PCA Airport Thickness Design

EB050 is PCA's 1973 airport method, built on Packard's PDILB program. Some engineers use it for heavy industrial and port equipment because it handles any wheel layout. It is a sound statement of interior-load theory and a useful framework, but it was written for airfield pavements with no free edges in traffic and adequate load transfer assumed at every interior joint, and it leaves out the conditions that most often govern an industrial floor.

Similar inputs, different problem. Accepting any wheel layout is not the same as fitting the problem. A design method is a stress calculation tied to a performance criterion calibrated against a kind of distress. EB050's safety factors and Table A4 were calibrated to airfields: wandering aircraft on pneumatic tires, no free edges in traffic, frequently stabilized bases, outdoors. Airfield slabs also tend to be thicker than flexure requires, often because owners want larger panels and less joint, and their maintenance is driven mostly by foreign object debris and durability. Industrial floors are indoors, curl at their joints, and carry small, hard wheels whose impact on a deflected joint edge spalls it, a mechanism airfields do not see. That is why floor guidance designs against other limits: ACI 360R-10 ties thickness to a fatigue curve and suggests limiting differential deflection across joints to about 0.010 in. for small, hard wheels and 0.020 in. for larger, cushioned wheels. EB050 has no joint deflection criterion, and its Table A4 allows unlimited repetitions at a stress ratio of 0.50, where PCA's 2008 floor table allows about 760,000.

A high but unknown factor of safety. Floors designed this way have often performed well, because a safety factor of about 2 on an interior stress covers a lot, and it absorbs what the method leaves out. For the load alone, the same slab has about 2.0 / 1.77 = 1.13 at a free edge using AirPave's free-edge factor. Curl cannot simply be stacked on that number (see below), and the capacity side is missing too: the modulus of rupture is a beam strength, and a slab-on-ground carries more than a beam. The factor of safety is high in the interior and unknown at the joints until curl and load are solved together, with loss of support, load transfer and the loads sharing a panel, and compared with the slab's capacity rather than the beam's.

Interior load on an infinite, fully supported slab. PDILB integrates Pickett's extension of Westergaard's interior solution over the tire imprints on a dense-liquid foundation. There are no joints, no slab size and no edges in the model. EB050 justifies this by requiring adequate load transfer at all joints, so that "a paved area acts as a continuous large slab." Sawcut, undoweled floor joints that open with shrinkage do not meet that condition.

Joints and free edges only through AirPave's factors. EB050 has no joint or edge stress. Both the 1955 and 1973 editions required load transfer at every joint and, where there was none, thickened the joint edges about 20 percent, which is worth about 1.2² = 1.44 in stress. A PDILB stress carried into a floor design usually leaves that requirement behind. The interior joint values on this page use the joint factors ACPA's AirPave program added (1.39 single, 1.33 dual, 1.00 single tandem, 1.14 dual tandem). They came from a limited number of ILSL2 finite element runs on one 25 x 25 ft slab at three k values and three thicknesses, with fixed E, contact area and pressure. They include a fixed 25% reduction for load transfer taken from FAA practice, so joint type, dowels and joint opening play no part in them. The 1.00 for single-wheel tandems means a joint is no worse than the slab interior, which is physically doubtful. The page also shows the implied free-edge value (factor divided by 0.75). No joint factors exist for dual tridem or larger gears.

No curl, warp, shrinkage or temperature. As with every Westergaard-based method, the stress is from the load alone. ACI 360R-10 reports curling stresses of 200 to 450 psi for interior floors. These cannot simply be added to the load stress: a curled slab lifts off its support at the joints and corners where loads are highest, the load pushes it back down, and the contact, the stress distribution and the face in tension all change with the load. The combined stress can be above or below the sum. The computed stress here is tension at the bottom of the slab; the top tension of a cantilevered edge is not considered.

The bottom of the slab only. PDILB computes tension at the bottom of the slab under the gear. Top-down cracking, which governs curled, jointed floors between closely spaced loads such as back-to-back rack legs and along the joints, is outside the method. A method that cannot model a stress state cannot design against the cracking it causes; it can only rely on the factor of safety.

Support. A higher k lowers the computed interior stress, but a stiff or cement-treated base also restrains the slab and increases curl, which the method does not see. The k helper reads EB050 Figs. 11 and 12. EB050's Appendix B adjustment to the radius of relative stiffness for multiwheeled gear on thick pavements is not included.

Tire contact. Contact area is taken as wheel load divided by inflation pressure, as for pneumatic tires; EB050 notes actual areas are somewhat smaller. Small solid and polyurethane wheels do not make a tire imprint: they bear on a narrow strip across the wheel, closer to a short line load, and are not represented by this assumption. The vehicle list comes from AirPave and includes 1990s-era equipment that should be checked against current manufacturer data.

Fatigue. Table A4 is PCA's fatigue curve of the period, summed with Miner's hypothesis, with unlimited repetitions at a stress ratio of 0.50 or less. The load repetition factors in Table A1 come from aircraft wander studies; channelized industrial traffic should use 1.0. Mixed traffic has to be summed load by load as in Table A3.

A stepping stone. EB050 and PCA's floor bulletin were both revised several times, each edition adding what the last lacked. The 1955 edition, a revision of the 1950 original, offered its method as safe "until the application of the elastic solid theory is developed," noting that "a practical solution soon will be found." The airfield community has since moved on: the FAA computes rigid pavement edge stresses with three-dimensional finite elements in FAARFIELD, and ICAO retired the PCA program for aircraft and pavement ratings in 2024. Floor design stalled at the joint factor.

Not covered. Rack posts, bearing and punching shear under plates, and distributed storage loads. For those, use the PCA Slab Thickness Design tool.

A design that has to account for joints, curl and loss of support needs an analysis that models them together and compares the result with the slab's capacity rather than a beam's, which is what SLABS does.