Does $f_n(x) = \sqrt[n]{1+x^n}$ converge uniformly on $[0,\infty)$?
First, let's investigate point-wise convergence:
- For $x>1$: $\lim_{n\to\infty} \sqrt[n]{1+x^n} = x$
- For $x\le 1$: $\lim_{n\to\infty} \sqrt[n]{1+x^n} = 1$
So there is a point-wise convergence only for $x\ge 1$.
Basically, the question asks if there's a uniform convergence at $[0,\infty)$ so the answer is no because there isn't even a point-wise convergence at $[0,1)$. Am I right?
But (assuming I'm right so far) let's assume the question asked about uniform convergence at $[1,\infty)$:
We should look for $$\sup_{x\in [1,\infty)} \left( \sqrt[n]{1+x^n} -x \right)$$
Let's find extremum points of $f_n(x)$:
$$f_n'(x) = (1+x^n)^{\frac{1}{n}-1}x^{n-1}$$
Here I got a little stuck, tough I think there's only minimum point at $x=0$ so $f_n(x)$ grows with $x$.
Could you help me finish up this proof? (And validate the first part.)
Partial answer.
Use $$(u-v)(u^{n-1}+u^{n-2}v+\cdots + uv^{n-2}+v^{n-1}) = u^n-v^n$$ With $u=\sqrt[n]{1+x^n}$ and $v=x$. This will give you a good estimate for $f_n(x)-x$ when $x\geq 1$. In particular, $f_n(x)-x \leq \frac{1}{n}$ for $x\geq 1$.
So the question reduces to whether there is a uniform convergence on $[0,1)$.