Of the many forensic investigations of warehouse and industrial slabs on ground we have conducted throughout the years, the majority of the problems have involved joint deterioration, including edge spalling due to lift-truck traffic. This problem is commonly a result of the slab designer having relied on aggregate interlock to transfer the wheel shear load across the joint. This practice is ineffective because the aggregate interlock is nearly always lost over time due to horizontal shrinkage and curling of the slab.

Joint spalling in undoweled joints is caused by lift-truck wheel impact on the higher joint edge.
Joint spalling in undoweled joints is caused by lift-truck wheel impact on the higher joint edge.

When aggregate interlock is lost, the slab edges on either side of the joint become free edges. Free edges have significant vertical differential movement. This movement is made worse by the slab losing base contact at the joint due to curling.

When a lift-truck's wheels travel across the joint, the unloaded slab panel edge is exposed and damaged by the lift-truck wheels (see drawing on page 37). Once the edge spalls, the lift-truck wheels can be damaged when the vehicle travels repetitively over the spalled area. If the joint spall is not quickly repaired, this repetitive traffic often causes the spall to become wider, with a corresponding increase in damage to lift trucks.

As we discussed in a previous article (Ref. 1), the trend in lift trucks to use harder, smaller wheels has made this problem worse. Joint spalling often begins only after most of the concrete shrinkage and curling has occurred—typically 12 to 18 months after slab construction and the owner has taken possession of the facility. The cost for the repairs, therefore, is hidden often in the facility owner's maintenance budget, or as we will discuss, the contractor pays. We have collected data that shows it to be cost effective to use a small portion of these “hidden maintenance costs” to provide dowels in the initial construction to minimize the joint deterioration and equipment repair costs. The payback period for this investment in dowels is as short as 18 months.

Should aggregate interlock be used?

Joint spalling is just starting to occur due to lift-truck traffic.
Joint spalling is just starting to occur due to lift-truck traffic.

Many slab designers rely on aggregate interlock even though American Concrete Institute (ACI) publications for many years have cautioned against this practice and have recommended doweled joints when load transfer is required. ACI 302.1R-04 “Guide for Concrete Floor and Slab Construction” states that “Doweled construction and contraction joints are recommended when load transfer is required ...” The 1996 edition of ACI 302R.1 had similar wording regarding dowels.

ACI 360-06 “Design of Slabs on Ground” states “if the designer cannot be sure of positive long-term shear transfer at the joints through aggregate interlock, then positive load-transfer devices should be used at all joints subject to wheeled traffic.” We have shown in previous articles (Refs. 2 and 3), that it is impractical to rely on aggregate interlock for long-term load transfer at the contraction joints for floor slabs subjected to lift-truck traffic. Even as early as 1956, ACI Committee 325 on concrete pavements cautioned against relying on aggregate interlock: “Experience indicates that aggregate interlock may be satisfactory as a means of load transfer only under unusually favorable conditions of joint opening and foundation support. It is not satisfactory under a large volume of heavy commercial traffic.” (Ref. 4)

Another potential problem in relying on aggregate interlock is what we call the “dominant joint” issue. Saw-cut contraction joints are intended to control the location and width of shrinkage cracks. Hopefully, if done properly, the shrinkage cracks will occur below these saw-cuts. But due to uneven base restraint below the slab and differences in slab thickness, not all saw-cut joints actually crack or “activate.” Because some joints do not activate, or activate very little, this causes those joints that do activate to open much wider, since the shrinkage is concentrated at these “dominant joints.” These dominant joints are the first to lose aggregate interlock, and if they occur in traffic areas, are the first to spall.

Many times the slab panels on either side of these dominant joints will curl differentially, producing a differential elevation between the panels and creating tripping hazards and potential lawsuits. This dominant joint behavior is made worse by the ever-increasing use of vapor barriers/retarders, which reduce base friction and make the dominant joints more noticeable and problematic in exposed concrete floors.

Given the cautionary statements in the ACI documents, other published information, and our experience with the problems inherent in relying on aggregate interlock, we recommend against it. To rely on aggregate interlock for slabs with wheeled traffic, slab designers would need to have great confidence in their knowledge of the local long-term shrinkage potential of the concrete, the slab base restraint, the anticipated construction tolerances, the corresponding joint spacing, and other factors. If the slab designer cannot be sure of positive long-term shear transfer at the joints through aggregate interlock, then the prudent and cost-effective approach is to specify dowels in the contraction and construction joints.

Passing on the hidden costs

In building and operating a large warehouse or retail store that will have lift-truck traffic, there are typically two teams involved. The first is the design/construction team whose responsibility is to design and build the best facility for the given budget and to minimize initial cost. The second is the maintenance team whose responsibility is to maintain the floor and the lift trucks during operation.

Our experience is that the design/ construction team often makes design decisions without input from the maintenance team, which could provide valuable information on the long-term performance of the design/construction system. The construction team typically does not like to ask the maintenance team for input, fearing that they would ask for quality items that would increase the initial cost. The design/ construction team gets no benefit or reward for initial cost increases and may be penalized for the increase. Therefore, there is a disconnect between these two teams' objectives, even though they may work for the same company.