AN EXPERIMENTAL APPROACH TO MEASURE PROPELLER CAVITATION NOISE ON- BOARD A RESEARCH VESSEL

Transcription

1 AN EXPERIMENTAL APPROACH TO MEASURE PROPELLER CAVITATION NOISE ON- BOARD A RESEARCH VESSEL Spyros Leivadaros, Newcastle University, United Kingdom Batuhan Aktas, Newcastle University, United Kingdom Mehmet Atlar, Newcastle University, United Kingdom Patrick Fitzsimmons, Newcastle University, United Kingdom Noriyuki Sasaki, Newcastle University, United Kingdom The quest towards achieving sustainable marine transport has recently targeted the contribution of the shipping to underwater radiated noise. Commercial shipping has experienced a rising trend both in terms of the number of vessels sailing through the world seas as well as the gross tonnage. This trend over the last half century resulted in elevation of the ambient noise levels of the world oceans with an average of 3 db per decade. Furthermore underwater radiated noise dominates the low frequency ambient noise levels which may overlap with the frequency region of communication for some of the marine mammals causing various forms of behavioural disturbance or even mortality. Considering that noise emission trials in big merchant vessels could constitute serious economical and time loss constrains, this project aims at an experimental audio-visual study on collapsing cavities of a propeller cavitation in full-scale. By means of hydrophones (Bruel and Kjaer, type 8105 and 8103) and a high definition, wide angle camera (GoPro Hero 3, black edition) appropriately fitted on a customized submersed device, called Hydropod ; noise emission and tip-vortex cavitation are studied on a research catamaran. By taking advantage of the propeller replacement occasion for research vessel, The Princess Royal, the Hydropod was used to measure the acoustic performance and to observe cavitation development of her propellers in full-scale. This has enabled to make comparison between the old and new propellers of the vessel in terms of acoustic performance and cavitation development. 1. Introduction It is widely accepted that the recent years, manmade underwater noise has significantly increased in proportion with maritime transportation, oil exploration, seismic surveys or military sonar operation globally. This subsea radiated sound disturbs some marine mammals and interferes their communication frequencies causing various forms of behavioural disturbance, irreversible injury or even mortality [1]. The marine species use low frequency sound band for hunting, mating, migration, 1

2 communication and navigation over great distances. Unfortunately the low frequency ambient noise levels are dominated by the contribution from the commercial shipping [2]. Furthermore the absorption rate of the sea water as a fluid medium is very low, which results in the contamination of large masses of water with radiated noise levels from a single vessel [1]. Although there are numerous noise sources present on board of a commercial vessel, the cavitation dominates the overall under water radiated noise levels once it incepts [3]. Hence, there is an increasing interest to cavitation induced noise in recent years with a plethora of technical reports and papers covering the issue especially focussing on the cavitation phenomena from commercial shipping. Whilst the issue is clear, the assessment procedure for commercial shipping is cumbersome and costly. The amount of time required for conducting the ANSI standard survey method (i.e. least precise method) underwater noise trials can constitute significant amount of time for a vessel off the trade. Whilst there are various standards available for conducting full scale noise trials such as ANSI, ISO and STANAG [4] [6], the inherent nature of the underwater noise trials require change of course from the trade route and sailing to the trial site as well as time consuming trial preparation. In case of a potential enforcement of limits to the radiated noise levels of a commercial vessel, current standards for noise survey may be deemed to be too costly in terms of economic losses. Within this framework this study presents an experimental method for underwater noise assessment of a vessel by means of a practical measurement device deployed and attached on-board of the vessel. The developed and engineered equipment can be named as hydropod and referred with this name here onwards. Hydropod measurements ensure cost effective and accurate noise assessment whilst requiring minimum amount of time off the trade, less change of course of the route and significantly reduced preparation requirement. Furthermore the introduced method also includes an underwater camera system enabling the observation of the cavitation experienced hence enhancing the understanding of the noise creating mechanisms. Following this introductory section, the hydropod and other equipment used during the sea trials are described in Section 2 of the paper. The details of the cavitation observations with the old and new propellers are presented in Section 3 while those of the underwater noise measurements are given in Section 4. The results of the measurements are discussed in Section 5 and main conclusions are presented in Section 6. 2

3 2. Equipment and trials For the sea trials, Newcastle University s research catamaran RV Princess Royal was used (Figure 1). During the first set of trials, an omnidirectional hydrophone (Bruel and Kjaer, type 8105) and a high definition, wide angle camera (Gopro Hero 3, black edition) were used. While for the second set, the omnidirectional one was replaced by a miniature hydrophone due to technical malfunctions (Bruel and Kjaer, type 8103). Thus, cavitation inception and development were observed visually and acoustically. To achieve improved results, it was wiser to immerse this gear, using a hydro-dynamically shaped cowling in order to mitigate drag and underwater self-noise. A symmetric, vertical glass fibre hydrofoil was designed, constructed and named Hydropod. At the upper end it was properly mounted on vessel s hand rail, while at lower end it was shaped into a convenient pod on which the camera and hydrophone were accommodated (Figure 2). Ultimately, the full scale trials were held at North Sea by the beginning and the middle of summer It has to be taken into consideration that the two hydrophones, which were used for the recordings, have not got identical characteristics. In detail, their value of vertical directivity or uncertainty in the axial plane has a 2 db difference: for the omnidirectional 8105 is ± 2 db at 100 Hz while for the miniature 8103 is ± 4 db at 100 Hz. Fig. 1 Newcastle University research vessel The Princess Royal 3

4 Fig. 2 - Hydrophone, camera and LED lighting attached to the Hydropod. The method employed for the sound measurements is for a prototype; hence it cannot be entirely subjected to regulations and formats used in various noise trials standards. Even though, several other full scale noise trial standards are present, ANSI recommendations were adopted in order to ensure experiment s reliability [4]. Once the acoustic signals are collected in their raw format, analysis had to be carried out using Fast Fourier Transform (FFT) and Constant Percentage Bandwidth (CPB). Therefore, the data were converted from time domain files to frequency domain. Bruel & Kjaer (B&K) PULSE Type 3023 data acquisition system with a 6/1 LAN interface through a dedicated software were implemented for that process. Certain guidelines were used in both trial runs (which were conducted before and after dry docking to swap the old propeller with the new set) as per ITTC s speed and power trials procedure (ITTC, 2012). Particularly, the double run format was adopted with 5 minute (instead of 10) increments between every differentiation in propellers rpm. This amendment took place for two reasons: i) the battery life of the camera constituted a major constrain during the trials and ii) 5 minute recording was providing a tolerable period for the sound measurements. Additionally, the minimum water depth below the keel of the vessel was maintained 30 m at all times, in accordance with the formula for minimum water depth without shallow water correction: h = 3 B T and h = 2.75 Vs2 g (1) 4

5 Where: [h] is the water depth in [m], [B] is vessel s breadth in [m], [T] is vessel s draft in [m], [Vs] is vessel s speed in [knots] and [g] is the acceleration due to gravity. Furthermore, the total wave height during runs satisfied the following formula: where: [H] is the total wave height in [m] and H 1.50 Lpp 100 (2) [Lpp] the length between perpendiculars in [m]. Moreover, for the subject vessel (Lpp < 100 m) the sea state has to be below 5 in the Beaufort scale and the amplitude of rudder angle to be lower than 5 degrees while on run. Although Hydropod method is not using an official experimental format, similarities may observed with the grade C measurement set-up described as Survey method (ANSI, 2009). The major difference is that the hydrophone is attached to the subject vessel rather than located beneath a floating buoy anchored away from the vessel under test. Therefore, certain formulas couldn t be applied in this particular case. Background noise and sensitivity adjustments have been performed following the standard sequence recommended by ANSI: where: = L p+n L p (3) is the signal plus noise-to-noise ratio, Lp + n is the sound pressure level in [db] related to the subject vessel 5

6 Ln is the background pressure level in [db] uninfluenced by the vessel under test. Hence, three conditions are experienced: a) when is greater than 10 db no alterations are needed, b) when is less than 3 db the data should be discarded and c) when is between 3 and 10 db and no fluctuations occur in the background noise, adjustments are required using: Where: L p = 10log (10 (L p+n 10 ) 10 (L n/10) ) (4) L p is the modified background noise sound pressure level (SPL) of the subject vessel, measured in 1/3 octave. Furthermore, as far the sensitivity is concerned, variations need to be accomplished as follows: Where: L p = L p + A SEN (5) L p is the sensitivity-corrected sound pressure level and ASEN is the adjustment for any hydrophone sensitivity e.g. amplifier gain, cable sensitivity, directivity etc. For the present experimental work ASEN is 1 db. In addition, the distance of the hydrophone to the acoustic centre i.e. the subject propeller in every run, had to be normalized as described in the equation below: Where: 2 2 d Total = d Horz + d Vert (6) dtotal is the total normalized distance, dhorz is the horizontal distance of the hydrophone from the acoustic centre i.e. the hub of the subject propeller and dvert is the depth of the hydrophone. For these trials, dtotal is always 2, 25 m 6

7 Finally, the subsurface sound source level for every single run is determined by: L s(r) = L p + 20log (d Total d ref ) (7) Where: Ls(r) is the subsurface sound source level at 1 m reference distance, as a function of run number (r). During post-processing, the acoustic data derived from the port and starboard propeller measurements were averaged. This was carried out after the transformation from 1/3 octave band to an equivalent 1 Hz bandwidth sound source level in db relative to 1µPa (db re 1µPa at 1m) from 20 Hz to 20 khz. 3. Cavitation Observations The hydropod enabled the accompanying information on the cavitation experienced by the propeller throughout the measurements. The cavitation observations were most useful to interpret the measured noise levels and understand the reasons behind the differences in the measured levels. Although the subject research vessel operated with a high speed diesel engine the observations were still highly informative and insightful. The shutter imposed a limit as employed camera is a mass market underwater camera which is not a specialized scientific equipment. However the camera used for this prototype will perform much better at conditions that are common for commercial shipping. Water clarity during the 2 nd sea trials was not ideal for video recordings. Nevertheless, filming of propeller slip-stream development was carried out for the entire range of propellers rpm. Comparing the two sets of propellers, it is clearly observed that the new NPT (New Profile Technology) propellers [7] perform better in terms of tip-vortex cavitation progress (figures 3-5). Although sea water s blur, it s visible that for the new ones, well-formed slip-stream does not exist before 1200 engine rpm. While at 1500 rpm, the New Profile Technology based propellers slip stream is quite similar with the old propeller at 1200 rpm. This big difference in cavitation performance, between the two types of propellers, might be caused not only by the technological supremacy (advance blade profiles, higher skew angle etc.) but also by the blade surface condition during the trials. On the first trials, secondary fouling was observed on the hull and propellers reducing the overall performance of the vessel. On the other hand, the second set of trials took place almost right after the dry docking with very clean hull allowing greatly improved roughness characteristics. 7

8 Fig. 3 - Slip-steam comparison between old [A] and new [B] propellers at 600 engine rpm. Fig. 4 - Slip-stream comparison between old [A] and new [B] propellers at 900 engine rpm. 8

9 Fig. 5 - Slip-stream comparison between old [A] and new [B] propellers at 1200 engine rpm. By further visual analysis on the slip-stream signatures, distinct outcomes are exhibited: very weak tip vortex cavitation is witnessed with the old propellers, even at 600 rpm (figure 3). On the upper bladetip section, scattered vortices are visible while no such occurrence is experienced by the new propellers. At 900 rpm, well-developed steady tip-vortex cavitation characterizes the old propellers whereas rather intermittent tip vortex cavitation on the higher blade section of the new propeller is apparent (figure 4). Finally, similar observations are revealed at 1200 rpm where a much stronger and continuous tip vortex from each blade forming the slipstream is generated by the old propellers while on the other hand, less strength of slipstream by much weaker tip-vortices is produced by the new propeller (figure 5). 4. Noise Measurements Firstly, in order to verify if the hydropod technique is a useful on board alternative method to measure underwater noise radiated mainly by the propeller of a ship, comparisons are made with the results of the off board noise measurement made with The Princess Royal by Brooker and Humphrey at another trials conducted in 2013 [8] (referred as SOTON) (figures 6-8). Secondly, ambient and engine noise level are measured in order to identify their contribution to the overall underwater radiated sound (figures 9). Thirdly, evaluations for the old and new sets of propellers, at the entire tested engine rpm range are performed (figures 10-13). 9

10 Fig. 6 - Hydropod and SOTON noise data at 600 engine rpm. Fig. 7 - Hydropod and SOTON noise data at 900 engine rpm. 10

11 Fig. 8 - Hydropod and SOTON noise data at engine 1200 rpm. By studying figures 6-8, it is noticed that minor disagreements encountered in the noise levels between the classical off-board underwater sound measurement techniques and the Hydropod method. Figure 6 is comparison of the only sub cavitating condition. The initial observation is that the trends of the measured spectral levels are not matching and the hydropod measurements are more effected by the background noise levels. This may be attributed to the dominating acoustical sources. Namely the condition is dominated by the dipole and quadrupole sources as there is no cavitation present. Therefore due to the inherent acoustical analogy the nearfield measurement of the hydropod is dominated by these sources, but far field off board measurement is much less affected by these sources due to their higher absorption rate with distance. Figure 7 and 8 are comparisons of the cavitating conditions which shows good agreement except the hump in the mid frequency region. The literature research indicated that the hump may be due to the self-noise of the hydrophone [9]. The hump moves towards the higher spectral levels and lower frequency region with the rising speed of the vessel following the theory proposed by Haddle based on the experimental study carried out with different size and diameter hydrophones that are tested at different speeds. 11

12 The net noise levels are achieved by the extraction of the background noise levels as given in equation 4. Furthermore while doing the background noise measurements declutched engine noise measurements are also made as shown in figure 9. Although in the declutched condition there is no load on the engine this measurements given further insight to the contribution of the engine over the spectra. In addition to that, comparison of the background levels with the 600 RPM and 900 RPM visualizes the significant elevation of the spectral levels with the inception of the cavitation. The only remark that has been left unanswered after the execution of the background levels was the peak at 50 Hz in the background noise measurements which resulted in the discontinuity of the measured data after the background noise correction. The detailed investigation carried out leaved the electric generator as the only suspect but this cannot be confirmed as it required measurement without the electric source which was not possible during the test schedule Fig. 9 - Ambient and declutched engine rpm noise levels are demonstrated. Great divergence is noticed when tip-vortex cavitation is developed. Following the above described measurement campaign the vessel was dry docked in June 2015 and her propellers were replaced with the New Profile Technology (NPT) propellers and the similar trials were repeated with the new propellers. This enabled the comparison of the performance of the old and new propeller in terms of the cavitation and noise performance. The full scale trials were repeated at identical conditions except two differences. The first one was the hull condition. While the hull was 12

13 fouled when the trials conducted with the old propellers, this was in the contrary during 2 nd trials with very clean hull. Second was an unfortunate malfunctioning of the B&K 8105 hydrophone during the pre-test calibration. This issue was rectified by swapping it with B&K 8103 hydrophone in the hydropod. As a drawback of this substitution; the self-noise of the hydrophones were different showing humps at different frequencies due to their significant size difference. This is further elaborated and investigated through Figure 10 to 12 Fig Comparison between old and new propellers at 600 engine rpm, using the Hydropod. 13

14 Fig Comparison between old and new propellers at 900 engine rpm, using the Hydropod. Fig Comparison between old and new propellers at 1200 engine rpm, using the Hydropod 14

15 Fig Comparison between old and new propellers at 1500 engine rpm, using two different techniques (New propeller; Hydropod, Old propeller; SOTON data) It is noted that at non-cavitating speeds the old and new propellers operate quite similarly in terms of their noise levels and overall characteristics (figure 10). On the other hand, when the cavitation is incepted and develops further, significant divergence of the noise spectra of the two sets of propellers are spotted. In detail, acoustic results of the old propeller demonstrate poor noise performance at cavitating conditions, compared with the NPT propellers. At 900 and 1200 engine rpm, in the entire frequency range, there is a clear precedence for the recently installed propellers (figure 11-12). The comparisons of the cavitating conditions has shown up to 30 db difference between the old and new propeller indicating potential to improve the propeller design for better noise performance. Overall the measurements have clearly indicated the improvements in the noise characteristics of the vessel with the new propeller and also shown promising results for the hydropod to be a useful and practical technique in comparison to the classical off board measurement technique which are more cumbersome and costly. The accuracy and measurement quality can be further enhanced by carrying out towing tank or cavitation tunnel tests in order to develop transfer functions for the hydrophone selfnoise. 15

16 5. Discussion The results obtained using the Hydropod are proved to be consistent and thus, the method is claimed to be reliable for future noise recordings on board merchant or navy vessels. More practical, flexible and cheaper compared to classic three-hydrophone array, the Hydropod provides similar acoustic values coupled with relevant cavitation observations. Hence, classification societies may review and study this technique for upcoming amendments to existing standards. Furthermore, it is a self-accomplished method without requiring any auxiliary vessels or shore facilities because all recording and analysing equipment is accommodated within the subject vessel. In terms of tip-vortex cavitaton inception and development, it is apparent that the new propellers of the vessel establish a superior difference compared to the old ones. The slipstream expansion and growth of the old propellers are characterized by much larger in thickness and extension related to the NPT propellers tip-vortex. As far as the noise characteristics are concerned, overall it might be said that the NPT propellers appear to be much quieter compared to the old propeller. At non-cavitating speeds, where singing may be present, this difference is not great, but when cavitation is incepted and developed further, clear divergent results are obtained. Moreover, ambient and engine noise contribution appear to have an important part in the overall underwater sound emission, but propeller noise in cavitating speeds dominates the entire frequency spectrum. 6. Conclusion In this paper a practical on-board device, named as Hydropod, has been proposed to measure the underwater radiated noise of a ship propeller and observe its cavitation as well as to demonstrate the functionality of the hydropod on board a research catamaran. The device is used to assess the underwater noise and cavitation assessment of two different type propeller on the research vessel. Based on the study presented it can be concluded that: Hydropod based on-board technique can be a practical and attractive alternative to classical off board noise measurement technique for ball-park noise and cavitation assessment 16

17 Hydropod device was able to measure the underwater noise spectra of the research vessel with comparable sound pressure level and frequency distribution of the noise spectra obtained by the off-board measurement. The device was able to clearly distinguish the different noise and cavitation characteristics of the two different types of propellers used on the research vessel in favour of the new NPT based propellers. Hydropod can be an attractive cavitation observation device depending on lighting conditions Hydropod device can be further developed to apply other commercial vessels and its selfinduced noise can be filtered based on further investigations 7. References 1. J. A. Hildebrand, (2005) Impacts of anthropogenic sound, Marine mammal research: conservation beyond Crisis, vol. 124, pp G. M. Wenz, (1962) Acoustic Ambient Noise in the Ocean: Spectra and Sources, The Journal of the Acoustical Society of America, vol. 34, p D. Ross, (1987)"Mechanics of Underwater Noise". California, USA: Peninsula Publishing. 4. ANSI, (2009) Quantities and Procedures for Description and Measurement of Underwater Sound from Ships, Part 1: General Requirements, vol. American N. 5. ISO, (2012) Acoustics Quantities and Procedures for Description and Measurement of Underwater Sound From Ships, NATO, (1995) Standards For Use When Measuring And Reporting Radiated Noise Characteristics Of Surface Ships, Submarines, Helicopters, Etc. In Relation to SONAR Detection and Torpedo Acqusisition Risk, North Atlantic Treaty Organization, vol. STANAG N. Sasaki and G. Patience, (2005) Evolution of high efficiency propeller with new blade section, in Motor ship Conference. 8. A. Brooker and V. Humphrey, (2014) Measurement of Radiated Underwater Noise from a Small Research Vessel in Shallow Water, in A. Yücel Odabaşı Colloquium Series 1 st International Meeting - Propeller Noise & Vibration 6th 7th November 2014, Istanbul, Turkey. 9. G. P. Haddle, (1969) The Physics of Flow Noise, The Journal of the Acoustical Society of America, vol. 46, no. 1B, p