Q.: I have been asked to repair an area of about 10 feet by 10 feet in a 6-inch-thick floor on grade. This floor, in a building used by an oven manufacturer, had performed well for 15 years. Recently the company tested out a new experimental oven. I believe the oven did not have a bottom, but they set it on some insulation and heated the oven to 1100 degrees F. After some period of time the whole 10x10-foot area exploded. Concrete came off for depths ranging from 2 1/2 to 3 1/2 inches. None of the rubble from the explosion was larger than 3/4 inch diameter. I expect to patch this by sawing out the remainder and replacing the area to the full depth, but I am curious about what caused the concrete to blow apart.

A.: The behavior of the concrete was similar to the way concrete performs in a standard fire test. The insulation, incidentally, does not prevent heat from going into the concrete; it just reduces by a small amount the maximum temperature in the concrete.

If the room temperature is 70 degrees F and we overlook the effect of insulation, the temperature rise is 1100 minus 70 or 1030 degrees F. The concrete in the heated area tries to expand greatly, because of the higher temperature. This potential expansion is almost completely restrained, however, by the large surrounding area of unheated concrete. Multiplying the temperature change by the coefficient of thermal expansion, 0.0000065, gives 0.0067 inch per inch potential expansion.

Assume this potential expansion were completely restrained by the very large area of surrounding cool concrete, and that the concrete had a modulus of elasticity of 4,000,000 psi. (In a 15-year-old slab on grade, which would probably be fairly moist, the value could be even higher.) The maximum compressive stress would be 0.0067 x 4,000,000 = 27,000 psi at the top surface. This stress diminishes rapidly from the hot surface to the cooler depths of the slab, but the gradient in compressive stress causes great shear stresses. Together the compressive and shear stresses destroy the affected layer of the slab.

There are a considerable number of other factors that are thought to contribute to what happened, but this is perhaps the main cause.