HYDRAULIC MODELS 675
ment of water based systems (Abbott, 1991). Although hydraulics is a
central component of hydroinformatics it itself is being influenced by this
new area (Abbott, Babovic and Cunge, 2000).
In a further development artificial neural networksattempt to simu-
late the working of a human brain by passing information from one
‘neuron’ to all others connected with it; although a large set of data is
required to ‘train’ the network, the method has been applied successfully
to a number of flow situations.
The use of experiments in the solution of hydraulic problemscan be
traced back over many centuries, but it was not until the second half of the
19th century that the idea of using scale models to solve engineering prob-
lems was evolved and gradually put on a sound basis. In 1869 W. Froude
constructed the first water basin for model testing of ships, and in 1885
O. Reynolds designed a tidal model of the Upper Mersey. The turn of the
century saw the establishment of two pioneering river and hydraulic struc-
tures laboratories by Hubert Engels in Dresden (1898) and Theodor
Rehbock in Karlsruhe (1901). These were followed by many new laborato-
ries all over the world, with the major expansion occurring during the first
half of the 20th century.
The increasing use of mathematical techniques and computers since
about 1960 may have led to a reorientation, but not necessarily diminished
use, of the hydraulic laboratory in solving hydraulic engineering problems.
The reasons for this are manifold: the continuously increasing size and
complexity of some schemes requires new, untested designs which defy
mathematical solution; hydraulic scale models are increasingly being used
as aids in the solution of environmental problems by studies both of basic
physical phenomena of, for example, sediment and pollutant transport and
of design applications; the development of the theory of similarity led to
the realization of the inevitability of scale effects in the use of scale models
which, in turn, became the impetus both of wider use of field studies and
of specially designed laboratory investigations; mathematical models
require data which are often derived from physical models – leading also
to an increasing use of hybrid modelscombining the advantages of both
modelling techniques.
In this text only a very brief overview of hydraulic models can be
given; for further details see references quoted in the following sections.
16.1.2 Mathematical, numerical and computational modelsAs stated above the basis of numerical and computational models is a
mathematical model, i.e. a set of algebraic and differential equations rep-
resenting the flow. These are e.g. continuity and Navier–Stokes equations
for turbulent flow (including terms for turbulent stresses) and/or shallow