Ian Hume discusses a powerful weapon for Conservation officers faced with buildings which, calculations show, should have fallen down long ago.
Many of us who are involved the conservation of historic buildings have long been convinced that, equal and opposite to the force of gravity is the force of habit. if it does not exist, why am I, as a structural engineer, unable to use my computer and my knowledge of the theory of structures to prove the safety and stability of many historic structures that are surviving perfectly well. The reason of course is that I am unable to analyse all the complexities of the buildings, I underestimate the permissible stresses and the materials are much stiffer than I imagine thus reducing deflections. What can I do about this?
Load testing is a powerful weapon that the Structural Engineer can use as a means of justifying the structural adequacy of historic structures. It is particularly useful when, as suggested above, structures or individual elements of structures cannot be justified by calculation when the primary evidence presented by the building itself shows that there are adequate reserves of strength.
Perhaps the best form of load test is the examination of the past history of the building or structure. The following matters should be addressed:
What loads it had to carry in the past?
What distress has been caused by these loads?
Has there indeed been any serious distress at all?
Is the continued existence of the structure a better justification for its future life than evidence produced by calculations proving that it should have collapsed a long time ago?
What loads is it intended that the building should carry in the future?
If these loads are similar or less than past loading, are any strengthening works really necessary?
Consideration of these questions together with perhaps some repairs to local areas of weakness caused by the effect of time, weather, beetle and corrosion, etc. can often be used as a long term load to test to justify confidence in the future use of a historic building.
Load tests to destruction, although capable of giving research data, do not fit in with conservation philosophy and therefore will only very rarely be of interest to conservation. The load applied during a test for serviceability, that is the way the structure behaves in service, should be representative of the real life situation of the structure. Loads applied during load tests may well be the loads likely to be carried by that structure with a 25% increase as a factor of safety.
Case studies
Four case studies of load tests in the historic buildings will serve to illustrate the value of load tests. Members of the Conservation Engineering Team of English Heritage carried out all four tests and several others of a similar nature.
A Timber framed barn in Essex
Neglect, storms and death-watch beetle had cause some decay to the barn and this had resulted in a partial collapse in the central section of the roof. A mixture of traditional carpentry repairs, steel plating and replacement timber, together with a little resin repair work had resulted in the successful conservation of the principal framework. However there was considerable doubt as to the adequacy of some of the rafters that had been much attacked by beetle and which, in some instances, had to have a new ends scarfed on. Quite simple calculations showed that in high winds or under snow loading the rafters would be over-stressed and suffer excessive deflections.
It was therefore decided to load test a number of the more borderline rafters and one of the new rafters for comparison. Clearly, some rafters were totally beyond redemption and many were largely unaffected by beetle but a large number were in the grey area between these two extremes. It was these "grey area" rafters which were the subject of the tests.
The load test was simple. Two plasterer's trestles provided end bearings and a series of two-gallon buckets of water provided the load. A scaffold tube was fixed between the trestles to carry a number of dial gauges to record deflections. Screws drilled with a small hole to accept the point of a Demec strain gauge allowed strain to be re~ corded and thus stress to be calculated. All rafters were tested to their design working loads with a 25% overload.
A number of conclusions were drawn from the tests:
In all cases the actual measured deflection was considerably less than that predicted by calculation using the actual measured sizes of the rafters ignoring parts missing due to waney edge or minor beetle decay.
The original rafters, both repaired with scarfed on ends and unrepaired alike, all behaved satisfactorily and this gave confidence to replace them and others like them in the reconstructed structure.
The rafters behaved considerably better than their apparently decayed state suggested.
Simple tests can be of great value both in saving of historic fabric and on economic grounds.
The tests have given confidence in the use of similarly decayed timbers in many other structures.
A timber staircase in a major London buildingThe lower flights of this stair are frequently used by large numbers of people attending functions on the principal floor of the building and the stairs can be crowded with people waiting to take their places at dinners. This had led to some concern over the structural adequacy of the lower flights. The staircase, although grand, is of typical timber construction and, like many timber staircases, was quite springy in use.
A standard steel tube and fitting scaffold was used with screw jacks to allow a support scaffold to be erected to within about 25 to 30 millimetres below the staircase and instruments were attached to measure strains and deflections. 56-pound steel weights were used for the load. As the building was still in use at the time of the test, the carpet was protected by double-sheeted polythene and a layer of scaffold boards below all scaffolding and underneath the steel weights. The load was applied in stages and the maximum load represented approximately 100 people standing on the staircase. This was very close to the value of four kilonewtons per square metre (80lb/sq ft) as suggested by BS 6399 as the design load. In practice it is unlikely that the actual load would approach anything like this figure.
The conclusions drawn from this test were that the staircase was safe for the number of people who used the staircase. Although the maximum recorded deflection of 22 mm was in excess of the BS 5268 permissible value of 10.8 mm it was not considered excessive as the finishes were of a flexible type and not easily damaged.
This building was built largely between 1148-1179. The intention was to cover the top of the vaulting with a 100 millimetre concrete slab and 50 millimetre thick York stone paving slabs in order to waterproof it. There was concern that this additional load would overload the masonry construction. The complexity of this type of structure does not lend itself to accurate analysis hence the load test was considered to be the most practical method of proving its adequacy to withstand the additional loads.
The proximity of the structure to a river made water the obvious medium for applying load. A contract was let for the supply and erection of a temporary dam constructed from polythene sheeting, scaffolding and timber. A safety scaffold was erected beneath the vaulting and a separate scaffold was erected to enable the structural engineers to read the strain gauges and dial gauges beneath the vaulting. Theodolite readings were also taken to check spreading of the vaulting in a horizontal direction. The vaulting was loaded, by stages, and measurements were taken as the load increased. Measured deflections were considered to be well within acceptable limits and after removal of the load, it was noted that the recovery of the deflections was in excess of 85%, and therefore only a very small residue of deflections was left. The test was regarded as a success and the proposed work was put in hand.
A stone 'cantilever' staircase.
This test was referred to in my article in Context 55 (September 1997). The opportunity was taken to load test a stone 'cantilever' or 'hanging' staircase to destruction.
As with the timber staircase discussed above, a scaffold framework was erected to within a few millimetres of the underside of the staircase and this one was also instrumented. It was loaded by stages to its working load of
2.5 kN/sqm, to its design load of 4 kN/sqm and onwards with the intention of making it collapse so that the failure mode might be better understood. At 10 kN/sqm (200 lb/sqft) there were no more weights left with which to load the stair but even at this very high loading, the structure had shown little signs of deflection. These stairs are very difficult to justify by calculation but the test showed them to be immensely strong.
Conclusions
Particular care must be taken to ensure that damage to the historic fabric is not caused either by the installation or the removal of the test equipment or by a failure during the test itself. In most other aspects the testing of an historic structure or building is similar to the testing of a more modern structure. Load tests should only be carried out under the supervision of a structural engineer.
Load tests can be a very useful way, perhaps the only way, of proving conclusively that a structure or part of a structure is adequate to carry the loading demanded of it by its future uses.