B2.18

Methods of Mathematical Physics
Part II, 2003

Let y^(p)\widehat{y}(p) be the Laplace transform of y(t)y(t), where y(t)y(t) satisfies

y(t)=y(πt)y^{\prime}(t)=y(\pi-t)

and

y(0)=1;y(π)=k;y(t)=0 for t<0 and for t>πy(0)=1 ; \quad y(\pi)=k ; \quad y(t)=0 \text { for } t<0 \text { and for } t>\pi

Show that

py^(p)+keπp1=eπpy^(p)p \widehat{y}(p)+k e^{-\pi p}-1=e^{-\pi p} \widehat{y}(-p)

and hence deduce that

y^(p)=(k+p)(1+pk)eπp1+p2\widehat{y}(p)=\frac{(k+p)-(1+p k) e^{-\pi p}}{1+p^{2}}

Use the inversion formula for Laplace transforms to find y(t)y(t) for t>πt>\pi and deduce that a solution of the above boundary value problem exists only if k=1k=-1. Hence find y(t)y(t) for 0tπ0 \leqslant t \leqslant \pi.