QUESTION: Our business has become more focused on polishing concrete retails floors, and we are concerned how grinding and polishing affects a floor's flatness and levelness. How can we be sure we are meeting the correct specifications? Do any studies exist showing the effects on FF and FL numbers?

ANSWER: Grinding and concrete floor tolerances go hand-in-hand. More than 30 years ago, using gasoline-powered, single 10-inch diamond disc grinders, my father and I started spot grinding the wheel tracks of narrow-aisle lift trucks to improve their dynamics.

By the mid-1980s, we were planing off all the tracks superflat in a single pass using an aisle-width, rail-mounted, concrete milling machine he built. While this sophisticated “tolerance grinding” work has traditionally been done by just a few specialists, recent improvements in both the profiling equipment and supporting software have now opened this lucrative and still expanding market to anyone who is suitably equipped.

Measuring ground depth with a 6-foot wing-arm jig mounting a D-Meter.
Allen Face Measuring ground depth with a 6-foot wing-arm jig mounting a D-Meter.

Sharply focused on lowering long-term maintenance costs, an increasing number of retailers—both low- and high-end—are abandoning the use of traditional floor coverings in favor of chemically hardened and polished exposed concrete (often also stained and/or supplemented with special decorative aggregates). Since a large percentage of commercial concrete floor projects employ the F-number system to control the slab's as-built flatness and levelness, the question naturally arises as to what effect the subsequent grinding and polishing will have on the floor's delivered FF flatness and FL levelness numbers.

Since the standard concrete floor grinder/polisher is manually steered and employs articulated, or floating, abrasive discs, the floor's local high spots always tend to be ground deeper than its local low spots, regardless of the operator's attempts at uniformity. Fortunately, this asymmetrical material removal always benefits the finished flatness and levelness.

While there are no studies concerning the improvement in FF and FL as a consequence of cosmetic grinding, the grinding contractor can always confidently guarantee that his work will in no way diminish the floor's local flatness and levelness. On the contrary, both FF and FL should always be improved.

MLFF and MLFL grinding

To provide equitable remedies for out-of-tolerance results, ACI 117-06 requires the specification of both Specified Overall Average (SO) and Minimum Local (ML) F-numbers. The specified SOFF and SOFL numbers are strictly general in character and apply only to the entire project floor when completed (and no sub-division thereof).

In contrast, the specified MLFF and MLFL values are strictly local in character and apply only to the rectangles defined by the floor joints, or by the column and half-column lines, if they are smaller. While the specified SOFF and SOFL numbers define the “average profile quality” to be delivered over the entire floor, the specified MLFF and MLFL numbers define the “minimum usable profile quality” to be delivered at every location.

So there are two different ways that a concrete flat-work contractor can violate his F-number specification: by missing the overalls and/or by missing the minimum locals. The remedy to be exacted will also differ considerably according to which of the two possible violations has occurred.

If the entire floor fails to average both the specified SOFF and specified SOFL, then the installer will have failed to deliver the general profile value required, and commensurate liquidated damages will then be forfeited. But if any one of the minimum local rectangles measures below the specified MLFF and MLFL, then that rectangle will be deemed unusable, and its surface will then have to be physically upgraded to exhibit not less than the specified MLFF/MLFL.

Owing to the same logic that ensures only positive tolerance effects from cosmetic grinding, a crayon and water hose are sufficient to determine where to start the corrective grinding on any defective minimum local rectangle. Just flood the area, let the water recede, and outline the highest spots as they appear. Grind each spot into a plane and fan the grinder around the perimeter to take out the knuckle. After all the spots have been done, re-measure the rectangle for MLFF and MLFL per ASTM E-1155, and repeat the entire process until compliance is verified.

F-min Wheel Track Grinding

For profile quality control purposes, every concrete floor falls into one of two categories according to the type of traffic it is intended to support. If the traffic is constrained by guide wires or guide rails to run along specific paths—as in the case of a narrow-aisle warehouse serviced by turret trucks—then the floor is said to be of the Defined Traffic type. Any other type of traffic puts the floor in the Random Traffic category. Note that ACI explicitly restricts using the F-Number System (the “FF/FL system” set forth in ASTM E1155) to Random Traffic floors.

An entirely different F-number system called the F-min System is used to control the wheel track profiles on defined traffic floors. The F-min System is much more sophisticated than the FF/FL System, since it simultaneously controls the floor's impact on the truck's angular displacements, angular velocities, and angular accelerations around both its pitch and roll axes. (Motions about the yaw axis are negligible, since the truck always operates along a straight path.)

For a given wheel pattern, the specification of a single F-min number translates into a complex of six different wheel track tolerances, all of which must be brought into conformity simultaneously. Since no human can visualize the associated 3-D correlation problem well beyond any human's visualization capacity, in the past the corrective grinding of wheel tracks to the various F-min tolerances has been a frustrating, circular, trial-and-error process based on the correction of just one tolerance at a time.

Unfortunately, since the correction of any one tolerance may adversely affect all of the other five, the grinding mechanic often ends up just “chasing his tail.” Even seemingly small defects can result in long lengths of ground wheel track.

In 2005, a special software package was developed for the D-Meter to support the direct measurement and analysis of any combination of wheel track profiles relative to any F-min number. The concrete floor profiles in many new defined traffic warehouses have now been controlled by local testing laboratories at very low cost using this software.

In early 2009, the software package was finally amended to produce an optimum set of grinding instructions based on the simultaneous solution of the six-tolerance problem. Comparisons between the amounts of grinding required by this new optimization program and those required by the old trial-and-error method typically show about a ten-fold savings. This can equal many tens of thousands of dollars on large retrofit projects.

The D-Meter displays the aggregate depth that has been removed.
Allen Face The D-Meter displays the aggregate depth that has been removed.

F-min grinding reports

The optimum F-min grinding report consists of a table showing the depth of surface to be removed (in thousandths of an inch) at each foot down each wheel track. For example, a value of 117 at the 41-foot station on the center wheel track listing would indicate that 0.117 inches had to be removed at that spot. Using these tabular instructions, the corrective grinding crew simply goes in order from spot to spot down the aisle.

Each spot is initially ground to the required depth, making no effort to blend it into the adjacent concrete. Once all the spots have been brought separately to their required elevations, a second pass is made down the aisle to fair everything in.

The depth ground is controlled using a simple 6-foot swing-arm jig mounting a specially programmed D-Meter. The jig is positioned so that its probe rests on the spot to be ground. The D-Meter's display is then reset to 000, and the arm is swung out of the way to allow the grinder to access the spot.

Whenever the probe is returned to the spot, the D-Meter displays the aggregate depth that has been removed. With this instant feedback, the crew quickly learns the grinding times required to remove various depths on a given slab.

Allen Face is a regular contributor to CONCRETE CONSTRUCTION, is an ACI Fellow, and a long-time member of ACI committees 302, 360, and 117.