An initial-boundary value problem for \(\begin{aligned} \left\{ \begin{array}{ll}u_{tt} = \big (\gamma (\Theta ) u_{xt}\big )_x + au_{xx} - \big (f(\Theta )\big )_x, \qquad & x\in \Omega , \ t>0, \\[1mm] \Theta _t = \Theta _{xx} + \gamma (\Theta ) u_{xt}^2 - f(\Theta ) u_{xt}, \qquad & x\in \Omega , \ t>0, \end{array} \right. \end{aligned}\) is considered in an open bounded real interval \(\Omega \) . Under the assumption that \(\gamma \in C^0([0,\infty ))\) and \(f\in C^0([0,\infty ))\) are such that \(f(0)=0\) , and \(k_\gamma \le \gamma \le K_\gamma \) as well as \(\begin{aligned} |f(\xi )| \le K_f \cdot (\xi +1)^\alpha \qquad \hbox {for all } \xi \ge 0 \end{aligned}\) with some \(k_\gamma>0, K_\gamma>0, K_f>0\) and \(\alpha <\frac{3}{2}\) , for all suitably regular initial data of arbitrary size a statement on global existence of a global weak solution is derived. By particularly covering the thermodynamically consistent choice \(f\equiv id\) of predominant physical relevance, this appears to go beyond previous related literature which seems to either rely on independence of \(\gamma \) on \(\Theta \) , or to operate on finite time intervals.
