You are here

Development of operational monitoring sensors for the Veolia racing monohull

News International-French

15 Mar 2011

The use of composite materials has been significantly improving racing sailboats performances for more than fifteen years. Hulls, masts and appendages are now made of carbon prepreg materials that consist of Nomex honeycomb and often high-modulus fibres to reduce the weight and increase structural strength and stiffness.

(Published on September 2007 – JEC Magazine #35)






Anumber of problems have been observed in conjunction with the improvement of sailboat performance, notably keel instability at high speed and the poor resistance of core material to slamming. This article reports on three series of measurements carried out during sailing to:


  • analyse the keel’s dynamic behaviour in order to make the boats more reliable,
  • monitor rudder lift in real time in order to adjust the rudder blade’s relative angle of attack,
  • quantify the strain rate of the sandwich core for a more accurate characterization of the materials.


The keel’s dynamic behaviour

A number of racing sailboats have a carbon keel spar. The choice of carbon saves considerably on weight, but gives rise to vibration resonance linked to the coupling of the hydrodynamic forces and the keel’s twist/bend deformations. When the boat sails at high speed for the first time, the bulb end of the keel starts to yaw violently. This phenomenon is called “flutter” in the aviation sector. Here, it is due to the coupling of the keel’s fundamental twist and bend modes of vibration when the two modes are too close because of incorrect design stiffness and an insufficient number of ±45° plies used with respect to the keel length. Those plies are needed to give the keel stiffness in torsion. After reinforcement, strain gauges were bonded to the keel to measure its twist/bend deformations during sailing. An example of a result for fifteen seconds of recording is shown below (Fig.1).



The first curve corresponds to bending strain, which occurs mainly when the boat sails through a swell. This gives a highly variable curve. The ratio shown is the bending deformation ratio between values recorded during sailing and those obtained during a regulatory test that consists in heeling the keel to one side at a 40° angle while the boat is motionless, making the boat at a maximum 10° angle.


The second curve shows the twist deformation caused by bending load. The yawing caused by the sailboat’s trajectory makes the keel vibrate at its fundamental twist mode. A frequency of about 5 Hz appears each time the boat yaws significantly, and is rapidly damped.


Adjusting the rudder blades



Strain gauges are bonded to the carbon-epoxy structural main frame embedded in the rudder blades, and wired in a full bridge configuration to the boat. The signal is recorded by an HBM Spider 8 data acquisition system. Then, the sensors measure the bending moment due to hydrodynamical lift on each side. Sample data are shown in the figure below. (Fig.2) The recordings show a lot of scatter, due to the difficulty in controlling the boat in the swell. However, with this type of recording, it is possible to reach the average lift in a short time. The lift was smoothed in this way for various adjustments to the rudder parallelism (X-coordinate in the figure below). On the left side, there is too much lift of the starboard rudder blade, producing an opposite lift on the port-side rudder blade. In the middle, the adjustment ensures that the starboard rudder blade will carry the total lift, while the one on the port side is set at zero. An extreme point appears on the right side of the graph where the wrong rudder blade undergoes the lift. The appropriate adjustment was chosen, helping to reduce drag.



Shear deformation in the core of a sandwich hull



The local deformations to the sailboat hull’s sandwich core were measured using two sensors placed in the bow, where the hull is subject to slamming. Each sensor consists of pieces of foam, cut and reassembled using vacuum bonding to form a cylinder where one or several planes are sensitive to deformation. Bearing in mind the expected type of stressing on a sandwich core, the bonding directions and type of wiring used for the gauges were chosen in order to measure the shear strain while eliminating the effects of other stresses and temperature. These sensors were inserted into the shear web at the place of measurement, and a test according to the NF T 54-605 standard was conducted to assess their sensitivity.


More Information
The boat: The skipper:
Length: 18.28 m Roland Jourdain
Displacement: 9.5 T 1st: Route du Rhum 2006 St Malo-Pte à Pitre
Height from deck: 26 m 1st: Calais Round Britain Race 2005
Sail area upwind: 290 sq m 2nd: Transat Jacques Vabre 2005
Sail area downwind: 600 sq m 2nd: SNSM 2006 speed record
Draught: 4.50 m 3rd: Vendée Globe 2000
Materials: carbon-Nomex sandwich World Champion: IMOCA 2000-2001-2002
Speed: 25-30 kts World Champion: FICO 2002


During the sea trials, the boat travelled at a speed of 8 knots and was subjected to the impact of waves one to two metres high – relatively mild conditions for the type of boat.



The shear strain values are relative to the zero of the data channels used in the measurements carried out on land. The load level remains negligible in relation to the foam’s maximum shear stress. However, nothing indicates that the sensors were where the greatest shear deformation occurred.


The boat entering into contact with the waves produces shear strain peaks with a period of about 2.5 seconds. The waves have different shapes, which causes random maximum deformation peaks as shown by the readings below (Fig.5): The measurements also serve to locate the impact zones. In Figure a, the front sensor undergoes peak load before the rear sensor does. In Figure b, the opposite is true, and the load is lower than in scenario a. In Figure c, the type of stressing corresponds to a local impact in the bow, eliciting a random response from the sensors. The level of peak deformation is low during this measuring period, although the shear strain rate is 118 %/s. These results indicate that the mechanical characterization of the core materials used for structures of this kind should be done at high speed, whereas currently, only quasi-static tests are conducted.



The measurements at sea provided crucial information that served to make the boat more reliable and refine its settings. The information also suggests that a next step for future projects would be a characterization of the materials that is consistent with the strain rates encountered during sailing. Based on the results obtained and the analysis of them, the objective is now to instrument the boat in order to sail in the correct stress range, i.e. measure the main stress flows in order to be able to adapt sailing manoeuvres to limit extreme loads on the structure in dynamic and fatigue conditions.