surmounted by the observatory proper. First, the ground sill is a
square of 20 ft., made of yellow pine sticks mortised together and
pinned with stout trunnels. The sill of the observatory is made
likewise of heavy timbers, 12 ft. long. The two sills are joined
together by four stout yellow pine corner posts, which in turn are
mortised into both sills. The posts are 26 ft. in length. Five feet
above the lower sill is the sill which supports the floor of the first
room. Ten feet above this is the sill which supports the upper
room. Both these sills again are mortised into the corner posts.
The structure is sheathed outside with German siding, and inside
with rough boards covered with felt, and again by tongued and
grooved yellow pine boards. The observatory proper, octagonal
in shape, is securely mortised into the top sill and covered with a
corrugated iron roof conical in shape. The cellar is floored with
3 in. wood, and boarded all round on the inside of the posts. A pit
was first dug in the sand about 6 ft. deep and fully 20 ft. wide on
the bottom. The cellar sill was laid on this bottom, and the structure
built upon it; thus the whole depth of cellar is sunk below the top
of the hill or the level of the sand. The cellar was then filled up
with sand and packed solid all round, consequently the building is
anchored in its place by the load in the cellar, about 100 tons in
weight.
 |
| Fig. 13.—Cantilever Foundation over Railway Tunnel. |
The subject of foundations, being naturally of the first importance, is one that calls for most careful study. It is not of so much importance that the ground be hard or even rocky as that it be compact and of similar consistency throughout. It is not always that a site answers to this description, and the problem of what will be the best form of foundation to use in placing a building, more especially if that building be of large dimensions and consequently great weight, on a site of soft yielding soil, is one that is often most difficult of solution. The foregoing article indicates in a brief manner some of the obstacles the architect or engineer is required to surmount before his work can even be started on its way to completion.
Authorities.—The principal books for reference on this subject are: A Practical Treatise on Foundations, by W. M. Patron, C. E.; Building Construction and Superintendence, part i., by F. E. Kidder; Notes on Building Construction, vols. i. ii. and iii.; Aide Memoir, vol. ii., by Colonel Seddon, R.E.; Advanced Building Construction, by C. F. Mitchell; Modern House Construction, by G. L. Sutcliffe; Building Construction, by Professor Henry Adams; Practical Building Construction, by J. P. Allen. (J. Bt.)
FOUNDING (from Lat. fundere, to pour), the process of casting
in metal, of making a reproduction of a given object by running
molten metal into a mould taken in sand, loam or plaster from
that object. To enable the founder to prepare a mould for the
casting, he must receive a pattern similar to the casting required.
Some few exceptions occur, to be noted presently, but the above
statement is true of perhaps 98% of all castings produced. The
construction of such patterns gives employment to a large
number of highly skilled men, who can only acquire the necessary
knowledge through an apprenticeship lasting from five to seven
years. A knowledge of two trades at least is involved in the
work of pattern construction—that of the craft itself and that
of the moulder and founder. Patterns have to be constructed
strongly. They are generally of wood, and they thus require
skill in the use of woodworking tools and the making of timber
joints, together with a knowledge of the behaviour of timber,
&c. Some few patterns are made in iron, brass or white metal
alloys. They have to be embedded in a matrix of sand by the
founder, and being enclosed, they have to be withdrawn without
inflicting any damage in the way of fracture in the sand. Since
cast work involves shapes that are often very intricate, including
projections and hollow spaces of all forms, it is obvious that the
withdrawal of the patterns without entailing tearing up and
fracture of the sand must involve many difficult problems that
have to be as fully understood by the pattern-maker as by the
moulder. It is from this point of view that the work of the pattern-maker
should be approached in the first place. No closed mould
can possibly be made without one or more joints, for if a pattern
is wholly enclosed in a matrix of sand it cannot be withdrawn
except by making a parting in the sand, and it is not difficult to
conceive that the parting in the pattern might advantageously
be made to coincide, either exactly or approximately, with that
of the mould. Nor must obstacles exist to the free withdrawal
of patterns. They must therefore not be wider or larger in the
lower than in the upper parts; actually they are made a trifle
smaller or “tapered.” Nor may they have any lateral extensions
into the lower sand, unless these can be made to withdraw
separately from the main portion of the pattern. Finally, there
are many internal spaces which cannot be formed by a pattern
directly in the sand, but provision for which must be made by
some means extraneous to the pattern, as by cores.
 | |
| Fig. 1. | Fig. 2. |
A single example must illustrate the main principles which have just been stated. The object selected is a bracket which involves questions of joints, of cores, of pattern construction and of moulding. The casting, the pattern, and its mould are illustrated. Fig. 1 illustrates in plan the casting of a double bracket, the end elevation of which is seen in fig. 2; the pattern of which presents obvious difficulties in the way of withdrawal from a mould, supposing it were made just like its casting. But if it be made as in fig. 3, with the open spaces A, B, in fig. 2, occupied with core prints, and the pieces A, A in fig. 3 left loosely skewered on, everything will “deliver” freely. Moreover the pattern might be made solidly as shown in fig. 3, or else jointed and dowelled in the plane a–a, as in fig. 4, or along the upper faces of the prints b–b, fig. 3. The
