|
Tom
Martin gives his opinion on why the Tay Bridge collapsed.
He
studied mathematics and worked as a mathematician in the British
Steel Corporation in Scotland before moving into the IT sector.
He has worked closely with Professor Ian MacLeod of Strathclyde
University, Glasgow, to investigate the structure of the Tay
Bridge.
How
did you come to investigate the Tay Bridge disaster?
"I can remember, whilst at secondary school, reading
a book entitled The High Girders by John Prebble. He made
the whole thing very interesting and I wondered if good construction
could have saved the bridge. Later on, when I was working
as a mathematician at the British Steel Corporation, one of
the metallurgists there handed me a copy of the Court of Inquiry
Report on the Tay Bridge disaster. He had been interested
in the performance of the materials used in the construction
of the bridge, especially the cast iron. I read the report
a couple of times and wondered if modern analysis techniques
could shed new light on the disaster.
 Things
lay quiet for a couple of years and then I noticed a flyer
from Strathclyde University, Glasgow (where I had studied
mathematics) advertising an evening class on computer analysis
in structural engineering. That caught my eye and I thought
that I’d maybe attend. I enrolled on the course, which was
run by Professor MacLeod. After the course was finished, I
told him I was interested in investigating the Tay Bridge
disaster.
After
that we both collaborated on the project because he also found
it very interesting. It is fourteen years since we started
looking at the disaster.
Back
to top
Theory Summary
We investigated the disaster by utilising modern computer
analysis techniques in conjunction with a modern approach
to wind loading. The bridge was examined with and without
the train on the bridge to see what effect it had on the performance
of the pier structure when subject to wind loading. A pier
was analysed under various load conditions with a view to
proposing a collapse mechanism.
We found
that the bridge was simply not strong enough to withstand
the wind on that night. The train marginally increases the
overturning effect.
Back
to top
What
do you think happened on the night of the disaster?
The collapse scenario given by our model goes along these
lines. When the train reached the high girders there was a
particularly strong gust of wind. This increased the overturning
force enough to cause the base of the windward column to lift,
in turn causing the diagonal ties to begin to fail, starting
at the second level and developing upwards. This weakens the
second level causing the failure of the bolted connections
in the column at that level. Simultaneously, the bracing failure
extends upwards. The column support on the leeward side would
then become ineffective. That side would start to drop and
the whole pier would start to rotate about the second level.
As it falls there would be a kickback on the first level causing
it to be demolished but retaining most of the first level
wreckage on top of the foundation. That would be our collapse
scenario
As all the high girders piers are of the same design only
one requires to be considered when analysing the bridge. The
presence of the train increases the lateral wind loading and
therefore the train makes a difference when it is on the bridge.
But, given the wind force (force 10/11 on the Beaufort scale)
on that evening, the lateral loading on the other piers could
have made them also.
I would put my money that the pier with the train on it was
the first one to collapse.
As far as I’m concerned the wind loading is the primary reason
for the collapse of the high girders. Once one pier goes it’s
would drag the rest with them. It works like a trigger point.
The low girders that remained were standing were much shorter,
145 feet long compared with 245 feet for the high girders.
So you’re looking at significantly lower wind loads on them.
Also, they weren’t as high, and therefore suffered less overturning
effect due to the wind. Computer analysis indicates that these
piers were safe for the wind loading that night, which matches
up to what actually happened.
Back
to top
What are the key factors that led you
to reach this theory and mechanism of collapse?
The key factor is our ability to model the various scenarios
using the three-dimensional elastic frame computer model of
the pier. Because you have a mathematical model you have flexibility
of examining various loading scenarios.
To build up the model of the pier we used a combination of
the engineering drawings and also the Court of Inquiry report
and the technical report of Henry Law which gave information
such as the weight of the girders. In addition, we used the
results of the component testing that was performed for the
Inquiry.
These test results were critical for our study as they make
such a big difference. For example, the actual bolts that
were used for holding down the columns at the base were taken
away and tested. We have real figures. I can’t over emphasise
what a difference the test results makes to the analysis.
It allows one to incorporate the bad construction inherent
in the wind bracing members - i.e. they fail at a much lower
load than they should have done. So, we really had very good
to build the model - presuppositions were kept to a minimum.
Back
to top
How
does this theory match the actual evidence on the ground?
The pier photographs show uplift on a number of piers, which
is of course critical to our theory, which predicted the uplift
as the first trigger point. By uplift I mean the windward
column lifts two courses of the stone masonry it’s attached
to, due to the massive wind loading. And the reason is Sir
Thomas Bouch had only anchored the bolts through the two top
courses of masonry. He should have anchored into the caissons.
A very thin layer of cement was all that held the courses
of stone together below the two top courses. You can see a
number of piers where this lifting has taken place. The photograph
of pier 5 shows clearly the uplift on the west side, for example.
Back
to top
Does the speed of the wind play a
factor in your theory?
The wind speed is the critical factor as it determines the
lateral loading on the bridge. The gale, which was blowing
along the Tay estuary that night, was at right angles to the
bridge, which is the worst scenario. We’ve estimated the wind
force at 10 to 11, which was recorded by the local met office,
naval officers (stationed on the Tay ) and some local meteorologists.
I don’t think there’s much doubt about the strength of the
wind. The main controversy at the time was over the quantification
of wind pressure. The science of wind effects on structures
was just in its embryonic stage of development then. You need
to get from wind velocity to pressure using a drag coefficient,
but they had no way of estimating this, especially for reticulated
girders used for the Tay Bridge. With modern wind loading
codes the drag coefficient is easily calculated, which then
allows one to calculate the wind loading on the girders. This
is the approach we used for calculating the wind loading on
the high girders and the pier structure.
The beauty of a computer model is that you can look at the
full range of wind velocities. We studied wind speeds from
44 miles per hour right up to 85 miles per hour.
Back
to top
Do you have an ethos that guided your
investigation?
A good theory is the one that can be tested. The degree to
which it can be tested will determine whether it is a weak
or strong theory. If you can’t test a theory then it would
be classed as non-scientific. The Tay Bridge pier model is
a mathematical computer model which allows one to load the
structure under different scenarios and simulate what happened
that night. The techniques used, by their nature, facilitate
a quantitative view of the collapse, which is essential to
having the theory accepted by the engineering/scientific community.
Theories, which take a qualitative approach, are by nature
more speculative. Tests were done on a limited number of wind
bracing members to ascertain their tensile strength. It could
be that that some of them on the bridge were weaker than the
ones that were tested. Our approach would only work with available
tensile test data and not use speculative data.
There could have been other factors involved that we can’t
model analytically. But regardless of any factors that may
or may not have been involved, based on the force of the wind
that evening, the bridge design and the test results for the
holding down bolts and wind bracing it was going to collapse.
The fundamental cause in our opinion is that the bridge was
blown over due to being under designed for the wind loading.
Obviously other factors, such as train derailment, could have
been involved. However, these factors would act as secondary
causes for the collapse.
Back
to top
How
do you rate the quality of the BOT inquiry, both against today’s
standards and also those of the 1880s?
I think for its time it was definitely an excellent report.
There was a lot of work put in to it. The experts called at
the Court of Inquiry were the best in their field and they
had testing carried out at Kircaldy’s labs in London.
But having said that, I think that today much more time would
be taken over recovering wreckage and subjecting it to more
rigorous testing. Instead of testing 14 wind bracing members
they’d test a huge amount. They’d go to an awful lot of trouble,
and as a consequence I think the report would take much longer
to produce.
It was such a big disaster - world famous - that the government
got involved and said ‘try and get to the bottom of this’.
I think it was a ground-breaking report. However, the court
of inquiry report stated that the bridge could have survived
if it had been properly constructed. But our modern analysis
shows that really it couldn’t have withstood the wind forces
experienced on the night of the disaster.
I think there are two dimensions to a disaster. There’s the
technical dimension, but underlying this there is the human
dimension that is of great importance.
If the building of the Tay Bridge had been a better financed,
Sir Thomas Bouch may have not compromised the strength and
stability of the redesigned piers. Due to the project being
behind schedule and over budget, his professional judgement
was probably clouded, for example he had intended to use sets
of 8 columns instead of 6 for each pier. If one looks at his
other bridges it is hard to believe it’s the same man that
designed the Tay Bridge. I think the Tay Bridge disaster has
a perennial message: professional engineers need to operate
at a high level of integrity."
For
more details of Tom Martin's theory about the Tay Bridge disaster,
visit his website: http://www.tts1.demon.co.uk/tay.html
Back
to top
|
