The stresses which in theory can occur in a simply supported structure owing to temperature distributions are, as demonstrated by the author, significant in terms of the total load effects. However, since the stresses are caused by the self-restraint of the section, it would seem logical to expect some reduction of stress with the onset of cracking, whether caused by the effects of the temperature distribution itself or by external loading. The method by which allowance for cracking should be included in the temperature stress calculations is not clear, although the technique
adopted would seem to be a logical approach. Indeed, if the stresses obtained (Fig 5(h)) are compared with those calculated assuming an uncracked section, there is a reduction in the top surface compressive stress of about 13% using the cracked section. However, if the stresses calculated at a depth of 360 mm using the two methods are compared, it can be shown that using an uncracked section results in a tensile stress of 0.6 N/mm2 compared to 1.1 N/mm2 for the cracked section. It would
seem that there is a need for further research in this area.