The detection and analysis of wave slamming from full-scale ship measurements.

• Main Author: Saunders, Clinton Frederick Wood
• Other Authors: Bekker, Annie, Muiyser, Jacques, Stellenbosch University. Faculty of Engineering. Dept. of Mechanical and Mechatronic Engineering.
• Format: Thesis
• Language: English
• Published: Stellenbosch : Stellenbosch University 2018
• Subjects: Hydrodynamics; Vibration > Measurement; Deviation (Mathematics); Wavelet analysis; UCTD; Full-scale ship measurements; Southern Ocean; Maak
• Online Access: http://hdl.handle.net/10019.1/105084

Thesis (MEng)--Stellenbosch University, 2018. ENGLISH ABSTRACT: Bow and stern wave slamming has proven to be a persisting and concerning problem on the S.A. Agulhas II. Various full-scale vibration measurements have been conducted over several voyages in the Southern Ocean. Past research has concluded that vibration has reached levels where damage was possible in the stern and is probable in the bow during open water navigation (Soal, 2014). In order to better understand the drivers behind slamming, an algorithm was developed to detect and count slamming events in full-scale signals. Detection was done in the time domain using acceleration peaks, as well as the frequency domain using a wavelet transform. Statistical thresholds were then determined using a median absolute deviation. The data in the frequency domain provedmost accurate for counting slams. Slamming counts were then compared with environmental and operational variables for the voyages. It was determined that these variables have a significant effect on the frequency of slamming events. AFRIKAANSE OPSOMMING: Skeepslag op die boeg en agterstewe van ’n skip is ’n aanhoudende en kommerwekkende probleem op die S.A. Agulhas II. Verskeie volskaalse vibrasie metings is al geneem op die skip tydens verskeie reise in die Suidelike Osean. Vorige navorsing het bewys dat die vibrasie wat ervaar word skade kan aanrig in die agterstewe en waarskynlik in die boeg wanneer die skip op die oop see vaar (Soal, 2014). Om beter te verstaan wat skeepslag veroorsaak, is ’n algoritme ontwikkel om branderklap in die vibrasie sein te identifiseer en te tel. Skeepslag is geindentifiseer in die tyd domein deur middel van versnellingspieke, asook in die frekwensie domein deur middel van golfomskakeling. Statistiese drempels is dan bepaal deur van ’n mediaan absolute afwyking gebruik te maak. Daar is gevind dat die data in die frekwensie domein meer akuraat is om skeepslag gevalle te identifiseer en te tel. Skeepslag is ook vergelyk met omgewingsfaktore sowel as operasionele ...