I am concerned by the search for scientific models that have no impact on empirical standards and thus protect neither men nor industrial products against the phenomena they are supposed to explain.    For example, how can we avoid the explosion of glass, fertilizer or cereal storage sites, high-voltage cables, touch screens, or archaeological treasures, wherever the materials used are INSULATORS? Why are only a few museums connecting their treasures with a wire to the ground? Why unbreakable glasses explode in operation and under very slight constraints? Why is the explosion of a high-voltage cable leading to the explosion of insulators in their neighborhood? Do these accidents mean that damage models are wrong or only applicable in particular stress value? The question is current and the insurance world is beginning to worry about it. ©