Lies and Truths about the M/V Estonia Accident
Chapter 3


About us


Contact info


Order books


Read also the book Disasterinvestigation


Chapter 3 Technical Descriptions of Visor, Ramp and Damages

3.1 Visor Design

The visor of the 'Estonia' was of very simple and basic recognised design. See figure 3.1. The visor consisted only of steel plates and stiffeners supported by three tiers of horizontal girders and the upper deck and a bottom horizontal support girder and some vertical web frames. The total weight was about 55 tonnes.

The visor was held in place against the superstructure and #2 main deck by three locks - two side (sido) locks and one bottom (botten) - Atlantic - lock. The function of the locks was generally to transfer the vertical wave/buoyancy load on the visor to the superstructure and the #2 main deck. Without the three locks the visor would otherwise flip up around the #4 deck hinges by the vertical wave/buoyancy load. The locks were of basic design. Each lock consisted of a lug on the visor, which fitted between two bushes of a locking pin assembly on the superstructure/#2 deck. The two bushes of the locking pin assembly were held in place by lugs. A pin held by the bushes would lock the visor lug. The pins were hydraulic operated. When the pins were engaged, they activated limit switches, which in turn activated the green light on the control panel to the effect that the locking pins had been pushed into the locks and the visor was locked. Evidently the locks could not transmit sideways loads.

The visor was therefore also held in place against sideways and longitudinal motion/forces by locating cones (horns) on the superstructure which fitted into suitable pockets in the sides and at the bottom of the visor. Longitudinal loads on the visor would of course also be transmitted to the superstructure by the rubber seals and other vertical contact points between visor and superstructure.

Fig. 3.1

3.2 External Loads acting on the Visor

External loads act on the visor at sea, when the visor is submerged into the waves, when the ship pitches and heaves. The principal vertical load is buoyancy, when the visor is submerged by the ship's motion. The vertical buoyancy load is mainly a function of the volume of the visor (about 165 m3) and to a less extent the velocity of submerging the visor. Another vertical load, which may act on the visor, is an impact (slamming) load of transient nature. It depends a lot on the speed and course of the ship and the shape of the visor and the angle between visor and waterline.

The vertical load P on the visor can only be transmitted to the superstructure via the three locks. As the vertical buoyancy load is applied on the visor about 2.5 metres forward of the locks and the vertical distance between the side and Atlantic locks is about 2.0 metres, it is simple to show that a vertical upward load P (generally buoyancy minus visor weight) tonnes on the visor is transferred to the superstructure as a compressive horizontal force 0.625P via each of the two the side locks and as a tensile horizontal force 1.25P via the Atlantic lock. See figure 3.2

A vertical load pushing up the visor will always push the side lock visor lugs towards the superstructure and will always try to pull away the Atlantic lock visor lug from the #2 main deck. Therefore the side visor lugs were always in compression and the Atlantic visor lug was always in tension in service at sea.

Depending on the clearances of the three locks and the clearance of the deck hinges some of the vertical load acting on the visor may be transmitted to the #4 weather deck via the deck hinges. Then the load on the locks is reduced. However, for all practical purposes the only function of the deck hinges was to enable to open/close the visor.

Fig. 3.2

3.3 The Function of the Visor

The function of the visor is very simple. The hydraulic locking pins are pulled out by remote control and the visor can be pushed up and be opened by two hydraulic pistons acting on the visor lifting arms connected to the deck hinges. See figure 3.3. To push up the visor you need initially a total force of the hydraulic pistons of abt. 293 tonnes (147 tonnes per piston). The force acting on each deck hinge is then about 119 tonnes, even if the hinge was designed to easily handle 200 tonnes. As the visor is lifted up, the required lifting force is reduced (as the bending moment in the lifting arm is reduced) and thus also the force on the hinge is reduced. Thus - the maximum force acting on each deck hinge - 119 tonnes - is applied, when the opening starts or the closing ends.

Closing the visor is evidently the reverse operation. Just before the visor comes to rest on its supports, the maximum force on the deck hinge is again experienced. The load on the hinge becomes zero, when the visor rests on its supports. When the three visor locks are thereafter engaged, the visor should in principle have been pushed against the rubber seals around the visor and there should have been no clearances in the three locks, while there should have been a little clearance at and in the hinges. All the vertical (upwards) load on the visor would then be transmitted to the superstructure only via the locks at sea.

3.4 The Bow Ramp

The bow ramp inside the visor was also of simple basic recognised design. It consisted of a strong plated frame grid with four hinges at the lower end. The scantlings of the ramp were decided by the tyre pressure of the trailers rolling over the ramp. The ramp also acted as the inner, weathertight door protected by the visor leading to the enclosed superstructure of the car deck (the garage) and should be able to withstand a certain water pressure.

The ramp was, like the visor, hydraulic operated. When closing the ramp, it was lifted up by two hydraulic pistons. Then two hooks were moving out of openings in the port and starboard front bulkheads and engaged mating lugs at the two sides of the upper part of the ramp and pulled the ramp against the rubber seals around the ramp opening. Then two locking pins each side moved out of the side, one after the other, into mating pockets at the ramp side. In fully extended positions the pins activated limit switches connected in series and when all four pins were engaged the green indicator light on the control panel on the car deck was activated.

Fig. 3.3

3.5 The Control Panel

The panel for visor and ramp operation and control was located at port side inside the ramp on the car deck. During closing operations you evidently had to close the ramp first as outlined above and then you had to close and lock the visor.

3.6 The Visor in Service

The normal vertical load acting on the visor is, as described in 3.2, a function of the visor volume and weight. You would expect the maximum load to be about 165 tonnes of buoyancy minus 55 tonnes of weight, i.e. 110 tonnes to which you could add, say 30%, to account for dynamic effects and the fact that the visor might be submerged below its upper part. The total vertical upwards load on the visor would then be P = 143 tonnes when the ship puts the bow into a wave up to the top of the focsle. This load should then be transmitted to the superstructure/#2 deck via the three locks - 1.25P = 179 tonnes (horizontally) via the Atlantic lock (tension in the visor lug) and 0.265P = 90 tonnes via each side lock (compression in the visor lugs). No load should have been transmitted via the deck hinges (as there should have been a positive clearance between the visor hinge bush and pin). See figure 3.2.

Model tests carried out by the JAIC generally confirms the magnitude of the vertical load acting on the visor. However, transient (shorter life = milliseconds) loads of higher magnitude - slamming - were also recorded. The slamming load was always perpendicular to the visor side and its sideways and longitudinal components were of course transmitted to the superstructure via the horns/pockets and the vertical contact points. The vertical component of the slamming load was naturally transmitted to the superstructure only via the locks.

Note August 2000 - read the analysis of the JAIC model test report.

3.7 The JAIC Damage Allegations - the Atlantic Lock

The JAIC alleges that a high amplitude, transient impact load (360 tonnes 4.12) on the visor first ripped apart the Atlantic lock at about 01.00. There is no proof for this. Nobody heard (a) the impact itself or (b) that the visor lock was ripped apart. The energy of the impact should have been noticed as a shock wave aboard the ship (unless this was event 1 in 2.2 - 20 minutes earlier).

The Atlantic lock was damaged as follows:-

The two longitudinal #2 main deck lugs supporting the port pin bushing had been torn apart in the 8 and 2 o'clock positions. The port bush had been torn away from the remains of the lugs and was missing. The longitudinal #2 main deck lug supporting the starboard pin bush had likewise been torn apart. It should be noted that the starboard bush was also connected by welding to the #2 main deck by a transverse bracket and that the bush had been torn away from the transverse bracket

The locking pin was found connected to the hydraulic piston rod, which was in the pushed out position, and which had been bent upwards and sideways to port.

The visor lug was bent to starboard (see figure 8.10 in the Final Report (13) and 4.10 or fig. 10.5 from the Part Report right) and its connection to the lower support girder was damaged:- the support girder web and face plate were buckled on starboard side and were fractured on port side.

JAIC has never commented upon the visor lug damages. If, as alleged by JAIC, the visor was damaged by an excessive vertical slamming load, it would have pulled out the lug from the lock without bending the lug (and without buckling and fracturing the lug support girder web and face plates) Appendix.

3.8 The JAIC Damage Allegations - the Side Locks

Immediately after the Atlantic lock fails as alleged by JAIC, the load transmission between visor and superstructure changes. A vertical load on the visor should now only be transmitted via the side locks and via the deck hinges.

It is simple to show that a vertical load P1 on the visor is transmitted as 0.42*P1 horizontally at each lock and hinge. The load is generally tension in the side visor lugs and compression in the visor hinge arms. See figure 3.8. Actually, you can see that the Atlantic lock was not required in the first place as the visor is held in place by two side locks and the deck hinges. This was the way bow visors were built in the beginning (1960's) - two side locks and two deck hinges. The Atlantic lock was added later to un-load and to reduce wear and tear of the deck hinges. As the JAIC claims (see paragraph FR15.3 of Final Report (13)) that the accident was caused by a too weak designed bottom lock, it is easy to counter this statement by suggesting that the bottom lock was in fact not required at all to keep the visor in place. The function of the Atlantic lock was to un-load the very strong hinges and to make the visor connection to the superstructure stronger. Even if the load was reduced in the side lock after the Atlantic lock failed, its direction in the side locks was reversed, and JAIC alleges that the visor side lugs now were ripped off the visor support plates.

We know that the side visor lugs were ripped away from the visor support plates. JAIC alleges that the lugs were ripped off, when the visor was pushed upwards by another impact load at about 01.01 hrs (the German Group of Experts for a long time alleged that the side visor lugs were ripped away when the visor was tipping forward, still connected to the Atlantic lock).

Fig. 3.8

Regardless, it should be clear that the Atlantic lock could not have been damaged by the same wave (load) as the side locks. First the Atlantic locks must fail by one very large impact, after which the load transmission pattern is modified, and then the side locks must fail by another very large impact. If one big impact was followed by another big impact, then of course all three lock could have been ripped apart within 10-12 seconds, but it could also be a longer time between the big waves. Nobody heard the second impact against the ship and that the side locks were ripped apart.

3.9 The JAIC Damage Allegations - the Deck Hinges

After the alleged side locks failure, the visor was only held in place by the deck hinges. However, the visor lifting arms were connected to the hydraulic lifting pistons and it is now assumed that the pistons restrained the motion of the visor. If the visor had not been restrained by the hydraulic pistons, a vertical wave load with enough energy exceeding the visor potential energy would of course had swung the visor around the hinge points, and the visor would have ended up upside down on the focsle deck in front of the deck house! This did not happen.

The JAIC alleges that the visor was now flipping up and down around the hinge points, when big vertical wave loads acted on the visor, and that there were heavy noises. The JAIC writes (on page 175 in (13)) that:

'it is beyond doubt that the sounds were caused by the visor moving and pounding on the forepeak deck'.

The JAIC says that:

'witnesses from several areas on board heard a repeated metallic noise from the bow area during a period of about ten minutes, starting after one o'clock'.

Say that this pounding took place at 01.02-01.12 hrs. According to point 2.2 in this book the 'Estonia' was already sinking at this time. When you read the 'detailed' testimonies of survivors in Chapter FR6 of (13) 4.7 it is difficult to see how JAIC could have concluded 'beyond doubt' that there were repeated metallic noises from the bow area, as very few survivors heard anything like them.

Say that the upward load was P2 (including the visor weight). The load P2 causes a tensile force 2.67*P2 in each hydraulic piston and a compressive load of 2.16*P2 in each hinge. See figure 3.9. JAIC assumes that the compressive load 2.16*P2 was enough to break a hinge and that the tensile load 2.67*P2 was enough to rip off a piston from its support at #3 deck. Nobody heard when the pistons' supports were ripped away and when the hinges broke.

When the deck hinges had failed and the lifting pistons were loose, JAIC alleges that the visor moved forward and rested on the inner ramp and tried to push open the inner ramp from aft.

Fig. 3.9

3.10 The JAIC Damage Allegations - the Inner Ramp (the weathertight door protecting the superstructure/garage)

If the visor was loose as shown above, it would probably move forward and rest on the ramp, which protrudes up into a recess built on top of the visor. See figure 3.10. Actually the bottom of the visor would rest on the extension of the #2 car deck - the fore peak deck - forward of the ramp, the top aft end of the ramp recess on top of the visor would rest against the top of the ramp and the lifting pistons would rest against the deck beam at fr. 159. If a wave load exceeding the weight of the visor was acting on the visor, it would lift the visor up on the focsle deck aft of the visor! This did not happen. As can be seen from the geometry of the arrangement the hydraulic lifting pistons upper eyes must eat through the #4 deck beam at frame 159, before the visor recess starts to touch the ramp top, and nobody heard this.

Note August 2000 - The Independent Fact Group, Stockholm, has in the spring 2000 presented an analysis to the effect that the visor could never have had cut through the #4 deck beam at frame 159. There are two reasons for this - (a) the cutting edges - the upper eyes of the lifting pistons on the visor arms are completely undamaged - not even the paint has been 'cut' off, and (b) from pure strength of material theory (cutting) the relevant cutting force (forward) cannot be achieved. The question if the deck beam is still undamaged remains unanswered. The Fact Group web page is - visit their site!

However, JAIC alleges that the recess of the ramp at the top of the visor now pushed forward the ramp, which then was dislodged from two hooks and four side locking pins/bolts. This should have been taken place at about 01.13 hrs. Nobody heard this either.

Fig. 3.10

The JAIC allegation is not supported by any facts or findings anywhere. If the visor ramp recess had dislodged the ramp from its locks, you would expect the recess to be damaged, but only some stiffeners on the port side are bent a little - the plate is straight and the aft, lower edge of the recess is straight and undamaged.

Furthermore, the #4 deck ramp hooks have not been salvaged and brought to the surface and to a laboratory for investigation. The JAIC states that the hooks were locked before the accident 1.15.5, so you would expect that they had been broken, but there is no proof for this. JAIC has also stated that the four ramp side locks were engaged before the accident, but it is not clear how and when they were damaged, e.g. dislodged by the visor pushing on the ramp top.

The longitudinal load pushing the ramp forward should have been of the order 10 tonnes (0.10 MN) 4.23 (to counter the moment trying to tip the visor forward) but probably less as the visor was kept in place also by the bulkhead at fr. 159. As there were six connection points between the ramp and the superstructure - two hooks and four locking pins/bolts the average load on the lock was only 3-4 tonnes. Evidently the visor resting on the ramp top could not have ripped apart the hooks or the pins/bolts. So what actually happened?


3.11 The JAIC Damage Allegations - Inner Ramp Opening/Loss of Visor

After the ramp had been dislodged from its locks, JAIC alleges that the ramp shifted forward to a partly open position and little water flowed in. See figure 3.11. The time for this alleged event is at 01.15 hrs, when it was observed by 3/E Treu on the VDU of the garage in the ECR 1.9, but when it was not reported by the engine crew to the bridge 4.23.

Soon thereafter the visor was lost and the ramp was pulled fully open. A lot of water could now enter on the #2 car deck. This is described in chapter FR12.6.2 of the Final Report.

According to JAIC the garage filled up with 300-600 tonnes per minute and after a couple of minutes the ship heeled 20° (when there was 1 000 tonnes of water on the #2 car deck). It is interesting to note that the ramp on the wreck is in the partly open or almost closed position as shown in figure 3.11.

Fig. 3.11

With the ship heeled 34° it is easy to visualise how an impact load on starboard side lifted the visor over the ramp 2.8. There is no mention at all about a sudden list 50° to starboard, uprighting and a new equilibrium at 15° starboard 3.16 and that many people were hurt already then.

(Addendum February 2001 - the JAIC states that the visor lifting hydraulics ripped open the deck plate in front the deck openings for the hydraulics port and starboard and also ripped open part of the front bulkhead - its rounded top part. However - no real pictures are shown of the alleged damages in the Final Report - actually no pictures of the starboard focsle deck and front, collision bulkhead are shown at all. In August 2000 private divers filmed a large opening in the starboard front bulkhead about three metres below the focsle deck - see picture to the right. This opening does not extend up to the foclse deck! Thus - the visor hydraulics could not have caused the damage to the front bulkhead).

3.12 The JAIC Damage Allegations - the Loss of the Vessel

After the ship had heeled 20° with 1 000 tonnes of water on the #2 car deck, the JAIC alleges that the vessel was lost and sank without tipping upside down. How these events happened are not fully explained in the Final Report, 4.20 and 4.21. The hull providing buoyancy was undamaged!

3.13 The German Group of Experts

The German Group of Experts has made an excellent job (11) to establish first the actual condition of the visor and the inner ramp and second what actually happened aboard by interviewing surviving crew and passengers and other people, who had previously worked on or travelled with the ship.

(Note August 2000 - the full German report was published on the internet in June 2000 ).

3.14 The actual Condition of the Visor

According to the Commission (13) the visor was in excellent and original condition without any modifications done to it during 15 years and without wear and tear. According to the German Group of Experts (11) the visor was not maintained properly. The Final Report does not mention the following.

The rubber packings were not renewed when worn and the visor was not weathertight. Also the pre-tension function of the rubber packings was lost, and this had the effect that the visor was vibrating/shaking at sea within the play of the three locks. The German Group of Experts did not point out clearly that this meant that the load transmission between visor and superstructure was modified and that more load was now put on the hinges 3.2.

Severe structural damages had been caused to the visor in the winter 1993/4. The result was that the whole geometry of the visor was changed and that nothing fitted anymore.

Note August 2000 - the writer now believes that the Atlantic lock did not fit at all and was thus not in use when the accident occurred.

The deck hinges had been manipulated. In fact the Germans showed that the hinge bushes had been replaced in such a manner that their load carrying capacity was drastically reduced. However, as shown in 3.3, the hinges were only required when opening and closing the visor, when the maximum load experienced was of the order 119 tonnes (in tension), when the visor was almost closed and the visor arms were horizontal. Had the hinges been broken in service, this would have taken place when the crew tried to open the visor in port. As the visor was clearly opened and closed at Tallinn on the 27th of September, the visor hinges could not have been so bad as suggested by the Germans - the load carrying capacity had been reduced by the burning marks to a mere 20% of the original load carrying capacity (page 16 of (11)).

The Atlantic lock had been renewed and reinstalled at a different location already in 1981/2 with only 3 mm welding seams between the bushings/lugs instead of 8 mm original. The lock had also been repair-welded in a very unqualified manner that resulted in considerable loss of it load-carrying capacity (X ... has cut off the upper parts of the three steel lugs holding the bolt .... bushings .... After having removed these parts he has welded the bolt ....bushing into positions, fitting .... the changed position of the visor lug...). The visor lug had also been manipulated in different ways obviously to make the bolt fit through the bore of the lug.

The Atlantic lock hydraulics were not working and the locking pin had to be hammered in and out of the lock. The inside of the visor was covered by a hydraulic oil film.

Note August 2000 - the Germans have not concluded, as this writer, that the Atlantic lock was probably damaged before the accident and was thus not even used on the fatal voyage.

The side locks had a play of 10 mm and it had to be assumed that the condition of the side locks were no more original.

3.15 The Actual Condition of the Ramp

According to the German Group of Experts the ramp was also not maintained properly.

The ramp weathertighness was not maintained. The Germans found that one hinge was definitely damaged and that two side locks were probably not lined up and could therefore not be used. This meant that the ramp was leaking. However, the leaking could not have been serious and the crew was aware of the problem and fitted temporary packings (by cloth and mattresses) to stop or reduce the leaking 2.21. The Germans make a big story about this defect, but the writer thinks that the defect was not particularly serious and did not affect the seaworthiness of the ship. The Commission in (13) says that the ramp was weathertight and undamaged.

3.16 The German Damage Allegations - the Visor

The German Experts allege that the primary causes for the sinking of the Estonia were (a) the extremely bad maintenance condition of visor and bow ramp, their hinges and locking devices, in connection with (b) a completely wrong loading of the car deck and a highly excessive speed under the prevailing wind- and seastate conditions, which had been forecasted (11).

With regard to the visor the Germans suggest that it was full of water, so the ship trimmed on the bow. Water also leaked onto the car deck through the leaking ramp. The Germans do not quantify the leakage in tonnes/minute or comment upon the fact that water leaking onto the car deck is drained out through the scuppers. The Germans then suggest that the visor hinges broke first (and not last as alleged by the JAIC 1.11 followed by the side locks. The visor then apparently moved forward and dislodged the ramp from its locks as early as 00.45-00.46 hrs and fair amounts of water started to enter the car deck. The visor was held back by the hydraulic pistons resting against the bulkhead/beam at frame 159. The inflow in tonnes/minute is not quantified.

The Germans then suggest - without published proof - that the crew observed the loose visor, reduced speed and turned the ship into the wind and tried to secure the loose visor (item (g) on page 37 of (11)). The ship was then upright, but at 01.02 hrs the ferry suddenly and abruptly heeled rapidly to starboard to an estimated angle of about 50°, however, almost uprighted shortly afterwards and then took a list of about 15° to starboard, which was slowly increasing. This is a very important observation backed up by the proof that many persons aboard hurt themselves and loose items fell when the ship heeled over. The Final report (13) does not mention anything of that.

Interestingly, the Germans then assume that the starboard side of the visor was pushed upwards due to buoyancy caused by the extreme heel, which also broke the Atlantic lock, and the visor was now only connected to the vessel by the two lifting cylinders.

3.17 The German Damage Allegations - Water on the Car Deck

The Germans then suggest that the list increased due to additional water quantities on the car deck streaming in through the partly open bow ramp, all of which accumulated at starboard side, and shifted the cargo and led to the increasing list.

Then the Germans make an incorrect suggestion. They say:

' ... at about 01.20 the visor was moving forward whilst the hydraulic cylinders broke through the front bulkheads and the visor separated from the vessel, the list then must have been 50°-60° and water streamed onto the car deck as well as onto the lower decks in increasing quantities'.

What the Germans fail to realise is that at 50°-60° list the vessel is never stable with water on the car deck - she will always turn upside down. At 50°-60° list no water can stream from the car deck onto the lower decks as the car deck is watertight and the openings to the lower decks are high above the alleged water on the car deck, 2.16 and 5.5.

(Note August 2000 - the Germans (Mr. W. Hummel) has later admitted in a newspaper interview that the ship should have turned turtle at this time).

As can be seen from figure 3.10 the visor can hardly dislodge the visor before the hydraulic cylinders broke through the deck beam at frame 159.

Therefore the German accident scenario is not convincing. So how could the 'Estonia' sink?

3.18 The Sauna was flooded and no. 1 Deck was flooded

The sauna is on the tanktop (deck no. 0) forward. The Germans say in (11) (page 36) that:

'According to statements of most survivors, in particular of the key witnesses ........ Passenger CÖ (cabin 1049 - 1st deck) ... Passenger MN - cabin 1027 - 1st deck, Passenger BN - cabin 1026 - 1st deck) ...... the sequence of events must have been somewhat different from what the JAIC has found and also what the authors Hellberg/Jörle assume in their book 'Katastrofkurs' (10), because ......... (b) there was water on the 1st deck, ... , in particular in the forward part, already before the sudden starboard heel occurred; (c) the sauna/swimming pool compartment on 0-deck ...... was flooded and under pressure, i.e. open to the sea, before the sudden starboard heel occurred .... '.

There are several other witnesses stating that there was water on deck no. 1. The Commission in its Final Report ignores all these testimonies, because it cannot explain how water flowed up on deck no. 1 before the sudden starboard heel occurred.

The writer finds it amazing that the JAIC ignores clear statements to the effect that 'Estonia' was leaking. These statements might not have been clear when the JAIC made its first statement on October 4, 1994 1.4 but must have been clear later. The statements about water on the car deck are not very convincing - the third engineer says he saw water on the car deck, the systems engineer says he didn't see any water on the car deck, but that the third engineer told him, that there was water on the car deck, and the motor man says that he saw water on the car deck (after the sudden listing had taken place (at 01.15 hrs?) and when all water should have been down in starboard aft corner), 4.7 and 4.23. Why the JAIC believes a third engineer saying that there was water on the car deck? Why believe him, when according the laws of nature, a ship with water on the car deck tips upside down, and the 'Estonia' did not do that? There are many questions that the JAIC does not answer in (13).


Chapter 4 Back to Content