36. JAIC Chapter 8.6.5 - The Ramp Attachment and Locking Devices :

This has been discussed already further up, but the JAIC apparently needs to be corrected again because they are using a different damage description, viz. the JAIC states: "The 2 port side hinges at the bottom of the ramp had failed because of tension fracture of the ramp mounted lugs." This is wrong: The outer lug of the heavier port outer hinge was found to be intact by the diver which is confirmed by the videos - see also Figure 8.11 of the JAIC Report where this lug can be seen at the upper right side of the picture. This lug slipped off the hinge bolt after it had worked itself far enough towards starboard side and the inner lug subsequently failed . All this occurred several months before the casualty. See Chapters 12.4.3 and 12.5.
The JAIC continues: "The mounting for the pins for the upper locking hooks (previously called by the JAIC: "lugs for the pull-in hooks) was heavily twisted." Actually only the landing with the starboard lug for the pull-in hook was heavily twisted as explained further up with the 2 wood pallets jammed in between ramp/bulkhead recess, the port one is just slightly bent.
The JAIC continues: "The locking hooks (pull-in hooks or ramp hooks) could not be inspected in detail, but were confirmed to be in locked position."
Actually the port hook was inspected in detail, found to be intact and in locked position. As the lug at the ramp side for this hook was also undamaged and intact the hook cannot have been engaged in this lug before as otherwise either the hook or the lug must have broken which they obviously did not. The reason why the hook was unable to engage the lug was given by the diver who found that there was a misalignment of 21/2 inches = ca. 100 mm, between the securing bolt and their mating boxes at the ramp, which was, of course, also existing in way of the port locking hook - pull-in hook = ramp hook. The starboard hook was not visible at all, it was not found in its location, neither by ROV nor by diver. Consequently the statement of the JAIC as to the pull-in hooks of the bow ramp is wrong.
The JAIC continues with wrong statements: "3 of the 4 side locking bolts were in extended position" - this is correct, but "and the mating boxes at the ramp side beam had been ripped open". This is completely wrong. As explained before:
starboard upper box:      Top plate burnt off the morning of 27 September upon arrival at Tallinn, because the bolt could not be disengaged, the top plate was not re-welded during the last day in Tallinn.
Starboard lower box:      Box crushed together to the effect that no bolt could enter.
Port upper box:              Box completely intact - the extended bolt could not enter due to ca. 100 mm misalignment between bolt/box.
The JAIC statement as to the port lower box is correct.

37. JAIC Chapter 8.6.6 - The Visor and Ramp Indicating Devices :

The JAIC does not mention that the cables as well as the mounting bracket (sensor plate) were cut off by the diver and thrown away to the sea bed, thus disappearing forever. Since divers do not cut off anything without having been instructed specifically, it has to be assumed that Stenström instructed the diver to cut off these important pieces of evidence and then, instead of putting them into the net hanging next to them in the crane, to throw them away.
Note: Such communication is not recorded on the publicly available videos, but the respective instructions would have been passed to the diver on his 2nd earphone.
Stenström had, of course, realised from the pictures on the monitor during the diving operation that the bolt holes of the sensor plate were corroded indicating that the sensors had been dismounted some considerable time before and, further, that the sensor cables had obviously been cut, all of which would have been documented after close examination ashore. Therefore he instructed the diver accordingly and, subsequently, threw the bolt of the bolt of the Atlantic lock back into the sea because also this bolt would have told its own, for owners and crew, negative story, after having been properly examined ashore.

38. JAIC Chapter 8.7 - Condition of the Interior :

The JAIC states among other things: The car deck was not surveyed due to the hazards related to divers working in the area." This is wrong for a number of reasons known to the JAIC, viz.:
- The diver Dave Mawston, probably other Rockwater divers as well, was inside the bow ramp up to the lower flaps of the bow ramp, however, he was obviously instructed not to inspect and report on the quite extensive damage to both longitudinal bulkheads to both sides of the bow ramp leading into the car deck. This damage becomes briefly visible when the diver moves his head while climbing up the starboard side of the ramp, along the lower flaps to the port lower side near the control panel below of which the bulkhead was also noted to be broken open. This damage as well as damage to the car deck plating commencing right behind the ramp flaps obviously extends into the car deck. The diver did not pay any attention to these very considerable, unusual and in no way mechanically explicable damages. See also Chapter 32.
- Other divers - apparently not belonging to the Smit Tak/Rockwater teams - were busy inside the car deck. These divers were apparently working without umbilical, thus self-sustained and did not produce bubbles.
It thus has to be concluded that the car deck was very closely inspected, but obviously not for the eyes of the public. The JAIC does not mention that the starboard stern ramp was open.

39. JAIC Chapter 8.8 - Observations on the Navigation Bridge :

The investigation of the bridge was obviously carried out very unprofessionally, which was possibly done deliberately.
The detailed analysis of the available videos by Disengage - see Chapter 34 - revealed that there were 4, probably 5, bodies on the bridge - which were obviously deliberately not identified by their uniforms or plain clothes. Also the absolute must of any bridge examination was not carried out: What were the headings of mother gyro and the repeaters and what was the position of the switch: rudder on auto-pilot or manual.
As also the pitch and engine control readings were reported quite differently by the two divers, it has to be concluded that the bridge inspection did not bring anything.

40. JAIC Chapter 8.11 - The EPIRB Beacons :

The statements of the JAIC have been commented partly in Chapter 23.2 and are considered to be correct. The apparent fact that the EPIRBs were found in switched-off condition has to be considered serious.

41. JAIC Chapter 8.12 - Other Observations :

(1) The almost zero position of the pitch propellers can be explained by the statement of the makers KAMEWA, viz. that in case of power failure the pitch returns to zero.
(2) The hard astarboard position of the rudders can be explained by the 120° degrees list of the vessel to starboard. As soon as the hydraulic pressure disappears, which takes place several hours after power failure, the rudders turn to hard astarboard due to gravity.
(3) The watertight doors on the 1st deck closed from starboard to port and were kept in open position by hydraulic pressure. After this hydraulic pressure had disappeared, the doors would close by gravity due to the starboard list of the wreck same as the rudders.
In summary of items (1) to (3) it can be concluded that the as-found conditions of the pitch propellers, the rudders and the watertight door on the 1st deck do not allow any conclusions as to their position during and after the casualty.
The JAIC states further: "The lugs of the manual lock on the starboard side (the lugs are hooks) of the visor were heavily twisted due to a blow to the front bulkhead of the vessel."
This is pure speculation. The hooks were twisted in such a way which cannot have been caused mechanically. See Chapter 32.
The Chapter closes by the following remarks of the JAIC:
"The Finnish police have taken several paint samples from inside the visor. TLC (thin layer chromatography), LC (liquid chromatography) and spot tests analysis of these revealed no vestiges of explosives."
The paint samples were taken already in November 1994 upon request of the Chairman of the Finnish Part of the JAIC, Kari Lethola. At that time the visor had been under water for almost 8 weeks and possible traces of explosives on the paint had long since disappeared. The only way to prove the effect of explosives on the structure would have been and still is to examine particular parts of the visor, by means of a scientific method developed in Sweden. This apparently has not been done. See also Chapter 32.

42. JAIC Chapter 12.1 - Determination of Sea Loads on the Visor by Model Tests :

The following Chapters 12.1 - 12.3 will be commented by Professor Petershagen from the Institute of Naval Architecture of the University of Hamburg, however to begin some remarks from the authors of the this Report:
The JAIC states: "The main purpose of the test program was to determine the wave impact loads on the visor at the speed, on the heading, and under the wave conditions, the 'Estonia' was likely to have been operating at the time of the visor failure." ... and further "The static weight of the visor (55 ts) was excluded from the measurements." This implies that the water quantity inside the visor, which must have been at the time of the casualty in the range of 150 ts, was also not taken into account. It is obvious what effect already this gross neglect has on the results which, according to the JAIC itself must be judged with care (page 159, 1st §). Professor Petershagen:
"12.1 Determination of sea loads on the visor by model tests
Model tests have been carried out at SSPA laboratories with the main purpose to determine the load on the bow visor for the probable conditions of the visor failure. The model test results have been compared with the results of computer simulations. The test program included head sea and oblique sea conditions. The model was self-propelled. The long-crested irregular waves were generated according to the JON - SWAP wave spectrum. The model tests have been commented by experts of the Hamburg Towing Tank (HSVA) and found to be carried out according to the state of testing technique.
12.2 Numerical simulation of vertical wave loads on the bow visor
Seaway and wave loads can be described on a statistical basis only. This can be done by means of tests or numerical simulations. It must be emphasized that local pressures acting on the hull cannot be calculated by numerical simulation with great accuracy at present.
Computer simulations of vertical wave loads have been carried out. They are based on the non-linear strip theory together with the JON - SWAP wave spectrum. In the simulation the vessel's behaviour in an irregular seaway and forces acting on the vessel and especially on the bow visor are calculated in subsequent short time steps. This simulation has been carried out by means of accepted methods.
The essential results of the simulations are probabilities of exceedance for the vertical wave force acting on the visor.
The wave load on the bow visor has been found to be highly non-linear with regard to wave amplitude. Small waves even do not reach the visor.
The calculation results are intended for comparison with the model test results and for extrapolation of these results to longer time periods and correspondingly higher peak loads. The calculations were carried out up to 36 hours total simulation time corresponding to about 30.000 wave encounters. The maximum forces / moments on the visor obtained have thus a probability of exceedance of 1/30.000, i.e. will probably be exceeded once in 30.000 wave encounters."
Note: Bearing in mind that the vessel was exposed to those conditions only for about 2 hours, the probability of meeting such a wave is very small.
"The results from the numerical simulations agree qualitatively with those from the experiments. However, on a quantitative basis the results differ from each other significantly with the forces from the numerical simulations being generally lower than the forces received experimentally.
Possible reasons are given for the divergence between measurement and calculation, which is growing at exceedance periods longer than 30 min.
In the Commission Report it is stated that "Due to the non-linear and random nature of the bow impact loads, the absolute quantitative measured loads must be judged with care" (report 12.1.2 p. 154). The same must be stated with regard to the loads received from numerical simulation.
Summarizing it has to be stated from a scientific point of view that conclusions from the tests and the numerical simulations can only be drawn with great care."
Note: The scientist's statement, "that conclusions can only be drawn with great care" means from the practical point of view that no realistic conclusions can be drawn from the results produced.
"12.3 Estimate of maximum wave loads on the visor for the conditions at the accident
The estimate given in the Commission Report is based on the following assumptions. The significant wave height at the accident site is estimated by different meteorological institutes to be 4.o - 4.1 m at 01.oo h at the accident site. The vessel's speed at that time is assumed to have been about 14 knots. It should be recalled that the load on the bow visor depends largely on the ship speed. As a prime basis for the estimate the model test series with long duration carried out in a seaway with a significant wave height 4.5 m have been used. Weibull probability distributions have been fitted to the measured load data. This procedure is based on the results of the numerical simulations.
In Table 12.4 of the Commission Report most probable maximum values are given for a 30 min period. Values for exceedance probabilities of 5 % and 95 % are also given. These values show ratios for the individual load components between 2.3 (Z force) and 11.6 (X moment). These large scatter ranges underline that the load data have to be used with great care only and that reliable quantitative conclusions regarding the bow visor failure can hardly be drawn.
The loads for the accident condition were estimated by reducing the model test loads extrapolated to 30 000 wave encounters (corresponding to 36 hours simulation time) by 30 % (forces) and 50 % (moments). It is agreed that the figures so obtained are rough estimates only. Moreover, no detailed explanation is given for the selection of the reduction coefficients.
The estimated most probable values are (Table 12.5):
Visor forces direction
X force (longitudinal) 3,6 MN aft
Y force (side) 1,o MN starboard
Z force (vertical) 3,6 MN upward Deck hinge moments direction X moment 1,7 MN . m upward on port side Y moment 7,5 MN . m opening around hinges Z moment 1,0 MN . m forward on port
These data will later be compared with the design loads applied." JAIC Chapters 12.4, 12.5, 12.6 are not commented.

43. JAIC Chapter 12.7 1

The JAIC states under item 12.7.1 - General - what they did and would like to have done, namely:
"Analytical and finite element calculations were used for estimating the strength of the different attachments taking account of their actual geometry and material. The welds as well as any deficiencies found were assessed where pertinent.
The combined strength of the visor attachment system was estimated with balanced reaction load calculations using variation of external loads."
However, the JAIC did not calculate the visor structure and the distribution of the reaction forces on the fixing points of the visor by means of the finite element analysis, although used by the JAIC for other purposes as stated above. Had they done it, they would have discovered that as long as the hinges are intact the load on the Atlantic lock was relatively small, only in the range of 50-60 ts, the force direction was aft upwards.
Only after the failure of the hinges did the force direction change by 180° to forward and the Atlantic lock become exposed to very heavy loads - most import for the JAIC in the light of their casualty scenario - nevertheless it has not been examined - see Chapter 34.2. Had the JAIC performed the above explained finite element analysis of the visor structure they would have realised that the Atlantic lock could only have failed if the hinges had failed before. Since the lugs of the Atlantic lock obviously broke due to the force direction having been upwards/forward it has to be concluded that the hinges must have been broken already when the Atlantic lock failed which is completely in accordance with the most likely casualty scenario.

44. JAIC Chapters 12.7.2 and 12.7.3 -

The comments of Prof. Petershagen are cited as follows :
"12.7 Investigation of visor attachments
This section of the Commission Report contains mainly the results of material investigations and observations with regard to the condition of individual parts, especially welds. The findings are summarized in Table 12.6 of the Commission Report.
The material of parts recovered was identified to be mild steel. For comparison the locking bolt material of sister ship "DIANA II" was tested and found to be high strength steel (UTS = 760 - 785 Mpa)." Note: This bolt had been renewed according to the previous chief mate.
"Fatigue cracks having developed before the accident with significant influence on the ultimate load are reported to have particularly been found in the visor actuator mounting platforms. This is contrary to |1|, where significant corrosion fatigue cracking has been found also at the hinges.
Investigations of individual attachments:
Bottom lock attachment and visor lug:
The lug fractures were of the ductile type indicating failure by overload. Some small cracks close to the primary fracture surfaces of the welds between plates and locking bolt housing are considered to be of small influence on the load carrying capacity of the lock.
Fillet welds between housing and lock bushing are reported to have shown signs of poor fusion and lack of penetration. The average throat thickness of these welds is reported as 4 mm (suppl. No. 516), measured in the fuison line. This would correspond to a throat thickness of less than 3 mm, the usual minimum throat thickness according to the Rules of Classification Societies. This is contrary to the Building Yard's practice, according to which the throat thickness should have been 8 mm and should also have been criticized by the Classification Societies supervisor during newbuilding inspection."
Note: A further indication that the lugs were not original.
"Various investigations are reported to estimate the load carrying capacity of the forepeak deck lug assembly. A load of about 1.5 MN is derived from the shape of the visor lug after the accident. However, with a view to the inspection report of TURBO TECHNIK for MS "DIANA II" dated 03.10.94 the findings in suppl. 511 regarding wear in the Atlantic lock visor lug and hence the load of 1.5 MN appear questionable.
It is further concluded that the strength of the bottom lock could theoretically not have exceeded 1.8 MN, the estimated load for shear fracture of the lug tip.
More reliable than these estimates appear the results of fracture tests carried out with full-scale mock-ups at the Institute of Naval Architecture in Hamburg.
Contrary to what is stated in the Commission Report these models were built from mild steel with the exception of the bolt (see results of material tests reported by ISSV 06.04.1998)."
Note: See Subchapter 34.3, Enclosure 34.3.431
In the first test the visor lug fractured at a load of 2043 KN.
With a reinforced visor lug and a light continuous fillet (a = 3 mm) weld between lugs and housing resp. bushing the lugs fractured at 2040 KN. Only when these fillet welds were intermittent welded or omitted at all, considerably lower fracture loads were obtained. It is concluded that a fracture load of 2.0 MN is a realistic figure. It is also noted that the tests are not correctly described in suppl. 511."
Note: See Subchapter 34.3., Enclosure 34.3.431
"Samples of coatings from the forepeak deck starboard and from the visor lug were analysed. These samples consisted of different numbers of paint layers with different colours of the first layer. It is questioned whether all attachments were still original or whether they had been replaced during the vessel's service."
Note: See also Subchapter 34.3, Enclosure 34.3.431
"Side locks:
In the Commission Report it is concluded that failure of the side locks took place by separation of the lugs by shearing of the horizontal stringer plate, the vertical stiffener at the related weld and through the visor aft plating.
A failure load of 1.8 MN has been received from one out of four mock-up tests. This test is the only one considered relevant for actual failure load reading.
Failure load estimates have been carried out by means of a rather simple calculation model (suppl. No. 511). Further, a non-linear finite element calculation of the failure load has been carried out (suppl. 508). The ultimate load of about 1.6 MN found in this calculation is in an acceptable agreement with the test result (1.8 MN).
However, in the discussion of the possible failure mode of the visor (section 15,10 of the report) a failure load of the port side lock of 1.2 MN only is assumed (fig. 15.5). This reduction is explained by a defect in the fillet weld between plate and horizontal stringer (fig. 12.22). However, the reduction is taken from the simple calculation (suppl. No. 511), which moreover is not very clearly described, and no attempt has been made to implement the defect into the more refined calculation (suppl. No. 508) or the mock-up-test. Therefore, the estimated failure load of the port side lock remains questionable."
Note: According to eye-witness reports - see Chapters 12.4.3/ 12.5 - the lugs of the side locks were frequently cut off the visor plating by flame cutting when the bolts could not be disengaged from the lugs. This was done from the visor inside and respective burning marks are clearly visible on the visor plating around the previous lug position - see Fig. 8.19 and 8.20 on page 127 of the JAIC Report See also the lower picture on page 1020 of this Report which shows the inside of the port visor bulkhead at the previous location of the lug. Burning marks can be seen all around the opening which are certainly to some extent originating from the burning when the plating part was cut out after the casualty, but by far not all.
It has further to be noted that the welding seams on both lugs are exceptionally thick - see Fig. 8.18 on page 127 of the JAIC Report and the photos on pages 1011 and 1022 of this Report. Actually there are several seams on top and/or beside of each other, i.e. after the lugs had been burned off the remains of the initial weld were not ground off the plating before the lug was rewelded. Therefore the thickness of the seams was growing, which is very clearly visible on the right picture on page 1011 of this Report, and while the thickness was growing the load carrying capacity was reduced, all of which was not taken into account by the JAIC.
The diver reported, when examining the lugs of the side locks still attached to the wreck, that there was only a small white piece of bulkhead attached to the lug - see the lower picture on page 1022 - while to the larger parts of the lugs no parts of the bulkheads were attached, although there were respective holes found in the visor plating. This can only mean that these holes had been previously cut into the plating and the lugs were stuck through these holes and welded to the plating from the inside and the outside subsequently. All this was obviously done without approval of the Classification Society, as demanded by their rules Had they had a chance to inspect and approve this repair.as demanded by the rules, this repair would never have passed.
This is another item rendering the ferry unseaworthy since some time before the casualty.
"Hinges: It seems that the failure load of the hinges has attracted less attention in the investigations described in the Commission Report than that of other attachments.
Extensive cracking in the roots of welds at the hinge bushings is reported. It is also stated that these cracks may have to some extent propagated under corrosion fatigue load. Pertinent investigations are described in suppl. No. 518. There it is, however, summarized that the results of the investigation do not support a significant role of pre-existing fatigue cracks in the failure of the hinges. This is contrary to conclusions drawn in |1|, where it is stated that the hinges failed by fatigue cracking and subsequent plastic overloading.
The information given in the Commission Report is not sufficient for a further examination with regard to a possible influence of pre-existing fatigue cracks on the failure of the hinges.
A calculation of the load carrying capacity of the hinges is given in suppl. No. 511, which can, however, be considered as a rough estimate only.
It should further be noted that the Commission Report is not taking the actual condition of the hinges into account. In suppl. No. 518 it is stated that "the inside surface of the holes in the visor plates for the bushings has been formed by flame cutting with fairly deep cut markings".
Taking the usual fabrication standard into account, it is doubted whether the visor plates mentioned were still original or had been replaced during a repair."
Note: It is proven without doubt by eye-witnesses from the Meyer Werft, the von Tell representative Todsen and the B.V. inspector, all attending during the hinge installation at newbuilding, that the hinge plates were not burned. See Subchapters 2.2.2 and 12.5 and Enclosure 2.4.2.21.
Actuator Attachments
These attachments and, in particular, the mounting (fundament) of the port actuator on B-deck "had cracking, significant for the strength". As a matter of fact, the mounting was only attached to B-deck by welds at the inside of the mounting lugs to which the actuator was connected. According to the truck driver Per-Arne Persson, who made many trips onboard the ESTONIA and who survived the casualty, the port actuator had once been totally cut off B-deck. A statement is not yet available.
Attachment System
The JAIC discusses the combined strength of the visor attachment system, however, draws attention to the obvious fact that the results of their assessments of the different failure load levels are more indicative than conclusive.

45. JAIC Chapter 13.1.

The comments of the JAIC as to the wind and wave height are generally accepted.It is however repeatedly stated by them - see Fig. 13.2 as example - that the wind was coming from SW = 225°, i.e. on a course of 262° the wind blows 37° from port, which is not close to head sea. After the course change at the waypoint to 287° the wind was said to be still SW = 225°, i.e. now came in about 60° from port and not 30°.
Also the statement that the vessel had encountered similar weather conditions only once or twice before is not acceptable. According to information available to this 'Group of Experts' it was 7 times since December 1992. See Subchapter 12.2.

46. JAIC Chapter 13.2.1 :

The JAIC claims to have analysed the 258 statements from the 134 survivors, which is true, though this was done in a rather restricted way, basically focusing on the hypothetical casualty scenario developed during the first 1 or 2 weeks after the casualty. The JAIC has never analysed the statements with a view to a probable alternative casualty scenarios as indicated by some survivors. Moreover, the JAIC has only spoken to a few crew members but never to passengers from the 1st deck or from the forward parts of decks 4, 5, 6. See also Chapter 21 and item 20 above.
On the other hand, this 'Group of Experts' does accept the comments of the JAIC about the survivors' statements in general, which have always to be borne in mind. Also the necessity to put more emphasis on earlier statements is shared, in particular of those of crew members, the "loyalty" of whom to ESCO was not so extensively developed. On the other hand, experience has shown that the first statements were frequently taken in a hurry and by policemen untrained to investigate maritime casualties. Furthermore it has turned out that almost none of the survivors were given the chance to read what policemen had summarised of their statements, and to correct wrong or misunderstood remarks. The best example is the passenger from the 1st deck, Carl Övberg, who categorically denied ever having stated that he had left the 1st deck at about 01.20 hours as stated and as quoted in the JAIC Report. He saw his statement for the first time, as did many other survivors, at the exhibition of this 'Group of Experts' in Stockholm in June 1997. Also the other comments about the JAIC's treatment of passenger evidence and statements in general should be borne in mind. See Chapter 21.
The JAIC's attitude to grant more credibility to crew members than to passengers with regard to their estimations and their judgements of sounds is, however, not shared by this 'Group of Experts', because a possible cooler approach to the prevailing circumstances by crew members and, therefore, subsequently better estimations, are by far outweighed by the unavoidable "loyalty" of crew members to their owners, which has considerable influence on their statements. This refers in particular to Silver Linde and Margus Treu and the behaviour of Hannes Kadak when he withdrew his earlier described observation that water had penetrated the bow ramp already before the big heel.

47. JAIC Chapter 13.2.2 -

The JAIC states :
"Heavy vehicles seem to have been loaded on the car deck without sufficient account of the athwartships weight disposition, resulting in the ship leaving port with the port side heeling tank almost full and the starboard one empty. Due to this cargo disposition and the wind pressure on the port side the ESTONIA, gaining the open sea, had a starboard list of about one degree .."
This was not confirmed by survivors having had own cars or trucks on the car deck. They stated that the trucks were more or less equally distributed on both sides of the centre casing, thus the severe weight concentration on starboard side necessitating the flooding of the port heeling tank must have had other reasons, which was with all probability the flooding of empty tanks or void spaces on the starboard side through holed bottom platings as explained and documented by images in Chapter 29.5. Consequently ESTONIA should not have been loaded differently but better inspected and maintained.
The repeated statements by the JAIC that the trucks and trailers were properly secured are not confirmed by several truck drivers. See Subchapter 17.2. The following comments by the JAIC to overcome these statements show the utter lack of seriousness in dealing with such questions of far reaching importance as seaworthiness:
"It is claimed to be common practice that securing of vehicles is not finished when the vessel leaves port but is completed during an early stage of the voyage. All indications are that the cargo was secured to normal standard."
The securing of most of the trucks was impossible because they were positioned so close to each other that it was not possible to secure chains /other lashings between them to the deck fittings.
According to two eye witnesses from onboard the "SILJA FESTIVAL" having been berthed opposite the ESTONIA the visor was half opened when the vessel turned in the harbour basin and was still half open when the ferry proceeded through the breakwater into the open sea. This is not mentioned by the JAIC, although the respective statement is in the files of the Finnish JAIC. See Enclosure 19.236.

48. JAIC Chapter 13.2.3 :

The JAIC states that the port lower securing bolt of the bow ramp was most probably not in extended position at the time of departure and thus did not enter its mating box. This is true, but not only for the port lower bolt, but also for the port upper bolt and for the port pull-in hook. Consequently the entire port side of the ramp was not secured when the ESTONIA set sail and left the port of Tallinn on her last voyage. The JAIC states further:
"Even if this defect had existed at the time of departure it has not been possible to find out whether any action was called for. This potential deficiency would have had no effect on the development of the accident, as the ramp would have been forced open by the visor even if all the locking bolts had been in their proper positions."
This is wrong. If the bow ramp would have been properly closed against intact rubber packings with intact locking devices the empty visor would not have been able to force open the ramp.
Concerning the mattresses, bed linen, blankets and rags visible on the ROV videos of 02.10.94 and 09.10.94 which were obviously jammed in between the ramp, deck coaming and side bulkhead, the JAIC "considers it likely that mattresses and rags were washed into the area from nearby storage spaces during the final flooding of the car deck", although the crew had admitted to having used the mentioned material as sealing material because the bow ramp was open at this area. The mattresses and bed linen had been seen in place several weeks before the casualty by passengers and, reportedly, also observed by Gunnar Zahlée and Åke Sjöblom during their inspection shortly before the last departure of the ESTONIA.
Moreover, it remains the secrete of the JAIC how a floating mattress can end up in a position as visible on Fig. 8.11/page 124 of our Report. See also the video images on pages 847 -851 of this Report This sealing material was still in place when the divers commenced their work in December 1994, however, it was gone when they had finished.
The problems to open the ramp locks by means of tools (this could include flame cutting equipment) were again treated with lack of seriousness namely by simple reference to other vessels, although truck drivers/ passengers had frequently reported to the Swedish and Finnish JAIC that the crew had increasing problems opening the visor and/or bow ramp, which was sometimes not even possible by means of flame cutting, where after the ESTONIA had to be towed away from the berth by tugs, turned around and berthed again "stern first". No small maneuver with a vessel the size of ESTONIA in the limited space of the port Tallinn. No mention is made of these repeated incidents in the Report of the JAIC.
Next the JAIC discusses the missing sensors of the Atlantic lock and goes so far to state:
"However since the distance from the magnet to the nearest sensor was a few centimetres a small chance remains that the pounding visor did detach the sensors."
Lack of seriousness is not always funny. First of all, there was no pounding of the visor and secondly, holes in the sensor plate for the bolts to fix the sensors to the plate were corroded and thirdly there was no damage whatsoever.
Given the attitude of JAIC in the conduct of their investigation as described above the reader will not be astonished to find the following conclusion in the Report of JAIC:
"The Commission's conclusion, which is supported by the failure pattern, is that the visor had been properly closed and secured at departure and that there were no deficiencies in the ramp affecting the development of the accident."

49. JAIC Chapter 13.2.4 :

Although according to their own investigation results the speed had little influence on the wave loads, the JAIC does everything to reduce the speed of the ESTONIA as much as possible. Assuming that ESTONIA passed the Tallinn breakwater already on full speed at 19.30 hours, which the JAIC does but which is certainly too early, but anyway let's assume it, she covered the 73 nm to the passing of the "AMBER" - see Chapter 30 - in 3 hours 45 minutes, which corresponds to an average speed of 19.5 kn. This passing between "AMBER" and ESTONIA took place almost abeam of Takhuna Nina, i.e. ESTONIA had left the shelter of the Estonian coast already and due to the increasing forward swell and sea the speed went down.
The passing times stated in the JAIC Report, i.e. Osmussaar at ca. 22.00 hours and the Apollo buoy correspond +/-5 minutes to the passing times based on an average speed of 19.5 kn, however the problem for the JAIC is the waypoint and further on the casualty position and the alleged visor position. If the visor was indeed lost at the position stated in the JAIC Report, it is certain that the vessel was even further West, let's assume just 0.5 nm. This means that after having passed "AMBER" at ca. 23.15 hours she had to cover 18.3 nm to the waypoint and another 11.5 nm to a position 0.5 nm West of the alleged visor position, where she must have been at 01.02 hours when the big heel occurred, i.e. she needed 1 hour 47 minutes for 29.8 nm, which corresponds to a speed of 16.7 kn, slightly more than the JAIC did find out.
Consequently the ESTONIA had been by far the fastest of the ferries heading for the Söderarm entrance to the Stockholm archipelago.

50. JAIC Chapter 13.2.5 :

It is wrong that the first indication of something being wrong in the bow area was only at about 00.55 hours, because
(a) the watch A.B. Silver Linde has made at least 7 different statements as to this time which varies from 00.35 to 00.55 hours - see Subchapters 21.2.2 and 21.5 and
(b) passengers on the 1st deck noted these indications already at about 00.30 hours, and
(c) passenger in cabins and bars on upper decks at about 00.40/45. The JAIC thus based their hypothetical scenario on the frequently changing evidence of the watch A.B. Silver Linde who obviously adjusted his evidence to the changing requirements of his superiors, JAIC continues as follows: "Shortly after one o'clock a low wave impact on the visor caused the visor attachments to fail completely. The visor started cutting openings in the weather deck plating and associated structures. Soon the back wall of the visor housing came into contact with the ramp, hitting its upper edge and thus breaking its locks. The ramp fell forwards and remained resting inside the visor. In a few minutes the visor started falling forwards."
The JAIC has after all understood that the hinges failed at first, subsequently followed by the side locks, and that the Atlantic lock failed last. This scenario is similar to that explained in Chapters 29/30, viz.
- at first the starboard - so much weakened - hinges failed which caused the visor to move forward by 120 mm only on starboard side until the lugs underneath the visor arm struck against the transverse beam and the visor was pushed back by the waves until the broken hinge plates struck against the hinge parts on the forecastle deck. It is possible that the crew attempted to hold the visor by the hydraulic actuator at this time. The visor moved several times forward/aft, which caused the heavy metallic banging noises.
- at 01.02 hours the vessel heeled very wide to starboard and came back to almost upright position (probably twice) and thereafter took a heel of about 10°-15°, whilst rolling.
- Next the port hinges failed followed by the port side lock and the whole visor now moved forward/aft between the transverse deck beam and the hinge parts on the forecastle deck.
As long as the deck beam was not cut through there was no contact between the visor ramp house and the bow ramp It has to be assumed that this forward/aft movement of the visor continued for some time while the starboard heel increased and the centre of gravity of the visor shifted from forward to starboard.
- At some time the lugs underneath the visor area had cut through the deck beam - at least at the port side - and the visor moved forward. The lugs were now cutting through the deck plating (the outside was not in contact with the deck plating or deck beam) until the forward end was reached. The lug plates now did not prevent the visor anymore to fall to starboard and hence that is what he now did until the port inner bulkhead was resting on the upper part of the bow ramp which itself was lying inside the visor on the vertical and transverse beams thereof. Simultaneously the forepeak deck had penetrated the visor bottom and thus the visor side platings were also supporting the visor against falling off to starboard as the bow ramp alone would not have been able to carry the entire momentum created by the weight of the visor. See Chapter 34.10. The above-mentioned scenario is fully supported by the damages to the wreck and the visor.
The JAIC continues:
"2. The ramp then followed the visor in a forward, tumbling motion. The starboard side actuator was extended to its full length and was torn out of the hull during the final stage of the sequence. The visor subsequently tilted over the stem, left the ramp fully open allowing large amounts of water to enter the car deck and as it fell collided with the bulbous bow of the vessel. The failure sequence of the visor and the ramp is described in further detail in 13.5."
The visor never tilted over the stem and the ramp was never fully open, but the visor kept hanging on the ramp, and was mainly held by its port side plating while the forepeak deck was compressing the visor bottom by about 1/2 metre. The visor moved further forward until the ice knife of the bulbous bow crushed the stempost, while still hanging on the bow ramp.
Based on the damage to the visor it has to be assumed that it remained connected to the vessel until the heel was about 130°-140°, where after the visor sank to the bottom of the sea. The visor was unable to float, not even for a short time. See Chapter 34.11.
The time of the separation of vessel and visor was probably 01.40/45 hours. Survivors have seen parts of the visor when the heel was in excess of 90°.
The sequence of events of the JAIC is - contrary to the contents of the following statement - not based on witness statements but it happens to fit perfectly into the hypothetical assumptions which the JAIC formulated so early and followed so ardently.
"This sequence of events is supported by witnesses from several areas on board who heard a repeated metallic noise from the bow area during a period of about ten minutes, starting shortly after one o'clock. The detailed timing is, however, uncertain. The witnesses have given several good descriptions of these sounds and it is beyond doubt that the sounds were caused by the visor moving and pounding on the forepeak deck. Some of the metallic blows were associated with hull vibrations. The sounds from the bow area ended in a few loud, metallic crashes, caused by the final separation of the visor and its colliding with the bulbous bow of the vessel. This occurred at about 0115 hrs. The collision is documented by clear impact marks on the visor. The observations by the witnesses are described in detail in Chapter 6."
No one has seen the visor separating, thus no one noted the time. At 01.15 hours the vessel had a heel of 30°-40° and the main engines had stopped already while the auxiliary engines were about to stop. The damage to the starboard shell plating of the visor clearly consists of two separate damages, the upper, older one was overlapped by the fresher lower one, which probably occurred when the visor was in contact with the bulbous bow.The collision impact(s) felt by survivors, in particular of the 1st deck, occurred without doubt before or just after the big heel, which was at 01.02 hours, when the visor was definitely still attached to the vessel. Thus these impacts cannot have been caused by collision with the visor as the JAIC maintains.

51. JAIC Chapter 13.2.6 :

As the JAIC did not question the key witnesses from the 1st deck, themselves, but just interpreted the insufficient and incomplete statements taken by untrained policemen, and translated by lazy translaters they could not understand what had happened actually. Carl Övberg and Carl-Erik Reintamm have clearly testified that there had been quite some water on the 1st deck already before the big heel. See the statement of Carl Övberg - Enclosure 12.4.2.151 and Carl-Erik Reintamm - Enclosure 21.3.1.289.1. Övberg observed water escaping from two gooseneck pipes coming up from 0-deck and further, water at the door, while Reintamm observed water in the alleyway up to 30-40 high also apparently under pressure. JAIC's lack of seriousness in the evaluation and interpretation of witness statements becomes obvious again when they interpret the observations made by those in the engine control room (ECR) which they restrict to Margus Treu whilst there were also system engineer Henrik Sillaste and motorman Hannes Kadak in the ECR. Henrik Sillaste, who was reportedly working at the forward vacuum system of 0-deck, rushed to the ECR after the big heel, which according to the JAIC was caused by large quantities of water flooding on to the car deck after the visor had pulled open the bow ramp; however when Sillaste looked at the monitor in the ECR, he saw the bow ramp only slightly open and water pressing through at both sides. This was still the situation when he and Kadak left the ECR shortly afterwards.
In other words, the JAIC is of the opinion that the visor forced the ramp to fully open whilst the vessel was still at full speed, with the result that large quantities of water entered the car deck which fact in turn caused the vessel to heel to starboard. This is simply not possible, because
a.) if the vessel would have proceeded with full speed against bow seas of 4 m significant wave height, the ramp would have broken off - see Chapter 34.10;
b.) Henrik Sillaste, Hannes Kadak and Margus Treu in the ECR saw the ramp after the heel still closed or almost closed on the monitor and the water pressing through at the sides;
c.) the three also testified that they saw the water continuously being pressed through the openings at the sides, no matter whether the bow was up or down in the waves, which means that the fully water-filled visor was still attached to the vessel.
Everything is debatable, especially in accident investigation, with the exception of the laws of physics as described sub a) above. The JAIC seems to think otherwise. We are inclined to leave them alone with that opinion as everybody else seems to do.

52. JAIC Chapter 13.3:

This Chapter is based either on routine or on the changing statements of watch A.B. Silver Linde, of which the JAIC selected no. 7, the last known one, to be the base for their "action on the bridge" scenario which allegedly began when Silver Linde returned to the bridge now at about 01.05 hours. This was 3 minutes after the big heel when Silver Linde together with two other ABs walked along the port inside alleyway of deck 7 towards the forward stairway where he was seen and identified by Per-Erik Ehrnsten, a passenger from a 6th deck cabin, who made his way up very quickly after the heel.
Consequently Silver Linde had not only never been down on 4th deck nor had he been back to the bridge after his last round, but went straight to his cabin probably to fetch his valuables. It is also impossible that the main engines kept running at a heel of 35°, they stopped already at about 20° shortly later followed by the auxiliaries. It is indirectly confirmed by 2nd engineer Peeter Tüür that the whole scenario was substantially earlier than stated by the JAIC, because he testified to having seen Hannes Kadak outside his cabin on deck 7 already shortly after the heel. It is obvious that the sinking scenario of the JAIC is very much based on the last statements of Silver Linde and Margus Treu both from early 1996 and that the statement of passengers have been totally disregarded.
It is therefore useless to keep commenting on the JAIC scenario, because a complete new evaluation of all the passenger statements is required which has to be left to a new investigation.

53. JAIC Chapter 13.5 :

This was commented a few pages above already and does not need to be repeated. The JAIC's "failure sequence of the bow visor", which is part of the casualty scenario, is not in line with the survivors' testimony nor does it correspond to the damage picture. The JAIC has relied on some statements only but has disregarded others. The casualty scenario in accordance with the survivors' testimony and the damage picture of visor and wreck can be found in Chapter 32 of this Report.

54. JAIC Chapter 13.5 :

This part of the catastrophe has not been investigated by this 'Group of Experts', however, in the course of the casualty investigation naturally some circumstances were discovered which obviously did influence the sinking. These are:
(a) The observations of Carl Övberg and Carl-Erik Reintamm that water penetrated the 1st deck from the 0-deck below under pressure before the big heel.
(b) The open starboard stern ramp: The ramp is slightly open on the ROV videos from 02.10.94 and 09.10.94, i.e. must be unlocked. Since this cannot have been done during or after the casualty, it must have been done before the casualty. Therefore the starboard stern ramp must have been partly open before and at the time of sinking which is certainly the explanation for the vessel sinking by the stern.
(c) The starboard bottom and/or bilge strake area is holed at various locations, which means that an unknown number of void spaces/ cofferdams/tanks were flooded which are not part of the JAIC tank condition.
None of these items have been taken into consideration by the JAIC in their sinking scenario which, therefore, does not need to be commented any further.

55. JAIC Chapter 15 -

In the following the comments of Professor Petershagen are quoted again: JAIC Chapter
"15.1 Design basis and requirements for the bow visor
The design loads on the bow visor were calculated by the building yard according to Bureau Veritas regulations (Note Documentaire B M 2 dated 5.4.1976) as follows (suppl. No. 204). This calculation was carried out by the building yard as a control to the design load assumed by the supplier of bow visor operating and locking devices Von Tell.

Forces (MN)
X - force (long. / aft)	3,81 (3,6)
Z - force upwards)	5,36 (3,6)
Deck hinge moments	(MN . m)
Y - Moment (opening	) 16,2 (7,5)

Numbers in brackets denote estimated most probable loads under accident conditions. While X-forces are close to each other, the design Z-force exceeds that under accident conditions by about 50 %. Correspondingly, the design Y-moment is by far in excess of the accident condition figure. Taking the safety factors of ultimate load against design load (see section 15.12) into account, it is stated that the locking devices and especially the Atlantic lock should never have failed under the estimated most probable loads under accident conditions, provided they were in a proper maintenance condition."
Note: Not even if the hinges would have failed first.
"Different regulations were in force in the rules of different classification societies with regard to bow visor load at the design period of MV "ESTONIA". A comparison of design loads calculated by the building yard (suppl. No. 204) with those according to the rules of Germanischer Lloyd (1978) indicated much higher loads in the latter case.
In this comparison it has to be kept in mind that regulations of classification societies with regard to bow pressure loads were not yet well developed at the design period of MV "ESTONIA". As far as known to the author of these comments, Germanischer Lloyd at that time had anticipated results of discussions held at IACS (International Association of Classification Societies) regarding bow pressure loads, which came into force in 1982.
With regard to the technical state to be applied by the building yard in the assumption of bow design loads the following arguments are given. Different load figures could be derived in the design stage of MV "ESTONIA" from the rules of various classification societies as indicated above. However, no means were available to the shipyard from which a judgement upon the correctness of one or the other rule could have been made.
The new IACS bow pressure loads were still under discussion and no requirements were released. IACS unified requirements are not directly transferred to shipyards, but included in the Rules of Classification Societies. The discussion at IACS mentioned above was hence unknown to the building yard.
On the other hand, it did never and does not comply with the technical state to compare rules of different classification societies and use the most severe regulation when designing to the rules of one of these societies.
An even distribution of design load over the five existing supports (2 deck hinges, 2 side locks, 1 bottom lock) has been assumed by the building yard. Contrary to what is stated in the Commission Report this was in agreement with the practice not only of Bureau Veritas, but also of other societies as Germanischer Lloyd.
As also stated in suppl. No. 511, the visor attachment system is statically undetermined. This means that the load distribution between the attachments depends on the stiffness of the visor structure. This has also been found in the analysis of visors of other vessels known to the author of these comments.
In suppl. No. 511 it is concluded from measurements "that the visor is flexible enough that it may itself distort under load such that the loading distribution in the various lockings is not smoothened such as the design assumption of an even distribution of the sea loads onto the visor attachments would have meant".
An estimate of load distribution has been carried out in suppl. No. 511 by means of a beam model. It is concluded that "breaking the side lock at 1,2 MN local reaction occurs at a wave load level which may be insufficient to break the next attachment". A damage pattern at MV "DIANA II", sister ship of MV "ESTONIA", is cited, where in January 1993 side lock fracture and hinge damage have occurred.
The load distribution on locks and hinges has been calculated more accurately in |2| by means of a finite element model of the bow visor. The result was that the load on the visor is acting in the aft and oblique upwards direction. It follows that in case of all attachments being intact most of the load is taken by the hinges and by the side locks, while the load on the bottom lock is rather small. This is also the case with simulated side load. Only after initial failure of the hinges the load on the Atlantic lock becomes high. It should also be noted that with an intact sealing the loads, especially on the hinges, are reduced.
Taking these results into account, the sequence of events described in section 13.5 and shown in fig. 13.6 of the Commission Report appears very unlikely. While in suppl. No. 511 a recurring judgement about an attachment breaking sequence is considered as being not possible, the Commission Report describes failure of the hinges after failure of all locking devices as the most likely sequence of events. With a view to the load distribution and to the maintenance condition of the hinges the conclusion drawn in |2|, failure of bottom lock after hinge failure, appears much more likely.
15.2 Sea loads on the visor
It is stated in the Commission Report and agreed here that "there is considerable uncertainty in the estimates of the maximum loads".
The visor and especially the attachments are subjected to fluctuating loads by seaway and, in case of the deck hinges, by opening operations.
The number of cycles has been estimated from figures given in the report as follows:

Operation history (Report section 2.3 p.25/26):
h sign. > 3 m (wave period 6,8 s)	540 h 4765 cycles
h sign. > 4 m (wave period 7,7 s)	154 h 1200 cycles
	5965 cycles

 

Opening operations (Report table 2.2 p.25):
			crossings per day	years on route	days per year	crossings total
	Route	1	2	10.5	174	3654
		2	av. 3	2	222	1332
		3	1	1.7	359	610

 

Significant wave heights of 3 m and above are considered only because lower wave heights yield small loads on the visor only.
It can be concluded that fatigue load is in the low cycle range, where significant damage is connected with large loads and correspondingly high stress ranges. A crack growth rate of 0.6/1000 mm per cycle as estimated in suppl 518 p. 3 supports this.
A quantitive assessment of possible fatigue damage does not appear possible because of too large uncertainties in the determination of cyclic loads. Reasons for these uncertainties are large cov's in load determined in models tests, unknown speed and course of the vessel during relevant sea states and uncertainties in the distribution of loads between attachments.
Moreover, additional high cycle fatigue loading caused by the ship motions in moderate seas as well as by opening and closing operations of the bow visor has to be assumed with a view to the maintenance conditions of the visor and its attachments. This assumption is supported by the fatigue cracks found in the visor actuator mounting platforms (see comments to section 12.7 of the Commission Report).
15.3 Evaluation of the bottom locking device
Fracture occurred in the attachment to the forepeak deck. The three plate lugs carrying the bolt housing and the mating support bushing failed partly through the fillet welds and partly in the lug base material. The fillet welds are described as 3 mm thick with indications of root cracking or lack of fusion prior to the accident. As mentioned before, this is not in accordance with the fabrication standard of the building yard and the requirements of the Classification Society.
A total load carrying capacity of the bottom lock of about 1,5 MN as considered realistic by the Commission is too low. The mock-up tests described in |3| are not correctly cited. Failure loads below 2.0 MN were obtained only when intermittent or no fillet welding at all was applied.
The conclusion drawn by the Commission with regard to the paint test seem to neglect the fact that different early paint layers were found in different parts of the attachment.
15.4 Evaluation of the side locking devices
It is agreed that the side locks had a lower load carrying capacity than bottom lock and hinges. However, as already stated in the comments to section 12.7 of the Commission's Report, the low load carrying capacity assumed for the port side lock appears not justified. Moreover, it is stated in the conclusions to suppl. No. 511 that the port side lock may break at a load level insufficient to break the next attachment. Together with the comments given to section 15.10 the failure scenario given in the Commission's Report appears questionable again."
Note: The above considerations do not take into account the reduced load carrying capacity of the side locks due to their particular attachment to the visor bulkhead, i.e. they were not welded to the plating but stuck through holes in the plating and then welded on to already existing seams. "
15.5 Evaluation of the hinges on deck
The recovered hinge bushing has revealed cracking in the weld roots especially in the downward-facing area. In some places these root cracks had propagated to the weld surface. These findings again are not in conformance with the building yard's fabrication standard. Further, it goes without saying that bow visors in ferries are load-carrying parts of the ship structure subjected to inspection during surveys of the classification society.
As stated earlier, fatigue cracks, especially shortly after their initiation, are often difficult to detect.
In the present case it must be questioned why the fillet weld cracks observed by a student doing paint work had not been detected during a survey by the Classification Society.
It remains open whether the rough surfaces of flame-cut surfaces of the holes in the lugs are due to adjustment of the hinges during assembly or due to later repair work.
The Commission Report concludes from the generation of cracks during normal service an insufficient strength of the hinge lug rims and bearing bushing welds. This conclusion contradicts the Commission's statement regarding an only marginal influence of the cracks on the load-carrying capacity of the hinges."
Note: The JAIC writes on pages 191/192 among other things:
"The surfaces of the holes in the lugs where the bushings had been inserted had, along much of their peripheries, a very rough contour as from manual flame cutting. This applied to all four lugs, but more so in the starboard hinge. The burn marks were in the lug plates only with no corresponding marks on the recovered bushing. In addition, the forward contour of the holes in the lugs in the starboard side hinge were located about 10 mm further forward relative to the outer contours of the hinge beam than those on the port side. It has not been possible to find the reason for the rough surface, whether it was an adjustment of the hinges during assembly or remains from a later repair. No documentation from any repair in this area has, however, been recorded."
The starboard hinge plates show extremely deep burning marks and investigations by Professor Hoffmeister - see Chapter 34.8 - revealed that the inner surfaces of the bore holes, which were initially drilled, were destroyed by the inadequate application of flame cutting. The port outer hinge plate shows only very few light burning marks but still the drilling marks from the original drilling procedure at Meyer Werft.
The JAIC is surprised that there were burning marks on the surfaces of the bores in the hinge plate. The reason is, of course, that the old bushings were burned out, instead of being drilled out, which created the burning marks at both the hinge plates, however, while the bushing was scrapped and replaced by a new one, the hinge plates stayed in place with their deep burning marks, each of which being a crack starter.
The explanation for the very poor condition of the starboard hinges is given by the video film made by a passenger just before the departure of the ESTONIA from Tallinn on 17 September 1994. This video proves that at least the starboard hinge had been repaired in a very unprofessional way, as analysed by the British reconnaissance expert Bryan Roberts - see Chapter 12.5 - Enclosure 12.5.180. This video as well as the report of Bryan Roberts was ignored by the JAIC, who wrongly state " that it has not been possible to find the reason for the rough ( they mean covered with deep burning marks) surface."
Professor Hoffmeister has also examined the recovered bushing (most probably starboard inside) and found that existing cracks in the welding seams between hinge plates and bushings had been attempted to be closed by welding because welding material was found in the space between hinge plate/bushing. This implies the knowledge of shore management and crew of these cracks because either they were welded by the crew (there was a welder on board or by shore staff which would have to be ordered by the technical managers N&T).

56. JAIC Chapter 15.6 :

The manual locking devices were evidently not engaged on the casualty voyage. The JAIC states among other things:
"The fact that there were no instructions for their use has been taken as an indication, however, that they were not regarded as part of the operational locking system."
This is wrong because both crews of the ESTONIA knew that these devices had to be locked when heavy weather was forecasted. The 2nd engineer Peeter Tüür has testified in his statement of 03.10.94 - see Enclosure 12.5.183 - among other things "... and there were two fixing bolts , which the boatswain had to fasten". Also Henrik Sillaste had said that the boatswain Heino Leikk of the Captain Piht crew also engaged the manual locks when heavy weather was forecasted. Professor Petershagen continues:
"15.7 Evaluation of the visor actuators and their attachment
The Commission Report states that normal operating loads from the actuator appear to have been high enough to initiate the fatigue cracking of the platform plating and the welds found. Details are given in suppl No. 511, where also repair work is reported. No attempt is made for a fatigue assessment of the platform under normal operation load cycles. It remains open, therefore, whether increased load range and additional cyclic loading due to bad maintenance conditions (rattling) plays a role in the initiation and propagation of the fatigue cracks found. However, the conditions at the visor actuator supports do not seem to have significant influence on the failure of the bow visor as a whole."

57. JAIC Chapter 15.8 :

As stated before the securing bolts at the port side were unable to engage the mating boxes at the ramp due to a misalignment between bolts and boxes of about 100 mm as found by the diver. As a result of this misalignment also the pull-in hook at the port side could not engage its mating lug. This is confirmed by the obvious fact that the pull-in hook as well as the mating lug are both intact, thus did not fail.
At the starboard side of the ramp the lower mating box is totally crushed, thus the bolt could not enter the crushed steel at all and the upper box has no top plate, because it was cut off by the crew on the morning of 27 September 1994 when after arrival in Tallinn, the visor again could not be opened. The starboard pull-in hook is not visible on the publicly available videos and where it was supposed to be there is just a big black hole.
The load considerations of the JAIC are thus purely hypothetical.
Professor Petershagen continues:
"15.9 Other damage to the visor
No comments except that several dirty "waterlines" inside the visor are reported indicating that water had been standing inside the visor for some time with some oil, presumably leaking from the hydraulic pipes of the bottom lock, on top. These findings contradict a general statement that the vessel was in a seaworthy condition at the time of the accident."
Note: The JAIC admits with a view to the water marks visible inside the visor: "The sealing on the forepeak deck had clearly not always been in a condition to keep the lower part of the visor watertight."
In other words, the visor was no more watertight due to the damaged/missing rubber packings on the forepeak deck, by which the inside of the visor was connected to the outside. Due to the physical law of the communicating pipes this means that the water level inside the visor always corresponded to the water level outside the visor, i.e. in calm sea up to the height of the bow wave, which is just above the 3rd stringer level and about 300 mm above the car deck level. During the heavy weather, however, which was prevailing during the casualty voyage, the bow was submerging according to the wave period (ca. 8 sec.) and from time to time even green water was taken on the forecastle deck. This means that the inside of the visor was filled 3/4 or more, i.e. with approximately 150 ts of water, which were moving around inside the visor, a condition like a partly filled tank wit free surfaces, creating additional loads on hinges and locking devices, but also increasing the total weight of the empty visor from 55 ts by about 150 ts to ca. 200 ts with a consequential change of the trim of the vessel.
This has neither been taken into account by the JAIC when determining the wave induced forces and moments acting on the visor, nor did they consider the consequences caused by a water-filled visor on a non-tight bow ramp in respect of the open car deck behind this bow ramp which became necessarily filled with water which did not leave the vessel again via the scuppers.
The fact that there had been water on the car deck at sea for many months before the casualty had been observed by many witnesses, for example Bo Söderman - see Enclosure 12.4.4.161 - Chapter 12.4.4 - and others. Professor Petershagen continues:
"15.10 Failure modes and combined strength of the attachment devices
The external forces and moments given in fig. 15.4 are much higher than the most probable ones given in section 12.3 and hence have a low probability of occurrence. Values from fig. 15.5 and most probable values are compared in the following table (most probable values in brackets):
Visor forces direction
X force (longitudinal) 5,4 (3,6) MN aft
Y force (side) 2,o (1,o) MN starboard
Z force (vertical) 5,4 (3,6) MN upward
Deck hinge moments
direction X moment 5,o/2 (1,7) MN.m upwards on port side
Y moment 15,5 (7,5) MN.m opening around hinges
Z moment 2,o (1,o) MN.m forward on port
With the higher forces the ultimate load of the port side lock of 1.2o MN as estimated by the Commission is reached (fig. 15.5).
Although being very unlikely to occur, the commission argues with the higher values in the conclusions to the section. Contrary to this it should be concluded that the most probable wave loads at the time of accident do not offer a reasonable explanation for the visor failure unless additional factors as maintenance conditions and pre-damage are taken into account.
The Commission states that the ultimate strength of the visor attachment system was exceeded for a load level about equal to the design load used. With a design load of 1 MN per attachment (suppl. No. 204), this statement in the Commission Report is not correct (see comments to section 15.12).
15.11 Design considerations
Most of the statements appear acceptable taking today's technical state into account. It is stated in the report. However, and this is underlined here, the state of knowledge available at the time of the design of MV "ESTONIA" was different to that of today. This refers e.g. to the load assumptions as described above.
Mention should also be made here of the technical state with regard to a fatigue assessment of ship structures.
The state of the art with regard to a fatigue assessment of ship structures is comprehensively described in the Proceedings of ISSC (International Ship and Offshore Structures Congress). This Congress is held every three years. Its work is carried out in Committees, of which the Committee on loads and on fatigue and fracture are relevant here.
It was, however, not possible to derive a procedure for a fatigue assessment of the bow visor details from these Proceedings at the design stage of MV "ESTONIA" (1978/80). This is due to the fact that pertinent data for fluctuating pressures on shell and especially on bow areas are not given and that methods for a fatigue assessment under given load distributions for the structures in question were not available. Such methods have to be based on local stresses and are still under development today.
Sections on fatigue were first included in Classification Rules of Germanischer Lloyd in 1978, but considered as recommen-dations rather than as requirements. These recommendations were generally based on the crane standard DIN 15018, but partly a transfer of this standard to the Rules of Germanischer Lloyd was not possible, e.g. with regard to load assumptions.
Especially, detailed knowledge about fluctuating hull pressure was lacking in the GL recommendation, which is the relevant load on a bow visor.
Moreover, the fatigue regulations of Germanischer Lloyd (Vol. 1 /1978, Section 19) refer to numbers of loads higher than those derived from the operation history of the vessel (see comments on section 15.2).
From the present state of knowledge it can be concluded that with the number of load cycles estimated in section 15.2 of the final report fatigue cracking in the bow visor structure and attachments appear unlikely under regular conditions.
The fact that fatigue cracks have been found in load carrying parts of the visor structure (see report H. Hoffmeister September 1996) indicates maintenance conditions, which differ from those assumed in the initial design. Bad maintenance conditions as damaged or missing rubber sealings and pre-deformations lead to additional fatigue loading, e.g. due to rattling of the visor during opening /closing operations and in seaway.
The Rules of Bureau Veritas did not contain regulations or recommendations for a fatigue life assessment at the design period of MV "ESTONIA".
Also, no IACS (International Association of Classification Societies) unified requirements with regard to fatigue assessment existed at that time. Today, a Working Group within IACS is working on such unified requirements.
15.12 Comparison of design requirements and actual installation
An ultimate failure load of 3 MN per attachment for a design load level of 1 MN and a simplified analysis is quoted in the report. It has to be stated that design against ultimate load has not been and is still not a standard procedure for the attachments in question as well as for most other ship structures.
No figures for permissible stresses seem to have been quoted by Bureau Veritas for bow visor attachments at the time of the vessel's design (see suppl. No. 205 and 206).
Actual safety factors of ultimate against design load were between about 1.6 (side lock) and 7 (deck hinges) in the present case. As already explained in the comments to sections 15.1 and 15.10 of the Commission's Report, the safety factors between ultimate load and most probable load under accident conditions are considerably higher.
15.13 Class and administration implementation require-ments
The Commission finds it unacceptable that retroactive upgrading of existing ships has generally been considered unacceptable within the shipping industry and that this attitude has been accepted by IMO as well as classification societies even in cases of serious safety implications.
Although not quoted in the Commission Report it is emphasized that such retroactive measures cannot be at the responsibility of the building yard, which usually will not be informed by shipowners of damages occurring when the guarantee period has expired.
Conclusions
The sections of the final report of the Joint Accident Investigation Commission commented in this contribution are based on a number of individual investigations. Most of these are carried out according to the present state of knowledge and analysis methods.
However, the results involve large uncertainties, especially with regard to the loads on the bow visor at the time of the accident.
Moreover, the report contains misinterpretations and neglects possible influence factors on the visor failure such as maintenance conditions of the structures in question and pre-existing damage.