Survey of HMVS Cerberus, December 2012 of HMVS Cerberus, December 2012 Roger Neill, John Gilbert, Peter Graham, Peter Mart, Clare Grandison, Martin Rowan and …

UNCLASSIFIED FOR OFFICIAL USE ONLY

FOR OFFICIAL USE ONLY UNCLASSIFIED

Survey of HMVS Cerberus, December 2012

Roger Neill, John Gilbert, Peter Graham, Peter Mart, Clare Grandison, Martin Rowan and Neal Winter

Maritime Platforms Division

Defence Science and Technology Organisation

DSTO-CR-2013-0092 Client Bayside City Council Authorised by: Stuart Cannon (RLSPS) Date: 4/4/2013 Approved for Release: Janis Cocking (Chief MPD) Date: 4/4/2013

ANNOUNCEMENT LIMITATION The metadata of this document may be publically announced.

RELEASE LIMITATION

This report provides immediate documentation of results of work performed in Maritime Platforms Division. The report is written for the information of Bayside City Council and its duly appointed consultants. Secondary distribution within DSTO is subject to approval by Chief of Division: other requests must be referred to the client. This report may not be cited in the open literature without the express permission of the Chief of Division.

UNCLASSIFIED

UNCLASSIFIED

Produced by Maritime Platforms Division DSTO Defence Science and Technology Organisation 506 Lorimer St Fishermans Bend, Victoria 3207 Australia Telephone: (03) 9626 7000 Fax: (03) 9626 7999 © Commonwealth of Australia 2012 December 2012 Conditions of Release and Disposal This document is the property of the Australian Government; the information it contains is released for defence purposes only and must not be disseminated beyond the stated distribution without prior approval of the Releasing Authority. The document and the information it contains must be handled in accordance with security regulations, downgrading and delimitation is permitted only with the specific approval of the Releasing Authority. This information may be subject to privately owned rights. The officer in possession of this document is responsible for its safe custody.




Executive Summary DSTO was approached by the Bayside City Council to undertake a survey of the historic naval vessel HMVS Cerberus, a breastwork monitor purchased by the Colony of Victoria in the 19th century for protection of Port Philip Bay. The vessel was scuttled to form a breakwater in Half Moon Bay, Black Rock in 1926. When scuttled the ship’s armour belt and breastwork remained above the waterline, but during a storm in 1993 the hull collapsed, with the result that in large measure only the breastwork and turrets remain above the waterline. The condition of the wreck continues to degrade, to the point there are concerns the structure supporting the 175 tonne turrets may collapse without notice. The approach to DSTO was made on the basis that it would be unsafe for divers to enter the vessel, particularly in the vicinity of the turrets. Bayside requested that DSTO use a Remotely Operated Vehicle (ROV) to undertake a survey of the wreck. Of particular interest was information that could be provided, either in the form of video imagery or direct physical measurement, on the condition of the beams that support the turrets. In addition DSTO was requested to provide information regarding the amount of debris that underlies the two turrets. It was agreed that the following activities would be undertaken:

• An attempt would be made to record video imagery of the support beams underlying the two turrets;

• An attempt would be made to measure the thickness of these beams using an ultrasonic thickness gauge;

• Assessments would be made of the nature of debris inside the ship, particularly in the vicinity of the turrets;

• A general survey would be undertaken external to the ship; • A general, environmental assessment would be made of the site; • If possible Corrosion Potential measurements would be made of the wreck.

These would give some measure of the effectiveness of the currently installed, passive cathodic protection system.

The survey was scheduled for the week 3-7 December 2012. In the event bad weather caused operations to be cancelled on the 4th and 5th, but three day’s work was undertaken. The loss of two days impacted the range of operations that could be undertaken in so far as, where obstacles were found inside the wreck, there was insufficient time to progressively work a passage around these obstacles.



The general survey of the vessel’s exterior confirmed that the armour belt surrounding the upper part of the hull is intact. When the hull collapsed, the ship settled reasonably evenly (with a final list in the order of only 4~5 degrees). There is considerable evidence of the ‘concertina effect’ which occurred during this collapse, particularly on the port side of the ship. Given the condition of the armour belt precluded any attempt to enter the vessel directly at lower deck level, attempts were made to insert the ROV via four different hatch openings. These surveys showed that:

1. There is a great deal of debris in the vicinity of the forward turret. This prevented the ROV from being used to directly visualise the under-turret beams;

2. The bulb-iron beams which are forward and aft of the turret appear to be intact, with heavy concretion over their surfaces. The shape of both the bulb at the bottom of the beams and of the riveted, angle iron webs at top could be defined underneath the concretion;

3. Access to the aft turret proved to be possible from the Aft Breastwork deck companionway, thus it was possible to image the first ‘I’ beam that supports this turret. This beam shows evidence of a break at one point, perhaps where one of the supporting pillars has broken away from it during the collapse, but otherwise it is intact, with webbing appearing to be substantially complete. It proved impossible to use an ultrasonic thickness gauge to measure the thickness of this beam;

4. The lower deck is substantially silted up in this area, so it was not possible to pass under the first ‘I’ beam;

5. Approaching the aft turret from the stern, there is a reasonable amount of debris, although much less than under the forward turret. It was judged that, but for the loss of two days due to bad weather, it would be possible to approach the aft turret from this direction;

6. Imaging of the bulb-iron beams before and abaft the aft turret revealed similar results to those for the forward turret: the beams are concreted but appear to be substantially intact;

7. Environmental assessments of the site produced no evidence of biological populations of any significance;

8. Corrosion Potential measurements were consistent with recordings made by other researchers. The cathodic protection provided for the wreck appears to be working reasonably effectively.

It is suggested that experts be consulted to determine whether measurements of the thickness of bulb-iron beams towards the aft end of the ship would be representative of the probable thickness of the under-turret ‘I’ beams. Divers may be able to enter this part of the ship safely, hence enabling direct physical measurements to be undertaken. While it was not possible to make direct measurements of the residual thickness of the turret support beams, the nature of the site, and the general condition of the vessel is better understood, which should assist relevant authorities in planning the next step in the long-term management of HMVS Cerberus.

Contents

1. INTRODUCTION............................................................................................................... 1

2. PRE-TRIAL PREPARATORY ACTIVITIES .................................................................. 2 2.1 Definition of the Scope of the Survey .................................................................. 2 2.2 Reconfiguration of the LBV ROV.......................................................................... 3 2.3 Diver Rehearsals ....................................................................................................... 6 2.4 Regulatory Compliance of the Survey and Assessment of Ecological

Considerations........................................................................................................... 6

3. DETAILS OF PLANNED SURVEY ACTIVITIES ........................................................ 7 3.1 General Approach to Undertaking the ROV Internal Surveys........................ 7

4. RESULTS ............................................................................................................................ 10 4.1 Day-by-Day Sequence of Events.......................................................................... 10

4.1.1 Monday 3rd December 2012.................................................................. 10 4.1.2 Tuesday 4th and Wednesday 5th December 2012............................... 11 4.1.3 Thursday 6th December 2012................................................................ 12 4.1.4 Friday 7th December 2012 ..................................................................... 13

4.2 External Survey ....................................................................................................... 13 4.2.1 Observations of the Environment in the Vicinity of Cerberus ......... 13 4.2.2 External Video Survey of the Wreck................................................... 17

4.3 ROV Survey through Forward hatch .................................................................. 25 4.4 ROV Survey through Aft hatch............................................................................ 28

4.4.1 Observed Deck Beams in the Aft Part of the Ship ............................ 41 4.5 ROV Survey through Boiler Room (Stoke Hold) Outlet................................. 45 4.6 ROV Entry Via Aft Breastwork Deck Companionway ................................... 63 4.7 Interpretation of the Ecology of the Site ............................................................ 74 4.8 Corrosion Potential Measurements ..................................................................... 74

4.8.1 Results ..................................................................................................... 76 4.9 Miscellaneous Imagery.......................................................................................... 78

5. DISCUSSION .................................................................................................................... 88

REFERENCES 92

ACKNOWLEDGEMENTS .................................................................................................... 93

APPENDIX A: LETTER OF REQUEST FROM BAYSIDE CITY COUNCIL............ 95

APPENDIX B: CHIEF MPD RESPONSE TO LETTER OF REQUEST FROM BAYSIDE CITY COUNCIL..................................................................... 98

APPENDIX C: PERMIT TO UNDERTAKE SURVEY OF HMVS CERBERUS ..... 100

1

1. Introduction

HMVS Cerberus (hereafter alternatively called ‘Cerberus’), a decommissioned Harbour Defence Monitor, was launched in 1868 and served for 53 years as part of the Victorian Colonial Navy and the Royal Australian Navy. Having been decommissioned, in 1924 it was sold for scrap. The hulk, including turrets and four cannon, was sold to Sandringham Council (now Bayside City Council) and, in 1926, was sunk as a breakwater off Half Moon Bay, Black Rock. The HMVS Cerberus wreck falls under the ownership responsibility of Bayside City Council (hereafter called ‘Bayside’) and Heritage Victoria has jurisdiction with respect to controlling access to the HMVS Cerberus Prohibited Zone. In 1997 the National Trust classified the vessel as being of national significance, and subsequently the Cerberus was the first vessel to become a registered historic place in the National Heritage List. During a storm on 27th December 1993 the lower part of the ship’s hull substantially collapsed under the combined weight of its eight inch armour belt, the breastworks and the turrets. This was followed by another significant movement in 19991. The condition of the vessel continues to degrade. The intact part of the ship’s hull substantially comprises the armour belt (now submerged), the weather deck (which is substantially underwater), the breastwork and breastwork deck. The turrets are currently still in place. They are supported upon a series of ‘I’ beams underlying the weather deck. Despite the fact the turrets have had their cannon removed, they weigh somewhere in the vicinity of 175 tonnes each [BMT, 2011]. Bayside commissioned BMT Design and Technology (‘BMT’) to propose concepts for the preservation of the vessel. BMT identified the need to provide support for the turrets as the most pressing issue [BMT, 2011]. Bayside determined they could not proceed towards commissioning a detailed engineering design for a support system until more data could be made available to potential tenderers. In particular, more detail was required regarding the internal state of the vessel. Given the unknown state of the turret supports, Bayside determined that sending divers inside the vessel to undertake a survey represented an unacceptable risk. Hence enquiries were made to DSTO regarding the possibility of using a Remotely Operated Vehicle (ROV) to undertake an internal survey. This resulted in a formal request being made for assistance from DSTO. The letter of request is included as Appendix A. DSTO agreed to provide limited support to this survey. In particular it was determined that survey infrastructure and personnel would be provided, but that DSTO would not use the survey results to undertake detailed structural analysis. DSTO’s letter of response to Bayside’s request is provided at Appendix B.

1 It was beyond the remit of this report to review the relative severity of the two collapse events. In the following text references to the 1993 collapse take into account additional movement which has subsequently taken place.

2

The following report provides: details of the preparatory activities that were undertaken prior to the survey; a description of the survey itself; and a summary of the products that were generated as a result of the survey.

2. Pre-Trial Preparatory Activities

2.1 Definition of the Scope of the Survey

Once it had been agreed that DSTO would support this survey, a meeting was arranged between DSTO and various organisations with significant interest or potential interest in the survey. These included representatives of Bayside City Council, Heritage Victoria, the National Trust (Victoria), the ‘Friends of the Cerberus’ community group and the engineering firm BMT Design and Technology. The purpose of the meeting, held at DSTO Melbourne on 10th October 2012, was to gain a clear understanding of the purpose of the survey and of what DSTO could realistically aim to deliver. At the meeting Bayside explained that it is attempting to address two, almost competing, considerations: (1) it must protect the safety of members of the public who enter the wreck, even though they do so illegally and (2) must preserve the heritage value of the asset, which could be compromised by certain approaches to securing the wreck. Both Bayside and BMT made the point that, if it is practical to do so, providing a support system that is internal to the vessel may have the least detrimental effect upon the heritage value of Cerberus. Given the progressive degradation of the vessel since the 1993 collapse, little is known of the vessel’s current internal condition. At the meeting it was agreed that DSTO would attempt to redress this lack of knowledge by using its LBV ROV (‘Little Benthic Vehicle’ remotely operated vehicle) to undertake a video-based survey of the wreck, particularly in the parts of the vessel which underlie of the two turrets. The following were agreed outcomes of the meeting:

1. This was seen to be a preliminary activity which would inform the longer-term planning for the conservation of the wreck;

2. DSTO could only commit to one week’s survey activity. While every effort would be made to select a week in which the weather window was predicted to be favourable, there would be no option to shift the trial if unfavourable conditions arose;

3. The main deliverable of the survey would be video imagery. DSTO would provide information about the locations on the wreck where the imagery was recorded, but no structural interpretation would be undertaken, nor advice given;

4. It was understood that, if possible, particular emphasis would be placed upon inspection of the pillars and beams supporting the turrets, and any debris in the vicinity;

5. An attempt would be made to make measurements of the residual thickness of the turret support beams using an Ultrasonic Thickness Gauge mounted on the ROV. It

3

was pointed out to the meeting that DSTO’s previous experience with this gauge, on the AE2 submarine wreck, did not generate a great deal of confidence in a favourable outcome being achievable from these measurements.

6. It was suggested that an approach be made to the local Royal Australian Navy Reserve dive team, seeking its support in undertaking this survey. Specifically, the dive team would be requested to provide general support for insertion, removal and tether management of the ROV and to provide a video-based survey of the external condition of Cerberus and her surroundings.

7. It was also agreed that an approach would be made to MPD’s Environmental Research and Biotechnology Group and Corrosion Group for limited support in relation to their respective fields of expertise.

Subsequent to the meeting an approach was made to LCDR Rafael Fabris, the officer in charge of the local Australian navy reserve dive team (ANRDT Six) seeking diver support for the survey. This was subsequently granted. 2.2 Reconfiguration of the LBV ROV

For the purposes of this survey, it was decided to install an external, high definition video camera (GoPro ‘Hero’ camera in a waterproof housing) and a Cygnus MiniROV Ultrasonic Thickness Gauge onto the LBV. This vehicle has an external ‘bumper’ frame upon which external instrumentation is generally attached. The normal practice is to install such hardware underneath the main body of the vehicle, but because the primary focus of this survey was upon the turret support beams, which would be overhead relative to the vehicle, in this case it wasn’t appropriate to do so. Hence brackets were made to install these instruments high up on the vehicle, as shown in Figure 1. In addition, a significant challenge was to provide a means for the vehicle to safely manoeuvre in the vicinity of the overhead beams and impress the measurement head of the thickness gauge onto the face of a beam.

4

Figure 1 The LBV vehicle with instrumentation installed

The Cerberus was built with three-foot2 centres between the beams. The beams were of riveted construction, 0.5 inch by 12 inches on the side and with angles riveted each side top and bottom. The angles on the lower edge were 2.5 inch angle iron [Nicholls, 2001, page 155]. This yielded a net opening between the beams of about 30 inches (~75 centimetres). Given that the vehicle itself has a length of approximately 55 centimetres and the measurement head of the thickness gauge projects at least 10 centimetres in front of the vehicle, there would be insufficient space to manoeuvre the vehicle between the beams. The approach which was adopted, therefore, was to install the thickness gauge well above the top of the vehicle, and to provide a pair of ‘beam hooks’ which would latch onto the lower web of the beam that was targeted for measurement. These hooks can be seen in Figure 1. The vehicle is fitted with a pilot’s camera that can be rotated around a lateral axis. Therefore the simple expedient of adding four cable ties to the tines of the beam hooks enabled the camera to be rotated into a position where the pilot could simultaneously visualise both the web of the beam and its position relative to the hooks. Once the hooks were engaged onto the web of the beam, the concept was for the vehicle’s vertical thruster to be engaged, which would raise the aft end of the vehicle, causing the measurement head of the thickness gauge to engage onto the face of the beam.

2 The Cerberus was specified and built according to the Imperial measurement system. Where direct reference is made to the structure of the ship dimensions will be quoted in the original units.

5

The system was tested in DSTO’s Noise and Vibration water tank. A mock up was made of two beams and the overhead deck, as shown in Figure 2. This was suspended in the tank and a list of approximately ten degrees was applied to it, hence approximating the reported current disposition of the Cerberus3. A single, short practice session proved sufficient to enable the pilots to reliably latch onto the beam. While this arrangement worked well from a pilot’s perspective, and it enabled the sensor to remain at a fixed location on the beam, there were inconsistencies in the resulting test measurements. Hence the ultrasonic sensor was serviced and calibration rechecked. After this the Cygnus system was deemed to be operational.

Figure 2 The ROV in position to take ultrasonic thickness measurements.

In Figure 1, an external camera can be seen. This is the GoPro Hero high definition video camera in its waterproof housing. The camera is fully self-contained in that, once recording is commenced, it simply records whatever it is pointed at. This camera was installed because the standard on-vehicle camera is only PAL definition. Because the camera is mounted in a relatively exposed position, a protective ‘top hat’ was installed on the vehicle.

3 Estimates made at the site place the true list at somewhat less than ten degrees – it is actually close to five degrees.

6

2.3 Diver Rehearsals

When permission was granted for ANRDT Six to support this survey, an arrangement was made for them to visit the Fishermans Bend site to undertake a familiarisation session with both the ROV and the video system that they would be using to undertake the external survey of the Cerberus. This visit took place on 27th November and was attended by the full dive team. Divers undertook both land-based and in-water training related to the LBV vehicle and the underwater video equipment. The divers were also able to arrange for DMS to provide a dive support vessel for the trial, so details of the foreseeable support requirements were also agreed at this meeting. 2.4 Regulatory Compliance of the Survey and Assessment of Ecological Considerations

The HMVS Cerberus is a historic shipwreck and as such is protected under the Victorian Heritage Act 1995. Furthermore the Cerberus is located within a protected zone which extends 25m each side of the centreline of the wreck and 5m beyond the bow and stern. Protected zones are declared for a small number of fragile and highly significant shipwrecks: they prohibit boating, fishing, anchoring and diving within the zone. In order to enter the zone or undertake work inside a shipwreck approval in the form of a permit must first be obtained from Heritage Victoria. Permission was sought and granted for the survey to be undertaken. A copy of the issued permit is included below as Appendix C. Prior to the trial it was identified that the Cerberus is listed as a national heritage place and falls under the protection of the Environmental Protection and Biodiversity Conservation Act (EPBC) 1999 (as well as the auspices of both Victorian and local council heritage provision). This Act is also applicable to the actions of Commonwealth agencies such as DSTO. The EPBC Act is specifically concerned with actions producing a significant impact on the environment and should it be likely that significant impact will occur, permission for the action to progress must be sought from the appropriate Minister. Assessment of the aims of the trial suggested that the likelihood of significant impact was highly unlikely. The non-destructive nature of the hull condition survey was unlikely to have any impact whatsoever on Cerberus. An examination of local ecological communities produced no species or ecological communities of significance (under the EPBC) likely to occur in the area and/or be impacted by the trial activities. This assessment was performed using the Australian Government Department of Sustainability, Environment, Water, Population and Communities Environment Protection and Biodiversity Conservation (EPBC) list of threatened species and ecological communities using the EPBC Protected Matters Reporting Database and selecting an area surrounding HMVS CERBERUS with the inclusion of a one kilometre buffer [Commonwealth of Australia, 2012].

7

3. Details of Planned Survey Activities

The agreed plan was to undertake a survey lasting up to five days at the Half Moon Bay site of HMVS Cerberus. The trial was undertaken in the week starting 3rd December, 2012. The plan was for the first day to be spent undertaking preliminary activities in the form of external HD video survey of the wreck and its surroundings, gaining familiarity with the environment from the perspective of operating the ROV, and undertaking a calibration check of the Ultrasonic Thickness (UT) gauge, by finding a part of the ship’s hull where UT measurements could be complemented by actual physical measurements of residual shell thickness (using callipers). The results of the first day were to be used to guide planning for the remainder of the week, which would concentrate upon inserting the ROV into the wreck for the purpose of undertaking the internal survey. In the course of the week, the plan was that a number of additional activities would be undertaken including:

• An interpretation of the general ecology and environment of the site (see Section 4.7);

• Using diver support, make measurements and log the Cerberus hull potentials at various sites and, if possible, the cannons on the seabed using a hand held CorrMon SpotCheck bathycorrometer (discussed in Section 4.8);

• Undertake dipping measurements of hull potential of Cerberus. This activity was an adjunct to the bathycorrometer-based measurements (discussed in Section 4.8);

• Undertake video and stills-based survey by inserting a pole camera through appropriate deck openings (see Section 4.9);

• Locate and assess the condition of the anodes on the hull and four cannons (discussed in Section 4.8).

3.1 General Approach to Undertaking the ROV Internal Surveys

The BMT’s 2011 report included some images recorded by divers during a survey undertaken in August 2002. These indicated there is a considerable amount of debris inside the wreck, in the form of twisted or broken stanchions, torn metal, etc. It was anticipated, therefore, that for the ROV to get into a position where it could inspect the under-turret beams, it would need to ‘thread a path’ into the vessel. A simple position plotting scheme was adopted to enable the path the ROV followed to be logged. This was based upon the video imagery recorded off the vehicle’s cameras. A relatively coarse, three dimensional grid was laid over the vessel. This comprised a numeric grid along the length of the ship, an alphanumeric grid across the ship’s breadth, and a deck-by-deck vertical grid. Starting at the bow, as a general rule every third beam underlying the upper deck was taken to define a grid boundary. Hence Grid Zone 1

8

extended from the bow to Beam 34, Zone 2 extended from Beam 3 to Beam 6 etc, as shown in Figure 3. There are a couple of exceptions to this rule. Most of the beams were placed at three-foot centres, but in places the inclusion of watertight bulkheads caused the spacing to be staggered slightly. Thus Zone 10 actually includes four beams within its scope. There were 26 grid zones along the length of the ship. In the across-ship direction there were only four grid zones (plus one outside each side of the ship). These comprised (starting at the port side) P2, P1, S1, S2. The line of the keel corresponded to the P1/S1 boundary. In the vertical plane, starting at the breastwork deck, the spaces were identified as Breastwork (i.e. on top of the breastwork deck), Upper Deck (below the breastwork deck and above the upper deck), Lower Deck (below upper deck and above the lower deck) and Hold Deck (Below the lower deck). According to this scheme the survey objective was to operate the ROV in the Lower Deck space. As the survey progressed an effort was made to estimate and log the heading of the vehicle relative to the ship’s bow (in approximately 45 degree increments – Forward, Port or Starboard 45, Port or Starboard 90, Port or Starboard 135, Aft). The video files recorded by the ROV system were either sourced from the on-board PAL definition camera or by the GoPro external camera. In both cases the files carried a time code relative to the start of the file. Hence at any one time the vehicle’s location could be identified as: File name; source camera (PAL or GoPro); timecode (minute:second); location (along ship, across ship, deck); orientation. Thus, as a hypothetical example, a vehicle location may be logged as:

• GO01068; GoPro; 10:20; 9, S1, Lower; Port 90 This is saying that the vehicle was under the forward turret (in Zone 9), on the starboard side of the ship (across-ship zone S1), looking towards the port. The camera was the GoPro high definition camera and the time was 10 minutes 20 seconds from the start of the file named GO01068. Using this logging scheme, the progress of the various ROV missions were recorded on Excel spreadsheets, along with any points of note. Each significant change in vehicle attitude or circumstances was duly logged. Throughout this document, where images are derived from ROV video footage, a ‘Relocation Descriptor’ is included in the figure captions. This will facilitate relocation of the image in the relevant files. One reason for providing this detail is that, at times, ‘live’ review of video footage conveys much more information to the reviewer than a still image capture is able to do.

4 The beams have been numbered sequentially according to a convention that the first beam after the bow stem (which is also the forward perpendicular) is beam number one.

9

Figure 3 The grid system used to log the ROV survey of the ship.

During the post-processing of the video imagery it was found that, while it was very difficult to exactly determine the across-ship location of the ROV, it was frequently possible to determine the overhead beam in closest proximity to the vehicle. Thus in the following section reference is sometimes made to the beam number. This is the number of the beam which underlies the upper deck, as can be determined from Figure 4.

Figure 4 Showing the relationship between Grid Zones and upper deck support beam numbering.

Two of the insertions were undertaken through openings that were below the surface. These are described in Sections 4.3 and 4.4. For these, a diver was kept on standby to monitor the entry of the umbilical into the ship. The other two surveys were undertaken with the vehicle deployed through openings in the breastwork deck, as described in Sections 4.5 and 4.6. In these cases DSTO personnel acted as cable handlers on board the Cerberus.

10

4. Results

4.1 Day-by-Day Sequence of Events

4.1.1 Monday 3rd December 2012

Two factors contributed to the first day being relatively productive. The first was that Heritage Victoria provided a workboat to supplement the DMS vessel. This gave the survey team great flexibility in conducting the first day’s operations. The second factor was that the weather conditions on the first day of the trial were reasonably favourable. Wind was westerly, estimated at approximately 15-20 knots at 0830, but dropping to 5-10 knots by midday. Wind conditions deteriorated again later in the afternoon. The following activities were undertaken:

1. The dive team set up safety lines upon the Breastwork Deck of the Cerberus, as shown in Figure 5. Through the use of harnesses, this enabled DSTO personnel to safely work off this deck, as shown in the photograph. This provided the survey team additional options for deploying the ROV into the interior of the vessel.

2. The dive team undertook a video survey, external to the hull and of the surroundings. This is discussed in Section 4.2.

3. Preliminary external survey was also undertaken by the ROV. The purpose of this survey was twofold: (1) to determine what local operating conditions were like relative to operation of the vehicle – particularly with respect to swirl effects caused by wave action in the vicinity of the wreck; (2) to complement the divers’ video survey. The ROV video files for this survey confirmed that the hull has collapsed virtually down to the armour belt, with the result there were really no options for entering the wreck from below upper deck level. The main outputs from this activity complement the divers’ material which is covered in Section 4.2.

4. Entry into the forward part of the wreck, by the ROV, via the hatch opening immediately aft of the capstan. This is described in Section 4.3.

5. Investigation of the wreck, using probes that could be deployed from the Breastwork Deck, was initiated

6. Using diver support, the corrosion potential measurement survey was initiated. Unfortunately it was not possible to undertake external UT calibration because no convenient and appropriate location was found. Previous advice was that the rudder would be an accessible and appropriate test site, but this has now been covered by sand. One site at the bow was investigated, which was being considered for use on the second day of the trial, but ultimately bad weather prevented this activity from being undertaken.

11

Figure 5 Safety lines were deployed, by ANRDT Six, upon the Breastwork Deck.

4.1.2 Tuesday 4th and Wednesday 5th December 2012

Weather conditions were deemed to be highly unfavourable on these days, to the extent that it was deemed unsafe to work. Wind was estimated at 30 knots and waves were breaking over the top of the vessel, including over the top of the forward turret on occasions. Hence operations were paused until conditions abated on Thursday morning. The loss of these two days impacted very heavily upon the operations for the last two days of the survey. As indicated above the intention was to use results of the first day to guide the planning for the remainder of the week. By the end of the first day four potential points of entry had been identified. The hope was to spend each of the next four days undertaking surveys based upon each of those entry points. This would allow the team to adopt a slow, cautious and thorough approach to each of the four surveys. The loss of two days compromised this plan. There was no question of the team becoming careless, so the only way to compensate for lost time was to undertake surveys that were less thorough than what we would have liked.

12

Figure 6 Weather conditions on Tuesday and Wednesday prevented operations taking place.

4.1.3 Thursday 6th December 2012

Weather conditions abated considerably during the evening of 5-6 December. At 0830 wind was estimated at 10-15 knots westerly and abating. By 1000 this had eased to 5-10 knots and swung around to the southwest and from 1200 onwards the winds were 5 knots southerly. Seas were estimated as seastate 2 early, abating to seastate 1 by midday. Operations were able to resume. For the divers, the first job of the day was to remove reinforcing mesh from the boiler room outlet – i.e. the opening through which the exhaust gases were ducted to the funnel. This was judged to be the most promising access point for the ROV to survey the forward turret. Meantime the ROV was flown into the aft hatch through the upper deck. The results of this part of the survey are discussed in Section 4.4. After lunch the vehicle was inserted into the hull via the boiler outlet opening on the breastwork deck. This part of the survey is discussed in Section 4.5. In addition to providing general support, during the course of the day the divers continued to undertake video surveys, covered in Section 4.2, plus anode checks and corrosion potential measurements as discussed in Section 4.8.

13

4.1.4 Friday 7th December 2012

Weather conditions remained favourable on Friday, being generally the same as for Thursday afternoon. On this occasion the ROV activities all involved deployment of the vehicle through openings in the breastwork deck, as detailed in Section 4.6. DSTO personnel were therefore able to act as cable handlers. Corrosion Potential measurements were also undertaken from the breastwork deck, using a dipping probe. Because water clarity was very good on this day, the divers undertook a final visual survey of the site. They also completed inspection of the anodes and provided general support to operations. Survey activities concluded at approximately 1500. 4.2 External Survey

The external survey was undertaken by the divers from ANRDT Six, with supplementary imagery recorded from the ROV. Both sets of imagery were complementary, so only the diver-derived video imagery will be described here. The divers were asked to undertake a general survey of the site, and once having done so undertake a survey of the wreck itself. For the latter, the divers were asked to concentrate their primary attention on the lower part of the visible hull, in particular (a) to determine the approximate current seabed level relative to the bottom of the armour belt. (b) to look for potential sites for inserting the ROV into the wreck and (c) to identify an appropriate location for undertaking the proposed calibration of the UT gauge. Divers were asked to periodically return to the surface to record a video marker of their current position relative to the ship. The reason for this was to enable the approximate position on the wreck of points of interest to be determined. The technique appeared to work quite well. (A more formal survey would require installation of either tape measures or the use of acoustic tracking transponders.) 4.2.1 Observations of the Environment in the Vicinity of Cerberus

The divers described the seabed as being of a firm, course sand (one described is as being ‘Portland Sand’), relatively uniform in texture around the whole vicinity of the wreck. The divers undertook penetration tests using dive knives, and they reported no evidence of striking bedrock. An example of one such test is shown in Figure 7, in which a firm downwards stroke followed by three open palm strikes on top of the knife produced approximately 20 cm burial of the knife. This is consistent with the divers’ descriptions of a relatively uniform seabed material.

14

Figure 7 Diver penetration test of sea bed immediately aft of the vessel stern (Filter applied:

Program Irfanview [2010]: Image/Colour Corrections/Green -20).

There is little evidence of either heavy build up of sand on either side of the wreck, or of significant scouring. This even applied in the vicinity of the counter at the stern of the ship, as is implied by the image included at Figure 8. At most the seabed level under the stern differed from the surroundings by an amount of the order of only 30 cm. This is apparently not always the case, however, because MacLeod and Steyne [2011] report that in 1999 approximately 2 metres of material was scoured from under the stern, exposing the bedrock upon which the keel rests. With the list of the vessel, this has exposed some of the counter, showing that the shell plating below the armour belt is substantially gone, but that the ribs at this location appear to be in reasonable condition. This can be seen in Figure 9 Note that, in environments featuring more ‘aggressive’, perennial currents, it is quite common to find quite significant scour features at one or both ends of a shipwreck. For example, the images included in Figure 10 are of the Currajong shipwreck off Bradleys Head in Sydney Harbour (data recorded in support of the 1999 Shallow Survey Conference [Neill 1999]). These include a multibeam echo sounder image at lower left and a sidescan sonar image at lower right, both of which clearly show a scour/deposition complex near the bow of the ship. In addition, there is significant deposition of material along the port side of the ship. This can be seen as a dark5, high intensity return in the sidescan sonar image and the multibeam sonar image shows a rising seabed adjoining the hull.

5 In the sidescan sonar representation used in this report, very high-intensity reflection of the sonar ‘ping’ produces a black trace, absence of reflected energy is painted as white and a more-or-less linear greyscale represents intensity in between.

15

Figure 8 The stern of the Cerberus is at upper right. The log at left is at the level of the

surrounding seabed. There appeared to be very little scouring in the vicinity of the stern (Filter applied: Program Irfanview [2010]: Image/Colour Corrections/Green -80).

Figure 9 The stern of the Cerberus below the level of the armour belt shows that much of the

shell plating has corroded away (Filter applied: Program Irfanview [2010]: Image/Colour Corrections/Green -30).

16

Figure 10 The Currajong shipwreck features significant scouring in the vicinity of the bow. The

Cerberus site has no comparable feature.

The divers reported that there were a number of baulks of timber lying on the seabed, in various locations around the wreck. In all probability these have fallen off the timber breakwater which adjoins the bow of the ship. One of these timbers at the portside stern of the wreck can be seen in Figure 8, and another is located at the starboard bow, having become ensnared with two ‘wheelie bins’ as shown in Figure 11.

17

Figure 11 The wreck site is relatively clear of detritus, except for two ‘wheelie bins’ (Filter applied:

Program Irfanview [2010]: Image/Colour Corrections/Green -60).

4.2.2 External Video Survey of the Wreck

Divers undertook two video ‘sweeps’ of the wreck. The first took place on Monday 3rd December and the second took place on the 6th December. The first survey was essentially undertaken at the seabed level, with the video camera looking along the line of the ship’s hull. The second survey was undertaken from higher up in the water column, with the camera looking down towards the seabed adjacent to the ship’s side. The following is a summary of the reports from the two sweeps. The divers’ reports, supported by video evidence, were that the starboard side of the ship is lying with the lower edge of the armour belt essentially at seabed level. There is very little debris external to the ship on this side, although there is some evidence of the ‘concertina’ effect of the collapse adjacent to the aft turret. This can be seen in Figure 12.

18

Figure 12 Evidence of the lower hull bulging outwards as the ship has collapsed can be seen adjacent to the aft turret, starboard side (Filter applied: Program Irfanview [2010]: Image/Colour Corrections/Green -75).

There is much more external evidence of the collapse along the port side of the wreck. During the collapse the vessel acquired a list of approximately five degrees. In various places along the port side there is some indication that, during the 1993 collapse, in addition to acquiring a list the top half of the vessel may have toppled slightly towards the starboard side. The video grabs in Figure 13 show two views of the Cerberus’s stem, the first looking directly across the bow from the port side and the second looking forward towards the bow, again on the port side. It can be seen that stem has sheered off, with the lower, buried portion lying 0.3~0.5 metres to port of the exposed part.

19

Figure 13 Two views of the stem of the Cerberus indicate that it has sheered off, with the upper

part ending up to starboard of the buried portion (Filter applied: Program Irfanview [2010]: Image/Colour Corrections/Green -60).

The ship dropped quite a long way, however, in the order of 2 metres [MacLeod and Steyne, 2011], so in large measure the collapse process appears to have been the aforementioned ‘concertina effect’. Figure 14 shows an estimate of the approximate level of

20

the seabed, before and after the 1993 collapse6 and the probable vertical extent of the affected structure.

Figure 14 Estimated posture of the Cerberus before and after collapse. The reference plane is along

the vessel’s midline; list is not accounted for here.

The frames which collapsed had to go somewhere to accommodate the drop, and on the port side much more material is evident outside the line of the hull than can be seen on the 6 It proved to be very difficult to find ‘hard numbers’ regarding the pre-collapse disposition of the vessel. The bulk of the information the authors were able to source was photographic in nature, and the state of the tide wasn’t recorded to accompany the photos. Hence, based upon comparison of a number of pre-1993 photographs, an estimate was made of the mean waterline of the wreck. This was then scaled off plans. The attitude of the wreck as it now rests, in terms of both its list and bow-down posture, were estimated from photographs taken at the site. It is estimated that the ship is pitched down about one degree, with the bow having dropped about 2.6 metres and the stern 1.4 metres (although there are big error bounds on these estimates – probably + 0.5 metres!). The list is between four and five degrees to starboard. Average depth to the seabed during the survey was 2.5 metres.

21

starboard side. Figures 15 through 22 show snapshots taken from various locations along the hull.

Figure 15 Debris immediately aft of the bow, portside (Filter applied: Program Irfanview [2010]:

Image/Colour Corrections/Green -60).

Figure 16 Debris in the foredeck vicinity, portside (Filter applied: Program Irfanview [2010]:


22

Figure 17 Another view of debris in the foredeck vicinity, portside (Filter applied: Program

Irfanview [2010]: Image/Colour Corrections/Green -60).

Figure 18 Adjacent to the fore turret, portside (Filter applied: Program Irfanview [2010]:


23

Figure 19 Amidships, portside (Filter applied: Program Irfanview [2010]: Image/Colour

Corrections/Green -60).

Figure 20 Adjacent the citadel, portside (Filter applied: Program Irfanview [2010]: Image/Colour


24

Figure 21 Adjacent to aft turret, portside (Filter applied: Program Irfanview [2010]: Image/Colour


25

Figure 22 Two views taken adjacent aft end of breastwork deck, portside (Filter applied: Program

Irfanview [2010]: Image/Colour Corrections/Green -70 (upper image), -80 (lower image)).

4.3 ROV Survey through Forward hatch

The forward-most hatch, highlighted in a pre-collapse aerial photograph, is unobstructed and currently approximately one metre below the mean water level. This was deemed to be an appropriate site for the first ROV-based internal survey. It was hoped that it would be possible to fly the vehicle aft, to inspect the support structure of the forward turret.

Figure 23 Location of the forward hatch on Cerberus (Photograph courtesy John Rogers).

This survey was undertaken on 3rd December. The Heritage Victoria workboat was tied off immediately to the portside of the hatch which enabled the vehicle to be directly flown across to the hatch opening with minimal deployed umbilical. Tether management was substantially undertaken from the workboat, although divers remained stationed nearby to render assistance if required.

26

The potential benefit of entering through this hatch was that it provided direct access to the lower deck level of the ship. The ship’s capstan was also directly forward of the hatch, which represented a unique, identifiable reference landmark to aid the ROV pilot’s navigation. The distance from the hatch to the centre of the forward turret was approximately nine metres. Upon entry into the forward messdeck it quickly became apparent that the watertight bulkhead in the vicinity of Beam 19 had still been in place prior to the collapse of the ship. This is now heavily broken up, plus there is a great deal of structural debris within this part of the ship, as can be seen in Figure 24. This represented a significant fouling hazard for the ROV. It was decided, therefore, that alternative routes would be sought to access the underside of the forward turret.

Figure 24 Debris and broken bulkheads in the forward part of the wreck represent a significant

fouling hazard for ROV operations. (Relocation Descriptor: Monday Tape2_1; PAL; 13:13; 6, P1, Lower; Starboard 90).

It was possible to image the underside of the deck in this area. Figure 25 is an image of Beam 17, the beam which underlies the forward-most extremity of the ship’s heads. As the

27

survey progressed it was found that the description of this beam, reported by the ROV pilots, applied to virtually all of the beams that were visualised: the beam is heavily encrusted by marine growth, it appears to be complete (i.e. no holes corroded through), the bulb at the lower extremity of the beam appears to be complete (in so far as the concretion forms a bulbous shape) and the concretion over the angle iron at the top of the beam appears to form a straight, relatively uniform line.

Figure 25 Beam 17, showing heavy encrustation over its entire surface. (Relocation Descriptor:

Monday Tape2_1; PAL; 8:31; 6, P1, Lower; Starboard 135).

The capstan is still in place, and its lower part, including the holes in which bars were inserted to enable hand operation, can be seen on the lower deck. Figure 26 is a screen capture of the capstan, taken off the ROV’s PAL video system.

28

Figure 26 The Capstan at lower deck level. (Relocation Descriptor: Monday Tape2_1; PAL; 5:14-

6:41; 5, P1, Lower; Starboard 45).

4.4 ROV Survey through Aft hatch

The aft hatch, highlighted in a 1953 pre-collapse aerial photograph, is unobstructed and currently approximately 0.5 metres below the mean water level. This is a more challenging point for an ROV to enter the vessel than the forward hatch (Section 4.3), because the shallow water makes tether management more difficult. In addition, the hatch was once divided by a fore-and-aft bulkhead, some of which remains intact, as can be seen in Figure 28. This represented a significant fouling hazard for the tether. Nevertheless it was hoped that it would be possible to fly the vehicle forwards to inspect the support structure of the aft turret. As for the previous section, entry through this hatch opening had the potential benefit of inserting the ROV directly into the required, lower deck space. The description that follows doesn’t necessarily follow the exact sequence in which the survey was undertaken. This section details the parts of the ship which were inspected, not the order of inspection. The latter can be accessed via the log in spreadsheet file HMVS CERBERUS LOG.XLS, included on the data disk.

29

Figure 27 Location of the aft hatch on Cerberus (Photograph from Picture Victoria).

Figure 28 Showing the aft hatch with the remains of a longitudinal divider still in place.

The aft hatch is not shown on any of the original plans. The suspicion is that this hatch opening was installed during the time Cerberus was serving as a secondary submarine tender for the Royal Australian Navy, although no hard evidence has been found to support this. Figure 29 shows the approximate location of this opening relative to the original layout of the vessel. It would appear the bulkhead that bisects the hatch may be the remains of the wardroom wall.

30

Figure 29 Estimated location and size of the hatch, relative to the upper deck (red square) and the

original layout of the lower deck (dotted red square).

Once the ROV descended into the space below this hatch it became apparent that the deck layout was not as shown on the ‘as built’ plans, which made navigation more difficult than initially anticipated, because there were very few obvious landmarks to follow. Presumably, like the installation of the hatch opening, these alterations were undertaken late in the working life of the ship. The original deck layout is shown in Figure 29. While it appears the partition wall between passageway and the Wardroom may have been left in position, an additional partition appears to have been installed inside the Wardroom space (see Figure 30), perhaps making a new passageway.

31

Figure 30 The remains of the bulkhead at left may be the original Wardroom partition. To the

extreme right of the above pair of images (actually to port as this image is looking aft) is a partition wall not shown on the plans. The purpose of the ‘stringer’ highlighted by an arrow is unknown. (Relocation Descriptor: GOPRO069; GoPro; 40:40; 21, S1, Lower; Aft).

Approximately in line with the athwartships centreline of the hatch opening, the new partition wall appears to have a ‘kink’, perhaps to make it run directly fore-and-aft. Just forward of this kink, the remains of part of this wall support several severed pipes, as shown in Figure 31.

32

Figure 31 Two views of the remains of a rack of pipes (Relocation Descriptor: GO010069; GoPro;

07:50; 21, S1, Lower; Forward).

The ROV was flown towards the midline of the ship, through the gaps in this partition. A general survey was undertaken of this space, plus an inspection was made of the beams underlying the upper deck in this area (separately described in Section 4.4.1). The space is relatively clear of debris, and at least some of the support pillars remain in place and intact. For example Figure 32 shows a pillar which is in the vicinity of the aft skylight. It appears to be straight and still connected to the overhead beam.

33

Figure 32 Support pillar in vicinity of aft skylight (Relocation Descriptor: GO010069; GoPro;

03:13; 23, S1, Lower; Port 150).

An inspection of the lower deck revealed that there is a hole leading through to the hold. It is not clear whether this has been deliberately cut, or if it is the result of structural collapse. The space below appears to be reasonably full of sand/silt, as shown in Figure 33. This would correspond fairly closely with the level of the seabed outside the vessel.

34

Figure 33 Hole in the lower deck, revealing the level of silt buildup in the hold (Relocation

Descriptor: GO010069; GoPro; 07:32; 21, S1, Lower; Port 45).

The ROV was flown back into the ‘new passageway’ and an attempt was made to fly forward towards the turrets. Immediately forward of the hatch is a support pillar, which once again appears to be straight and attached. This pillar is attached to the beam that supports the leading side of the forward skylight.

35

Figure 34 Support pillar immediately forward of, and slightly amidships of the hatch opening

(Relocation Descriptor: GO010069; GoPro; 26:08; 20/21, S1, Lower; Port 110).

Just visible in the above image is a jumble of debris. This represents a very significant fouling hazard for the ROV. Figure 35 is a more detailed view of this debris. While it was possible to move forward past this debris, the limit of travel on this occasion was just short of the after end of the Breastwork.

36

Figure 35 Jumble of debris, forward of the aft hatch opening (Relocation Descriptor: GO010069;

GoPro; 25:12; 21, S1, Lower; Port 20).

Along the path of advance lay the remains of overhead pipework, although this stops abruptly at a break three beams in front of the hatch (see Figure 36).

37

Figure 36 Overhead pipe terminates at a break, just visible as highlighted. (Relocation Descriptor:

GO010069; GoPro; 16:53; 20, S1, Lower; Port 20).

The officers’ heads were over the top of the location where the vehicle reached the limit of its travel, and some of the remaining pipework appears to have survived. This can be seen in Figure 37. There was no single obstruction which blocked the vehicle’s path; the route was challenging, but not impossible. The main reason that this passage to the turrets was not pursued further was one of time. It was apparent that meticulous planning and a progressive approach to the insertion would be required to ensure the tether would not foul. The loss of two days due to bad weather resulted in the team having insufficient time to pursue this approach, so alternative routes were sought.

38

Figure 37 The limit of forward travel was underneath the officers’ heads. The overhead pipe is the

one which terminates at a break in Figure 36. (Relocation Descriptor: GO010069; GoPro; 14:24; 20, S1, Lower; Forward).

A second attempt was made to move forward from this hatch opening. This time the vehicle moved across to the starboard side of the ship, and then was flown forwards. Figure 38 is a video capture of the starboard side end of the beam which is immediately forward of the hatch opening (beam number 61). Note, by comparing this image with the photograph below it, the level of concretion over both the beam and the frame to which it

39

is attached is very evident. The end of one of the bolts that secure the armour plate to the ship is also visible in the video capture, as highlighted.

40

Figure 38 Image of starboard side, showing the end of a deck beam and one of the bolts securing the

armour belt to the ship. (Relocation Descriptor: GO010069; GoPro; 12:11; 21, S2, Lower; Starboard 45). Photograph taken from Peter DeGruchy’s Cerberus Pictures Slideshow, http://www.cerberus.com.au/degruchy_slideshow.html.

Moving forward it was possible to image a number of the beams, as is separately described in Section 4.4.1. As with the previous attempt to approach the turret from this hatch opening, tether management became a real issue once the vehicle had proceeded forward to the vicinity of the aft end of the breastwork deck. Once again it was decided that more time would be required to successfully pursue this route than was available. Hence the vehicle was withdrawn from the hatch and an alternative point of insertion, described in Section 4.5, was tried. A final, general image capture is included, which gives some clue to the difficulty associated with interpreting the material recorded from this survey. The image shows what appear to be two support pillars. Upon review of the video imagery, it is believed that the ‘straighter’ vertical object is in fact a support pillar, but the bent structure behind it is most likely to be a piece of overhead pipework that has collapsed. This pipe may be a continuation of the one shown in Figure 36.

41

Figure 39 Showing two vertical structures. The one nearest the camera is believed to be an intact

support pillar and the other may be length of trunking pipe. (Relocation Descriptor: GO010069; GoPro; 13:26; 20, S2, Lower; Port 135).

4.4.1 Observed Deck Beams in the Aft Part of the Ship

During this survey it proved possible to visualise a number of the bulb-iron deck beams underlying the upper deck in the aft portion of the ship. While these beams were of different configuration to the beams supporting the turrets, as with the latter they were half inch thick. The bulb-iron beams were nine inches deep x half inch thick, with a pair of 3” x 3.5” x 1/2” angles riveted along their upper edge. The under-turret beams were twelve inches deep by half inch thick, and they had the same size angle riveted to their upper edge. In the case of these beams, however, they had additional angle iron pair riveted to their lower edge, being 2.5” x 2.5” x 7/16”. Thus both sets of beams presented a similar expanse of exposed half inch wall to the environment (5-6 inches). While no attempt was made to measure the wall thickness of the bulb-iron, where possible a visual inspection was undertaken. In no case was there evidence of the beams having corroded through. In almost every case they had a reasonably thick concretion layer. It was very rare to see a patch of exposed rust. The following are video grabs of various beams. Every effort was made to correctly assign beam numbers to images, but it wasn’t always possible to keep beams visible when navigating inside the ship, so in some instances the beam numbers may be misidentified by one digit.

42

Figure 40 Beam 60. (Relocation Descriptor: GO010069; GoPro; 14:06; 20, S1, Lower; Port 45).

Figure 41 Beam 61, towards starboard side of ship. (Relocation Descriptor: GO010069; GoPro;

13:21 ; 20, S2, Lower; Forward).

Figure 42 Beam 61, amidships. opening (Relocation Descriptor: GO010069; GoPro; 25:12; 21, S1,

Lower; Port 20).

43

Figure 43 Two views of what is believed to be Beam 63, corresponding points highlighted.

(Relocation Descriptor: GO010069; GoPro; 04:53-05:02; 21, S1/P1, Lower; Port 45).

44

Figure 44 Believed to be Beam 64 (right of picture) and 65, almost amidships. (Relocation

Descriptor: GO010069; GoPro; 07:10; 22, S1, Lower; Port 90).

Figure 45 Probably Beam 65, almost amidships. (Relocation Descriptor: GO010069; GoPro; 07:05;

22, S1, Lower; Port 135).

Given the size of the hatch, the close proximity to under-deck beams and the distance from the turrets, the aft portion of the ship is one location where it may be possible for divers to enter safely. If it proves impossible to measure the beam thickness of the beams that underlie the turrets, it may be worth having a diver measure the wall thickness of the bulb-iron beams near the aft hatch. Macleod [2013, personal communication] has confirmed that previously-taken measurements indicate there is going to be a relatively uniform corrosion environment inside the vessel, and therefore that measurements taken in ‘the accessible area would provide a good to a very good guide as to the stability or degree of decay of the other structural members in the ship’.

45

4.5 ROV Survey through Boiler Room (Stoke Hold) Outlet

When machinery was stripped out of the vessel, a large void space was made in the vicinity of the Stoke Hold and Engine rooms. This was judged to be a good candidate for insertion of the ROV into the vessel, to provide the vehicle with access to the underside of the forward turret. The opening through which the vehicle was inserted can be seen in Figure 46. In this case the work boat was anchored about 20 metres from the side of the wreck, as shown in Figure 47. The ROV was launched off the stern of the workboat and flown across to the breastwork of Cerberus, whereupon it was lifted across the deck and deployed into the stoke hold. This means that there were two places requiring a cable handler – at the stern of the workboat as shown in Figure 48 and on the breastwork deck at the boiler room outlet, of which Figure 49 is a close up view.

Figure 46 Location of the boiler room outlet on Cerberus (Photograph courtesy John Rogers).

Figure 47 The workboat was moored approximately 20 metres from Cerberus.

46

Figure 48 Cable handling at the stern of the workboat.

Figure 49 Stoke hold entry was via the space previously occupied by the funnel uptake.

47

Once inserted into the upper deck space within the breastwork, the vehicle was flown straight down to determine the state of siltation in the hold of the ship (being in the engine room, this is an open volume). It was found that the seabed level inside the hold is essentially the same as what was measured outside – about 2.5 metres depending on the state of the tide. This can be seen in the screen grab of the ROV pilot’s display, showing a vehicle’s depth reading of 2.4 metres.

Figure 50 Showing vehicle depth as 2.4 metres immediately above the seabed, in the stokehold.

(Relocation Descriptor: Thurs 3_3; PAL; 10:10; 13, S1,Hold; Starboard 135).

Following this, the vehicle returned to the surface and was flown forward inside the shelter deck (i.e. inside the breastwork, over the top of the upper deck) towards the forward turret. Immediately forward of the stokehold are the remains of a series of companionways, the remnants of framing for the Crew’s Galley, and ventilation shafts. Some of this structure can be seen in Figure 51. These represented the first of a number of potential fouling hazards for the ROV umbilical that were encountered in this operational sequence.

48

Figure 51 Looking forward from the funnel uptake opening through which the ROV was deployed.

It was decided to fly as far forward as possible on this level to determine whether openings in the decking may provide access to the lower deck, and to visualise the roller race supporting the turret. It was known from a photograph in the ‘virtual tour’ of the vessel, available on the Friends of the Cerberus website [Friends of the Cerberus, 2013], that there was at least one hatch opening through the deck immediately adjacent to the turret. What could not be determined from the photograph was the size of this opening (see Figure 52). If it was sufficiently large to admit the vehicle it would be an ideal access point. The vehicle was flown around the starboard side of the turret. Only one opening was detected close to the turret (also shown in Figure 52). This was on the starboard side of the turret, just aft of its midline. The opening is not shown on the plans, but proved to be too small to admit passage of the ROV. The position of this opening is almost directly below the gap between the turret and breastwork deck. It may be possible to safely visualise at least two of the under-turret beams from the breastwork deck by inserting a pole camera through this opening.

49

Figure 52 (a) The highlighted hatch opening is between shield and lower decks (see main Figure

caption for more detail).

50

Figure 52 (b) This deck opening is not shown in the ship’s plans, but can be seen in the

photograph taken by Mr John King in 1968. (Photograph courtesy of Friends of the Cerberus; Relocation Descriptor: GOPR0070; GoPro; 32:35; 9, S2, Upper; Port 90).

Figure 53 is a photograph of the lower half of the ship’s forward turret, taken in 1971. Note that it was supported by a roller bearing comprising 26 rollers. The two ROV images show that the roller cage is still in place and appears to be in very good condition, and the rollers themselves appear to be sound. The 1971 photograph also shows one of the loading chutes that enabled ammunition to be passed into the turret from the shell room (in the hold) via lower and shelter decks. The junction between the smaller-diameter turret trunk and the upper turret was reinforced with webs. These webs are visible in the two ROV images below. They are clearly heavily encrusted, but still appear to be complete.

51

Figure 53 Lower half of Cerberus’s forward turret, photographed in 1971. Photo courtesy of Leigh

Doeg, Friends of the Cerberus.

52

Figure 54 Showing the cage that retains the turret support rollers. (Relocation Descriptor:

GOPR0070; GoPro; 29:12; 9, S2, Upper; Port 135).

53

Figure 55 One of the rollers that support the forward turret. (Relocation Descriptor: GOPR0070;

GoPro; 28:48; 9, S2, Upper; Port 45).

54

Figure 56 Looking into the turret loading chute. Note webs at top of view are heavily encrusted.

(Relocation Descriptor: GOPR0070; GoPro; 30:10; 9, S2, Upper; Port 45).

55

Figure 57 The turret loading chute opening imaged in Figure 56.

56

Figure 58 The lower part of this images shows two heavily encrusted turret support webs. Note the

ROV was virtually on the surface, so the upper part of the image is a mirror image reflection off the water-surface interface. (Relocation Descriptor: GOPR0070; GoPro; 26:29; 9, S2, Upper; Forward).

Once the base of the turret had been inspected, the vehicle was flown back to the upper deck-to-lower deck ladderway immediately aft of the forward turret, as seen in Figure 59. The vehicle then descended one level, so that it was at the lower deck level.

57

Figure 59 The ladderway which provided the means of access to the lower deck is immediately

below the ROV. The forward turret is at extreme left of the image.

58

The ship’s plans show a watertight bulkhead approximately 1.2 metres in front of the leading edge of this ladderway, and the sides of the coal bunkers to either side of it. These bulkheads were clearly still in place until the ship collapsed. Their remains, now broken and distorted, represented a very significant hazard for the ROV pilots to manage.

Figure 60 Remains of the watertight bulkhead in vicinity of the forward turret. (Relocation

Descriptor: GO010070; GoPro; 13:31; 10, P1/S1, Lower; Forward).

In the time available it was found to be impossible to plot a path under the turret. It was found to be possible, with some difficulty, to fly into the remains of the portside coal bunker. A convenient reference point for this excursion was the coal loading chute, easily and uniquely identifiable in both the ship’s plans and the ROV video imagery (see the highlighted feature in Figure 61). This revealed two bulb-iron beams, numbers 30 and 31, which are the second and third beams aft of the turret. Given the cluttered nature of this forward space, and the loss of two days, it was decided to settle on visualising these two beams. As discussed in Section 4.4.1, the condition of these beams is likely to be representative of the condition of the under-turret, ‘I’ beams.

59

Figure 61 Showing location of coal loading chute which was used as a navigation reference.

60

Figure 62 The circular hole is the coal loading chute highlighted in the ship’s plan above. The beam

running across the top of the image is Beam 31. (Relocation Descriptor: GO010070; GoPro; 29:01; 10, P1, Lower; Port 135).

The following are images of Beam 30. Both are views of the same part of the beam. Analysis of other parts of the video imagery, while not lending themselves to capture of still images, indicate the beam is intact with its bulb encrusted but apparently complete.

61

Figure 63 An image of Beam 30, indicating that walls are complete and the lower bulb are intact.

(Relocation Descriptor: GO010070; GoPro; 30:41; 10, P2, Lower; Port 45).

Figure 64 A repeat view of Beam 30, indicating that walls are complete and the lower bulb are

intact. (Relocation Descriptor: GO010070; GoPro; 31:25; 10, P2, Lower; Port 45).

Beam 31 appears to be in very similar condition to Beam 30. Once again the following images are recorded from the portside of the ship, in the general vicinity of the coal bunker.

62

Figure 65 A view of Beam 31, in the proximity of the portside coal bunker. (Relocation Descriptor:

GO010070; GoPro; 27:33; 10, P2, Lower; Port 135).

Figure 66 Another view of Beam 31, in the proximity of the portside coal bunker. (Relocation

Descriptor: GO010070; GoPro; 30:05; 10, P2, Lower; Port 135).

63

Figure 67 A view of Beam 31, looking towards amidships. Once again the bulb is intact. .

(Relocation Descriptor: GO010070; GoPro; 29:18; 10, P1, Lower; Starboard 135).

4.6 ROV Entry Via Aft Breastwork Deck Companionway

In the morning of Friday 7th December, it was decided to attempt to access the underside of the aft turret via the Aft Breastwork Deck Companionway, highlighted in the 2005 photograph below.

Figure 68 Showing the point of entry for the fourth point of insertion into the vessel

This companionway had the advantage of being in close proximity to the turret, and the area appeared to be relatively free of debris, as can be seen in the photographs included as Figures 90 and 91 in Section 4.9. The ladderway down into the lower deck space was directly below the hatch opening, so the tether management team on deck were able to provide direct feedback to the ROV pilots regarding the relative position and attitude of the vehicle.

64

Of the four entry points attempted, this one proved to be only one that provided reasonably ready access to the under-turret beams. The authors are not, however, saying access was easy. The ROV tether fouled twice during the course of the inspection, but on each occasion was able to be freed. A general observation made by the ROV pilots was that the lower deck in this part of the ship appears to be heavily silted. On the starboard side of the vessel there was barely sufficient headroom for the vehicle to manoeuvre underneath the beams. The vehicle was flown directly down into the lower deck, and then flown essentially aft, but with a slight offset heading towards the starboard side. There are three bulb-iron beams between the ladderway access and the ‘I’ beams that underlie the turret. Figure 69 is an image of the first beam, immediately aft of the ladderway. This appears to be in good condition, with both the wall of the beam and the bulb intact.

Figure 69 Bulb-iron Beam 45. (Relocation Descriptor: GOPR0071; GoPro; 24:37; 15, S1/P1,

Lower; Aft).

The next beam, number 46, also appears to be intact, as shown in the following photograph.

65

Figure 70 Beam 46. (Relocation Descriptor: GOPR0071; GoPro; 26:48; 15/16, S1, Lower; Aft).

Beam 47 is the final bulb-iron bearer before the support system for the turret begins. It had two support pillars attached to it. The pillar on the starboard side of the ship is intact and appears straight. While the one on the port side is in place, it is bent as can be seen in Figure 74.

66

Figure 71 Beam 47, the final bulb-iron beam, which underlies the leading edge of the aft turret,

with a support pillar still in place. (Relocation Descriptor: GOPR0071; GoPro; 27:16; 16, S1, Lower; Starboard 135).

67

Figure 72 Close-up view of Beam 47, the final bulb-iron beam which precedes the leading edge of

the aft turret, with a support pillar still in place. (Relocation Descriptor: GOPR0071; GoPro; 27:23; 16, S1, Lower; Starboard 135).

Figure 73 Attachment point of starboard side support pillar of Beam 47. (Relocation Descriptor:

GOPR0071; GoPro; 29:03; 16, S1, Lower; Starboard 135).

68

Figure 74 The portside support pillar for Beam 47 is in place, but bent. (Relocation Descriptor: GOPR0071; GoPro; 33:52; 16, P1, Lower; Port 135).

In the following image, looking along Beam 47 towards the centre of the ship, the line of the upper angle iron webs can be inferred under the concretion layer.

69

Figure 75 Build-up of concretion on the beam follows the line of the underlying angle iron

webbing. (Relocation Descriptor: GOPR0071; GoPro; 30:56; 16, S1, Lower; Port 135).

Towards the port end of this beam were areas of exposed metal, showing rusted surface. There is an airtight pocket here, with an amount of gas trapped between Beam 46, Beam 47 and the longitudinal carling between them. Concretion, which generally provides a degree of protection to iron, forms when full immersion of an object, combined with an appropriate degree of water movement, allows larvae to settle and then to grow. In this area the presence of an air pocket has inhibited concretion, hence resulting in surface corrosion.

70

Figure 76 Towards the portside end of Beam 46 and 47 is an entrapped air/gas pocket. This

appears to have inhibited concretion processes, hence perhaps encouraging surface corrosion of the beam. (Relocation Descriptor: GO010071; GoPro; 0:19; 16, P1, Lower; Port 135).

A clear passage was found for the ROV to be flown under Beam 47, and was therefore able to image the first of the ‘I’ beams that support the turret. While in this location the vehicle was relatively restricted in its ability to manoeuvre (once again, the loss of time due to bad weather prevented the team from determining the optimal course of approach to this part of the ship). Immediately evident was the fact there is a great deal of sedimentation in this part of the ship, as can be seen in Figure 77. It was not possible to fly the vehicle underneath this beam, hence Beam 48 was the only ‘I’ beam which could be imaged. Upon first inspection the beam appeared to be in good condition, with both lower and upper angle iron webs evident. Figure 78 is a closer view of the lower web.

71

Figure 77 Beam 48, the first 'I' beam underlying the aft turret has a heavy buildup of silt

immediately behind it. (Relocation Descriptor: GO010071; GoPro; 04:32; 16, S1/S2, Lower; Aft).

Figure 78 Close-up view of the riveted angle irons that comprise the lower web of Beam 48.

(Relocation Descriptor: GO010071; GoPro; 04:31; 16, S1/S2, Lower; Aft).

When post-processing analysis of the imagery was undertaken, however, it was found there is a portion of the beam broken or missing. Unfortunately the break wasn’t noticed at the time the survey was undertaken, and hence all views of it are peripheral, as can be seen from the montage of images included below (Figure 79). The best-centred image of the break is derived from the ROV’s PAL video camera, but unfortunately a problem with

72

the vehicle’s tether was generating a lot of noise in the imagery, as can be seen in Figure 80 – probably the reason the pilots didn’t notice the break at the time the survey was being undertaken. The break in Beam 48 is approximately amidships, or at most within a metre of the ship’s centreline. In this vicinity the plans show a diagonal carling leading to Beam 49. While it is conceivable that loads translated through the carling, due to relative motion between the two beams, could lead to a break in one or the other beam, no evidence could be seen of the remains of a carling in the video imagery. A more likely possibility is that this had been the attachment point for a support pillar, which broke out a section of the beam when the ship collapsed. No formal plans have been found specifying the location of the pillars. Indeed, the ship’s specification (included in Nicholls [2001], page 157) states that ‘Under the Breastwork and Turrets wherever practicable solid or wrought-iron Tubular Pillars are to be fitted as the Overseer may direct’. Thus the pillars could have been positioned at any convenient location along the beam. Given the level of sedimentation in this part of the ship, a bent or broken pillar would very readily be hidden.

Figure 79 A montage showing all images of the broken Beam 48, with a common location

highlighted. (Relocation Descriptors – clockwise from top left: Friday Tape 2_1, PAL, 03:03; 16, S1/P1, Starboard 135; GO010071; GoPro; 04:32; 16/17, S1/S2, Lower; Aft; GO010071; GoPro; 04:36; 16/17, S1/S2, Lower; Aft; GO010071; GoPro; 04:29; 16/17, S1/S2, Lower; Aft).

73

Figure 80 ROV image of the break in Beam 48, at the right of the image. (Relocation Descriptor:

Friday Tape 2_1, PAL, 03:03; 16, S1/P1, Starboard 135).

One, brief view was taken of Beam 48 on the port side of the break. This part of the beam is intact. It is apparent that the break described above is quite localised.

74

Figure 81 An image of Beam 48, looking towards the port side of the ship. (Relocation Descriptor: GOPR0071; GoPro; 34:05; 16, P1/P2, Lower; Port 135).

An attempt was made to take an ultrasonic thickness measurement of this beam. The UT gauge was successfully ‘locked onto’ the side of the beam, but despite several attempts it was not possible to take a reliable thickness reading. 4.7 Interpretation of the Ecology of the Site

The hull interior is home to a mature ecological community with extensive tertiary biofouling over the majority of exposed surfaces, as well as mobile species such as seahorses (possibly Hippocampus abdominalis ‘big belly seahorse’) and blennies and/or gobies as well as other smaller unidentified fish species. During the visual inspection, no species of significance were apparent on the ROV visual outputs in the areas examined. Fouling communities and mobile species observed were common species occurring regularly in Port Phillip Bay waters and on structures within the bay. Due to the extensive nature of the fouling combined with the need to have close contact between the inspected surface and the ultrasonic thickness measurement instrument, it is considered that it would be extremely difficult to find a suitable area to achieve reliable thickness measurements. 4.8 Corrosion Potential Measurements

A CorrMon SpotCheck bathycorrometer manufactured by RhoCraft R&D (BC, Canada) was used to measure corrosion potentials of the HMVS Cerberus hull and the four Armstrong guns lying on the seabed to the north of Cerberus.

75

The SpotCheck is a hand held instrument with a sharp steel probe, an adjacent silver/silver chloride reference electrode, and a data collection system that allows multiple readings to be made at each of up to 99 “sites”.

Figure 82 CorrMon SpotCheck bathycorrometer

Readings were taken by the diver scraping marine fouling, calcareous deposits and corrosion products from the hull with a dive knife, then pressing the probe tip firmly onto the exposed metal substrate while finning to maintain contact, and initiating the reading using the switch on the SpotCheck. Momentary activation of the switch allowed selection of different data storage locations for measurements taken at different sites. As readings were taken they could be viewed by the diver on a small display, and correct operation of the unit was verified by taking calibration measurements on a small piece of zinc anode. Perimeter hull potential readings were taken on the afternoon of Monday 3/12/2012 from approximately low tide, starting at the starboard bow and working aft, then on the port side working forward. Five readings were taken on each side: bow, fwd turret, amidships, aft turret, stern (Figure 83). To conserve limited diver air supplies, readings were taken at constant depth approximately midway between the seabed and the surface i.e at about 1.5m depth and 1.5m from the seabed. Corrosion potential measurements were also taken mid-barrel on each of the four Armstrong guns, located in pairs on the seabed on the starboard side, approximately 20m from the hull and opposite the respective turrets (Figure 83).

76

Figure 83 Hull Potential Measurements. SpotCheck (orange) and dipping (yellow).

Dipping measurements of the hull potential of Cerberus were also made on Friday 7/12/2012 using a silver/silver chloride reference electrode and multi-meter, with a ground connection made to protruding fixtures on the edge of the breastwork deck forward of the Forward Turret (Site 1 on Figure 83) via a heavy duty galvanized battery clip. Measurements were made on the port side from abeam the forward turret to abeam amidships, with the reference electrode deployed outboard of the submerged main deck, adjacent to the hull at approximately 2 m immersion depth. One measurement was made directly beneath the ground point, with the reference electrode immersed 1 m outboard of the shield deck in the vicinity of what remains of the ship’s company heads.

4.8.1 Results

Software/firmware problems with the SpotCheck have to date prevented downloading of the measurements, which are held in non-volatile memory. The manufacturer has been contacted to determine whether the meter is still supported, and every effort will be made to download the measurements and to forward them to stakeholders receiving this report. ANRDT Six divers noted the condition of the following anodes on the Port side corresponding to the reading positions in Figure 83: #12 - anode clamped and in good condition as can be seen in Figure 84; #13 - worn anode; midway between #14 and #15 - good anode. They also noted that the anode on Gun 4 (#2 measurement) was not connected.

4 3

2 1

#2 (Gun 4) #3 (Gun 3) #4 (Gun 2)

#5 (Gun 1)

#6 #7

#8 #9

#10

#11 #12 #13

#14

#15

77

Figure 84 Anode at position #12 is visible with diver’s hand over, clamp and connecting cable are

in good condition.

The dipping hull potential measurements indicated in yellow on Figure 83 were: Position 1, -972 mV; Position 2, -958 mV; Position 3, -956 mV; Position 4, -952 mV (all measurements versus silver/silver chloride reference electrode). These readings indicate well protected hull structure and are generally consistent with the presence and condition of the sacrificial anodes on the adjacent hull. Dipping hull potential measurements provide an average reading over a wide area, and are unable to discriminate variations in hull potential over localised small areas, which requires a more “focussed” meter such as the SpotCheck where the reference electrode is more directional and located close to the probe tip. It should be noted that the accuracy, stability and consistency of both SpotCheck and dipping measurements is reliant on the integrity of the ground connection. Divers were instructed to scrape away deposits to achieve good electrical contact of the probe tip with underlying metal. It was difficult to achieve a good ground (for dipping measurements) on the heavily corroded superstructure, and important not to move the connection once established. It was noted that the four dipping measurements made were stable as indicated on the high impedance multimeter, but attempted measurements made aft of amidships were unstable in the short term, for the same ground connection. This could be consistent with an electrical discontinuity in the hull between the forward ground point and the aft hull structure, as was found in the continuity measurements by MacLeod and Steyne [2011], and which was attributed to a break in the hull structure. Further measurements with an alternate aft ground point would be required to confirm this possibility.

78

4.9 Miscellaneous Imagery

In addition to the imagery generated by divers and the ROV, a pole camera arrangement was used to photograph those parts of Cerberus that could be accessed through deck openings. A selection of these images follows.

Figure 85 Inside the forward turret.

79

Figure 86 Gun carriage viewed through the gunport, portside gun if trained forward, forward

turret (two views).

80

Figure 87 Gunport opening, aft turret.

When the ship was initially sold for scrap, an attempt was made to demolish the aft turret. This proved to be uneconomical and demolition work ceased, but in the process parts of the turret’s structure has been exposed. The photograph included in Figure 87 shows the 10 inch thick armour plate/East Indian teak/iron sandwich construction of the turret.

81

Figure 88 Forward breastwork deck companionway opening.

The companionway opening shown in Figure 88 serviced both the Shield Deck (the enclosed part of the upper deck) and the Lower Deck. The transverse frame which is visible under the water is at upper deck level, and it would have once supported a bulkhead separating the companionway from the funnel uptakes.

82

Figure 89 The Aft breastwork deck Companionway opening with Ventilation trunk in front.

83

Figure 90 Looking through the Aft breastwork deck companionway hatch, looking at the

Ventilation trunk which is immediately forward of it. Note the layered corrosion mechanism evident in the web of the I beam.

84

Figure 91 Looking aft past the ventilation trunk. Photograph taken through the coal loading chute

opening indicated in Figure 92.

85

Figure 92 Opening for a Coal loading chute, through which a camera was inserted.

86

Figure 93 Outlet of the ventilator, identified in the plans as the ‘wind machine’. (Relocation

Descriptor: GO010071; GoPro; 9:40; 15, S1/P1, Lower; Port 60).

Figure 94 shows detail of the riveted beams which support the breastwork deck. It can be seen that the support struts are straight and the steelwork, which is only intermittently submerged, is in reasonable condition. A small amount of white paint remains in places. The ladderway below the vehicle is the same as the one shown in Figure 59.

Figure 94 Beam structure supporting the breastwork deck.

DSTO investigated whether it was practical to deploy a pole camera through convenient openings in the breastwork deck, to generate images of accessible, submerged parts of the

87

ship. The object of this exercise was to confirm the camera arrangement could be used, not to image any particular object. Hence the positions the following images were recorded from was not logged. A ‘Snake Eye’ pole camera was used to record these images. They show what appears to be an adjacent pair of pipes (Figure 95) and, in Figure 96, a circular opening through the upper deck in the vicinity of the aft turret. The visual flare is from the ROV’s lights.

Figure 95 Possibly a pair of pipes.

88

Figure 96 A circular opening through to the lower deck in the vicinity of the aft turret.

5. DISCUSSION

The following discussion will consider the various agreed objectives of the survey (see Section 2.1) and the degree to which they were met.

• This was seen to be a preliminary activity which would inform the longer-term planning for the conservation of the wreck.

The activities that were undertaken were consistent with a preliminary investigation. Given that both external and internal surveys were undertaken, it was possible to take a good ‘snapshot’ of the condition of the vessel in December 2012. Future surveys by other organisations should be able to build upon the data which were collected.

• DSTO could only commit to one week’s survey activity. While every effort would be made to select a week in which the weather window was predicted to be favourable, there would be no option to shift the trial if unfavourable conditions arose.

89

Weather forecasts began to sound unfavourable for the survey during the week before it was scheduled. Unfortunately, the degree of fore-planning that had to be done precluded the possibility of postponing the trial. In the event two full days were lost, but the weather on the three remaining days was acceptable. In the circumstances the results of the three days work are judged to have been quite productive.

• The main deliverable of the survey would be video imagery. DSTO would provide information about the locations on the wreck where the imagery was recorded, but no structural interpretation would be undertaken.

Due to the fact undocumented changes had been made to the interior of Cerberus, it proved more difficult than expected to ‘geo-reference’ the imagery. The authors have made every effort to accurately identify the location and nature of features, but can not provide an absolute assurance no errors have occurred. Nevertheless, several hours of video imagery have been post-processed and it is believed other reviewers will have sufficient information available to them to facilitate the relocation, in the video imagery, of features of interest.

• It was understood that, if possible, particular emphasis would be placed upon inspection of the pillars and beams supporting the turrets, and any debris in the vicinity.

Four separate attempts were made to survey the support structure of the two turrets. In the event this proved to be a very challenging undertaking. A great deal of debris underlies the forward turret, approaching from both the front and rear of the vessel. The ship had watertight bulkheads, before and after the vicinity of the turret. These bulkheads appear to have still been in place at the time the vessel was scuttled. The collapse of the ship resulted in these bulkheads becoming badly broken and distorted. Hence it would be very challenging to successfully fly a ROV of the size of the LBV into the space under the forward turret. All is not lost, however, because the survey identified an opening in the upper deck which may be accessible by a pole camera deployed from above. If successful, such a camera could be used to visualise at least the two adjoining ‘I’ beams (probably Beam 25 and Beam 26). The aft turret presented less of a challenge to the ROV pilots. If two days hadn’t been lost due to bad weather, the authors are confident it would have been possible to approach the turret from aft. Approaching from the front of the ship, it was possible to record imagery of the fore-most ‘I’ beam (Beam 48). While this beam appears to be generally sound, it has a broken portion near its midline. The space aft of this beam is currently very full of sediment. There was insufficient space to fly the ROV under it. It is not known whether or not this level of sedimentation carries right across to the extreme port side of the ship. Again, perhaps two additional days would have facilitated a more extensive survey under this turret. It was not possible to take a formal count of the relative numbers of straight versus bent pillars. A number of pillars are still in place and still appear to be relatively straight. It is not known, however, what they are sitting upon.

90

• An attempt would be made to make measurements of the residual thickness of the

turret support beams using an Ultrasonic Thickness Gauge mounted on the ROV. It was pointed out to the meeting that DSTO’s previous experience with this gauge, on the AE2 submarine wreck, did not generate a great deal of confidence in a favourable outcome being achievable from these measurements.

Unfortunately DSTO’s initial lack of confidence in the likely success of the UT measurements was proved to be justified. Calibration checks in DSTO’s test tank had been favourable, but it proved to be impossible to record a realistic thickness reading on the Cerberus. There appears to be little information available on the practical use of this type of technology. Before attempts are made to use it again, a systematic study of concretion thickness versus instrument performance should be undertaken. A number of the bulb-iron beams that underlie the upper deck remote from the turrets were visually inspected. All of these appear to be intact. The area of exposed flat face of a 9 inch bulb-iron beam is similar to that of a 12 inch ‘I’ beam. It is suggested that expert opinion be sought as to whether measurements of the residual thickness of ½ inch bulb-iron, taken in one part of the ship, would be representative of the condition of the under-turret ‘I’ beams. Given that divers could perhaps safely enter the aft part of the ship via the aft hatch, which is well clear of the breastwork and turrets, subject to such expert opinion, it is suggested consideration be given to measuring the thickness of beams in this vicinity (say Beams 62-64).

• It was suggested that an approach be made to the local Royal Australian Navy Reserve dive team, seeking its support in undertaking this survey. Specifically, the dive team would be requested to provide general support for insertion, removal and tether management of the ROV and to provide a video-based survey of the external condition of Cerberus and her surroundings.

The ANRDT Six supported this survey very effectively. They were able to undertake a very comprehensive video survey of the ship’s exterior and they determined and reported upon the condition of the cathodic-protection anodes on both Cerberus and her guns.

• It was also agreed that an approach be made to MPD’s Environmental Research and Biotechnology Group and Corrosion Group for limited support in relation to their respective fields of expertise.

Environmental surveys using CoA databases and visual imagery suggest that there were no unique or significant biological colonies. This is probably not surprising since the vessel has only been a part of the local environment since 1926 and is contained in a busy embayment. The corrosion potential measurements were generally consistent with what had been recorded in a previous survey by MacLeod and Steyne [2011]. The August 2011 BMT study developed a relatively simple, three dimensional finite element model of the turret structure and weather deck beams. The conclusion drawn from the computer modelling was that ‘Simply put, if the assumptions made are correct,

91

the deck should have collapsed already’ [BMT, 2011, page 7]. BMT provided possible reasons why the collapse had not occurred as yet [ibid, section 2.5.7]. While the DSTO survey didn’t provide evidence to support or discount the potential reasons, at least some were able to be addressed to a greater or lesser degree: • ‘The twisted and bent stanchions and pillars inside the hull provide some support to the weather deck’ While some of the pillars are in place and appear straight, many appear to be either bent or missing. It was not possible to visualise the lower ends of the pillars in the vicinity of the turrets. Given that the whole ship collapsed by approximately 2 metres, it is not possible for any of the pillars to have a direct, intact connection with the ship’s lower hull. The presence of sediment inside the ship, particularly under the aft turret, may be a consideration, however, as regards to provision of support for the lower deck upon which the pillars are standing. • ‘The bulkhead walls and dividers provide some support to the weather deck’ There was no evidence that any of the bulkheads were still in place. This can almost certainly be discounted. There was, however, good visual evidence that bearers were in place and substantially intact. This included regions that underlay the breastwork. There was no evidence of the breastwork separating from the upper deck. Presumably the breastwork, while imposing a load on the upper deck, is also acting as a longitudinal structural member. • ‘The residual plate thicknesses are greater than those estimated using McLeod’s (sic) formulae’. While it was not possible to measure the thickness of the ‘I’ beams under the turrets, the suggestion of measuring the residual thickness of the bulb-iron beams in the aft portion of the ship may give a defensible answer to this speculated reason. The authors have noted there are at least two references, in the public domain, to apparent structural damage in the upper parts of the ship (comprising the armour belt and above). Electrical discontinuities between the forward and aft parts of the ship were reported by Macleod and Steyne [2011], which could be an indication of structural failure amidships. In the current survey careful visual inspection, both above and below the waterline, was unable to pinpoint any specific site at which structural failure was evident. The Maritime Archaeology Association of Victoria report that ‘the bow has dropped, and a large crack has appeared across the deck forward of the number one turret’ [Charlesworth, 2010]. The referred-to crack had no direct relevance to DSTO’s ROV-based survey, so no attempt was made to locate and photograph it. The diver-based survey of the exterior of the ship indicates that the armour belt is unbroken and shows no obvious signs of buckling. It has been noted above, however, that there is a great deal of bent and twisted debris inside the forward part of the ship. Perhaps this has some relationship with the cracked foredeck.

92

To draw a general conclusion: as a result of this survey the current status of the wreck site and the general condition of the vessel itself is better understood. This should assist relevant authorities in planning the next step in the long-term management of HMVS Cerberus.

REFERENCES

BMT (2011) HMVS Cerberus Preservation Concept Solutions Report, OPS/1/7264/R0409, August 2011. Charlesworth, P. (2010) H.M.V.S. Cerberus 1867 – 1926. A page on the Maritime Archaeology Association of Victoria website: http://home.vicnet.net.au/~maav/hmvscerberus.htm. Last updated April 2010. [Accessed 1st April, 2013]. Commonwealth of Australia (2012) Protected Matters Search Tool [Online]. Canberra: Department of Sustainability, Environment, Water, Population and Communities. Available: http://www.environment.gov.au/arcgis-framework/apps/pmst/pmst.jsf [Accessed 14 December 2012]. Friends of the Cerberus (2013) Virtual tour of the vessel, labelled ‘Inside Shield Deck Next To Forward Gun Turret’, http://www.cerberus.com.au/virtualtour/location10s_1024.html, [Accessed November 2012, re-accessed 2nd April 2013]. Irfanview (2010) Irfanview for Windows, Version 4.27. by Irfan Skiljan, http://www.Irfanview.com. MacLeod, I.D. (2013) Personal Communication – E-mail dated 19th march 2013. MacLeod, I.D. and Steyne, H. (2011) Managing a Monitor - the Case of HMVS Cerberus in Port Phillip Bay: Integration of Corrosion Measurements with Site Management Strategies, Conservation and MGMT of Arch. Sites, Vol. 13 No. 4, November 2011, 334-61. Neill, R.A. (Editor) (1999) Proceedings of Shallow Survey ‘99, Sydney Australia. Nicholls, B. (2001) The Three-Headed Dog – Towards the First Battleship. Bob Nicholls, Bowral, Australia. Stepanow, L. (2006) http://www.cerberus.com.au/image_library.html BirdsEye View 06. Cerberus in 2006. Photo with kind permission of Lindsay Stepanow.

93

ACKNOWLEDGEMENTS

The authors would like to acknowledge the very considerable assistance provided by Mr John Rogers of the Friends of the Cerberus organisation. John has contributed a great deal of material, time and knowledge, in support of this report. Dr Ian MacLeod also provided significant background support for the survey. His guidance was greatly appreciated. In addition, Heritage Victoria, and particularly Rhonda Steel, provided very valued support to the survey. On the first day of the survey this extended to provision of a workboat. Peter Harvey made the appropriate arrangements and attended the first day of the survey. One of the Heritage Victoria volunteers, Callum Harvey, also gave freely of his time to support the first day of the survey on board the workboat. Two additional members of MPD’s Automation and Unmanned Maritime Systems group, John Barber and David Taylor, provided essential shore-based support to the survey. Half Moon Bay is a very popular recreational site and the support of John and David was essential to ensure the operation could be undertaken safely and without undue hindrance. Finally LCDR Fabris and his team of divers from ANRDT Six provided professional, timely and comprehensive support for the survey. Without it, the results may have been very different and much less favourable.

94

95

Appendix A: Letter Of Request from Bayside City Council

96

97

98

Appendix B: Chief MPD Response to Letter Of Request from Bayside City Council

99

100

Appendix C: Permit to undertake survey of HMVS Cerberus

101

102

103

DISTRIBUTION LIST


Roger Neill, John Gilbert, Peter Graham, Peter Mart, Clare Grandison, Martin Rowan and Neal Winter

No. of Copies Task Sponsor (CMPD) 1 Bayside City Council 1 DSTO Research Library (Report Distribution Officer) 1 RLSPS 1 Friends of the Cerberus, c/- Mr John Rogers 1 Heritage Victoria, c/- Mr Peter Harvey 1 Dr Ian MacLeod, WA Maritime Museum 1 LCDR Rafael Fabris, OIC ANRDT 6 1 Authors 1 All copies of this report are to be distributed on CD.

Survey of HMVS Cerberus, December 2012 of HMVS Cerberus, December 2012 Roger Neill, John Gilbert, Peter Graham, Peter Mart, Clare Grandison, Martin Rowan and …

Documents