Optimal upper bound for the maximum of the \(k\)-th derivative of Hardy's function (Q2343176)

We quote the author's abstract: ``We give an approximate functional equation for the derivatives of Hardy's \(Z\)-function, uniform with respect to the order of the derivatives, from which we deduce an optimal upper bound for the \(k\)-th derivative of \(Z\).'' For \(\zeta\) the Riemann zeta-function, Hardy's function \(Z\) is, by definition, \[ Z(t)=e^{i\theta(t)}\zeta\left({1\over 2}+it\right), \] where \(\theta(t)\) is defined as \[ \theta(t)=\arg\left(\pi^{-i{t\over 2}}\Gamma\left({1\over 4}+i{t\over 2}\right)\right). \] (See equation \((1)\) and the details in the paper.) As the author declares in the abstract, he gives an approximate functional equation (uniform w.r.t. \(k\), the derivative order); this comes from work of \textit{A. A. Lavrik} [Anal. Math. 17, No. 4, 257--279 (1991; Zbl 0774.11049)], that gives an explicit formula for the derivatives of the Riemann zeta function on the line \(\mathrm{Re}(s)= \frac{1}{2}\). Then the author joins some combinatorial formulæ, see \((6)\) for an example, to the standard environment of analytic number theory calculations; one noteworthy feature here is Lemma 4, estimating exponential sums, due to the author. The present ``work stems from an observation of \textit{A. Ivić} [``On some reasons for doubting the Riemann hypothesis'', in: P. Borwein (ed.) et al. The Riemann hypothesis. A resource for the afficionado and virtuoso alike. New York, NY: Springer (2007), see also \url{arXiv:math/0311162}]''.