3.7 The Accident according to the Commission - Sequence of Events - the Atlantic Lock
How the visor was detached from the superstructure is not clearly described in the Final Report (5) and there is no evidence for any 'events' proposed by the Commission leading to the ramp being ripped open. Furthermore there is no logic between the various 'events' and their causes and the times when they occurred. It is therefore very probable that the visor was attached to the ship when it sank. Regardless below follows an analysis of the suggested sequence of events proposed by the Commission.
The commissione has produced a video MS Windows MediaPlayer v.7 CIF 352x288, 256 Kbps, 6,9 Mt of the events below - or a variation of it. There is no evidence for any essential statements in the video. Normal waves do not even lift the visor! That the port side lock failed first is not proven. That the visor can pull open the ramp is not proven - the ramp is locked. But if the visor is lost and the ramp is down and the speed is forward, water evidently enters the superstructure - it is well shown. Then the 'Estonia' should have capsized, turned upside down and floated. Not sink as shown in the video.
Ten Minutes of Noise
In chapter 1 The Accident (page 22 of (5)) it is stated that
"a metallic bang"
was heard shortly before 01.00 hrs and that
"Further observations of unusal noise, starting at about 01.05 hrs, were made during the following 10 minutes"...
"Shortly afterwards (i.e. shortly after 01.00 hrs) (Linde) was sent down to the car deck to find out the cause of the sounds reported by telephone to the bridge"...
"At about 01.15 hrs the visor separated from the bow ... The ramp was pulled fully open ... Very rapidly the ship took on a heavy starboard list".
Five Minutes of no Noise
Chapter 13.2.5, page 175 of the Final report (5) states that at 00.55 hrs a sharp metallic bang was noted from the bow area and that seaman Linde on the car deck in the superstructure at that time waited five minutes (sic) behind the ramp and did not note or hear anything unusual - no further noise or impacts or 'bangs' in spite of the fact that the ship continued at 15 knots speed in the heavy weather and model tests indicated >100-tons impacts every minute. Then Linde maybe quietly continued his patrol round to decks 1 and 0 down into the hull (it is not clear) and returned to the bridge on deck 9 where he arrived at 01.00 hrs (sic), witnessed the change of watch and the reception of a telephone call to the effect that Linde later was sent down to the reception on deck 5 at about 01.05 hrs to investigate noise and water in the ship. Then the Final report states that
"Shortly after one o'clock a few wave impacts on the visor caused the visor attachments to fail completely",
i.e. the bottom Atlantic lock on deck 2, side locks in the superstructure and the hinges on deck 4 were broken.
According to surviving passengers 2.1 there were two sharp bangs just prior to 01.00 hrs and then the ship suddenly listed at 01.02 hrs but let's assume that only the times are wrong.
Then (according to the Commission)
"the visor started cutting openings in the weather deck (deck 4) plating and associated structure."
(which is further discussed in 3.10 and 3.11 - there is no evidence that the visor was completely lose and cut openings in the weather deck). Then
"Soon the back wall of the visor housing came into contact with the ramp, hitting its upper edge and thus (sic) breaking its locks. The ramp fell forward and remained resting inside the visor. In a few minutes the visor started falling forward."
There is no evidence that the lose visor ever hit the upper edge of the ramp or that the ramp locks were broken, but let's assume that it happened. Therefore the ramp became fully open -
"allowing large amounts of water to enter the car deck ..."
This should have occurred at about 01.15 hrs when the speed was still 14-15 knots straight into the waves. Then the ship listed slowly (sic) and sank until 01.50-01.55 hrs. You would have expected that the ship with an intact hull would have capsized and floated upside down with 1 900 tons of water 'loaded' in the superstructure, but it never happened. While all surviving passengers testified about a sudden list >30 degrees (at 01.02 hrs), the star witness of the Commission, 3/E Treu, suggested that the vessel slowly listed at 01.15 hrs, when water leaked in at a virtually closed ramp - the ramp was seen in the up position! Linde was either at the reception on deck 5 or somewhere else, when the ship suddenly listed >30 degrees.
In the Part Report (16) the only events between 00.45-01.15 hrs were:
'At about 00.45 hrs (sic) several unnormal signs were noted in the ship. Metallic noise was heard in the ship'.
That was all. No other times for the different events were given in the Part report, and in the Final report the starting time is at 00.55 hrs. But according to the Part Report noise was heard 30 minutes (!) before the visor actually fell off.
Repeated metallic Noise
The Final Report (5) however states that
"This sequence of events is supported by witnesses from several areas on board who heard repeated metallic noise from the bow area during a period of about ten minutes starting shortly after one o'clock." 2.1.
Now the noise starts at 01.05 hrs. Unfortunately and unexplainable the Final report does not quote any witness hearing ten minutes of noise, because it is an invention by the Commission, but let's listen to the Commission to the end. The Commission thus states in the Final Report (page 175) that
"The failure sequence ... is described further in (chapter) 13.5".
However, in chapter 13.5 (pp 180/1 of (5)) the failure sequence during these 10 minutes is not described in any great detail supported by any evidence.
Initially it is stated in 13.5 that
"The attachments (i.e. three locks and two hinges) may (sic) have failed in one or, possibly a few steps. The partial initial failure may have coincided with a single metallic bang observed by the AB seaman " (i.e. Linde at 00.55 hrs).
It does not sound very convincing! What failed initially? Was the visor actually locked at this time or was it only secured by ropes as suggested by, e.g. the Germans? The Commission does not give any clear answers. Then it is stated that
"The main failure (what is it? - the writer's remark) is believed (sic) to have happened in a subsequent wave impact, shortly after the metallic bang" (at 00.55 hrs).
"... In this main failure the remaining locking (?) devices failed completely ...",
i.e. first the Atlantic lock or the port side lock was broken according to the Commission - the bang at 00.55 hrs or 01.10 hrs - and then the remaining (side) locks (and maybe the hinges - they are not locks) were broken - the main failure? - but in chapter 15.10 (page 194) it is stated that
"Most likely (sic) the port side lock failed first",
i.e. before the Atlantic lock, in spite of the fact that the side locks were always under compressive loads when the Atlantic lock was closed and could not be pulled apart. The hinges were broken later, because it is stated further down in chapter 13.5 that
"Once the visor had lifted off its locating horns (i.e. all three locks holding the visor down were broken), the port side hinge failed under the overload generated by the high twisting and yawing moments and the vertical force".
There is no evidence at all for any high twisting and yawing moments acting on the lose visor at this time - it is another invention of the Commission - and that the visor actually fell off, but the hinges must break for it to happen. Then
"The starboard side hinge failed as a result of twisting ...".
No evidence for that either; it is just an unproven statement by the Commission. Because in the Part report (16) the Commission had stated something completely different.
False Statements in the Part Report
In the Part report (16) page 24 (published 4 April 1995) the Commission reported that the hinges were broken by forces in the upward and forward directions resulting from the hinge arms being in contact with the weather deck as follows:
"The fracture at the lugs of the hinges occurred in tension. The fracture surfaces show that failure occured during one short overload. The hinge lugs and the weldings have been examined by metallurgic and strength of material analysis. The results of the examinations will be published in an addendum to the Final Report.It is considered probable that the forces that broke the hinges occurred, when the hinge arms hit against the weather deck (deck 4) as a result of the bottom part of the visor being compressed (the bottom is not compressed). The lever effect which then developed in combination with inertia forces, when the visor hammered on the fore peak deck (deck 2 - totally undamaged), was sufficient to pull apart the hinges".
The visor bottom part is really not very compressed, and if it 'hammered' against the fore peak deck 2 - not proven, the fore peak deck 2 is undamaged.
An addendum with a metallurgic and strength of material analysis report dated before the 4 April 1995, that the hinges were broken in tension and during one short overload, does not exist.
No evidence how the Hinge Arms were ripped apart
In fact there are no reports or supplements in the Final Report (5) at all (e.g. Supplements nos. 510, 511, 516, 517 and 518) to the effect why and how the hinges had been pulled apart, e.g. in tension. Only Supplement no. 518 discusses the fracture surfaces of the hinge lugs - the only conclusion is that available information does not support the conclusion that previously existing fatigue fractures should have played a part in the failure.
No Results of Examinations published
The statement in the Part Report (16) that "The results of the examinations will be published in an addendum to the Final Report" is not correct. It was probably intentional lie 1995 - no examinations or analysis had been done! And no reports could later be presented to substantiate the false statement.
Chapter 15.5 in the Final report (page 191) states:
"The lower rims of the hinge plates had generally failed under tension and the upper ones under bending ... with clear signs of bending overload. ... The lugs and one recovered hinge bushing have undergone metallurgical and strength investigations as described in 12.7 and in more detail in the Supplement;"
As the Supplement does not describe the surfaces of the fractures in the hinge lugs, we have to look in chapter 12.7 - (page 168) in (5):
"The studied failures (of the hinge fracture surfaces) were of ductile character and signs of fatigue were not seen"
Suddenly the existing fatigue fractures had disappeared and no information how the fracture occurred is given.
Chapter 15.5 thus states (page 192)
"It is most likely (sic) that the forces to cause the hinges to fail were generated when the visor, moving upwards around its hinges and having lost support from the locating horns, was exposed to twisting and yawing moments".
Why the visor was exposed to twisting and yawing moments during the very short time it had lost support from the locating horns is not explained. The visor was only a few meters below water 1-2 seconds every 7-8 seconds, when the ferry pitched down into the waves.
That the Commission does not know what it is talking about is shown in the next sentence:
"... it is also possible (sic) that the initial failure of the port hinge was caused by high reaction (sic) forces before (sic) all locking devices failed."
Now suddenly the 'initial failure' was in the port hinge before all locks failed - caused by a reaction (?) force. But it could have been the port side lock ... or the bottom lock that failed first.
Actually, with all three locks locked, there were no loads/forces at all acting on the hinges!
As shown in the previous chapters the hinges were not transmitting any loads at all from the visor to the superstructure under any conditions as long as the locks were applied. The hinges were only used to lift the visor in port.
Fractures in the Hinge Lug before the Accident
Fractures had been observed in the hinge lugs before the accident -
"... the cracks generated during normal service ..."
are noted in the next sentence but are never analysed. Cracks during normal (sic) operation reduce strength and are due to either fatigue, overload or design/manufacturing fault 1.18. These must be repaired and/or be reported to the Class. But of course - the only time the hinges were under tensile load was when the visor was lifted up/down ... in port - by the hydraulics. If waves tried to lift the visor the force in the hinge was compressive!
In the Final Report (5) chapter 8.12 the Commission stated that cracks had been seen in the hinges before the accident without being repaired and the Commission had no comment, except that the crack were in an area not visible, when the visor was closed. The Commission had stated that the 'Estonia' was in good shape and seaworthy and considers that cracks in the visor hinges are 'normal service'. And not to forget one theory - "the initial failure of the port hinge was caused before all locking devices failed"! The question whether the accident was caused by the port visor hinge fractures is hanging in the air.
But how could wave loads on a locked visor (two side locks and the Atlantic lock) damage a hinge that did not transmit any loads at all?
Constructive Weaknesses cannot be detected
The Swedish NMA boss - Johan Franson 1.16 has an interesting comment to the above (in the Swedish daily Göteborgs Posten/Debatt on 28 February 2001):
'The constructive weaknesses which together with severe weather caused (the accident) cannot be detected by ... a periodic inspection ... . It is something that must be detected in connection with a newbuilding inspection'.
Thus - professional Swedish inspectors of ships at inspections cannot detect cracks and deformations due to overload or fatigue after 15 years of normal operations !!!
However - there are several theories how (and when?) the hinges broke - either they were bent sideway or pushed up and pulled forward 3.9 but the Final Report (5) has no definitive conclusion. Maybe the port visor hinge was finally pulled apart under water, when the ship sank, and the starboard hinge was ripped apart even later - when the visor was finally removed by explosives under water and pulled away from the ship?
But let's assume like the Commission that the hinges broke first and the locks later. According chapter 13.5 the visor was then
"constrained in the longitudinal direction"
by the hydraulics, but in spite of this ...
"Impact marks indicate violent transverse movements"
No photos of impact marks due to transverse movements are shown in the Final report (5). Actually there are no such marks! Then the visor cut the weather deck (deck 4) plate and structure (a deck beam) and the collision bulkhead of the superstructure and - the visor pulled open the ramp, when the visor fell off. However
"The exact timing of this development cannot be determined (sic) ..."
and
"The many uncertainties involved make detailed calculations of this development meaningless (sic)" (page 181 in (5)).
It was thus meaningless - after three years of investigations - to show how the ramp protecting the superstructure was pulled open and caused the flooding of the deck 2 in the superstructure. Note from the above that it is not clear at all when any locking device of the visor ruptured and then we are not told why, how and when the ramp was ripped open.
Actually the Commission has presented at least three versions/proposals/hypotheisis how the visor became lose by the waves:
1. the port hinge could have broken first (at 00.55 hrs), or
2. the port side lock (in compression!) could have broken first (at 00.55 hrs), or
3. the bottom-Atlantic lock could have broken first (at 00.55 hrs).
None of the three proposals are followed up by what happened then, e.g. if the port hinge broke, why would the locks fails between 01.05 and 01.15 hrs, etc.
Finally the Commission states that
"the course of events described is fully possible.",
i.e. as no course of events is really described at all, it is clear that there is no evidence for any events including the allegation that the visor actually fell off! Not even the times are established - the Commission stated that
"The time for the full failure sequence ... may have been 10-20 minutes",
i.e. as the 'full failure sequence' ended at 01.15 hrs, it must have started 00.55 hrs or 01.05 hrs during a period when the crew did not do anything other than to ask Linde to go down and have a look. Full speed was maintained during the complete 'full failure sequence'.
Passenger Observations ignored
It is at this time 00.55 hrs that passengers observed water on deck 1, before the sudden listing, and when the sudden list occurred 2.1 at 01.02 hrs, which the Commission conveniently decided to ignore.
The writer thinks the above course of events of the Commission is fascinating fantasies. Not one word of it is true - but it is interesting to study the disinformation. Read on! How did the Atlantic lock fail? Was it an impact on the visor? Or was the lock damaged long before the accident?
Model Tests of Wave Forces on the Visor
In chapter 13.5 of (5) the Commission states:
"... the maximum resultant force (on the visor was) between 4 and 9 MN,"
in the severe weather. They refer to chapter 12.1-12.3, where this should be shown and to chapter 15.2, where it is summarised.
1 000 tons Force hits the Vessel
In chapter 12.1.3 is stated that once during model tests a force on the visor was measured which was 7.7 MN in x-direction, 7.4 MN in z-direction and 2.2 MN in y-direction, i.e. the total resultant force was 10.9 MN, i.e. say about 1 090 tons, which is more than stated in chapter 13.5. If such a force actually hit the real ship 'Estonia' is not selfevident. It should have been noticed aboard - a 1 000 tons impact hitting the visor/ship during less than 0.5 seconds would result in enormous noise and structural damage and the vessel would have stopped immediately and everybody standing aboard would have fallen to the deck! Actually, there is no way that such big impacts could have hit the little visor. The model tests cannot be re-done by an independet model test basin!
In Figure 12.3 in (5), reproduced below, is shown that only the vertical force (in z-direction) was >4MN (>400 tons) six times during 1 000 seconds of model test, i.e. every third minute the visor experienced a vertical (z-direction) impact force >400 tons. These big impacts apparently occur when the ship encounters big waves >6 meters (not very big actually) - the normal waves are about four meters high but seven, eight times in 1 000 seconds the waves are more than seven meters high - and every time there is an impact. The writer has serious doubts about these model tests Appendix 2 and considers the written report of the test as pure falsifications (the forces have been enlarged 5-10 times or did not exist at all).
There exists no method to extrapolate full scale forces from forces measured in model tests. The full scale Z force in figure 12.3 from (5) below is not proven (and as stated before - the model tests and the extrapolation to full size cannot be re-done by any independent scientist!)
In chapter 12.2.3 (pp 157/8) it is stated that
"There is a chance of about 1/20 that during 30 minutes of exposure the extreme load was larger than the the value corresponding to 10 hours mean exceedance period".109
The extreme loads in z-(vertical) direction on the visor, with 30 minutes respectively 10 hours mean exceedance period are, according table 12.3 (in 150° heading, 15 knots speed, 1 meter bow wave and 4.0 meters significant wave height) in (5) only 2.95 respectively 4.20 MN, i.e. much less than was measured in the model tests. In chapter 12.3 the Commission summarises above (table 12.5) that the most probable vertical (z-direction) load was 3.6 MN (about 360 ton) during 30 minutes. In chapter 15.2 - the summary - The Commission repeats that the maximum resultant force (on the visor was) between 4 and 9 MN, but adds that the maximum vertical load was only between 3 and 6 MN, in spite of a load of >4MN every third minute in model tests. No corrections were made for the weight of the visor or water inside the visor.
Figure 3.7.1 - Figure 12.3 in (5) Example of time series from model tests
The Commission thus suggests that a vertical impact force (P - figure 3.2) of 360 tons (3.6 MN) pulled open the Atlantic lock at about 00.55 - 01.05 hrs - it might also have been the port hinge or a side lock that broke first even it the former did not carry any load, when the Atlantic lock was intact, and the latter was in compression, when the Atlantic lock was intact.
Frequent Impact Loads - but with little Energy
There is evidently no evidence for a sudden impact load on the visor causing a force breaking the Atlantic lock. The model tests above show that there were impacts on the visor causing a vertical Z-force >400 tons every third minute and >300 tons every minute. But AB Linde had heard only one big 'bang' - the 360 tons impact? - sometime before 01.00 hrs and then it was quiet for five minutes. Most passengers only heard two big, sharp bangs and then there was a sudden listing. But the model tests show clearly that there were wave impacts of >300 tons every minute! What shall we believe? The evaluation of the model tests is apparently manipulated, i.e. when extrapolating the full scale Z and X forces from the model tests.
How the impact load was transmitted via the locks and hinges to the hull is not known and never explained. There were unknown wear and tear in all attachments but the five attachment points should however, in the writer's opinion, have easily transmitted impact loads to the supporting structure because of their short duration.
There was never enough energy in the impacts to damage the locks.
And evidently, if there were not enough energy in the impacts to damage the locks, there was not enough energy to lift the visor itself up from the supports. If there were enough energy to lift the visor up from its supports, the same energy would ensure that the visor came crashing down on its lower supports, resulting in a lot of damage to the hull: but the forepeak deck 2 on the wreck below the visor is completely undamaged.The Commission suggested, as one possibility, that it was the Atlantic lock that was pulled apart in tension first. The Commission assumed that the Atlantic lock was in perfect and undamaged condition before the accident, even if there is no evidence for that.
|
Actually there is no evidence at all that the Atlantic lock was undamaged before the accident or that it was in use. There is of course the possibility that the Atlantic lock had been damaged before the accident and was not in use. The Commission never examined these possibilities. The Atlantic lock looks as follows (figure 15.2 in (5)) right: |
Figure 3.7.2 - Figure 15.2 in (5) Bottom locking device |
The Damages of the Atlantic Lock
All three lugs holding the two bolt bushes had been torn apart in the 8 and 2 o'clock positions. The starboard bush had been torn off the transverse lug and had disappeared. The port bushing, welded to two lugs, had also disappeared, in spite of the fact that it should have been attached to the locking bolt. The locking bolt itself was seen attached to its hydraulics in the pushed open, i.e. locked position. The bolt was salvaged and later thrown back into the water without being photographed.
|
The lug on the visor itself can been seen on figure 10.5 in the Part Report (16) reproduced right: it was bent to starboard and its welded connections to the visor lower stringer were damaged - the horizontal stringer web plate and the face flat were buckled (!) on the starboard sided and were fractured on the port side of the lug. Such damages rarely occur, if you pull in the lug - the damages are clearly due to a force from starboard to port on the visor! |
Starboard is right in the photo |
(A) The lug of the visor (or its welded connection to the visor stringer) should have been pulled apart first, as it was the weakest part of the lock - see fig. 3.7.3 above. Now the lug was bent and the connection to the visor buckled and fractured.
A longitudinal pulling force can hardly bend the lug sideway and buckle/fracture the visor stringer.
On the other hand if the visor were stricken off sideway after the sudden list, the lug would bend and the connection buckle/fracture (see also the German ideas about the bending of the lug in 1.22). However, it is also possible that the visor had been stricken sideway before the accident and that the lock was damaged before the accident - see below.
Below figures 3.7.4-5 show how the visor lug is pulled apart in model tests paid for by the Germans, where the lock bolt bushings are connected with 3 mm welding to the lugs. The lugs are welded to the fore peak deck 2. The longitudinal pulling force is about 210 tons, when the visor lug is broken! The force is easily transmitted to the forepeak deck via the bolt, bushings and their lugs.
The visor lug is the weakest part in this condition!
The Final report (5) page 167 states that the visor lug (original hole diameter 85 mm) was found to be plastically deformed >6 mm and that you need a pulling load of 150-180 tons to make that deformation (but >210 tons when the lug is pulled apart). This means of course that the the three three lugs welded on the forepeak deck would have withstood >180 tons.
Figure 3.7.4 - close up of broken visor lug and deformed locking bolt (B) If the visor lug for any reason did not
break (if it had been of stronger material), the
locking bolt should have sheared off (see figure
3.7.4 above), as it was the second weakest part.
According the Commission the bolt was undamaged,
when it was examined - found at the divings 2-4
December 1994 |
Figure 3.7.5 -
the visor lug breaks at 210 tons |
|
Evidently the visor lug itself will not bend, when the fore peak deck lugs are ripped apart. The damages to the forepeak deck lugs would look as seen in figures 3.7.6-7 below (where the bushings are completely incorrectly welded with only 1 mm to the lugs). Now on the wreck all three lugs were found ripped apart - and rusty! - and you wonder how it could have taken place. A logical explanation may be that the visor lug had been pushed against the port bushing by a transverse load, which pushed the bushing out of the lugs, and the lugs were later pulled apart as found. The weldings of the lugs to the bushings were also found rusty. The initial damage could very well have occurred before the accident on 28 September 1994. The Commission has evidently never made above analysis. When it saw on an underwater video on 2 October 1994 that the lock was damaged, it made up the story that it had been pulled apart by one big wave impact on the visor. They could also see the undamaged locking bolt on the ROV film, but the Commission did not then know that the weakest part of the lock was the visor lug, which was intact (the visor had officially not been found). When the damages later indicated that the lock probably had been damaged by a sideway force, the Commission ignored this possibility. It is quite likely that the Atlantic lock was damaged before the accident. It does not change any conclusions in this book, as the visor had nothing to do with the accident - except as being used by the Commission as a patsy. The visor lug was bent and its connection to the lower stringer was buckled and the lock could not be used - it did not fit - more pictures. Several parts were heavily rusted. Old Damages discovered All this was of course discovered by outside experts (in this case from the Royal Institute of Technology, KTH, Stockholm - see Supplement no. 517 in (5)). |
Figure 3.7.6
Port fore peak deck lug (right)
Fig 3.7.7 The
Atlantic lock broken - the port |
It is interesting to see that the German report 3.18 published in June 2000 stated (again) that there were many findings to the effect that the visor did not fit and that the condition of the visor was bad. In spite of this the Germans have never stated that the Atlantic lock did not fit. However ...
The Bottom (Atlantic)-Lock was probably damaged before the Accident took place!
It is a reasonable assumption. Contacts with ice had dislocated the visor hinges a little and the Atlantic lock visor lug had been bent to starboard and was, say 20 mm out of line with the bushings - it didn't fit. Probably the bushings were not even there! What did you do? You put the lock assembly on the repair list and didn't use the lock at all. It was not necessary to use the Atlantic lock as the visor was kept in place by the side locks and the hinges.
If that were the case, the side locks would always have been in tension due to wave loads. And the side locks and the hinges were enough to keep the visor in place. The Atlantic lock was just an extra attachment to unload the side locks and the hinges.
The conclusion of this chapter is that the Commission never established the condition of the Atlantic lock prior to the accident and how, when and why it later was found damaged. However, in order for the visor to become detached the other attachments must fail. We shall in the next chapter look at the side locks.
---
109 What it means is the following: if one maximum load is probable during 30 minutes and another higher load is probable during 10 hours (20 times longer), then it is 5% (1/20 probability) that that the higher load occurs in any given period of 30 minutes.