Cantilever stone staircases have been used in all sorts of buildings for more than 350 years. Russell Taylor cites Inigo Jones' tulip staircase in Queen's House at Greenwich (1629-35) as the first British example, 1 and William Talman's 2.16m-wide stairs at Chatsworth House in Derbyshire (circa 1699) as the widest ever.
But cantilever stairs have never been reserved for grand designs, and thousands of more modest examples of principal and service stairs can be found throughout the country. Unfortunately, they are also subject to the most modest of maintenance considerations.
On 17 March last year in Bedford Square, London's only completely intact Grade I-listed Georgian Square, part of a stone staircase suddenly collapsed after more than 150 years of use.
A tread near the bottom of an intermediate flight fell out shortly after the simple removal of a screw-fixed aluminium nosing. The tread fell on to the flight beneath, fracturing two further treads.
HANGING BY A TREAD 'Cantilever' stone staircases rarely act in a pure cantilever mode, as the roots of the treads are usually only built 75-100mm into the stairwell walls - too little to resist cantilever stresses. There are notable exceptions, such as the 225mm embedment of the best-quality stairs of Georgian architect Thomas Hopper, who was renowned for his elegant and robust staircases (for example, at Leigh Court in Bristol, Grade II*, 1814).
Cantilever stairs primarily achieve their stability by the nose of each tread resting on the tread beneath it, while being prevented from tipping backwards by torsional restraint from the wall embedment. Occasionally, stairs fly past windows, where the tread roots are held by wrought-iron frames fixed between the window reveals, substituting the torsional restraint of walls.
The landings and half-landings from which each flight springs are seldom single slabs of stone, but usually a collection of slabs with rebated joints, perched on each other and fixed into wall corners. Sometimes landings have a stone surface with a timber carriage beneath. The metal balustrade acts as a safety belt if anything goes wrong, distributing the load from one tread to those adjacent. The balustrade is therefore of paramount importance to the safety of the stairs in the limit state.
Tread cross-sections vary, from rectangular and serpentine to the weaker and more commonplace triangular shape. Regardless of section, treads that have rebated joints between them have secondary ways to acquire stability: by plate-action and/or by compression in the plane of the flight.
2BANISTER FLEXURE The heel-drop test on every tread and landing (simply dropping one's weight from tip-toe on to the heel of one's foot) is a valuable first indication of any hidden fractures or loose joints that might justify the removal of finishes for further inspection. Wobbling the handrails will show which balusters are loose in their leadcaulked pockets. In our survey of many such staircases, defects had arisen from normal wear and tear, historic bomb damage, the relaxation of the structural fabric and its supporting walls, and overloading.
The structural problems that can ultimately lead to the collapse of treads are:
? yielding of landings and/or half-landings that support the bottom tread of a flight of stairs, due to differential settlement of the walls carrying the landing, the joggling of joints between multi-piece stone landings and fracturing of single slabs, loss of kentledge thickness of stone landings where new doorways have been formed, and possible creep deflection of the landing (in the case of timber construction);
? loosening of the lime-putty joints between the treads (indicated by cracked soffits and/or the distinctive thud arising from heel-dropping), allowing the treads to hammer against each other under a live load;
? worn or damaged tread-nosings that have been repaired by letting in a new veneer of stone, thus reducing the structural section of the tread and also acting as a stress-raiser/crack-inducer.
Similarly, the modest screw fixings of remedial metal stair nosings can (surprisingly) initiate tread fractures, as at Bedford Square;
? the spalled back edge of triangular-section treads, due to excessive point loads from lead or slate shims in the rebated joint of the tread above;
? individual treads cracked across their section, which are then solely reliant on the balustrade to prevent them falling, as at Bedford Square; and - various balustrade deficiencies that allow cracked treads to fall, such as cracked baluster pockets, loose lead caulking between the tread and baluster, missing balusters, a loose fixing between the baluster and top rail, loose fixing between the top rail and handrail, 'S' scroll balusters (see figure 7 ), and loose W1 window frames (where stairs fly past windows).
STAIR NOSING Fractured treads can stay in place for years, providing they are prevented from twisting and dropping out by the balusters. Some types of baluster fixings are stronger than others, but should they fail, then the very heavy tread can fall, with potentially disastrous consequences. It is therefore essential that fractured treads are repaired, balustrades are maintained and other defects are remedied to maintain the customary margins of structural safety.
If a landing is fractured or sagging, the kindest repair is the insertion of a small box-section steel beam underneath the leading edge, with intumescent paint for fire protection. But beware of the possibility of flues in the landing walls where such beams will need to bear. And beware of conservation contraints.
Treads that are fractured can be repaired by inserting a post-tensioned bolt, so long as the end of the tread is not obstructed by balustrading (see detail above). Loose tread-joint mortar should be sawn out and repacked tightly, as long as the joint is first temporarily wedged to prevent loosening of the tread joints further up the flight. Balustrade deficiencies can be left to a good blacksmith, but 'plastic' repairs of damaged stonework need the specialist attention of conservators (see figure 7).
Occasionally, stairs are so derelict that they cannot be repaired safely without temporary support and so special scaffold 'trees' should be erected up the stairwell.
However, before we all panic about an epidemic of collapsing stairs, performance in use is the ultimate test of any structure, and in this respect cantilever stairs pass with flying colours. They only seem to fail if maintenance and repairs are neglected, leading to the fracturing of treads and ultimately their collapse.
Clive Richardson is a structural engineer and technical director of Cameron Taylor, engineer to the dean and chapter of Westminster Abbey, technical secretary of the Engineers'Conservation Accreditation Register for Engineers and author of many technical books, including the AJ Guide To Structural Surveys. Email: clive. richardson@camerontaylor. co. uk