Why did the number 1 funnel collapse

Not open for further replies.
As Titanic sank down by the head, the angle of the boat deck began to tip forward. At about 2:00am the sea reached the boat deck and the bridge began to flood. Moments later, the forward funnel collapsed and fell into the sea, crushing the bridge and killing many people swimming in the water.

I've seen animation of the sinking, and I've played with the angle of the ship during the sinking myself, and I still can't understand why the forward funnel suddenly collapsed.
The funnels were not top heavy that I'm aware of, and they were all secured to the superstructure with 6 heavy cables. (I think its 6, but not positive about that)
Even though the angle of the deck reached about 20 degrees, thats not unusual for a ship to endure especially during rough weather, right?
And why didn't the other three funnels also collapse at that point?

So what was different about the forward funnel?
Why did it tear loose?

Yuri Singleton

Paul Rogers

Hi Yuri.

May I have a stab at this one? No. 1 funnel was sited in front of the forward expansion joint. However, some of the cables supporting the funnel were afixed aft of the expansion joint.

When the expansion joint began to open as the bow sank under the sea, the cables were stretched to breaking point. When they finally snapped, the weight of the funnel caused it to collapse, as it was no longer supported from behind.

I read somewhere that the anchoring points for the #1 funnel cables on Britannic were moved forward of the expansion joint after Titanic sank, as their original siting was seen as a (minor) design flaw.

Hope this helps.


I hate to contradict you, but I have run some modeling on this. It wasn't the failure of the guy wires that caused the funnel to collapse, but rather the differential of pressure between the water around the base of the funnel uptakes and the air inside. Think of an inverted empty glass being pulled under the water as an over-simplified example. There was enough slack in the guy wires to accommodate movement in the expansion joint; otherwise, you'd be snapping cables during heavy weather. I believe the wires failed when the funnel was swept off its base.


Couldn't the stress on the decks owing to the weight of the water in the bow (which culminated in the breaking away) have at least partly caused the funnel to fall? What Paul has said seems quite reasonable to me. Not that the other factors Parks suggests weren't there too. But the stresses on the submerged portion of the ship, if they were as great as is believed, could not be compared to that caused by a normally heavy sea. It just seems that the wires, after the slack was taken out by the bending decks, would have finally snapped.


Paul Rogers


Thanks for your post. Yet again I have to learn the lesson not to believe in everything I read, but to go back to the source and check, check, check! I wish I could remember where I read that info on the cables now...

Was I wrong that the Britannic had her cables moved to forward of the expansion joint? (I'm sure I read this somewhere!) If so, what was the given reason for the change?

Perhaps Randy has got it right. One question I have, which touches on what Randy says, is as follows:

When I believed the "snapping cables" theory, I was working on the assumption (careful!) that when the funnel collapsed, the expansion joint had opened A LOT further than the original design had envisaged - thus causing the stress and final failure of the cables.

I based this assumption on the state of the expansion joint as we see it today. After all, the ship was near breaking when the funnel collapsed and, after she did break, there should have been a reduction in the bending force on the joint. Thus, the expansion joint must have opened a LONG way by the time the funnel fell....further than the cables could stretch, perhaps?

What do you guys think?


I think you're alluding to the same forces I am, but instead of decks, think of the comparatively weaker steel of the superstructure and funnel uptakes, rather than the decks themselves. The "weight" (I think in terms of pressure) of the water building around the empty (air-filled) uptakes as the bow filled and submerged pressed in until a portion of the uptake gave way, allowing water down the uptakes and causing the upper structure of the funnel to collapse. The wires finally snapped when the funnel fell forward.

Could the funnel have fallen of its own weight had the funnel stays snapped? I count 4 out of 12 wires supporting the funnel straddling the expansion joint break. So presumably, if those four wires broke, the funnel is free to fall forward. Or is it? Normally, the funnels has an aftward rake of about 8.5 degrees (I'm pulling that number off my protactor. According to the model presented in the Hackett/Bedford analysis of the sinking, the ship trim was somewhere between 7 and 12 degrees at the time the funnel collapsed (between Conditions C-6 and C-7 in their analysis). This means that the funnel was either perpindicular or even leaning a little forward. Let's say those aft 4 stays failed because they didn't have enough slack to accommodate the opening of the expansion joint. That leaves two more stays which, given the new angle of trim, are supporting the funnel from falling forward. Why would the funnel fall? There was no extreme angle forward...at most, the funnel was leaning forward 4 degrees or so. With 8 stays supporting it, I just don't see the funnel falling due to simple gravity.

As some of you know, this runs counter to the argument that is presented in the final report on the disaster from the Marine Forensic Panel. This is an example of an area where, even though I am a member of the Panel, I take an alternate view.

The idea that differential flooding caused the collapse of the funnel was first put forward by Edward Wilding, a naval architect for Harland & Wolff at the BOT Inquiry in 1912. Like Parks, he proposed that the outside of the funnel and casing flooded faster than the inside, the water pressure causing the structure to collapse and bring the funnel down.

The broken stays/expansion joint theory was first put forward by one of the surviving deck officers (Lightoller I think — G. Behe would know for certain) and was given wide circulation by Walter Lord’s “Night to Remember”.

Between the two, I prefer Wilding’s partial flooding theory. Aside from his qualifications, the theory has better reasoning behind it and a non-spectacular ring to it.

The broken stays theory has several difficulties. First, funnels have to withstand extreme inclinations while rolling and pitching in a storm. (although funnels have broken off in storms, the cause was attributed to extreme acceleration on account of the ship’s cargo being loaded to make her very “stiff”, not degree of roll). Second, funnel guys are rigged to allow for the natural expansion of the funnel with heat, plus normal working of the ship. I feel that there was more than enough slack in the guys to accommodate deflection of the structure as the ship sank. Parks has already made the case for objections based on surviving stays afore the expansion joint supporting the funnel and the possibility that the trim for the bow section was not as dramatic as shown in movies for the sake of its horror factor.

Bill Sauder

20919. (Mr. Laing.) There is evidence that certainly one, and I think two, of the funnels fell? — (Mr. Wilding) Yes.

20920. I think the evidence is that the forward one and the after-one fell? - Yes.

20921. What would account for that? - The funnels are carried from the casings in the way of the comparatively light upper decks - that is, the boat deck and A deck. When these decks became submerged and the water got inside the house, the water would rise outside much faster than inside, and the excessive pressure on the comparatively light casings which are not made to take a pressure of that kind would cause the casing to collapse; would take the seating from under the funnel and bring the funnel down.

Testimony excerpts from the Board of Trade Inquiry curtsey of The Titanic Inquiry Project -- http://www.titanicinquiry.org/
A few notes on expansion joints and funnel guys.

The Olympic/Titanic’s superstructure was constructed with two expansion joint. This number was increased to three joints for the Britannic and the placement of the joints rearranged to equalize their spacing.

I suspect that this may have been done for one or both of the following reasons:

1. When the new gantry davits were designed for Britannic, it was realized that the superstructure needed to be strengthened to accommodate the additional load and so the weight of the deckhouse reduced as much as possible to reduce top hamper. A deckhouse built of even lighter material will require the introduction of additional expansion joints, to limit the fore-and-aft run of the deck house side and to prevent overloading and deformation in any one section.

2. It is possible that deformities / cracks were showing up in the deckhouse on the Olympic from ordinary working of the structure in a sea way, suggesting that an additional joint would help contain local failures with out resort to heavier material.

It should be noted that Titanic was unusual in that she had only two expansion joints leaving a midship section of about 220 feet — that’s a tremendous length for a midship’s deckhouse as the following small table will show. Cunard liners typically had three joints (as well as Britannic) and this helps reduce the midship’s sections of deck length considerably.

Mauretania midships deck section length : abt 100 feet each
Aquitania midships deck section length: abt 100 feet each
Britannic midships deck section length: abt 126 and 159
Titanic midship deck section length: abt 220 feet

I would also point out that even after the redesign on the expansion joints, Britannic’s funnel guys cross over them to adjacent deck sections. This was also the case on the Mauretania, Aquitania, and Queen Mary.

Bill Sauder
Parks & Bill,

I stand happily corrected and convinced. Thanks for the details. Those of us without the technical knowledge need it explained; at least I do. So thanks again.


Paul Rogers

Parks and Bill,

I echo Randy's words. And Bill, thanks for the additional info on Britannic's expansion joints. Fascinating stuff!

What amendments in regards to expansion joints and/or funnel guys were made to Olympic, BTW? Or did WSL just live with the cracks in the superstructure?

FWIW, I don't think the Olympic's basic structure was really altered. At least not to the point of adding an expansion joint. Such a modification would have required a nearly complete(Meaning expensive) rebuild of the superstructure and a prolonged dockyard period.

Of course, there were internal modifications, after the Titanic's loss, but that involved adding things to what already existed. Raised watertight bulkheads for example, double hull, extra lifeboat provisions and so on.

I'd go into greater detail if I had a more reliable source

Michael H. Standart
It's too long to go into here, but I can point you to Simon Mills' "RMS Olympic: The Old Reliable," which recounts the sad tale of Olympic's material decline during the late 1920s-early 1930s.

Sparks, THS offers that one don't they? I understand the cracking was serious enough that the ship was slapped with a speed restriction by the 30's.

Michael H. Standart

I can't say if THS carries that book, but I got my copy from iBS.

Olympic was placed on the BOT's Confidential list as early as 1927. In 1931, the BOT only granted a seaworthiness certificate for a six-month period while repairs were effected. In 1932, White Star itself informed the BOT that Olympic would be restricted to 21 knots.

Total replacement of the topside shell plating was required, and with a price tag of over 100,000 pounds Sterling, the work could not be justified for a liner of Olympic's age.

Paul and Randy:

Thanks for your recent kind comments about my notes on the Titanic's expansion joints and funnel rigging. There were a few loose ends that I wanted to tie down.

Paul asks "What amendments in regards to the expansion joints and/or the funnel guys were made to Olympic? Or did WSL just live with the cracks in the superstructure?"

I was only speculating about cracks showing up in Olympic's deck house because of metal fatigue through an insufficient number of expansion joints, but given the ship's history late in her career, I don't think it's unreasonable. Any cracks would most likely show up at the lower corners of the deck house windows and possibly where the deck channel passes through the deck house side (Its covered by a cast iron shoe in most photos).

Any cracks in the deck house side would be mostly a cosmetic concern, and so a small patch or "doubling" could be welded over to help arrest its advancement and hide the fault.

If you look at the side of the QE2 you will see dozens (maybe a hundred now) of similar patches on her promenade deck side screen windows (the large square windows on the sides of the ship, right under the Boat Deck). Either Cunard or John Brown decided to do away with expansion joints and make the entire superstructure cooperate in the overall hull strength. That meant that these square windows are BELOW the strength deck -- a major Bozo no-no. Square windows have a tendency to focus stress in their corners and cause cracks radiating from them in the metal. That's why portholes are round and shell doors have generous radii in the corners and heavy doublings at the top and bottom. There were several witless moments in the QE2's design, this is one of the more obvious.

Parks said : "the funnels has an aftward rake of about 8.5 degrees."

Close, but not quite. The Titanic's funnels had a rake of 2 inches in 12 or 9.5 degrees.

Also: "It's too long to go into here, but I can point you to Simon Mills' "RMS Olympic: The Old Reliable," which recounts the sad tale of Olympic's material decline during the late 1920s-early 1930s."

A summary from My Friend Simon's book (Did I really say that?):

1926 stem frame replaced
1926 number of small cracks starboard side of Bridge deck.
1927 similar cracks discovered on port side, Olympic placed on BOT's Confidential list.
1929 new stem frame heavily pitted
1930 pitting had advanced alarmingly and stem sheathed in white metal.
1930 cracks detected in 1927 in bridge deck advancing. Extensive welding and fitting doublers over the affected areas.
1931 strain deterioration in A deck support girders forward of Number 2 funnel -- large number of slack rivets near the aft expansion joint. BOT grants a certificate of seaworthiness for a period of six months; renewed six months later.
1932 Olympic limited to 21 knots to mitigate hull stresses
1932 cracks found in Olympic's engine bedplates, thrust block foundations and part of the crank shaft.

Mike writes: I don't think the Olympic's basic structure was really altered. At least not to the point of adding an expansion joint. Such a modification would have required a nearly complete (Meaning expensive) rebuild of the superstructure and a prolonged dockyard period.

I agree. Post IMM White Star was in serious financial difficulties. The money was not available in the depression for anything more than Band-Aids.

Bill Sauder
Not open for further replies.