Simulating Costa Concordia

Discussion in 'Boat Design' started by APP, Jan 17, 2012.

  1. CliffordK
    Joined: Feb 2012
    Posts: 24
    Likes: 2, Points: 0, Legacy Rep: 20
    Location: Oregon, USA

    CliffordK Junior Member

    There seem to be 3 Costa topics.
    Under Stability,
    http://www.boatdesign.net/forums/st...eg-list-really-scary-41331-21.html#post533120

    I made a short mental study of the stability of water (or fuel) ballast tanks which I posted there.

    Any water ballast tank that has the compartment above it flooded becomes ineffective at providing stability to the ship. Uncovered ballast tanks would still provide some stability, as well as heavy equipment such as low placed diesel generators.

    If the ship had a port-side list, it is quite possible that the crew tried to compensate with a ballast transfer to starboard. But, it is very difficult to compensate, especially if there is very little transverse compartmentalization. The goal should not be to bring the ship to vertical, but rather keep it from listing further to the port.

    My guess is that they overcompensated with the ballast transfer, allowing the ship to roll to the starboard when it caught the wind.

    Anyway, so while water ballast can provide stability to these top heavy ships while they are intact, they loose that added stability if the hull is compromised.

    Rather than using water ballast, if a large portion of the ballast was lead, or perhaps heavy steel armor on the bottom, then the ship would be much more stable when inundated with water.
     
  2. Starbuck1
    Joined: Feb 2012
    Posts: 30
    Likes: 1, Points: 0, Legacy Rep: 17
    Location: San Francisco

    Starbuck1 Junior Member

    Costa Victoria Propulsion Details & CSections

    Smartbight,
    Please see the following.
    http://www.sam-electronics.de/dateien/pad/broschueren/1.001.pdf

    Sam-electronics was the manufacturer of the main electrical control system. Their brochure has two cross sections and lots of other interesting details.

    It appears quite similar in layout to CConcordia, just 76000 GRT rather than 114000 GRT.

    Let me know if you have problems opening it.

    One other question. Both the Costa Victoria plan and the Costa Concordia photos have what appear to be numerous "doors" in the side of the hull low down in the 0 and A deck (3&4 Solas deck) area. No ship I was involved with had doors this low down. I'm guessing they are either welded up doors for installations, or access doors for loading of supplies since they open into the crew/utility-service area of the ship at a dock level. Could they have been part of the problems?

    Thanks,
     
  3. IEWinkle
    Joined: Jan 2012
    Posts: 97
    Likes: 15, Points: 8, Legacy Rep: 184
    Location: Glasgow

    IEWinkle Retired Naval Architect

    Revision of Post 58 Simulating Costa Concordia – a Scenario for Progressive flooding

    Analysis of the profile photograph and cabin arrangements (see PDF file) – particularly of Deck 1 (Olanda) suggests that the structure of the ship is made up of a series of equal bays of length 2.9 m (one overall cabin spacing) which corresponds to the spacing between the large transverse web frames supporting the longitudinal frames running throughout the lower structure, implying a transverse frame spacing of 725 mm through the double bottom (in the zone of damage). Multiples of these bays (4, 5 & 6) allow the position of pillars between lifeboats to be established giving the possibility to reference the damage location fairly accurately in terms of web and frame locations. This allows an estimate of the location of the major damage (of the soft side shell) on the port side as running from 37.7m forward of the AP to 74.4m forward of the AP. The AP appears to be 10 m forward of the transom at the Load Waterline, in line with the twin rudder stocks. The damage to the bilge area only appears to run forward to about 92.8 m forward of the AP. Careful analysis of the arrangement in 'profileCCONCORDIA9 Rev3.pdf' together with the photos of the early stage fabrication of the vessel, suggest that the deck heights can be deduced as:
    Deck 1 – 14.18 m (given as vessel’s depth)
    Deck 0 – 11.28 m + 4 longitudinal spaces of 725 mm = 2.9 m
    Deck A – 8.38 m + 4 longitudinal spaces of 725 mm = 2.9 m
    Deck B – 5.48 m + 4 longitudinal spaces of 725 mm = 2.9 m
    Tank Top - 2.03 m + 6 longitudinal spaces of 575 mm = 3.45 m
    The significant change here from my earlier scenario (post 58 – Simulating Costa Concordia) is the height of Deck A which modifies the analysis below.

    Overlaying this information on 'profileCCONCORDIA9 Rev3.pdf' identifies at least four (from a bulkhead around 31.85 m forward of AP to one at 89.125 m forward) and possibly five compartments along the length (up to 101.45 m forward) that will have been compromised. The ‘definitely flooded’ compartments would include Compartments 4 to 7, the two main diesel generator rooms and the motor room aft - the after generator and motor rooms subject to immediate inundation, with the forward generator room flooding more slowly (hence the blackout after 10 mins). Overall damage length therefore varies from 57.275 m (min) to 69.6 m and will include a substantial length of double bottom port wing tanks of some 14 sq m cross-section which were probably empty in this condition (no obvious oil leakage and no significant ballast). If we assume about 60 m of these tanks were flooded together with a rock of some 70 tonne (immersed weight) there would be a port listing moment of some 880 tonne at about 14.25 m from the centreline which would displace the vessel’s centroid by about 0.244 m. This would produce an initial list of about 7 degrees to port if the GM is assumed as 2 m.

    Analysis of the night photo as the vessel approaches the shore heeling to starboard shows a trim by the stern of 4.95 m between perpendiculars giving draft aft of 12.05 m and draft forward of 7.10 m (see post 57 Simulating Costa Concordia). Heel would appear to be about 13.2 degrees to starboard putting the bulkhead deck about 3 m underwater at its starboard after corner. The estimated inflow of water (lost buoyancy or added weight) for this condition based on an estimated WPA of 8339 sq m with an LCF 15.89 m aft of amidships would be about 14470 to 15325 tonne depending on the initial draft which probably ranges from the maximum load draft of 8.2 m down to 8.1 m (estimated from photos of the bow draft marks on 2 separate photos). If we deduct the 880 tonne loss of buoyancy in the double bottom to port, discussed above, that leaves 13590 to 14445 tonne to fill the compartments above the double bottom at an assumed permeability of 85%.

    Noting the presence of B deck at 8.38 m above base and assuming all water was contained by this, then the flooded length would be from 68.86 to 73.20 m. If this deck is assumed to only run for 25% of the breadth each side outboard of a machinery casing running from 56.5 m to 78.25 m forward of the AP which is flooded to an average depth of 8.687 m (mean local draft 10.687 m), the casing would contain about 776 tonne of floodwater and the flooded length would then be from 64.94 to 69.27 m. The latter figure almost exactly matches the assumed maximum 5 compartment length and provides the basis for a stable near upright condition which could be maintained as long as there is no progressive flooding into any of the wing areas of Decks A or 0. To provide the necessary extra floodwater to cause the starboard list it might be assumed that the flooding was contained by B deck in Compartment 8 as all damage in this area is only to the bilge corner. This would allow up to 1100 tonne of ‘spare’ flood water to permeate the starboard spaces to counteract the lost double bottom buoyancy and rock to port to give the condition seen in the night-time photo.

    Exactly how this happened has yet to be determined, but in evacuating the machinery spaces, one or more doors giving access to the starboard wing spaces may have been left open onto A deck (possibly around the engine control room which seems to be at the aft end of compartment 6) which allowed progressive flooding to establish itself in the 30 minutes or so before the vessel grounded, resulting eventually in the 13.2 degree list to starboard deduced from the photo. Continued acceleration of this progressive flooding would have resulted in the vessel’s increasing list to starboard experienced as the starboard lifeboats were launched which eventually immersed the passenger decks and led to complete capsize onto the rocky shelf. If this progressive flooding had been prevented the vessel could have probably survived, near upright, close to the final centreline draft deduced above despite her 5 compartment damage case – although this would not have been considered a SOLAS 90 survival case in the normal course of events.
     
  4. smartbight
    Joined: Dec 2006
    Posts: 112
    Likes: 8, Points: 0, Legacy Rep: 55
    Location: London

    smartbight Naval Architect

    Great find. This must be the only existing engine room profile on the Internet.
    After looking at it; We found one more reason to move the motors to comp. 5. We had forgotten the thrust blocks (keeps the wheels from pushing the motors fwd). Those 'mothers' are big enough to claim their own compartment 4.
    Those are standard side cargo/passengers doors, very strong. The only ones that were closed, for sure.
     

    Attached Files:

  5. IEWinkle
    Joined: Jan 2012
    Posts: 97
    Likes: 15, Points: 8, Legacy Rep: 184
    Location: Glasgow

    IEWinkle Retired Naval Architect

    Addendum to Post 78 Simulating Costa Concordia – a Scenario for Progressive flooding

    The sketch attempts to show the waterline W3L3 (in red) relating to the 13.2 degree heel as seen in the photo discussed as it intersects the Engine Casing lying between the two main generator rooms. At this point substantial flooding has occurred into the Starboard side of Decks A and possibly 0 thorough a point somewhere just above Deck A as shown.

    To make some sense of the sequence of flooding I have proposed 2 intermediate waterlines W1L1 and W2L2, the first of which represents the condition after initial stabilisation of flooding at which overtopping of an opening to starboard probably started. In this condition the vessel would still have a small list to port to support the extra weight of flooded double bottom + rock on the port side. As the flooding continued the list would slowly move to starboard (W2L2) as the extra floodwater counterbalanced the excentric port load and from here progressive flooding would begin to accelerate. Much beyond W3L3 and the loss of the stabilising influence of the port side of deck A would lead to significantly reduced stability - again accelerating the final capsize.

    In the initial phase of flooding it is probable that for a minute or so the floodwater would pile up to port as suggested by the 'Intermediate Flooding in Engine Rooms?' line as it attempted to rush across the ship past the fixed plant and equipment. During this initial phase it is likely that the ship rolled to about 10 degrees to port (as a continuation of the roll as it picked up the rock) and this would have slowly corrected itself as the water flooded up to waterline W1L1. During this period the loss of waterplane inertia would have been significant enough to knock about 4.25 m off the BMt and the ship would have been technically unstable (in a state of loll) which would have quickly been recovered as A deck hit the water. Once A Deck to port left the water, after W3L3, that instability would have been re-established through the massive free surface running across the machinery space further accelerating the capsize.
     

    Attached Files:

  6. Starbuck1
    Joined: Feb 2012
    Posts: 30
    Likes: 1, Points: 0, Legacy Rep: 17
    Location: San Francisco

    Starbuck1 Junior Member

    Wow!

    Nice Work! EIWinkle and Smartbight.

    I wonder if the starboard opening was a WT door, non WT door or a ventilation return.

    FYI, the Costa Victoria info came from one of the participants on gCaptain.com.

    I wonder how we could find/connect to an engineer from the CC that would participate.

    Thanks!
     
  7. janneke
    Joined: Jan 2012
    Posts: 4
    Likes: 2, Points: 0, Legacy Rep: 38
    Location: Belgium

    janneke New Member

    I find it quiet difficult to understand that the stabis were extended under a windforce of 12 kn, which is just at the lower end of force 4 scale, i.e. a "moderate breeze" where "small branches begin to move" (looked it up in wikipedia).
    I have been on cruiseships with more severe winds w/out the stabis extended!
    And the question still stands : why is the stabi NOT destroyed by the rock boulder? The stabi falls well within the turning circle of the ship when it was turning to st side.
    As Heiwa would say : MAGIC?
     
  8. IEWinkle
    Joined: Jan 2012
    Posts: 97
    Likes: 15, Points: 8, Legacy Rep: 184
    Location: Glasgow

    IEWinkle Retired Naval Architect

    See my posts 216 and 219 on page 15 of forum http://www.boatdesign.net/forums/stability/costa-concordia-80-deg-list-really-scary-41331-15.html
    It is not difficult to explain the fact that the stabiliser missed the rock - it would have cleared it by about 1 m! It does not require any MAGIC!
     
  9. janneke
    Joined: Jan 2012
    Posts: 4
    Likes: 2, Points: 0, Legacy Rep: 38
    Location: Belgium

    janneke New Member

    I'm still sceptic. But how do you come to that 1 m of clearance and in which direction : vert or hor?
    I'm more inclined to believe that the stabi was deployed AFTER the encounter with the rock. Notice also in which direction it is turned : to give max lift on port side! Coincidence???
    Also you used the word "probably" in post 216, which in my opinion means that you have no proof and hence not sure of this idea.
     
  10. IEWinkle
    Joined: Jan 2012
    Posts: 97
    Likes: 15, Points: 8, Legacy Rep: 184
    Location: Glasgow

    IEWinkle Retired Naval Architect

    The stabiliser lies roughly horizontal well above the line of the bilge keel which was the first point of contact with the 'rock' . The aft end of the ship also closed fairly rapidly with the 'rock' to give a deepening cut as the ship progressed and turned. The position of the stabiliser well forward of the first impact suggest that little more than its tip was close to the 'rock' and passed clear over it - the 1 m is but an estimate at the vertical clearance based on the geometry of the vessel.
     
    1 person likes this.
  11. Starbuck1
    Joined: Feb 2012
    Posts: 30
    Likes: 1, Points: 0, Legacy Rep: 17
    Location: San Francisco

    Starbuck1 Junior Member

    There is an interesting video out there from a camera in the ship of the moment of impact in a lounge or concierge desk. First a heel, then a sharp heel and all the furniture starts sliding back and forth, with people falling down and holding on to anything nearby with the ship rocking at least 10º each way for 6-8 gradually dampening rolls with stuff sliding vigorously.

    It may have been an automatic response by the ER crew to deploy the fins to stop the rolling in the first five minutes for passenger comfort while they still had power, but it wouldn't be any where near my first priority with my engine rooms flooding rapidly and power going down. Any ideas about how long they take to deploy?

    The AIS course modeling on gCaptain.com has a very good illustration of how the stern swings/skids 200-300' to port outside of the line of travel in a hard turn to starboard, with the center of turning and movement being about 30% aft of the bow. Think skidding a car and the rear wheels going off the road and whammo! hitting the end of a guardrail. No damage to the driver's door, but the rear panels and driving wheels are a mess.

    There were early reports that the captain tried to swing back to port to clear the stern once the bow was past the point, but the 53000 ton ship didn't respond quickly enough.
     
  12. CliffordK
    Joined: Feb 2012
    Posts: 24
    Likes: 2, Points: 0, Legacy Rep: 20
    Location: Oregon, USA

    CliffordK Junior Member

    I think it is in part a fault of a stern rudder which requires turning towards an obstacle to initiate a turn.

    Try pushing a shopping cart backwards through narrow store aisles. Push it up close to a counter, then see how easy it is to recover only pushing it backwards. Of course, in water it will tend to slip sideways in the direction of motion somewhat too.

    Bow & stern jets might help. Are they used in general navigation? Were they used by the Costa Concordia for obstacle avoidance?

    It would seem that a bow rudder could initiate a turn quicker than a stern rudder, although as I understand it, it is difficult to control the turn with the bow rudder. But, it might make a difference if one could initiate a turn 1000 feet earlier than the stern rudder.

    Or, perhaps a forward mounted below hull rudder that could be lowered like a sailboat centerboard to initiate turns, and raised to avoid taking excess draft in harbors. Two of them may even be able to be mounted at fixed angles with the turning force controlled by the depth of insertion.
     
  13. rwatson
    Joined: Aug 2007
    Posts: 5,852
    Likes: 288, Points: 83, Legacy Rep: 1749
    Location: Tasmania,Australia

    rwatson Senior Member

    So much anger with so little research done. A little knowledge IS a dangerous thing.

    If you go through the simulation videos at the beginning of this thread, it explains why the impact was behind the stabilizers very clearly.

    It was a case of stern drift after evasive action was taken, no magic necessary.
     
  14. smartbight
    Joined: Dec 2006
    Posts: 112
    Likes: 8, Points: 0, Legacy Rep: 55
    Location: London

    smartbight Naval Architect

    There is your answer, as quoted from the Telegraph:

    http://www.telegraph.co.uk/news/wor...ed-ships-speed-for-dinner-with-ex-dancer.html

    "The captain “slowed down the ship so that he could finish dinner in peace”, just prior to sailing close to Giglio in order to perform a ‘salute’ to an old colleague on the island, prosecutors alleged in a report.

    He then ordered the ship’s speed to be increased to 16 knots “despite the proximity of obstacles, the presence of shallow water, the conditions under which the ship had to manoeuvre and the night-time darkness,” prosecutors charged.

    As a result of the increased speed he was unable to maintain “an adequate distance” between the ship and the island. "
     

  15. rwatson
    Joined: Aug 2007
    Posts: 5,852
    Likes: 288, Points: 83, Legacy Rep: 1749
    Location: Tasmania,Australia

    rwatson Senior Member

    Just having re-read the post, I think I missed one of the points - that the stabilizer MAY have been deployed after the crash.

    Although the crash could have missed the stabilizer based on the course information, would it have mattered if it had been deployed after the crash ?

    Considering that the whole boat lost power very quickly after the engine room flooded, maybe the stabilizer couldn't have been activated anyway.
     
Loading...
Forum posts represent the experience, opinion, and view of individual users. Boat Design Net does not necessarily endorse nor share the view of each individual post.
When making potentially dangerous or financial decisions, always employ and consult appropriate professionals. Your circumstances or experience may be different.