The Lyapunov–Schmidt reduction is a procedure used to study solutions to nonlinear equations at singular points, where the implicit function theorem does not work. In several cases of interest, this procedure permits the reduction of infinite-dimensional equations in a Banach space to a system of equations in finite dimension. Once this system of equations has been obtained, singularity theory can be used to classify singular point.
We shall work with a Hilbert space , and
We consider a Lagrangian , where is the loading parameter. We shall suppose that is Frechét differentiable as many times as is needed, and we shall use the notation
Moreover, we shall use
We introduce the operator defined by
We suppose that is a set of solutions of the equation . This implies, in particular, that
For each we consider the linear operator define the operator defined by:
We assume that there exists a critical load such that the operator
has one-dimensional kernel, and that the range of is the orthogonal complement . Hence
is one to one.
We denote by the orthogonal projection. We select one element of and we define
The map
is bounded and surjective. Hence by the implicit function theorem, we deduce that
Moreover, the function is smooth in this neighborhood and vanishes at the critical point
Thus, for in a small neighborhood of we have the identity
We define the equivalent function
In a small neighborhood of the critical point , the solutions of the equation are in one-to-one correspondence with the solutions of the equation .
The following properties are important:
Although the function is not known explicitly, it is possible to compute its derivative at the bifurcation point .
Taking the derivative of with respect to we obtain:
Note that
Moreover,
Hence, becomes:
The last equation allows us to compute from the knowledge of by solving the linear equation:
Once this solution is known, we can substitute into
which is obtained from by differentiation with respect to .
A calculation similar to the previous one allows us to compute . Indeed, we find, differentiating with respect to :
where is the solution of the following equation
which is obtained by differentiation of .
Differentiating with respect to we obtain
where is obtained by solving the equation
which follows form .
Differentiation of with respect to yields
Differentiation of with respect to yields an equation for :
Differentiating
where the unknown is determined by solving
which follows from upon differentiation with respect to .
We first show that:
In fact, we recall that and that , hence implies in particular that solves the equation , and hence . On the other hand, also cf. ). Thus, the only possible solution of the above equation is
From we obtain the first of . A similar argument, based on the observation that yields
This implies the second of .
The fact that is singular at the critical point implies that the implicit function theorem cannot be applied at that point. However, it is not a sufficient condition to guarantee that the critical point is an actual bifurcation point. A sufficient condition can be formulated, however, which goes by the name of transversality condition.
Evaluating at and making use of and and on setting
we obtain
Recalling that for a system of one degree of freedom the transversality condition is we obtain the transversality condition .
We find from :
where can be computed by solving the equation:
which follows from . From this we deduce, using the definition of , and with some calculation, the formula:
Finally,