In the 19th century it was industrial practice alone that was responsible for technological progress – with all the negative consequences that entailed. Systematic experiments were unknown. The only method used was trial and error. If, for example, a steam boiler had proven to be reliable in practice, its technical specifications were copied for other boilers. If it wasn't reliable, BOOM! The disadvantages of this method manifested themselves in numerous steam boiler explosions killing many people. All that could be done was to reinforce the boiler casing for the next model. The same approach was also used for other machines and in the construction of iron bridges.
Thus increasingly urgent calls were made for reliable assessments of the properties of materials, for dependable detection of flaws in components and well-founded estimates of the life span of parts.