Q: We used an early-cut saw to joint a concrete floor immediately after finishing. The floor is 6 inches thick and the sawcut depth is 3/4 inch. Although the floor has no random cracks, the engineer says we now have to chase the joints with a conventional saw so the cuts are 1 1/2 inches deep (one-quarter the slab thickness). Is this necessary? We think it's a waste of time and money.

A.: A Portland Cement Association publication, Concrete Technology Today (November 1995), cites research showing that sawcut timing is much more important than sawcut depth in controlling random cracking of pavements. The researchers, in a field investigation by the Texas Highway Department, monitored joint formation and crack control for 13-inch-thick plain concrete pavement test sections placed directly on subgrade soils. Variables included differing concrete mixtures, coarse-aggregate types, curing methods and sawcutting techniques.

Joints in the test sections were sawed at 15-foot intervals using two different methods conventional water-cooled sawcutting to a 3-inch depth and early-age sawcutting to a 1-inch depth (without cooling water). The early-age sawcutting was typically done less than three hours after concrete placement.

Observations over a 10-month period showed that of all the transverse cracks that developed, only two occurred between sawcut joints, and both of these uncontrolled cracks originated at re-entrant corners for inlet drainage structures. This investigation seems to indicate that the shallower depth of early-cut joints doesn't reduce their effectiveness in controlling random cracking.

Original research results are reported in Transportation Research Record 1449, published by the Transportation Research Board, Washington, D.C. Call 202-334-3214 for ordering information. You can buy a copy of the Concrete Technology Today article by calling 800-868-6733 and asking for PL953.01.

Editor's Note

This item originally appeared in Troubleshooting Newsletter, published by the American Society for Concrete Construction. For more information about this quarterly newsletter, call ASCC at 800-877-2753.

Readers Respond

In the June 1997 Problem Clinic (p. 529) the reader wanted to know if it was necessary to chase ¾-inch-deep sawcuts, made by an early-cut saw, to a depth of 1½ inches (one-quarter slab thickness). I believe that many slabs on grade have large variations in slab thickness due to poor subgrade preparation and construction procedures, resulting in planes of weakness requiring joints that exceed the depth of early-age sawcuts.

You quoted results of a research project by the Texas Highway Department, where early-age sawcuts gave excellent results. I would guess that the project had good slab-thickness control, which is not found on many projects. I recommend that all slabs on grade be cut to one-quarter the slab thickness.

- Larry Asel, Conco Companies, Springfield, Mo.

The information on sawcut depth in the June 1997 Problem Clinic is wrong. About every 20 years, the United States has a rash of off-center longitudinal cracking in concrete pavements after a cool fall. In subsequent field surveys, it was found that longitudinal joints were sawed to a depth of less than one-quarter the pavement thickness. Most states now specify that the depth of the longitudinal joint be one-quarter the slab thickness plus ¼ inch, and many recommend a depth of one-third the thickness.

I assume that in the research you referenced, the concrete was not tested by nature. Contractors who saw joints to a depth of less than one-quarter the slab thickness will experience problems if the weather turns unusually cool after concrete placement. The thermal shock will cause cracking, although the cracks may not be visible until months later.

- Thomas J. Pasko Jr., Director, Office of Advanced Research, Turner-Fairbank Highway Research Center, McLean, Va.