Model Test and Full Scale Measurements of Whipping on Container Vessels in the North Atlantic

Abstract

Ships vibrate due to waves, and these wave induced vibrations can not easily be avoided by moderate changes to the hull lines. The waves may cause the whole hull girder to vibrate due to springing (resonance) and whipping (transient response), which increase the fatigue and extreme loading. Recently this has also become an industry concern. Modern hull monitoring systems in combination with model tests are the best tools to answer the key questions: How important is the wave induced vibrations, and does it have to be included in design? This paper addresses the effect of whipping on the extreme loading. Measurements have been carried out on two container vessels operating in the North Atlantic. An elastic model of the larger vessel has also been tested. Results are obtained at quarter lengths and amidships. From the measurements the increase due to whipping is considerable, even though the wave conditions are not extreme. The full scale measurements and model test show that IACS URS11 rule loads may be exceeded in less than extreme sea states, in particularly amidships and in the aft ship. The IACS UR S11 may need revision for container ship design. MAIB’s report based on the investigation of the MSC Napoli incident (vessel broke in two) also recommends increased requirements for container ship design and further research into the effect of whipping.

Sea state identification using machine learning—A comparative study based on in-service data from a container vessel

“…2. Additional details of the measurement campaign including a study pertaining to structural fatigue due to wave excitations, are described in Storhaug et al [24].…”

“…systematic) uncertainty within the data may be influential on the following machine learning approach. In the paper of Storhaug et al [24], it is stated that the typical transit draft is approximately 9.5 m and the transit trim lies between 0.5 to 1.0 m. In addition, the loading condition, and thus the transverse metacentric height 𝐺𝑀 𝑇 , is not included in the dataset adding further uncertainty due to the widespread operational profile of container vessels, in general.…”

“…(4) Moreover, four strain gauges are attached to longitudinal stiffeners at port and starboard in the aft section (50.3 m from AP) and amidships (118.7 m from AP). As described in [24], the port and starboard strain sensors are aggregated into two artificial or virtual sensors indicating vertical bending and axial stress in the aft and midship section. Lastly, (5) the propeller revolutions (rpm), (6) the rudder angle 𝛿, (7) the GPS position, as well as (8) the Speed Over Ground (SOG) were obtained during data acquisition period.…”

“…1. Case study in service taken from [24]. and frequency discretization will be determined in the frequency domain in another sensitivity study.…”

“…systematic) measurement uncertainty may be responsible for the minor deviation of both joint distributions. On the other hand, it is stated in [24] that the calibration of the Wavex wave radar is based on earlier studies and has not been performed for this particular vessel. Even in case of a slightly biased wave radar, it has no direct impact on the machine learning methodology itself, as the offset remained the same throughout the measurement campaign.…”