The summer of 2004 saw what seemed like an abnormal number of sudden hailstorms and downpours in the Denver area while concrete was being placed. We investigated concrete from two different projects that were halted shortly before a rainstorm. The cores were sent to a lab by concerned contractors who wanted to know what performance to expect from the concrete.

In Project A the concrete was not covered, and rain fell 1½ hours after placement. In Project B the concrete was placed and covered with plastic sheeting about ½ hour before the storm, but water got under the plastic.

Figure 1. This cross section shows the damaged concrete surface from Project A. The top 1-mm-thick blue-stained layer is the disturbed porous layer. The brown paste below is normal. A horizontal white spall fracture is visible near the bottom of the porous layer. Note how the sand particles protrude from the top surface due to erosion of the surrounding paste.
Figure 1. This cross section shows the damaged concrete surface from Project A. The top 1-mm-thick blue-stained layer is the disturbed porous layer. The brown paste below is normal. A horizontal white spall fracture is visible near the bottom of the porous layer. Note how the sand particles protrude from the top surface due to erosion of the surrounding paste.

In both cases the surface paste was eroded from ¼ mm to 1 mm deep, variably exposing the fine-grained aggregate and often leaving the aggregate in relief on the surface. The intended quality of the finished surface was reduced by varying degrees.

We cut thin sections perpendicular to the concrete surfaces for petrographic examination. In Figure 1, from Project A, a layer as deep as 1.2 mm (1/20 inch) was apparently loosened by battering of the surface. The damage includes an increase in paste porosity as shown by the blue color of the impregnated dye and by deformation of the air voids. The air voids in the disturbed zone are not spherical, as is normal, but had spread out under the fine aggregate particles, leaving them with poor adhesion to the underlying concrete. A spall fracture is shown in the disturbed layer.

Figure 2. This cross section shows the concrete surface from Project B. The porous zone is 2 mm thick and contains many irregularly shaped air voids. The black 0.1-mm-thick layer at the surface is unset, very porous, hydrated paste.
Figure 2. This cross section shows the concrete surface from Project B. The porous zone is 2 mm thick and contains many irregularly shaped air voids. The black 0.1-mm-thick layer at the surface is unset, very porous, hydrated paste.

In Figure 2, from Project B, the disturbed and loosened layer is about 2 mm thick. It occurred under a layer of water topped by the plastic sheeting. This disturbed zone will probably rapidly erode away with use and with freeze/thaw action. The surface is mottled with varying degrees of finish, or non-finish, and the fine aggregate varies in exposure and relief. But since the rain did not affect the concrete below the 2-mm-thick layer, no structural damage was incurred, and we would not anticipate any further erosion below the disturbed layer.

To fix the problem in Project A, the surface was diamond scored to a depth of about ¼ inch as shown in Figure 3.

Figure 3. On Project A, the contractor used a grinder to remove about ¼ inch of the porous surface layer of concrete. This is not always an acceptable solution from an appearance standpoint.
Figure 3. On Project A, the contractor used a grinder to remove about ¼ inch of the porous surface layer of concrete. This is not always an acceptable solution from an appearance standpoint.

In Project B, the contractor removed the plastic after the rain and swept the surface with burlap to re-texture it and to brush off the water. This produced the surface layer of washed, very porous paste that appears black in Figure 2. On the slab, this chalky layer is white, as shown in Figure 4. The resurfacing is shown in cross section in Figure 5 where a depression is bridged by the layer of washed displaced paste. Wear will eventually erode this layer and the underlying 2-mm-thick, porous paste zone, leaving a rough surface with exposed aggregate that is stable but not aesthetically pleasing. The blue air cavity at the bottom of Figure 5 is entrapped air.

Figure 4. The surface on Project B is white from a thin, chalky film of re-textured concrete.
Figure 4. The surface on Project B is white from a thin, chalky film of re-textured concrete.

We don't know precisely at what time after placement rain will no longer damage the surface, but we know it is more than 1½ hours. In Project B the re-texturing procedure used was not a viable long-term repair. The only successful repair method we have investigated is diamond grinding, but this is not an acceptable finish in many cases. We anticipate that hail damage would be similar but probably more severe and possibly deeper, all other things being equal.

Figure 5: A cross section of the Project B re-textured surface shows porous hydrated paste (thin black layer at top) that has bridged over a depression in the surface. The resurfacing smoothed the surface, but the bridge will likely disintegrate in a very short time as will the underlying 1-mm-thick porous zone. The blue air void at the bottom of the photo is entrapped air from the original placement.
Figure 5: A cross section of the Project B re-textured surface shows porous hydrated paste (thin black layer at top) that has bridged over a depression in the surface. The resurfacing smoothed the surface, but the bridge will likely disintegrate in a very short time as will the underlying 1-mm-thick porous zone. The blue air void at the bottom of the photo is entrapped air from the original placement.

Acknowledgement

Special thanks to Gary DeWitt of the Colorado Department Of Transportation, Evans, Colo., and Orville Werner, CTL Thompson, Denver, for their valuable contributions.

—Ted Paster is a consulting petrographer in Denver.