⚡ electrical engineering

Centrifuge modeling of single pile axial response subjected to near-field pulse-like ground motion in dry sand

This study utilizes 50g centrifuge shaking table tests to demonstrate that near-field pulse-like ground motions induce significantly more severe nonlinear pile–soil interactions, including intense interface shear, irreversible settlement, and uneven axial force distribution, compared to non-pulse-like ground motions in dry sand.

Original authors: Hongshuai Liu, Jingchao Gao, Longjia Liu, Tongda Li, Qiang Pei

Published 2026-07-15

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Hongshuai Liu, Jingchao Gao, Longjia Liu, Tongda Li, Qiang Pei

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine a giant, super-strong spinning top that can make a tiny toy world feel as heavy as a real mountain. That's what scientists call a centrifuge, and that's exactly what Hongshuai Liu and his team used to see how a single building pillar (a "pile") behaves when the ground shakes.

They weren't just shaking the ground randomly. They wanted to see what happens when the earth gives a specific, dangerous kind of shove: a near-field pulse. Think of a normal earthquake like a long, bumpy car ride where the bumps are small but constant. A "pulse-like" earthquake is different; it's like someone suddenly yanking the car door open with a massive, single, high-energy jerk.

Here is what the team discovered when they tested a single metal pile buried in dry sand under these two different shaking styles.

The "Yank" vs. The "Rumble"

The researchers compared two types of shaking:

The Pulse (PLGM): A big, sudden, long-lasting yank.
The Non-Pulse (NPLGM): A steady, bumpy rumble without the big yank.

They found that the Pulse was a total game-changer. When the ground gave that big yank, the sand around the pile and the top of the pile itself started shaking much, much harder than they did during the steady rumble. In fact, at a moderate shaking strength, the pulse made the soil shake 2.65 to 3.40 times more violently than the non-pulse version.

The Resonance Riddle

Why was the pulse so much worse? It turns out the pulse matched the "natural song" of the pile and the sand. Imagine pushing a child on a swing. If you push at just the right moment (the rhythm), the swing goes super high. That's resonance.

The pulse-like motion had a long, slow rhythm that perfectly matched the pile-sand system's natural rhythm. This caused a massive resonance effect, amplifying the shaking at the top of the pile. The non-pulse motion, however, was like pushing the swing at random, fast intervals—it just didn't sync up, so the pile stayed relatively calm.

The "Sinking" Problem

The most dramatic difference showed up in how much the pile sank.

Under the Pulse: The pile started sinking almost immediately (around 5.5 seconds into the shake) and dropped fast. By the time the shaking stopped, the pile had sunk permanently by a lot. At the strongest shake (0.4 g), the pile settled by 19.319 mm.
Under the Rumble: The pile barely moved. It sank slowly and settled by only 4.362 mm at that same strong shake.

The pulse caused the sand to lose its grip and slide irreversibly, like a shoe slipping on a wet floor. The steady rumble just made the sand wiggle a little, but it mostly held its ground.

The "Tug-of-War" on the Pile

The team also looked at the forces pulling on the pile. Under the pulse, the top part of the pile (in the loose, upper sand) got hit with a massive, uneven load. It was like a tug-of-war where one side suddenly pulled with all its might, while the bottom of the pile stayed relatively relaxed. The force wasn't spread out; it was concentrated right at the top, creating huge stress.

In contrast, under the steady rumble, the force was spread out evenly along the whole pile, like a gentle, uniform hug. The pile didn't feel any sudden, dangerous spikes in pressure.

What They Didn't Find (And What They Didn't Say)

It's important to know what this study didn't do. They didn't test wet sand or water-saturated soil, so they didn't see if the pile would liquefy (turn into soup). They also didn't test a whole building with many piles; they only looked at a single, lonely pile.

The results are based on measured experiments in a lab, not just computer guesses. The team is very sure that the pulse-like motion is significantly more damaging to this specific setup than the non-pulse motion. They found that as the shaking got stronger, the pulse's bad effects got even worse, while the non-pulse motion stayed relatively tame.

The Bottom Line

If you are building on dry sand near a fault line, this study suggests you need to worry a lot more about those big, sudden "yanks" in the ground than the long, steady rumbles. The pulse doesn't just shake the ground; it locks into a rhythm with your building's foundation, causing it to sink deeper and feel much more stress than you might expect from a normal earthquake.

Technical Summary: Centrifuge Modeling of Single Pile Axial Response to Near-Field Pulse-Like Ground Motion

Problem Statement
Near-field pulse-like ground motions (PLGM) are characterized by high-energy, long-duration velocity and displacement pulses that induce severe seismic damage to geotechnical structures. While historical events (e.g., Lushan 2013, Gaziantep 2023) and prior research have established the destructive potential of PLGM on above-ground and underground structures, the specific mechanisms governing axial load transfer in single pile foundations under such conditions remain unclear. Existing studies have largely focused on non-pulse-like ground motions (NPLGM) or general structural responses, leaving a gap in understanding how velocity pulses specifically alter pile-soil interaction, axial force distribution, and residual settlement in dry sand deposits.

Methodology
To address this gap, the authors conducted 50g centrifuge shaking table tests on a single pile embedded in a double-layered dry sand deposit. The study utilized the following experimental design:

Scaling and Setup: A geometric similarity ratio of 1:50 was employed using the DCIE-40-300 centrifuge system. The model simulated a 26.5m prototype soil profile consisting of a 30cm loose fine sand layer ( $D_r=50\%$ ) overlying a 25cm dense coarse sand layer ( $D_r=80\%$ ).
Model Pile: A 6061 aluminum alloy tube (prototype scale: 1.2m diameter, 26.5m length) was instrumented with high-precision strain gauges along its shaft to measure axial force and skin friction. A steel counterweight was installed at the pile top to simulate superstructure inertial effects.
Ground Motion Inputs: The study employed the Baker pulse decomposition method to separate a typical near-field PLGM (Hualien earthquake record) into two components: the original pulse-like motion and a residual non-pulse-like motion (NPLGM).
Testing Protocol: Comparative tests were performed at three seismic intensity levels (PGA of 0.1g, 0.2g, and 0.4g) for both PLGM and NPLGM inputs. White noise sweeps were conducted before each loading case to monitor system frequency changes.
Instrumentation: The setup included accelerometers in the soil and on the pile top, a vertical LVDT for settlement, and strain gauges to derive axial force and shaft resistance distributions.

Key Results
The comparative analysis revealed distinct differences in the dynamic response and load transfer mechanisms between PLGM and NPLGM:

Soil and Pile-Top Acceleration: PLGM significantly amplified acceleration responses compared to NPLGM. At a PGA of 0.1g, peak accelerations in the surrounding soil were 2.65 to 3.40 times higher under PLGM. PLGM induced strong resonance effects, causing the predominant period of the soil and pile system to shift toward longer periods as intensity increased. Conversely, NPLGM produced stable, low-amplitude responses with a predominant period shift toward shorter values.
Inertial Effects: The long-period velocity pulses of PLGM matched the natural period of the pile-soil system, triggering significant resonance that amplified the inertial response of the pile-top mass. This resulted in pile-top peak accelerations up to 0.543g at 0.4g PGA, whereas NPLGM induced negligible inertial amplification.
Pile-Top Settlement: PLGM caused early initiation and rapid development of pile-top settlement, characterized by a sharp unimodal response. Residual permanent settlement increased non-linearly with intensity, reaching 19.319 mm at 0.4g PGA. In contrast, NPLGM resulted in delayed, gentle settlement evolution with negligible residual displacement (max 4.362 mm at 0.4g PGA).
Axial Force and Skin Friction: Under PLGM, axial force and shaft resistance were highly non-uniform, with severe amplification concentrated in the upper loose fine sand layer (0–12.5m depth). The high-energy pulses caused intense interface shear and normal stress fluctuations, leading to irreversible plastic deformation. Under NPLGM, axial forces remained low and uniformly distributed along the pile depth, with minimal sensitivity to seismic intensity.

Key Contributions

Experimental Evidence: The study provides the first centrifuge-based experimental evidence clarifying the distinct axial load transfer mechanisms of single piles under near-field velocity pulses versus non-pulse motions.
Mechanism Elucidation: It identifies that the frequency-energy distribution of the ground motion is the primary driver of response discrepancies. Specifically, the concentrated long-period energy of PLGM induces resonance and severe interface slippage in loose soil layers, which is not observed with the broadband, high-frequency energy of NPLGM.
Quantitative Data: The research offers quantitative data on the non-linear growth of residual settlement and the spatial non-uniformity of shaft resistance under varying seismic intensities, highlighting the disproportionate impact of PLGM on pile foundations in layered dry sand.

Significance
The authors state that this study clarifies the pulse-induced pile-soil interaction mechanism, which has previously been unclear. The findings provide essential experimental references for optimizing existing seismic design methods and revising seismic design codes for pile foundations in near-fault regions. By demonstrating that PLGM causes significantly more severe damage and residual deformation than NPLGM, the study underscores the necessity of accounting for velocity pulse effects in the seismic design of pile-supported structures.