home *** CD-ROM | disk | FTP | other *** search
- Xref: sparky sci.math:12968 sci.math.num-analysis:2972
- Path: sparky!uunet!think.com!rpi!zaphod.mps.ohio-state.edu!moe.ksu.ksu.edu!math.ksu.edu!deadend
- From: bennett@math.ksu.edu (Andy Bennett)
- Newsgroups: sci.math,sci.math.num-analysis
- Subject: Re: Poisson kernel
- Date: 9 Oct 1992 10:10:41 -0500
- Organization: Dept. of Mathematics, Kansas State University
- Lines: 37
- Distribution: inet
- Message-ID: <1b47dhINNs5s@hilbert.math.ksu.edu>
- References: <1992Oct9.055416.18772@monu6.cc.monash.edu.au>
- NNTP-Posting-Host: hilbert.math.ksu.edu
-
- suter@fawlty8.eng.monash.edu.au (Mr D. Suter) writes:
- >The Poisson kernel allows one to evaluate a harmonic function
- >by knowing its values on a closed curve.
- >Can you do similar with a biharmonic function?
-
- Yes, you do exactly the same thing. A function in two variables is bi-harmonic
- if it is harmonic in each variable separately. So if you fix one variable then
- you have an ordinary harmonic function in the other variable and you apply the
- ordinary Poisson kernel along the curve defined by the intersection of the
- boundary of the region you are interested in with the plane defined by fixing
- one variable.
-
- More precisely. Let x,y \in R^2 and let \Omega\subset{R^2\times R^2}. If
- u(x,y) is biharmonic and (x_0,y_0)\in\Omega then u(x,y_0) is harmonic in
- \Omega\intersect\{y=y_0\} and
- u(x_0,y_0)=\int_{\partial\Omega\intersect\{y=y_0\}}u(t,y_0)P_{y_0}(t)dt
- where P_{y_0}(t) is the Poisson kernel for the domain \Omega\intersect\{y=y_0\}
-
- Obviously, p-harmonic functions are handled similarly.
-
- The main interest here is that we could obviously fix x=x_0 rather than y=y_0
- and so the value of u(x_0,y_0) can be determined from either knowledge of the
- boundary values along \{y=y_0\} or \{x=x_0\}. So, in contrast to the ordinary
- harmonic case, we don't have to know values along the whole boundary to
- determine the biharmonic function (or in pde lingo, the Dirichlet problem is
- overdetermined). This raises the question of determining the subsets of the
- boundary such that boundary values on those sets uniquely determine a
- biharmonic function. I don't know how much is known here. One easy result is
- that for the bidisk, \{|x|<1\}\times\{|y|<1\}, the values along the
- restricted boundary, \{|x|=1\}\times\{y=1\}, determine a unique biharmonic
- function.
-
- --
- Andrew G. Bennett bennett@math.ksu.edu If you count too
- Dept. of Mathematics Voice: (913) 532-6750 much you turn
- Kansas State University Fax: (913) 532-7004 purple. - SARAH
- Manhattan, KS 66502 STRICTLY MY OWN OPINIONS
-