About Us - intro background & philosophy

Arches into Archie

Bill Harvey always wanted to be a Bridge engineer. He began his real career on Humber Bridge perhaps the opposite extreme from the arches which have been the core of his life 30 years now. In 1977 he became a lecturer at the University of Dundee and began thinking about an area of research. The department was already active in Bridge Engineering and his mentors introduced him to the idea that Masonry bridges had not been brought into the modern world. The real trigger, though, was being asked by the Transport and Road Research Laboratory to devise and carry through a monitoring exercise on General Wade’s Bridge in Aberfeldy. A new barytes mine was to be opened north of the bridge and all the material would be carried over it. The barytes is a very heavy mineral (4.7 tonnes per m^3 in loose piles) and it seemed likely that the vehicles would be very heavy and probably overloaded. The bridge had been built in 1733 and was approaching its quarter millennium. As part of his first report bill sought to carry out an assessment of the structure using established tools, which in those days meant MEXE. The result was a much lower projected capacity than the then owners of the bridge had suggested. It became clear that there was room for huge variation in results from MEXE.

About that time, Jacques Heyman published a paper describing his assessment of Clare College Bridge in Cambridge. He had discounted MEXE and turned instead to older routines bolstered by the application of the plastic theorems. The hand calculation required for a thrust line analysis was tedious and complex but a pocket calculators were becoming ever more powerful and in 1982 bill produced what was the first version of Archie as a program for the HP 41 CV pocket computer. The output from this was graphical in the sense that it was a series of offsets of the line of thrust from the intercourse and these could be used to plot a thrust line. More importantly, the numbers themselves provided a sufficient picture to allow interaction between the user and the program. The user had to nominate him to positions and the offset values told him whether they were correct.

In 1983 Bill acquired the first personal computer in the University of Dundee an NCR DecisionMate V. It came with a colour screen boasting 640 x 480 dots and 16 colours, 10 MB hard disk and 640 K of memory. Compared to the little HP machine it was huge both physically and in the power it had available. Over a period of two weeks the first real version of Archie was built. It dealt only with circular arcs and a flat road surface. For numbers were required as input, span, rise, ring thickness and fill depth. The computer then drew an outline of the bridge on the screen. The load was positioned and the user nominated location of four hinges. Then calculation started and 25 seconds later a new picture appeared with a thrust line superimposed on the bridge outline. Computing and Archie have come a long way.

Exploration and interaction

In the early days it was clear that the fastest analysis could be achieved by appropriate interaction between the user and the computer. We chose to ask the user to nominate hinge to positions and get a result after 25 seconds, rather than allow the computer to iterate for five or 10 minutes. Interaction became and has remained at the core of everything we do. With every year that passes the power of computers increases. We believe that should be used to increase the speed of reaction rather than to remove operator interaction.
From the start we realised that truly analysing these Masonry bridges is not possible. There are so many unknowns and intangibles. Codes of practice don't help much because they capture belief systems rather than true understanding. What the engineer must do in assessing an arch bridge is to achieve a level of confidence that the structure is sound. (Ted Happold used to say that structural design was about achieving the confidence to build). That confidence is best developed by exploring possibilities.

Visualisation

We developed the concept of the zone of thrust in the mid-1980s and immediately incorporated it into Archie. The zone of thrust is best understood as the minimum bridge that can carry a particular set of loads (picture). Provided the minimum bridge is encapsulated within the real structure the bridge is sound. The speed of modern computers means that the zone of thrust can be presented on screen at the same speed as the user can drag a load across the model using the mouse.

Confidence

Confidence in the structure depends on understanding the potential behaviour, understanding the limits of your knowledge, and demonstrating that there is a viable path for all loads to reach the ground. This means that we should be able to model everything from the point of application of the live load to the point of reaction of the foundations. Unfortunately, our current state of knowledge means that most attempts to model foundation behaviour shows them to be inadequate. From the earliest times of arch bridge assessment this has been recognised and in all codes the user is advised to inspect abutments and assume they are sound unless there is real evidence to the contrary. In recent years Bill Harvey has developed an increasing understanding of the potential thrust distribution within large structures and the implications for behaviour. As we approach the second decade of the third millennium, and our fourth decade in business, we are working towards a new version of Archie-M which will make use of this new understanding.