We are pleased to announce our newest contributor to the SA Army, designer Tim Kernan, owner of Waterplane Design. Tim designed the very impressive 70' Peligroso, and will be contribting a series of YD 101 articles here. Enjoy.

The subject of designing to a handicap rule is not one that's typically covered in the text books, but every racer knows it's a critical component of racing yacht design. Designers spend many hours picking through the regulations and formulae of various handicapping rules, trying to outsmart the raters. Ironically, chances are good the designer had a hand in developing the very rule he's now trying to beat. Outsmarting yourself can be tough. I generally start by classifying the rule and determining what can be done.

Handicapping rules can be classified into two basic types:
1. Performance-based rules, such as PHRF, in which various factors affecting a boat's performance, including race results and subjective factors, are evaluated by a designated body or individual to determine a boat's rating.
2. Measurement-based rules, such as IMS, or Americap II (soon to be ORR), in which various parameters of the yacht are directly measured and input to a set of algorithms that, solved, generate a handicap. In its most modern form (as in IMS and Americap II) these algorithms take the form of a VPP or Velocity Prediction Program.
A subcategory of the above, Box rules, such as the TransPac 52 or Volvo 70 rules, define parameters (such as BOA, LOA, SA, Displacement, etc.) within a strict range such that every yacht which fits within these parameters or "box" may race level with the others.

For obvious reasons, Performance-based rules offer little opportunity for the designer to affect ratings. As in PHRF, a "conservative" initial rating is generally issued for new designs, and after the actual performance of the yacht has been observed and documented by race results, etc. the rating may be adjusted at the discretion of the rating board. Any initial rating benefit will typically not last for very long, and as often as not the designer finds himself in front of the rating board trying to convince them what a pig his design really is, just to get a fair rating. If you want to beat PHRF, get yourself a copy of "How to Win Friends and Influence People", and try to get on the rating board…

What about IRC? IRC uses a combination of measured and subjective factors to determine a rating, and so offers some opportunity for optimization. In the case of IRC, the treatment of measured factors is a highly-guarded secret, and designers are left with the "trial and error" approach for optimizing to that rule. During the design phase, and to a limited extent thereafter, a designer may submit data to the ORC and obtain trial certificates for multiple design iterations, which are then compared to the designer's performance predictions to determine the best configuration of measured factors relative to the rule. (Design offices are permitted 25 trial certificates per year, 6 per 2-meter range of length, and 6 trial certificates per year for currently rated yachts.) There are some substantial performance-affecting factors, most notably stability, which simply are not directly measured under IRC, and therefore easily exploited. Other factors not directly accounted for in the IRC rule that bear examination by the designer are: wetted area, heeled shape data, and appendage efficiency. An analysis of the existing IRC fleet which identifies common characteristics of winning yachts is a good approach to dissecting the "subjective" elements of the rule.

Undervaluation of important performance-dependent factors such as stability will inevitably lead to type-forming, and in the case of IRC the trend is clearly moving towards yachts with high stability-foreseeable yes, but not all bad. The fact is that any rule which attempts to predict a yacht's performance by measurement of objective factors will eventually lead to type-forming unless the rule is allowed to evolve. This is due to the inherent complexity of sailing dynamics in juxtaposition to the relative simplicity of rating systems. The problem is greatest at the measurement stage, as a relatively limited set of measurements must be taken to describe the complex geometry of hulls, appendages, and rigs. It is primarily this area where the designer focuses his efforts to gain rating advantages, and this has lead to some very rule-driven permutations in hull forms over the years.

In the infamous 1979 Fastnet race, 5 boats sank, 24 boats were abandoned, and 15 lives were lost. This tragic loss was largely attributed to the "tortured" hull forms that had arisen out of designers' attempts to manipulate key IOR measurement points and to the relatively low ultimate stability of the latter generation IOR yachts in comparison to traditional sailing yachts of the time and even those of a few rule-generations earlier. Designers had carried rule-beating to the extreme of producing less seaworthy yachts. What followed was the development of the VPP (Velocity Prediction Program) which would be used as the basis for the current IMS/ORC. The VPP that formed the basis of IMS arose out of the Irving Pratt Ocean Race Handicapping Project undertaken at MIT by J.E. Kerwin et. al., and was a monumental step forward for the sport as a whole.

So what can be done to optimize for VPP-based rules? By developing a computer-based LPP (Lines Processing Program) to input hull data, the developers of the IMS were taking a large step toward obviating the humps and bumps associated with taking hull data from set measurement points. The LPP generates a full integrated 3-dimensional representation of the hull form taken either from designer's offset files or via an ingenious (albeit imperfect) measuring "wand" that allows measurers to take offsets directly from an existing hull. Sounds good, however this 3-d shape is then transformed into a set of non-dimensional coefficients to be used by the VPP in its equilibrium equations. These formulations, if available to designers, can be studied and their limitations exploited. (For an example, see attached.)

When the MHS (Measurement Handicapping System, precursor to the IMS) was first used in the 1978 Newport-Bermuda race, it was a novel and technologically advanced tool. Today, it is not uncommon for design offices to run the IMS VPP along with far more sophisticated analysis tools including proprietary CFD codes (Computational Fluid Dynamics), more advanced and/or proprietary VPPs, weather routing packages, etc. As stated earlier, even the most complex measurement system in use today (the IMS VPP) must convert measured hull data into non-dimensional coefficients for use in the internal equations. The formulations used by the VPP are based on model test data which, although designed to cover a broad spectrum of typical hull shapes, are necessarily limited by the scope of testing. The designer of a new hull form has several options for more closely predicting the resistance characteristics of a new hull: tank test data, CFD analysis, and real-world proprietary data taken from previous generation designs. The designer can develop a far more accurate resistance picture of a given design, input this data into their in-house VPP and compare the results to those predicted by the rule, and thus determine the best configuration relative to the rule. This is approach is applicable to both hydro and aero models, and as a result this type of data is highly guarded.

There are many trade-offs in yacht design, and it is rarely possible to develop a design which will be at the top of its competitors in every true wind strength and sea state. Thus, an analysis of typical conditions in which the yacht will be competing can aid in developing the best "horse for the course". Current box rules attempt to discourage this in order to promote even competition and design obsolescence, yet to a large extent it is unavoidable, and often the design most closely optimized to racecourse conditions wins.

So the arms race rages on between designers and rating authorities. With sailors always looking for that edge over their competitors, the temptation to gain it by exploiting weaknesses in handicapping rules is just too juicy to overlook. The cold war drives development though, and fosters important advances in performance technology that benefits everybody. If nothing else it pays the rent for guys like me.

An example of optimization to IMS, formulation for "Second Moment Length"

Probably the most easily identified speed producing factor in a sailing yacht hull form is "sailing length". In simple terms, this is a measure of a yacht's waterline length, which due to the relative speed of entrained waves, governs a yacht's attainable "displacement mode" speed. Dynamically, however, the situation is more complex. Heaving and heeling motions, bow and stern overhangs, and volume distribution particularly at the ends, all effect sailing length. Prior to IMS, waterline length was generally measured as a centerline profile at a prescribed height, possibly with girth corrections made at the ends. IMS does not measure sailing length directly, but rather uses the concept of the "2nd moment length" Consider a yacht's sectional area curve (local sectional areas S (x) plotted in relation to length (x)). As shown in the graph, by taking the square root of the local sectional areas and plotting them along the length, volume in the ends is emphasized. This is linearized by the next step, in which the radius of gyration, or gyradius of the sectional area curve is taken. The gyradius is essentially a measure of the moment of inertia of the square root curve about the longitudinal position of its centroid, with any particular element represented by its value times the square of its distance to the centroid. Direct computation of the above yields a relatively small number, so it is multiplied by an additional factor to normalize the length to approximate the yachts actual waterline length. Second Moment Lengths (LSM 0-4) are calculated at 3 trim conditions and a weighted average applied to determine Sailing Length (L) which is then used by the VPP as a factor of hull resistance.
LSM0 is for the yacht in Measurement Trim floating upright
LSM1 is for the yacht in Sailing Trim floating upright
LSM2 is for the yacht in Sailing Trim with 2 degrees of heel
LSM4 is for the yacht in a deep condition sunk 0.025*LSM1 forward and .0375 * LSM1 aft, floating upright.

The formula for obtaining Sailing Length (L) is given as: L=.3194*(LSM1 + LSM2 + LSM4)

As can be seen in the above formula, L is essentially an average of LSM0, 1, & 4, with equal weighting given to each LSM. Since LSM4 is the only of these factors which truly account for volume in the ends (a speed producing factor) a hull with volume distributed towards the ends at locations above the LSM0 -LSM2 flotations will be advantaged under the rule. This is precisely the type - form that IMS has produced. Similarly, formulations for Beam Depth Ratio (BTR) and Effective Beam (B) used in the formulations determining residuary resistance of the hull have led to the slab-sided type form of today's IMS yachts.




Tim Kernan
Waterplane Design

12/12/05