hullform may not yet be fully worked out.  Similarly, it is possible to build a scale model of a design and measure its resistance in a tow tank, however, this is also costly and time consuming and typically is not done till later in a design when more details on its hullform have been set.  As such, several method's have been developed over the years to estimate a ship's resistance either from data on other similar ships and/or concepts that have been developed previously or through the use of model test data on Systematic Series of similar notional hullforms.

Systematic Series - In general a Systematic Series is a series of similar hullforms which have been developed and model tested.  These models typically have certain major characteristics varied over the family of hulls, such as block coefficient, LCB location, and L/B ratio, etc.  The model test resistance data is collected and analyzed and presented in such a way as to allow a user to interpolate an estimate of the resistance of his vessel, at its given block coefficient, LCB location, and L/B ratio, etc from the range of data provided.  Some of the more common typical Systematic Series that have been developed include;
  • The Taylor Standard Series -
  • Series 60 - Methodical Experiments With Models Of Single-Screw Merchant Ships
  • Series 62 -
  • Series 64 -
  • Series 65 -
  • The BSRA Series -
  • The SSPA Series -
  • The National Physical Laboratory (NPL) Series -
  • The Marwood and Bailey -
  • The Webb Trawler Series -
  • The NTUA Series -
  • The Naval Acadmey Hydrodynamics Lab Series -
  • etc

One potential problem with using Systematic Series  though, is that they are only useful over a limited range of vessel parameters investigated and may not cover the hullform shape or other hullform parameters of the design that you are interested in.  In order to account for this it is sometimes possible to combine the use a Systematic Series with data on existing vessels.  I believe that in the US Navy a method like this was used for many years, where restance data on existing vessels or designs were compared to the estimated resistance for those vessels as determined by the Taylor Standard Series .  The ratio of actual resistance to estimated resistance was then plotted over a range of speeds to develop a "Worm Curve" factor for that vessel.  If you then had a ship some what similar to an existing design, but for which some parameter, or group of parameters, were different (such as block coefficient or LCB, etc) you could then use the Systematic Series  to make a preliminary estimate of resistance but then multiply those results by the Worm Curve  factor for the similar ship to adjust the results to account for differences between your design and the hullform of the vessels in the Systematic Series.

Mathematical Regression Methods - In more recent years alot of effort has been made into using mathematical regression methods to condense this data down into relatively simple mathematical terms to ease in the estimation of early stage resistance estimation. These methods easily lend themselves to use in spreadsheets or simple BASIC or FORTRAN programs.  Additionally, more recently some individuals and research establishments have also made an effort at combining the data from several Systematic Series together sometimes including data for other specific designs or vessels for which model test or full scale data was available. One of the most well known of this type approach is the Netherlands Shipbuidling Model Basin's (NSMB) mathematical method, which is also known by the name of its two principal authors Holtrop & Mennen.

The NSMB/Holtrop & Mennen method is very powerful and useful because it incorporates data from a wide range of hullform types, but it is not always clear whether in doing so, if it is as accurate for a given hull type as a method that only addresses that specific hull type. For this reason, on this site right now I have decided to make use of data developed by Sui Fung specifically for transom hull ships representative of most typical modern frigate and destroyer type hulls, instead.

Resistance Scaling - Because many resistance estimating routines are based in part on model tests, it is important to understand how model test resistance is scaled when trying to estimate the full scale resistance of a ship.

For starters, it was discovered long ago that the wave pattern that a ship generates as it moves through the water is impacted greatly by the speed that the vessel is going in relation to the length of the vessel.  If a vessel is moving at a speed where the the waves generated result in the bow and stern resting on peaks of the waves then the ship will be at near even trim which is good for minimizing resistance.  However, if a vessel is moving at a speed where the bow is at a peak but the stern is in the trough of a wave then the ship will be operating at a greater than normal trim and the ship's resistance will be adversely impacted.  As such, the humps and hollows in the waves generated by the ship as it moves through the water can result in reltive increases and decreases or "humps" and "hollows" in the ship's resistance curve, as shown below. 

Humps-Hollows



Here the Red curve shows the estimated total resistance while the two major components of total resistance, the residuary and frictional resistance are shown in Blue and Orange respectively.  As can be seen in this figure the frictional resistance increases smoothly as speed increases, but there are humps (as denoted by the Purple arrows) and hollows (as denoted by the Green arrows) in the residuary resistance curves, which also are apparent in the overall total resistance curve.

The term Froude Number (Fn) has been defined as a means of reporting a ship's speed in relation to its length in a non-dimensional fashion.  Specifically;      Fn = v / sqrt (g * L)
     Where;

     v = the vessel's speed
     L = the vessels length
     g = the acceleration due to gravity
     (all in consistant units)

In general the wave pattern generated by a model at a given Froude number will be the same as the wave pattern generated by the full scale ship at that same Froude number.  As such, wavemaking resistance is said to follow Froude scaling rules.  Additionally Model Scale Ratio is defined as the ratio of the length of the full-scale ship in comparison to the length of the model and is sometimes called l. A s such the actual speed that a model will be operating at to give the same Froude number as a full-scale ship will be equal to the speed of the full-scale ship divided by the square root of the Model Scale Ratio (eg Vm = Vs / sqrt(l)).

However, because of the viscous properties of water it is not possible to scale all of a ship's resistance directly from a scale model, as the frictional components of resistance do not follow the same Froude scaling rules.  For frictional resistance a different non-dimensionalization of the vessel's speed is more significant.  This is called the Reynold's number (Rn) and is is defined as;

     Rn = v L / n

     Where;

     v = the vessel's speed
     L = the vessel's length
     n  = the kinematic visosity of water
     (all in consistant units)

Based on the size and wetted surface of the model, an estimate of the frictional (or in some cases viscous) resistance of the model is made based on a standard formulation.  In the US many early model tests and systematic series used a formulation put forward by the American Towing Tank Conference (ATTC) based on observations of measured frictional resistance of flat plates.  Here (I believe) an equation for a non-dimensional frictional  resistance coefficient (Cf) was derived as;
     0.242*sqrt(Cf) = log10 (Rn *Cf)
Because the frictional resistance of a vessel is a function of the vessel's Reynold's number, Frictional Resistance is said to follow Reynold's number scaling and if a model were run at a the same Reynold's number as the full scale ship then they would have the same non-dimensionalized Frictional Resistance coefficients, however the Froude number's would be different and hence the wave patterns developed by the model and full scale ship would be different, which is why the Frictional resistance and residuary resistance components are treated separately.

Later in 1957 the International Towing Tank Conference (ITTC) agreed on a newer formulation for calculating a Frictional Resistance coefficient as follows;
     Cf = 0.075/(log10 (Rn) - 2)2
I believe that is this Frictional Resistance Coefficient that is used by most (though not all) of the Systematic Series identified above.

More recently the ITTC put forward a revised methodology in 1978 where, instead of just considering frictional resistance based on data for flat plates, an effort was made to account for the form of the ship, as well.  In this 1978 formulation a vessel's viscous drag was defined as being equal to the the vessel's frictional resistance times a (1+k) term to account for the impact of hull curvature and form.  I believe that the NSMB/Holtrop & Mennen equations are based on this newer methodology.

As such, when scaling model test data for conventional displacement hulls it is typical to measure the total resistance and convert it into terms of a non-dimensional frictional resistance coefficient.  For each data point collected the calculated frictional resistance coefficient at that speed is then subtracted from the total resistance coefficient giving a "residuary" resistance coefficient which more or less includes all the other factors like wave making, wave breaking, and wave spray (if significant), etc.  This is the portion of the model test data that gets scaled to full scale by Froude scaling.  An estimate of the frictional resistance of the full-scale ship is then added onto the residuary resistance estimate (using a similar method as used for estimating the frictional resistance of the model), and then additions are also made for any other additional components (such as wind and appendage resistance).

Because most models do not include all of the "above water" components of a ship, the wind resistance of a ship is something that is typically added in later based on approximation equations or if available wind tunnel data for the ship.  However, depending on the towing tank and their standard procedures, sometimes there may be a correction to the model-scale data to account for the wind resistance of the limited portion of the above water hull that was included in the model tests.
Additionally, sometimes models are tested with small scale appendages (especially for fixed components, like bulbs etc).  However, since of scaling issues with items like rudders, fins, and bilge keels, etc, it is not uncommon that these are not included on the model (for resistance estimating) and that an approximation of their effects are considered later based on equations derived for the specific component type or simply by adding in an allowance based on previous experience.

Finally, typically an additional Correlation Allowance, based on the size of the full scale ship and its likely surface roughness, is typically also added in.

Appendage Drag - There are several methods of estimating appendage drag for a vessel.  The book "Principles of Naval Architecture" (Ref B-1) gives information on a couple of these methods.

Specifically, the book provides several equations relating the resistance of certain specific appendage types to the geometry of these items.  This includes equations specifically for;

  • Bilge Keels
  • Control Surfaces (such as rudders, shaft brackets, and stabilizer fins, etc)
  • Shafts & Bossings, and
  • Skegs

An alternate method to using these type of shape specific equations is instead to estimate the area of the appendages and multiply the area for each by an Effective Form Factor of the appendages (k2).  From this a total Effective Form Factor for the ship is calculated taking into the account the surface area and k2 of each of the appendages using the ITTC 1978 viscous resistance curves.  This type method is incorporated into the NSMB/H&M resistance estimation equations.

A third method is one outlined in the US Navy's Design Data Sheet 051-1 (DDS 051-1) entitled "Prediction of Smooth-Water Powering Performance for Surface-Displacement Ships" (Ref B-2).  In this method, curve fits through data on exisitng vessels are provided to allow the user to make an initial estimate of appendage r esistance for a given ship.

For early stage desgin, the first two type methods noted above can be a little cumbersome in that they require the user to estimate the size of all the appendages. As such, for early stage design I have made use of the method outline in DDS 051-1.

Wind Resistance & Still Air Drag - As a ship moves through the water it encounters resistance to this motion not only from the water it is moving through, but also the air that the portion of the ship above the waterline comes in contact with. This air resistance can be considered in three ways;

  • you can consider only the still air drag generated solely from the ship's forward motion (condition 1),
  • you can consider the still air drag generated from the ship's forward motion plus a certain amount of head wind acting on the front of the vessel (condition 2), or
  • you can consider the total air resistance of the ship taking into account any existing wind and the relative motion of the ship (condition 3).

In the first case, still air drag is the resistance that is generated by the ship assuming that there is no wind blowing other than the wind the ship is generating itself as it moves forward.  Thus in this case the speed of the self generated wind is equal to the speed of the ship, and it can be considered acting only on the frontal area of the ship.

In the second case you simply add a set amount of wind speed to the ship's speed, and assume that this acts on the frontal area of the vessel.  The final case however, is more complex and takes into account the relative direction of the wind to the ship as the ship moves through the water and acts on both the frontal and side areas of the ship.  It would be important to calculate for estimating actual vessel performance, but for design purposes, either assuming only still air drag, or still air drag plus a certain amount of head wind is (I believe) most typical.

The book "Principles of Naval Architecture" (Ref B-1) gives several different methods for estimating Wind Resistance & Still Air Drag.  These are typically of the form;

     Raa = coefficient * r * AT V2

     Where;

  • Raa = The total added air resistance
  • coefficient = an empirically derived coefficient
  • r = the density of the air
  • AT = the frontal area of the ship above the waterline
  • V = the total wind velocity (for condition 1 this would be equal to the forward speed of the ship, but in condition 2 this would include both the speed of the ship and any additional head wind)

Ref B-1 notes that in 1943 RADM D.W. Taylor derived a simplifaction to the above equation for ordinary ships where;

     Raa = 0.783 * * B2 VR2

     Where;

  • Raa = The total added air resistance
  • B = the Beam of the ship (in meters)
  • VR = the apparent relative wind velocity (for condition 1 this would be equal to the forward speed of the ship, but in condition 2 this would include both the speed of the ship and any additional head wind) (in meters per second)

This version of the equation is convenient for early stage design use as it doesn't require an estimate of frontal area, which may not be fully known in early stage design.

Powering

Once the resistance of a ship is estimated it is then necessary to make an estimate of the powering requirements for the ship.  In very basic terms if you multiply the resistance of a vessel times its speed (and make any necessary corrections for units) you get what is called the vessel's Effective Power requirement (or EHP  for Effective Horsepower).  If there were no efficiency losses or need for margins an engine capable of producing an amount of power equal to a ship's EHP would be able to propel the ship at the speed for which the EHP was calculated. However, there are many different efficiencies and margins that must be considered which drive the total installed power requirement for a vessel up to a value sometimes approaching twice the value of the calculated EHP at design speed.
Hull Form Effects
One of the first things that must be considered in determining full power requirements is the effect of the hull on the water around it.  In addition to the drag already considered in the resistance estimate, there are other factors that must be considered.  Amongst these are three terms called;

     - wake fraction
     - thrust deduction, and
     - relative rotative efficiency

Wake Fraction - as a ship moves through the water, the viscosity of the water will result in a layer of the water near the vessel being dragged more or less along with the ship. This results in the flow into a ship's propellers being overall typically a little less than the speed that the ship is traveling at, and depending on the level of detail that you are going into for your resistance and powering estimates it may become necessary to try and estimate this.  For early stage design this is trypically done using curve fits or rules of thumb based on data on similar ships.

Thrust Deduction - similar to wake fraction, there is also a term called the thrust deduction factor.  In simple terms, the total thrust that a propeller (or group of propellers) must produce to propel a ship tends to be a little more than the total resistance of the ship. This differnce is sometimes called an augment of resistance or reduction in thrust available at the propeller.  According to the book "Resistance and Propulsion of Ships" (Ref B-3) it is caused by a number of factors including the propeller accelerating the water into the stern which can cause an increase in frictional resistance, influences of the potential-velocity field in which the propeller operates and possible influences that the propeller may have on the stern wave pattern of the ship.  For our purposes its enough just to know that the total thrust required to be produced by the propellers is going to be a little more than the estimated resistance, and like the wake fraction, in early stage design it is usually estimated by means of using curve fits or rules of thumb based on data on similar ships.
Relative Rotative Efficiency -

Propeller Efficiency -
Shaft and Mechanical Efficiency -

Home