For fixed nonzero , expand the denominator geometrically:
The effective expansion parameter is , so the expansion fails when . In the inner variable ,
for fixed . Expanding the leading outer term toward gives at leading order; expanding the leading inner term for large and returning to gives . Thus the two one-term re-expansions match in . The next common contribution is when the corresponding next inner order is retained.
On , the relative denominator error is bounded by , hence
Integration by parts yields
For the last step split the remaining integral at one: , and replacing by one on that interval has a bounded error.
In the inner interval, gives an explicit error estimate:
The terms cancel between regions. Every displayed remainder is whenever and ; for example . Therefore the leading answer is independent of the arbitrary intermediate cutoff:

Articles by others on the same topic (0)

There are currently no matching articles.