BRENNEN

Industries / Transport and vehicles / Shipbuilding and repair / Boat building

Design

Designers decide a small boat's hull shape, structure and stability, and the sea conditions it is meant for, before any mold is built.

6

Milestones achieved, from 1960 to 2014.

1

To watch for: announced, not yet confirmed outside.

3

Open deliverables. Nobody has shipped them.

The road

AchievedWatch forOpen

Achieved

  1. 1960

    The deep-V hull wins

    A deep-V hull has a sharp V-shaped bottom that runs well aft, so it cuts through waves instead of pounding on them. C. Raymond Hunt built the first one, a 23-foot wooden prototype, in 1958. Dick Bertram then had Hunt draw a 31-foot version, Moppie, which won the rough 1960 Miami-Nassau powerboat race; Soundings gives its bottom 24 degrees of deadrise (the angle of the V) and lifting strakes, ridges along the bottom that help it plane and throw off spray. Bertram began building the Bertram 31 from 1961. Hunt had described the design to a magazine in 1958 and filed for a patent more than a year later, so he lost the patent and other builders copied the hull freely. Hunt Yachts' chief executive told Soundings it "still remains the hull shape of choice in the industry."

    maineboats.comsoundingsonline.com

    Credit

    C. Raymond Hunt; Dick Bertram

  2. 1964

    Planing speed by formula

    A planing hull rises onto the surface at speed and skims instead of pushing through the water. In October 1964 Daniel Savitsky of the Stevens Institute of Technology published "Hydrodynamic Design of Planing Hulls," a set of equations, fitted to model tests, that give a hull's lift, trim and drag from a few inputs such as weight, beam, deadrise and speed; Crossref counts 155 later works that cite it. The method is still in use: OpenPlaning, a free Python library released with a 2021 paper by E. L. Castro-Feliciano, is built on Savitsky's 1964 and 1976 methods.

    doi.orggithub.com

    Credit

    Daniel Savitsky, Stevens Institute of Technology

  3. 1979

    A storm tests stability

    A storm struck the 303 yachts in the Fastnet Race off southern England and Ireland in August 1979. By Yachting Monthly's account of the official inquiry, 15 sailors died, 24 boats were abandoned and five sank. The inquiry found that 48% of the fleet was knocked flat and 33% past flat, and five boats stayed upside down for between 30 seconds and six minutes. Racing then ran under the International Offshore Rule (IOR), a handicap formula that, by the account of the Royal Ocean Racing Club's former technical director, rewarded designers for keeping stability to the minimum allowed. An appendix compared a Contessa 32, which rights itself from up to 156 degrees of heel, with a 1976 Half Tonner that rights itself only from up to 117 degrees, and so is far more likely to stay upside down. Handicap rules then began to measure stability, with a screening system from 1988, and boats entering the Fastnet must now meet the ISO 12217-2 stability standard.

    Lessona rating rule that rewards low stability gets boats with low stability.

    yachtingmonthly.com

    Credit

    RYA and RORC inquiry (Sir Hugh Forbes, Sir Maurice Laing, Lt Col James Myatt)

  4. 1994

    Boats graded by sea

    The European Union's Recreational Craft Directive (94/25/EC) of 16 June 1994 set one set of safety requirements for leisure boats of 2.5 to 24 m sold anywhere in the EU. Each boat is given a design category: the roughest water it is built for. The 2013 recast, Directive 2013/53/EU, names four categories only by wind and wave height (the significant wave height, the average height of the highest third of waves). Category A is for winds that may exceed force 8 on the Beaufort scale and waves of 4 m and above, B for up to force 8 and 4 m, C for up to force 6 and 2 m, and D for up to force 4 and 0.3 m. Stability, buoyancy and strength must all be designed for the category chosen.

    legislation.gov.uklegislation.gov.uk

    Credit

    European Parliament and Council of the European Union

  5. 2008

    One standard for structure

    Scantlings are the sizes of a hull's structural parts: the thickness of its skin and the size of its frames and stiffeners. ISO 12215-5, first published in 2008 and revised in 2019, gives the design pressures and stresses from which scantlings are worked out for monohulls up to 24 m, built of fiberglass, aluminum, steel or glued wood, including small commercial craft and workboats. World Sailing, the sport's governing body, now reviews the plans of ocean racing yachts against parts 5, 8 (rudders) and 9 (keels and other appendages) of the standard, through notified bodies and classification societies it recognizes.

    iso.orgsailing.org

    Credit

    ISO Technical Committee 188 (small craft)

  6. 2014

    A bonded keel lets go

    On 16 May 2014 the British yacht Cheeki Rafiki, a Beneteau 40.7 cruiser-racer of a type Yachting World counted about 800 of afloat, turned over on an ocean passage; her upturned hull was found, but her four crew never were. The UK's Marine Accident Investigation Branch (MAIB) concluded that the keel came off and found that where a grid of stiffeners is glued (bonded) into a fiberglass hull, the bond can break down unseen, and a "light" grounding can damage it without anyone noticing. It asked the British Marine Federation to write industry guidance on inspecting and repairing such hulls. Yachting World reported that a keel working party of ISAF, now World Sailing, had listed 72 keel failures since 1984, in which 24 people died.

    Lessona hull's strength can rest on joints that nobody can see once the boat is finished.

    gov.ukyachtingworld.com

    Credit

    MAIB

Watch for

  1. Feadship trains hull AI

    The Dutch superyacht builder Feadship says its AURORA program, funded by the Dutch maritime innovation scheme, will run about 3,500 detailed computer flow simulations (computational fluid dynamics, or CFD) to train AI models that predict a hull's resistance and refine its shape.

    Achieved whenthe AI's resistance predictions are published next to towing-tank or sea-trial measurements of the same hulls, made or checked by an outside laboratory.

    feadship.nl

    Claimed by

    Feadship, March 2026

Open

  1. 01

    Keel failures in public

    an open table of every keel failure on a sailing yacht that can be found in public records since 1984, with the date, boat model and builder, year built, keel type and how it was fixed to the hull, the cause, and deaths, each row linked to its source.

    The brief
    Credit · unclaimed
  2. 02

    Test boats against Savitsky

    a public comparison of planing-speed predictions with real boats. Take 50 production outboard and sterndrive boats whose magazine tests give length, beam, weight, deadrise, engine power and measured top speed, predict the top speed with Savitsky's methods, and publish the errors.

    The brief
    Credit · unclaimed
  3. 03

    Design categories against real seas

    a public set of tables and maps showing, for 20 popular coastal and offshore routes, how often in each month over the past ten years the wind and waves exceeded the limits of each EU design category (for example, category C's force 6 and 2 m).

    The brief
    Credit · unclaimed

Open 01 · the brief

Keel failures in public

Credit · unclaimed

The first to publish is credited on this line, and the deliverable moves from Open to Achieved.

What to build

an open table of every keel failure on a sailing yacht that can be found in public records since 1984, with the date, boat model and builder, year built, keel type and how it was fixed to the hull, the cause, and deaths, each row linked to its source.

Why it is still open

The list compiled for ISAF's Special Regulations Sub-Committee by Mike Urwin and Simon Forbes was published only as a one-page summary, in November 2013: 72 incidents, 40 of them (55.6%) with no defined cause, 11 welded fin failures, 8 after grounding or collision, 8 hull or internal structure failures, 3 keel bolt failures and 2 canting keel failures. The spreadsheet itself is not public, and searches of the web and GitHub in November 2026 found no open keel failure dataset.

sailing.org

Built with what exists

accident reports from the UK MAIB, the US NTSB and Coast Guard, France's BEAmer and Australia's ATSB; yachting press archives; a language model such as Claude Fable 5.2 to pull the fields out of each report (see the Note "Three working devices"); a shared spreadsheet or a GitHub repository.

  1. First step

    list every MAIB yacht report since 1984 that mentions a keel and enter the first ten rows.

  2. Done looks like

    at least 72 incidents, each with at least one public source, sorted into the same cause classes as the 2013 summary, checked by two reviewers, with counts by cause and by keel attachment published.

Open 02 · the brief

Test boats against Savitsky

Credit · unclaimed

The first to publish is credited on this line, and the deliverable moves from Open to Achieved.

What to build

a public comparison of planing-speed predictions with real boats. Take 50 production outboard and sterndrive boats whose magazine tests give length, beam, weight, deadrise, engine power and measured top speed, predict the top speed with Savitsky's methods, and publish the errors.

Why it is still open

Savitsky's methods rest on towing-tank models. A 2026 review of such methods led by Dejan Radojčić of the University of Belgrade notes that full-scale trials are costly and that model and full-scale data are hard to combine, and OpenPlaning's documentation works its examples from Savitsky's own 1976 paper. A search in November 2026 found no comparison of an open planing code with published boat tests.

simlab.mas.bg.ac.rsgithub.com

Built with what exists

OpenPlaning (MIT-licensed Python); magazine test pages, such as Boating's 2026 Bayliner V22 test, which lists a 22 ft 5 in length, 8 ft 4 in beam, 3,731 lb displacement, 20-degree deadrise, a 225 hp maximum and its own test results; published ranges for propeller efficiency; Python with pandas.

  1. First step

    run OpenPlaning on ten boats from one magazine's 2026 tests and note which inputs the reports leave out, such as test load.

  2. Done looks like

    a table for at least 50 boats giving predicted and measured top speed and the mean error in percent, with every input and assumption published so others can rerun it.

Open 03 · the brief

Design categories against real seas

Credit · unclaimed

The first to publish is credited on this line, and the deliverable moves from Open to Achieved.

What to build

a public set of tables and maps showing, for 20 popular coastal and offshore routes, how often in each month over the past ten years the wind and waves exceeded the limits of each EU design category (for example, category C's force 6 and 2 m).

Why it is still open

The categories are defined only by wind force and wave height, and a builder's choice tells an owner which water the boat suits. Passage planning tools show the weather but not the categories: PredictWind's Climate Data, for example, shows ten-year monthly averages and wind and wave roses from the ECMWF ERA5 reanalysis (a reconstruction of past weather from observations and a weather model). A search in November 2026 found no published count of how often given routes exceed the category limits.

help.predictwind.comlegislation.gov.uk

Built with what exists

ERA5 hourly wind and wave data from the Copernicus Climate Data Store; Copernicus Marine wave products; the World Meteorological Organization's Beaufort-to-wind-speed table; Python with xarray; a small web map.

  1. First step

    download ERA5 hourly wind and wave height for the English Channel for 2015 to 2024 and count the hours above the category C limits, month by month.

  2. Done looks like

    exceedance tables by month and category for 20 named routes, with ERA5 waves checked against at least three wave buoys on those routes and the differences reported.