Bill Dow's theory on the Tay Bridge disaster (2/2)
HTML-код
- Опубликовано: 9 сен 2009
- Bill Dow has spent over a decade researching the Tay Bridge disaster, and here he shares his theory that the train de-railed through a combination of heavy wind and a weak girder.
(Part 2 of 2)
Playlist link - • The Tay Bridge disaster
Transcript link - media-podcast.open.ac.uk/feeds...
Read the article: The Tay Bridge Disaster: Design of the bridge www.open.edu/openlearn/scienc...
Study a free course on Tay Bridge disaster at the Open University www.open.edu/openlearn/scienc...
Study a free course on An introduction to design engineering at the Open University www.open.edu/openlearn/scienc...
Study Q65 Bachelor of Engineering (Honours)
www.open.ac.uk/courses/qualifi...
The Open University is the world’s leading provider of flexible, high-quality online degrees and distance learning, serving students across the globe with highly respected degree qualifications, and the triple-accredited MBA. The OU teaches through its own unique method of distance learning, called ‘supported open learning’ and you do not need any formal qualifications to study with us, just commitment and a desire to find out what you are capable of.
Free learning from The Open University
www.open.edu/openlearn/
For more like this subscribe to the Open University channel
/ @openlearn_ou
Like us on Facebook: / ouopenlearn
Follow us on Twitter: / oufreelearning
#OpenUniversity #engineering
Like with any disaster, we can safely assume that the collapse of the Tay Bridge was caused by a multitude of factors all working in unison. Yes, the design was flawed from the onset because Thomas Bouch neglected to properly account for the wind, the metal and the casting were sub-par, the ties of the diagonal bracing were loose and the lugs cracked by fatigue... The overspeed train derailing at the critical time on that stormy December night was merely the straw that broke the camel's back.
I am not qualified as an engineer, but speaking solely as a layman I feel this theory, as presented here, is in conflict with evidence presented at the 1880 inquiry into the disaster.
The video suggests that the collapse of the bridge was initiated by a shock created by the 2nd-class carriage which, having allegedly derailed, struck a plate at the junction of the fourth and fifth of the High Girders. However, since the train was in motion (I estimate its speed as 17mph, see further down for my workings), it would be expected that momentum should have carried this carriage some distance beyond the point of impact in the few seconds between the administration of this shock and the failure of the bridge. However it did not - the plans detailing the disposition of the wreckage show that the 2nd class carriage had only just reached the end of the fourth girder at the point the bridge failed, meaning that the only way this evidence matches with this theory is if either the train stopped dead at the point of the collision, or that the collapse of the bridge was instantaneous - the whole structure toppling the very instant the 'derailed' carriage struck the plate.
My second thought relates to the derailment - the video suggests that the van lifted clear of the track at a 'kink' in the roadbed where two of the High Girders met, presumably the third and fourth. This would mean the train travelled the whole 245ft length of the fourth high girder with a derailled vehicle, and that none of the train crew responded to this, even though the shock of a vehicle dropping off the rails and riding over the ballast would be felt throughout the entire train. Dugald Drummond, locomotive superintendant to the North British Railway, testified on this matter, and said that the first instinct of a driver who realises his train has derailled is to shut off steam and apply the brakes. However, when the wreckage of locomotive 224 was recovered from the firth, she was found to be in forward gear, her regulator wide open, with no application having been made of the handbrake. The fall of the bridge was thus sudden, with no build-up that would have given the driver time to react, as would be expected with a scenario shown here.
Testimony shows that the train was not speeding, having taken some three minutes to travel from the south shore to the fifth high girder. From Google Earth, I estimate this distance to be about .85 of a mile, giving the train an average speed of 17mph, or about 25 feet per second. The fourth high girder being one of the longest (245ft), this would have given the train crew nine or ten seconds to realise a vehicle had derailed and take action, yet apparently none was taken, even by the guard, who was riding in the van immediately behind the vehicle that is suggested to have derailed, and who could not have missed it.
Then there is the disposition of the recovered bodies. The fireman was found with his face blackened, meaning he was probably working at the firedoor when the train fell, the sudden displacement of the engine causing the firebox contents to blow back onto him. This indicates all was normal on the footplate right until the moment of disaster, inconsistent with the suggestion of a partially-derailed train.
The guard, when recovered, was found to have a calm expression on his face, except for his hands, which were tightly clenched. This suggests that the failure of the bridge was sudden, giving barely time to react, and does not suggest the consternation of a guard trying to stop a train in which a vehicle has derailed and is dancing over the permenant way.
Furthermore, Dugald Drummond testified that none of the wheels of the recovered train showed damage consistent with a derailment, such as damaged tyres or chunks taken out of the flanges.
The video also implies, somewhat scurrilously, that forensic investigation of the wreckage was lacking, but this is not the case. Although much of the ironwork was left in the firth due to time constraints and the limitations of diving gear of the era, all the vehicles of the train and the High Girders themselves were recovered and examined, and it was from examination of this wreckage that the derailment theory was discarded due to contradictory evidence - no damage to the permanent way, and no damage to the interior of the girders that could be attributed to the train prior to the collapse.
The inquiry also commissioned a photographer, Mr Valentine (whose firm later became a major postcard manufacturer) to photograph the wreckage on each of the piers with meticulous detail, taking pictures from each cardinal direction. This wreckage included not only the remaining uprights of the tower bases, but vast amounts of failed tie-bars, bolts and lugs that fell from the upper parts of the bridge during the disaster: many of these, together with other samples from the columns and girders, were sent for structural testing to determine their quality and strength. Indeed, its only thanks to the efforts that the 1880 inquiry took to gather evidence that we have the ability in the present day to debate and discuss the collapse mechanism that bought down the Tay Bridge.
It seems odd to me that the OU would involve themselves in a theory which is, as you point out, so much in conflict with the evidence.
Agreed with guy below, the wind was to much for the girders, as they didn't have very strong metal back then. One of the girders fell out the bottom a few days before, and the the wind came the middle of the bridge collapsed
Interesting theory. I would be interested to know which span the toppled girder became. You seem to be suggesting that it was the first one, accounting for the phrase 'a slight change in the direction of the rails as you passed from the low girders into this high girder'. Do you have evidence for the kink occurring there and the toppled girder being the first one in the Thomas Bouch bridge? Now without that evidence you don't have a sound basis for a theory that the second class carriage derailed at this point, nor is there a sound theory for the carriage striking the triangular gusset plate at the junction between spans four and five because from the evidence of the photograph the plates occur at more frequent intervals and if striking a gusset plate was the outcome of a derailment at the entrance to the high girders, the carriage would already have smashed at the first triangular plate, at span two; and the train would not have reached span five. For your theory to work (and the combination of carriage derailment and the carriage striking a triangular gusset plate is attractive), you have to demonstrate either, that the kink was indeed where you suggest it was, and that the train was able to safely negotiate three sets of triangular plates before coming to grief on the fourth; or that the kink was between spans three and four, where the carriage derailed, leaving the fatal plate where you suggest it was, between four and five. You also have to demonstrate that the striking of a triangular plate at either of these positions would result in catastrophic failure of the high girders. I am not an engineer, but my instinct tells me that the deficiencies in the design of the bridge, combined with the weight of the train and the ferocity of the storm, all acted in unison. I think there is something alarming about the combination of the low and high girders, the phrase 'overturning effect' comes to mind. On that stormy evening I can picture the train approaching the halfway point of the high girders, the point at which the maximum weakness in a structural system may be most likely to occur, with a sudden, unimaginably powerful blast of wind, creating the 'lift' effect you mentioned; perhaps a sharp application of the brakes with accompanying sparks. I picture the first column to fail was five, just ahead of the train, then four, at the junction between the fourth and fifth spans, sandwiching the last two vehicles of the train as it was passing, followed by three, then six, two and seven, one and then, domino fashion, eight to twelve, all within the space of about six seconds. Terrible
I'm a bit puzzled by this, the derailment theory was investigated at the original inquiry after the collapse, and was agreed to be unlikely. It WAS agreed that the bridge was badly built, with a lot of defects in the cast iron columns. It was also established that there had been a wind in excess of that allowed for in the bridge design. Given that, it seems a bit unnecessary to look to any other cause.
Bouch supposedly always leaned on the derailment theory. But no scoring marks from derailed wheels on the ties/sleepers but then American and British RRs were built differently then. Wood is expensive in Europe but iron rails were "right there". So British RR's had fewer ties and heavier rails.
The story here about the sunken girder's being reused is a first for me; I've never read that before! If the girder passed inspection upon being raised yes they'd use it again! But again, as an American I am suspicious of rivets--particularly British ones. British engineers' structures always have the look that maybe the designers haven't really thought through tension vs. compression and maybe muddled the two hence the British structural failures. The American practice of pin-connected structural joints practically requires that tension vs. compression members look VERY different from each other! Anyways, I will be hearing this one once I get my sound back. But is this the same theory that a change of attack where the low griders met the high girders, meaning they didn't meet straight on but there was a slight turn that Bouch made no attempt to handle and that that is where the collapse began.
Another bit a reading and watching back when I had sound a few months ago was that the towers/columns fell apart under the train and caused the now-unsupported spans to collapse, rather than the bridge's being blown over. The high girders fell right next to the piers rather than further out where intact towers would have taken them.
See? 143 years later--we're still discussing Bouch's Tay Bridge and its collapse!.
It was very spindly looking..
EVERY iron (and later, steel) structure looked spindly like that. You dont' overengineer because the client and his users can't afford it. Read the book "Structures, or Why Things Don't Fall Down", which treats of Compression and Tension and stress and strain. Compression structures have a way of looking very solid--yet it IS possible to design an arch that WILL fall down--as the author remarks, a Welshman named Jones did that at Pontypridd in 1751. Something called a "line of thrust" has to stay within the arch; if it doesn't the arch WILL develop hinge points and fold like a flexible ruler. I cannot too highly recommend "Structures".
A writer called William. T. Linskill also missed that ill fated train.
Interestingly, Oliver Wendell Holmes, Sr. of "Old Ironsides" and "The Autocrat of the Breakfast Table" fame--and father of the then 12-year-old future SCOTUS Justice ("O W Holmes, Jr.") also missed a New York to Boston train on May 6, 1853. Two hours later that train went through the open South Norwalk CT drawbridge killing 46. Supposedly many Bostn papers wrotes flattering obituaries--then his telegraph saying he had missed the train came through. History might have been different: Holmes Jr. might not have gone into Law; The Autocrat of the Breakfast Table would not be written for five more years.
I have always been interested in this disaster myself. I think it is because it is the only train disaster in this country where absolutely no one survived.
I was born and currently live in Dundee. I commute to Edinburgh, Glasgow and London alot and have only crossed this bridge once. I felt so uneasy and had a panic attack going across it as it was dark. Logically...I know it won't fall....but I stick to the mega bus....
The old stumps are sad to look at but a haunting reminder.
@@srehtaef1824I don’t blame you. I find images of the wrecked bridge a bit haunting if you look at them for too long.
THe bridge girders and tubes were flawed...this was proven at the time...the design was also flawed, as it didn't initially take account of either the softness of the bottom of the river or the high windspeeds in the area ... and safety was compromised in the interests of expediting the build...
I love how people like Bill Dow can spend decades pouring over "hundreds" of pages, when as a college student I have to research thousands of pages in a couple of weeks. And who pays him for this?! How do I get this gig to read a page every few weeks?
+J.D. Schriber You poor thing. You have it so tough. My suggestion is you just give up altogether.
I don't think anybody wants you getting their thousands of pages all wet with your pouring, thank you very much.
I'm here for homework I wonder how old I will be when I see this again
Depends how interested you get. Here I came upon my own comments of several years ago.
Yes, welcome. This is indeed RUclips.
bloody RUclips
Attempting to use the Bernoulli theory of lift for the derailment is rubbish science. In fact, the use of a wing to generate lift is more to do with the air being forced down by the bottom of the wing, and the wing moving upwards in an opposite direction because of this.
That's a never ending debate.