I need to express the solution of this initial value problem about vibration below using Convolution Integral;
$$my''+cy'+ky=f(t) \quad y(0)=0,\quad y'(0)=0$$
But don't have any idea where do i use the Convolution Integral. So how do I do it?
I tried to take laplace transform of both sides. $$ (ms^2+cs+k)Y(s)=L(f(t))\quad (assuming \quad L(y(t))=Y(s)) $$ I presumed; $$ f(t)=\int_0^t g(t-T)h(T) \,dT \quad L(g(t))=G(s) \quad L(h(t))=H(s) $$ So the Convolution theorem gives me the laplace transform of the right side; $$ L(f(t))=G(s)H(s) $$ and putting it into the equation; $$ (ms^2+cs+k)Y(s)=G(s)H(s) $$ $$ Y(s)=\frac{G(s)H(s)}{ms^2+cs+k} $$ $$ y(t)=L^{-1}(\frac{G(s)H(s)}{ms^2+cs+k}) $$ I don't know if this solution is enough or correct.
Method by Laplace transform (zero initial conditions):
$$my''+cy'+ky=f(t) \implies [ms^2+cs+k]y(s)=f(s)$$
$$y(s)=\frac{1}{m}\frac{1}{s^2+\frac{c}{m}s+\frac{k}{m}}f(s)$$
The inverse Laplace transform of the last expression is given as
$$y(t)=\frac{1}{m}g(t)*f(t)=\frac{1}{m}\int_{0}^{t}g(t-u)f(t)du$$ In which $g(t)=\mathcal{L}^{-1}\left[\frac{1}{s^2+\frac{c}{m}s+\frac{k}{m}} \right]=\frac{1}{a-b}\left[\exp(at)-\exp(bt) \right]$ and $a$ and $b$ are the roots of $ s^2+\frac{c}{m}s+\frac{k}{m}$.
So we obtain:
$$y(t)=\frac{1}{m}\int_{0}^{t}g(t-u)f(u)du=\frac{1}{m(a-b)}\int_{0}^{t}\left[\exp(at-au)-\exp(bt-bu) \right]f(u)du$$
Solution in state space form: First cast the system into first-order form:
$$\dot{x}_1 = x_2$$ $$\dot{x}_2 = -x_1-2x_2+f(t)$$
Note the system matrix is given by: $$\boldsymbol{A}=\begin{bmatrix}0 & 1 \\ -1 &-2\end{bmatrix}$$
The input matrix/vector is given by: $$\boldsymbol{B} =\begin{bmatrix}0\\1\end{bmatrix}$$
Using these expressions we can rewrite the system as $$\boldsymbol{\dot{x}}=\boldsymbol{A}\boldsymbol{x}+\boldsymbol{B}f(t)$$
The solution of such a system is given by the following expression $$\boldsymbol{x}(t)=\exp(\boldsymbol{A}(t-t_0))\boldsymbol{x}(t_0)+\int_{t_0}^{t}\exp(\boldsymbol{A}(t-\tau))\boldsymbol{B}f(\tau)d\tau.$$
While the integral on the right hand side is the convolution integral the expression $\exp(\boldsymbol{A}(t-t_0))$ is the matrix exponential which is defined as:
$\exp(\boldsymbol{A}(t-t_0)):=\boldsymbol{I}+\boldsymbol{A}(t-t_0)+\frac{(t-t_0)^2}{2!}\boldsymbol{A}^2+\frac{(t-t_0)^3}{3!}\boldsymbol{A}^3+\ldots .$