Stacker cranes used in AS/RS service travel on a bottom rail cast into the floor slab and a top guide rail fixed to the rack uprights; under seismic excitation, the mast, the base frame wheels, and the rail anchor bolts form one coupled system whose reactions must be resolved together [S5][S1].
A stacker crane consists of a base frame running on a floor rail, one or two masts, and a lifting frame, a configuration referenced in FEM-9101 and reproduced in current TUM rail/wheel predictive-design work [S3]. In seismic zones the design problem stops being a quasi-static lift-and-travel check and becomes a dynamic stability problem with uplift, lateral derailment, and mast ovalisation competing for the same capacity.
Where seismic loads enter a stacker crane
Earthquake excitation on a stacker crane enters through three distinct paths: inertial forces from the mast, top frame, and load; wheel/rail reaction redistribution including uplift at one rail; and P-delta amplification from the tall, slender mast geometry. The EnginSoft Mexico evaluation case used time-history analysis to establish site-specific seismic intensity levels and minimum design requirements, rather than relying on a single spectral coefficient [S1].
Seismic input is normally a horizontal response spectrum anchored to the local code (for example, the CFE Manual for Mexican territory, ASCE 7 in the US, Eurocode 8 in the EU); vertical pseudo-static amplification of roughly two-thirds of the horizontal spectral ordinate is then applied to the lifted load and to the mast self-weight [S1]. Twin-mast cranes pick up vertical inertial load on the cradle through both masts, while single-mast units transmit the same load as an eccentric moment into the bottom rail, which is the worst case for anchor pull-out.
Mast geometry vs. seismic vulnerability
Two rail configurations dominate. Two-rail systems apply the platform load radially to a round mast, which bends the cross section out of plane and drives the tube toward an ovalised, jam-prone shape over service life; this same radial loading path magnifies seismic bending and reduces the buckling margin of the mast [S2]. Four-rail masts transfer the load tangentially to the tube wall and avoid out-of-plane bending, which is also a more favourable seismic load path because lateral inertia is taken largely in shear through the guide rollers [S2].
A single-mast configuration is the default for light unit loads in box/miniload AS/RS, while twin-mast units handle pallet loads up to roughly 1.65 t per the MT-1 to MT-3 single-deep data; the taller MT-4/MT-5 reach 148 ft but are capped at 1.1 t, trading capacity for a slenderness ratio that has to be rechecked seismically because the natural period lengthens into the peak of the design spectrum [S5].
Floor rail and top guide rail: two different design problems
The bottom rail is grouted or chemically anchored into the concrete floor slab, while the upper guide rail is bolted to the rack profiles; the two rails are charged with stabilising the crane and absorbing lateral forces during put-away and retrieval, and the same interfaces have to absorb inertial shear and overturning moment under earthquake load [S5]. Anchor bolts for the bottom rail are normally headed studs or post-installed anchors sized for combined shear, tension (uplift), and prying from the base frame; in high-seismic zones, designers add shear lugs or embed the rail base in a continuous steel channel cast into the slab to control local concrete breakout.
Top guide rail attachment is governed by the rack uprights: seismic anchor capacity of the rack base, the upright section modulus, and the rack-to-rail interface all have to be on a common capacity path with the crane, because the rack is the only thing resisting lateral crane inertia above floor level. If the rack fails before the rail bracket does, the rail is useless. Interlake Mecalux operating data shows single-mast units running at up to 722 ft/min travel and 2 ft/s² acceleration, so seismic lateral demand has to be added to operational lateral demand on the same anchor group rather than treated as a separate load case [S5].
Mast base clearance and derailment geometry
Mast base clearance above floor level is normally set between 120 mm and 350 mm, set by aisle height, seismic zone, and unit-load capacity; tighter clearances cut the available rotation and lateral offset of the base frame before the mast fouls the rack structure, but looser clearances grow the seismic pendulum length and the impact load if the crane derails [S6]. Derailment under seismic load is treated as a low-probability, high-consequence event: tall single-mast units can lift one bogie off the rail by tens of millimetres without becoming unstable, because the crane is nearly centred between the two rails and gravity provides a restoring moment, but beyond a critical lift-off the system snaps laterally and the mast hits the rack [S9].
Modern stacker cranes mount on a base frame with two wheels running on the bottom rail [S3], so the designer has only two anchor points per bogie to resist seismic shear and uplift; this is why post-installed anchors are often supplemented by a continuous embedded rail plate, which turns the wheel reactions into a distributed moment couple rather than point loads in the slab.
Selection criteria: rail layout vs. seismic zone
Selection breaks along four criteria: seismic intensity at the site, mast configuration (single vs. twin), rail count (two vs. four), and anchor type (headed studs vs. post-installed mechanical/chemical anchors with shear lugs). For a greenfield high-seismic site, a four-rail twin-mast crane with continuous embedded bottom rail and rack uprights designed to a common seismic capacity is the most forgiving combination because the load path through the masts is tangent, not radial, and the bottom rail distributes wheel reactions [S2][S5].
For a low-seismic retrofit on an existing rack, a two-rail single-mast stacker with bolted bottom-rail plates and verified rack-base anchorage is normally sufficient, provided the mast cross section is checked for out-of-plane bending under the combined effect of eccentric platform load and seismic lateral load [S2]. Existing crane foundations in seismic zones should be re-evaluated against current spectra every time the spectrum is updated, not just at original design, because intensity maps in many jurisdictions are revised on a roughly decadal cycle and a crane qualified under the 1994 NEHRP map may not be qualified under current code [S1].
Operational constraints, failure modes, and what to inspect
Common failure modes in seismic events, in order of observed frequency, are: anchor-bolt loosening or pull-out at the bottom rail; rack-base anchor failure transmitting into the top guide rail; mast ovalisation jamming the lifting frame; and derailment with mast-to-rack impact on the unloading side. The EnginSoft case work treats these as separate but coupled check cases in the same time-history model, because in practice the first failure redirects load into the next weakest link within a fraction of a second [S1].
For overhead and stacker cranes more broadly, anchorage capacity under dynamic, cyclic, and impact loading is the dominant seismic design lever, and current 2026 practice guidance is to verify both the static pull-out capacity of the anchor group and the dynamic stiffness of the rail-to-slab connection so the crane's first mode period sits where the design spectrum expects it to [S10]. A practical next step for any seismic retrofit is to re-run the model with the as-built anchor group, not the original design assumption, because installer variability on edge distance and concrete compressive strength routinely produces anchor capacities well below the calculated value, a pattern that mirrors what we see in the broader crane-anchorage field, including structural-vs-stability limits in mobile crane load charts where the rated chart can diverge sharply from the as-built reality [S10].
Spec-level background on the components involved: stacker crane, pallet stacker, and checking and testing apparatus.
See also our earlier report, Structural vs Stability Limits in Truck Crane Load Charts.