Beruflich Dokumente
Kultur Dokumente
4-1
becomes
4-2
we have
, hence
4-3
In order to reach the expression of the integration constant C3 we can follows two ways.
a)
The first one consists in imposing the equilibrium of the beam in the vertical direction
b) Another way is to take into account the symmetric distribution of the internal shear force T
about the point where the external load P is applied, i.e. for
Considering that
4-4
leads to
where
4-5
Substituting C3 into the expression of the elastic curve we reach the following equation which
holds for positive values of x
where
4-6
Also in this case two ways can be followed to reach the expression of the integration constant C3
4-7
a)
The first one consists in imposing the rotational equilibrium of the beam
b) Another way is to take into account the symmetric distribution of the internal moment M
about the point where the external load P is applied, i.e. for
Introducing this condition into the equation
expressing the bending moment
we have
To apply this equation it is necessary to express the second derivative of the governing equation.
4-8
leads to
where
4-9
Substituting C3 into the expression of the elastic curve we reach the following equation that
holds for positive values of x
where
4-10
In the case of a large distance L between the applied loads and the edges of the strip footing
(i.e.
) the problem can be analysed adopting the solutions for foundation of unlimited
length.
This condition, however, seldom occurs in practice.
In the remaining cases it is necessary to work out the solution for the beam of limited length.
To this purpose, the superposition of suitable loading conditions on the beam of unlimited length
is adopted.
4-11
and let apply the same loads to a beam of unlimited length. This beam is also subjected to
external loads and moments (PA, WA, PB, WB) applied to points A and B that correspond to the
edges of the actual foundation of limited length.
In order to evaluate the loads applied to the end points of the unlimited beam it is sufficient to
enforce that at these points the internal shear force and bending moment vanish.
This leads also to the solution of the actual beam. In fact, in the portion AB, the same differential
equation governs both beams. In addition, the same boundary conditions apply. Hence, the
solutions of both problems coincide.
4-12
The following equations enforce that the internal forces (TA, MA, etc.) at point A and B of the
unlimited beam vanish.
,
, etc. represent the internal forces in the unlimited beam due to the applied loads.
(
) represents the internal shear force on the right of point A of the unlimited beam due
to a unit load applied load in A; etc.
The solution of the above system leads to the values of PA, MA, etc.
4-13
4-14
In the present context, the total potential energy has three contributions: elastic strain energy
of the beam B; elastic strain energy of Winkler soil W; work potential of the applied loads
WP. Note that the external loads do not vary with the deflection of the beam.
Rayleigh Ritz method is adopted to express the potential energy as a function of the admissible
displacement fields. In particular, these fields are expressed through series of functions that
fulfil the displacement boundary conditions and the strain compatibility relationships.
These functions depend on unknown parameters which will be evaluated by minimizing .
For sake of simplicity let divide the applied loads into their symmetric and non-symmetric parts.
4-15
. This definition of
implies
leads to
Considering that
4-17
. This definition of
4-18
leads to
4-19
where if i is even
, while if i is odd
4-20
where q is the distributed applied load and p is the reaction of Winkler soil.
It is necessary to express the derivatives of ( ) in terms of the nodal displacements.
First derivative:
Forward derivative:
; Backward derivative:
Central derivative:
4-21
Second derivative:
Finite difference approximation:
Third derivative:
Finite difference approximation:
Forth derivative:
4-22
Substitution of the expression of the forth derivative into the equation of the elastic curve
leads to
The solving system of linear equation is obtained by writing the equation of the elastic curve for
all nodes of the beam.
4-23
A drawback of the finite difference method concerns the imposition of the boundary conditions.
In the present case they correspond to vanishing internal shear force and moment at the ends of
the beam.
Consider the end node i of the beam,
Before solving the system it is necessary to eliminate the displacements of nodes i+1 and i+2.
To this purpose let impose the boundary conditions:
and
4-24
It should be observed that in the case of strip footings on Winkler soil the finite difference
solution does not present marked computational differences with respect to that based on the
finite element method.
In both cases, in fact, the behaviour of the beam is represented by a banded stiffness matrix K.
In the finite element context, the coefficients representing the stiffness of Winkler springs are
added to the terms of the main diagonal of the stiffness matrix of the structure.
This introduces constraints that eliminate the rigid movements of the beam (its stiffness matrix
would be singular otherwise) and allows for the solution of the system of linear equations.
4-25
4-26
The pressure exerted by the beam on the half space is seen as a series of distributed loads on
known areas
(side view)
(plan view)
The settlement of point j is expressed as
4-27
and settlement
where is the matrix of coefficients, deriving from Boussinesq solution, that represents the
deformability of the half space; is the nodal force vector corresponding to the pressures and
vector collects the nodal displacements of the foundation.
Note that matrix
is fully populated and, in general, non-symmetric. Note also that the same
relationship can be worked out for any structure resting on elastic half space.
fk
. . . . . . . ..
fpi
yj
4-28
Let
be the diagonal matrix which contains only the terms of the main diagonal of matrix
and that coincides with the flexibility matrix of Winkler soil.
];
acting on it is,
4-29
4-30
is evaluated,
and
(this corresponds to the analysis
exchanged between structure and soil.
4-31
c) The structural problem is solved again, without introducing the half-space stiffness matrix,
imposing to the structure the displacement of the foundations as nodal constraints,
are re-calculated.
Steps (b) and (c) are repeated until no appreciable changes of vectors
and
occur.
This iterative procedure could not converge when the global stiffness of the foundations is
markedly different from that of the structure (e.g. a very stiff structure on a very deformable
soil). In this case, a suitable provision could consist in averaging at step (c) the settlements
obtained from two subsequent iterations
(
where
4-32
4-33
Assuming
4-34
Substitution of the stress strain relationship into the equilibrium equation leads to
4-35
4-36
4-37
Mat or slab or raft foundations: these are continuous footings having a slab like shape. They
can have depressions or openings. In most cases they are chosen to reduce the differential
settlements. They could be economically convenient with respect to spread and strip footings
when the total area of the isolated foundations exceeds about 50% of the total area of the
building.
4-38
Cellular mats can be used also to host facilities, e.g. garages or technical plants.
4-39
4-40
Log scale
= unloading
. ..
(2)
(1)
(3)
4-41