Section BB.3.2 of BS EN 1993-1-1. Eurocode expressions such as BB.14 and
BB.16 (and their equivalents in BS 5950) can be immediately recognised in the
The reference to “Point A”, still used today, as the all-important junction
between the bottom flange of the haunch and the inside face of the column is
found in this paper, as illustrated in Figure 2 (Fig 1 from Horne et al)
The steel skeleton
The comment by Professor Bryan referring to Sir John Baker moves us back
another step to the early 1950s and the two volumes of The Steel Skeleton6,7.
Volume 1 covers “Elastic Behaviour and Design”. Volume 2 covers “Plastic
Behaviour and design” and perhaps it is no surprise that a co-author of
volume 2 was Professor Horne – described as “one of the leading protagonists
of plastic design”.
Both volumes are worth reading, containing some really interesting history.
Volume 1 looks back further to 1929 when an investigation was undertaken to
investigate the application of modern theory to the design of steel structures.
This review considered practice in the UK, New York, Germany, France, Spain,
Prussia and Belgium. The live loads to be designed for in different countries
varied, as they did in various UK cities. Edinburgh and Glasgow agreed that
halls, schools and churches must be designed for 180lb per sq.ft, (8.6 kN/m2)
whereas Newcastle was content with 112 lb per sq.ft (5.4 kN/m2). Perhaps
children and worshippers were not so socially distanced north of the border.
Volume 1 also records the live load reductions that were allowed in various
countries, linking to the reduction factors we find in modern codes.
The significance of the loaded area reduction is also evident in the
intensities of loading surveyed. Who would want to work in a (presumably
claustrophobic) small finance company, when the loading was measured at
11 kN/m2, compared to a structural engineering company at 4.5 kN/m2?
However, when measured over a larger area, the situation is more
comfortable; the loadings become 0.9 and 0.6 kN/m2 respectively. Over a
larger measured area, “consulting engineers” (as opposed to “structural
engineers”) fare the worst at 1.7 kN/m2. The survey also noted that
concentrated loads, such as fire cabinets and safes needed special attention –
another principle found in our modern loading codes.
Of equal interest in Volume 1 are the tests undertaken on real buildings. For
one building, a hotel under construction by Dorman Long, it was suggested
that a few platoons of soldiers from a nearby barracks could have been used
to provide a well-distributed load. In the event, point loads were suspended
from the beams. Strain gauge readings were affected by the riveting
operations - a hazard not experienced today. A summary of the findings is
that the measured effects in beams and stanchions were not as expected.
Professor Baker described the behaviour as “radically different from that
assumed in the design methods in common use”. One key difference was that
the riveted connections were relatively stiff, making the frame behaviour
more like a rigidly jointed frame than a pin-ended arrangement. It is worth
remembering that designers are modelling loading and modelling the
structure and its response.
The Steel Structures Research Committee produced “Recommendations for
Design” based on these studies, which was essentially a semi-continuous
design method, recognising the stiffness of the connection types used at the
time. Unsurprisingly, the connections had to be classified (based on the
detailing) – which was associated with the connection stiffness. We might
reflect on the current guidance in BS 5950 that the detailing of the
connections must be consistent with the assumptions made in the frame
analysis, and the explicit requirement in the Eurocode to classify connections
and allow for connection stiffness if the effects are significant. This is the same
principle advocated in the proposed design method some 90 years ago.
The proposed design methods, published in 1936, were complicated. Baker
comments that despite the constructional steel industry paying for the
research for a period of 7 years “neither the industry itself nor consulting
engineers generally felt any enthusiasm for the outcome of their labours”.
Baker noted that “whatever criticism could be levelled at the method of
design which had held the field for nearly fifty years, it certainly had the merit
of simplicity; in fact it would be difficult to imagine anything simpler”. Baker
also noted that the recommended procedures were laborious, and there was
“no advantage that the average client would appreciate”. The orthodox
method of design was shown to be safe, if quite conservative in some cases.
Once war broke out, steel was a very important commodity. A wartime
amendment was made to BS 449 which increased the permissible stresses by
25%. Baker notes that “this earned the taunt that the engineer had discovered
that steel was stronger in war than in peace, whereas all he was admitting was
Figure 2: “Point A” – where inner flanges meet (Fig 1 from Ref 5)