Bridging the Gap: A Cost-Effective Biaxial Shaking Table for Realistic Earthquake Simulation

Earthquakes are among the most devastating natural hazards that can claim thousands of lives each year and cause staggering economic losses on infrastructure. They are inherently unpredictable and their complex multi-directional forces make them uniquely challenging to study. Researchers in earthquake engineering have long relied on laboratory-based shaking tables to simulate seismic events in a controlled environment which allowed the responses of scaled models of buildings, bridges, and geotechnical systems to be observed and measured in detail. These experiments directly feed into improved design codes, retrofitting strategies, and disaster mitigation policies. However, the ability to replicate realistic ground motions in a laboratory is constrained due to the cost, complexity, and accessibility of the shaking table. Most existing shaking tables used in research laboratories fall into one of two categories. There are the uniaxial systems, which move the test specimen in a single horizontal direction. They are relatively affordable and simpler to install, however, they cannot capture the combined horizontal and vertical shaking that real earthquakes often produce—an omission that limits the fidelity of test results. On the other end are six degrees-of-freedom (6-DOF) systems, which are capable of simulating full translational and rotational motions but at a prohibitive cost (often exceeding several million dollars), and require extensive infrastructure such as deep reaction pits, reinforced strong floors, and specialized actuator arrangements. The vast majority of research institutions, particularly those with modest budgets or smaller laboratory spaces, cannot afford such systems which may result in a persistent gap between the ideal experimental setup and what is practically achievable. Vertical accelerations can amplify stresses in certain structural components, influence foundation performance, and exacerbate soil liquefaction, however, they are usually overlooked in purely horizontal testing regimes which will result in engineering designs that does not fully account for these effects and can leave infrastructure more vulnerable than it appears on paper.

To this account, in a new paper published in Earthquake Engineering & Structural Dynamics, researchers Dr. Rohit Tiwari, Dr.  Arturo Jimenez and Professor Adrian Russell (lead investigator) from UNSW Sydney in Australia developed a novel biaxial shaking table capable of delivering precise, independent, and simultaneous horizontal and vertical motions using only two horizontally aligned hydraulic actuators. One aspect of the innovation is in its compact scissor mechanism that translates one actuator’s horizontal movement into pure vertical displacement which eliminate the need for complex infrastructure like actuator pits or reaction walls. Moreover, the new system combines mechanical simplicity with an advanced iterative control strategy which allowed it to accurately reproduce complex earthquake motions even under nonlinear loading conditions such as soil liquefaction. The research team applied controlled uniaxial motions—sinusoidal pulses of varying frequencies and a time-compressed Kobe earthquake record and compared the commanded displacements with those actually achieved. The authors found the correspondence strikingly close, with deviations so small they fell well within the tolerances needed for high-quality seismic simulation. This early precision was an important first indicator that their mechanical design and control system were in harmony and capable of translating theory into motion without distortion.

The next stage brought in a heavy, dynamically complex payload: a 2.3-tonne laminated shear stack filled with saturated sand, designed to undergo liquefaction under shaking. Here, the real challenge emerged. Liquefaction introduces dramatic changes in stiffness and damping, often derailing even sophisticated control systems. However, the researchers managed to achieve accelerations and displacements that mirrored the targets with excellent fidelity by simply iteratively refining the drive signals using their feedback-based control algorithm. Even as the sand bed transitioned into a softened, fluid-like state, they found the table maintained control which proved its resilience under conditions where many systems struggle. Afterward, the authors introduced true biaxial motion—simultaneous horizontal and vertical shaking—again using the scaled Kobe earthquake record. They performed calibration with the heavy payload in place, allowing the control system to adapt to the specific mass and dynamic properties of the specimen. The authors findings showed an almost seamless match between input and measured accelerations across both axes, even for irregular, earthquake-like waveforms. Gains remained close to unity, and phase shifts were negligible within the target frequency range. This meant the table could not only reproduce complex motions, but do so in a way that preserved the intended timing and intensity—critical for simulating how real structures and soils experience seismic loading. The team also investigated potentially disruptive effects such as unintended rolling, pitching, yawing, and off-axis displacements and by attaching sensors to the table’s corners and running tests at different frequencies, they confirmed these motions were so minimal as to be inconsequential—fractions of a degree in rotation, and sub-millimeter in displacement.

In conclusion, the research work of Professor Adrian Russell and team successfully developed novel biaxial shaking table that delivers precise horizontal and vertical movements using only two horizontally aligned actuators. This innovation enables a much broader community of engineers and scientists to examine phenomena—such as the combined effects of vertical acceleration on structures or the triggering of soil liquefaction—that are often neglected in single-axis testing. We think the implications of the new research are both immediate and far-reaching. In the near term, the design offers smaller research facilities a pathway to conduct high-quality seismic experiments that were once the preserve of large, well-funded laboratories. This opens the door to accelerate the validation of new engineering models, and generate data from regions where seismic hazards are acute but resources limited. From a scientific perspective, the system’s ability to maintain stability and accuracy under nonlinear conditions—such as the liquefaction of saturated sand—opens new avenues for investigating failure mechanisms that are both sudden and catastrophic in the field. Additionally, the fine control over motion, combined with minimal unintended rotations or displacements, ensures that test results are trustworthy and repeatable, a quality essential for refining numerical models and informing building codes. The capacity to replicate earthquake motions with high fidelity, even in irregular and scaled scenarios, also enhances the value of physical modelling as a complement to computational simulations, which still struggle to capture some of the soil-structure interaction.