Hyperasymptotics and the Linear Boundary Layer Problem: Why Asymptotic Series Diverge (Q5317161)
From MaRDI portal
!
This is the item page for this Wikibase entity, intended for internal use and editing purposes. Please use the normal view instead:
scientific article; zbMATH DE number 2205857
| Language | Label | Description | Also known as |
|---|---|---|---|
| default for all languages | No label defined |
||
| English | Hyperasymptotics and the Linear Boundary Layer Problem: Why Asymptotic Series Diverge |
scientific article; zbMATH DE number 2205857 |
Statements
Hyperasymptotics and the Linear Boundary Layer Problem: Why Asymptotic Series Diverge (English)
0 references
16 September 2005
0 references
perturbation methods
0 references
asymptotics
0 references
hyperasymptotic
0 references
super asymptotic
0 references
exponential smallness
0 references
0.7661149501800537
0 references
0.7661149501800537
0 references
0.756611168384552
0 references
0.7560536861419678
0 references
0.7556807994842529
0 references
The aim of this paper is to explain how divergent asymptotic series appear in singularly perturbed problems. Why do some series diverge? How can one break the ``error barrier''? To discuss these questions the author considers the model problem \[ \epsilon^2 u''-u=-f(x), \quad u(-1)=0,\,\,u(1)= 0, \] where \(\epsilon>0\) is a small parameter. It is known that the solution of this case study is the superposition of a particular solution, the slow motion \(u_s(x,\epsilon)\), and solutions of the homogeneous equation which are boundary layers \(A\exp(-(x+1)/\epsilon)\) and \(B\exp((x-1)/\epsilon)\). A catalog of examples with explicit solutions shows clearly that in general we cannot choose \(u_s(x,\epsilon)\) to be a convergent series in \(\epsilon\). The author gives his view of why in this type of problems we have to use asymptotic series and why series diverge. This is based both on the intuition the author builds into such problems and on the Fourier integral solution. ``Superasymptotic'' approximations are defined as optimally truncated asymptotic series. The author describes then ``hyperasymptotic'' approximations as approximations that achieve higher accuracy than the optimal truncations. He indicates paths to obtain such improved approximations. This paper tries to describe the intuition behind these concepts so as to understand which phenomena are essential and which are the whereabouts of such problems. Such a goal is a difficult and somewhat hazardous task.
0 references