Stone 221
to be at different angles and so runoff velocity will vary. Even for a strong acid,
if there is little time to react, then little limestone will be dissolved. A surface
cannot be looked at in isolation. One surface will be connected to other sur-
faces also producing runoff and this will lead to flow between surfaces. The
acidity of solutions will change as reaction with the limestone occurs.
Likewise, acidity will change as solutions could become diluted by additional
runoff volumes. Runoff patterns across a building will therefore influence the
rate of dissolution on individual surfaces.
For sheltered surfaces, reactions with pollutants in the atmosphere can
only occur if there are no degradation products impeding the flow of pollu-
tants to the limestone surface. If a sulfation crust does form then direct reac-
tion between the atmosphere and limestone will depend upon the rate at which
gases can pass through the crust. Some researchers suggest that alteration of
limestone to gypsum occurs beneath an existing crust at the limestone/crust
boundary. This means that the diffusion coefficients for calcium, sulfate and
carbonate ions become important in determining the rates at which crusts can
form. Rates of alteration will, however, be affected by the thickness of the
existing crust and its permeability, illustrating a negative feedback within
the degradation process dependent upon form and so thus indirectly upon the
past interaction of material, process and environment.
The development of distinct patterns of degradation in space and time
means that there is some predictability about how degradational forms will
change. From the same initial starting point, variations in the rate of degrad-
ation, in the nature of degradation agents operating on the surface and even in
the timing of the actions of these agents, could be used to follow different
developmental pathways. The specific pathway followed by a specific surface
is almostimpossible to predict, particularly when the feedback between one
period of development and the next are considered. For a non-calcareous
sandstone, some possible developmental pathways are illustrated in Figure 5.
Some modelling of limestone degradation has also been undertaken using
the idea of a simple environment-to-weathering relationship. The production
of damage functions for limestone relates the ambient concentration of spe-
cific atmospheric pollutants to indices of stone damage, such as loss of mass and
surface loss. The damage function describes the various relationships between
atmospheric pollution and stone damage in Figure 6. The manner in which the
stone alters can be linear or non-linear, but the key concept in the figure is that
of a threshold above which critical and irreversible damage occurs. Identifying
this critical threshold is vital for developing any pollution policy to protect
buildings.
It could be argued, however, that the concept of a damage function is prob-
lematic for stone. First, all change is irreversible in the sense that stone lost
cannot be regained if atmospheric pollution is reduced. The relationship only