Soil liquefaction experiments are earth science and disaster-prevention experiments designed to observe how sandy ground containing a large amount of water can temporarily behave like a liquid during earthquake-induced shaking, resulting in a reduction in the ground’s bearing capacity.
In the experiment, a transparent container filled with sand and water is shaken to observe phenomena such as the upward movement of water to the ground surface, sand boils, settlement of model buildings, and uplift of lightweight objects. Conditions that make liquefaction more likely can also be compared by changing factors such as the degree of sand compaction, groundwater level, particle size, and shaking intensity.
This article is intended for students observing liquefaction models in university and high school laboratory exercises. It explains the mechanism of liquefaction, what should be observed, comparisons of experimental conditions, how to organize results, how to write a discussion, possible sources of error, and ways to improve the experiment.
Note:
This article is intended as a reference for discussing results obtained in earth science and disaster-prevention experiments at universities, high schools, and similar educational institutions. Always follow the instructions provided in your school’s laboratory manual and by your instructor or teaching assistant regarding the actual experimental procedures, sand, amount of water, shaking apparatus, models, and other materials used.
- What Is Soil Liquefaction?
- What Should Be Observed in a Liquefaction Experiment?
- Basic Conditions That Make Liquefaction More Likely
- How Sand Particles Support the Ground
- What Is Pore Water?
- What Is Pore Water Pressure?
- The Concept of Effective Stress
- Basic Structure of a Liquefaction Model
- Basic Observations in a Liquefaction Experiment
- What Is a Sand Boil?
- What Can Be Inferred from a Sand Boil?
- Why Does a Model Building Sink?
- Why Does a Building Tilt?
- Why Do Lightweight Objects Rise?
- Why Does the Sand Surface Settle?
- Effect of Groundwater Level on Liquefaction
- Experiment Comparing Different Groundwater Levels
- Effect of Compaction on Liquefaction
- Experiment Comparing Loose and Compacted Sand
- Effect of Particle Size on Liquefaction
- Effect of Fine Particles
- Effect of Shaking Intensity on Liquefaction
- Effect of Shaking Duration on Liquefaction
- Experiment Comparing Different Shaking Durations
- Experiment Changing the Weight of the Model Building
- Experiment Changing the Base Area of the Model Building
- Example Experimental Conditions for a Liquefaction Experiment
- Example Observation Results from a Liquefaction Experiment
- Example of How to Write the Results
- Indicators for Determining Whether Liquefaction Occurred
- Possible Reasons Why the Model Did Not Settle
- When No Sand Boil Was Observed
- Why Model Settlement May Vary Between Experiments
- How to Keep the Degree of Sand Compaction Consistent
- How to Keep the Water Level Consistent
- How to Keep Shaking Conditions Consistent
- Differences Between a Liquefaction Model and Actual Ground
- Considering the Limitations of the Model Experiment
- Relationship Between Liquefaction Countermeasures and Experimental Results
- Liquefaction Countermeasures Using Compaction
- Liquefaction Countermeasures Using Drainage
- Liquefaction Countermeasures Using Ground Improvement
- Main Sources of Error and Observation Problems
- Example Discussion When the Experiment Did Not Work as Expected
- How to Write Improvements
- Difference Between a Weak and a Strong Discussion
- Example Discussion Comparing Groundwater Levels
- Example Discussion Comparing Compaction Conditions
- Example Discussion of a Sand Boil
- Example Discussion of Uplift
- Useful Expressions for Laboratory Reports
- Checklist for Discussing Liquefaction Experiments
- Summary
What Is Soil Liquefaction?
Soil liquefaction is a phenomenon in which loose sandy ground saturated with groundwater temporarily behaves like a liquid when subjected to strong earthquake shaking because the water pressure between the sand particles increases.
Under normal conditions, sand particles are in contact with one another, forming a framework that supports loads applied from above.
However, when loose sandy ground saturated with groundwater is repeatedly shaken, the arrangement of the sand particles tends to change. If the water between the particles cannot drain quickly enough, pore water pressure increases, reducing the force carried through contact between the sand particles.
As a result, the bearing capacity of the ground decreases substantially, which may cause buildings to settle or lightweight underground structures to rise toward the surface.
What Should Be Observed in a Liquefaction Experiment?
In a liquefaction experiment, it is important to compare conditions before, during, and after shaking.
Main Items to Record in the Results
- Type of sand used
- Particle size of the sand
- Thickness of the sand layer
- Amount of water
- Water level corresponding to the groundwater table
- Degree of sand compaction
- Shaking intensity
- Duration of shaking
- Position of the model before shaking
- Settlement of the model after shaking
- Changes in water level
- Upward movement of water to the ground surface
- Presence or absence of sand boils
- Settlement of the sand surface
- Uplift of lightweight objects
If possible, recording the conditions before, during, and after shaking with photographs or video makes it easier to compare the changes.
Basic Conditions That Make Liquefaction More Likely
Liquefaction does not occur in the same way in every type of ground.
In general, conditions that make liquefaction more likely include the following:
- The ground is sandy.
- The sand is relatively loosely deposited.
- The groundwater level is high and the spaces between sand particles are filled with water.
- The ground is subjected to sufficiently strong shaking.
- The shaking continues for a certain period of time.
In a model experiment, the susceptibility to liquefaction can be compared by changing these conditions one at a time.
How Sand Particles Support the Ground
In sandy ground that has not liquefied, sand particles are in contact with one another and transmit forces between them.
Loads from buildings and other structures are transmitted downward and outward through these contacts between sand particles.
Because this framework of sand particles is stable, the ground can normally support buildings and other structures.
What Is Pore Water?
There are small spaces between sand particles. These spaces are called pores, and the water contained within them is called pore water.
In sand layers below the groundwater table, the pores are generally filled with water.
This condition is called saturation.
What Is Pore Water Pressure?
The pressure exerted by pore water on the sand particles is called pore water pressure.
When repeated earthquake shaking occurs, loosely deposited sand particles tend to move toward a denser arrangement.
However, because the pore water cannot escape sufficiently within a short period of time, the water pressure may increase.
As pore water pressure rises, the force transmitted through contacts between sand particles decreases, reducing the strength and bearing capacity of the ground.
The Concept of Effective Stress
In geotechnical engineering, the concept of effective stress is used when considering the force actually carried through contacts between soil particles.
Even if the total applied stress remains the same, an increase in pore water pressure reduces the effective stress carried by the soil particles.
During liquefaction, this effective stress becomes extremely small, causing the sandy ground to lose shear resistance and behave in a liquid-like manner.
Basic Structure of a Liquefaction Model
In student experiments, a common method is to create a ground model by placing sand and water in a transparent container.
| Material or Equipment | Role |
|---|---|
| Transparent container | Allows observation of sand and water movement from the side |
| Sand | Represents sandy ground |
| Water | Represents groundwater-saturated conditions |
| Model building | Used to observe reduced bearing capacity and settlement |
| Lightweight object | Used to observe uplift of underground structures |
| Shaking apparatus or manual shaking | Provides a simplified representation of earthquake motion |
The materials and shaking methods used vary depending on the experiment. Follow the specified experimental procedure.
Basic Observations in a Liquefaction Experiment
The sand is thoroughly saturated with water, and a model building or similar object is placed on the sand surface before the container is shaken.
If liquefaction occurs, the model building may settle, water may seep out from the sand surface, or sand and water may erupt from the ground surface.
A lightweight object buried in the sand may also rise toward the surface.
What Is a Sand Boil?
When pore water pressure increases because of liquefaction, water may move upward toward the ground surface.
The phenomenon in which sand particles are carried upward with this water and discharged onto the ground surface is called a sand boil.
In a model experiment, turbid water and sand may appear to well up from part of the sand surface.
What Can Be Inferred from a Sand Boil?
If a sand boil is observed, it provides one indication that water pressure increased within the sand layer and that pore water moved toward the ground surface.
Because sand is transported upward from below at the location of a sand boil, the surrounding ground may settle or voids may develop.
Why Does a Model Building Sink?
Before liquefaction, the framework formed by the sand particles supports the load of the model building.
However, when liquefaction reduces the bearing capacity of the sandy ground, the ground may no longer adequately support the model building, causing it to sink into the sand.
During actual earthquakes, liquefaction can similarly cause settlement of buildings, roads, and other structures.
Why Does a Building Tilt?
If the ground liquefies uniformly, a model may settle almost vertically. In practice, however, ground conditions and load distributions are not perfectly uniform.
As a result, one side of a building may settle more than the other, causing the building to tilt.
This phenomenon, in which settlement differs from one location to another, is called differential settlement.
Why Do Lightweight Objects Rise?
In liquefied ground, the resistance of the sandy ground decreases and the water-rich sand behaves in a more fluid-like manner.
Hollow objects or objects with a lower density than the surrounding material may then rise because of buoyancy.
During actual liquefaction events, underground structures such as manholes and underground tanks may rise toward the ground surface.
Why Does the Sand Surface Settle?
Sand that was loosely deposited before shaking may undergo a change in particle arrangement during shaking and move toward a denser condition.
After liquefaction, when pore water drains away, the sand particles come into contact again and may become more densely packed than before shaking.
As a result, the overall volume of the sand layer may decrease and the ground surface may settle.
Effect of Groundwater Level on Liquefaction
For liquefaction to occur, it is important that the spaces between sand particles are filled with water.
When the groundwater level is high and the sand is saturated close to the ground surface, the effects of liquefaction are more likely to appear near the surface.
In contrast, if the groundwater level is low and the sand near the surface is unsaturated, liquefaction may be less likely to occur under the same shaking conditions.
Experiment Comparing Different Groundwater Levels
| Condition | Expected Observation | Direction of Discussion |
|---|---|---|
| High groundwater level | Settlement and sand boils are more likely to occur | The saturated sand layer is thicker |
| Low groundwater level | Changes may be smaller | The sand near the surface is unsaturated |
When comparing groundwater levels, keep other conditions such as the amount of sand, degree of compaction, model weight, and shaking conditions as consistent as possible.
Effect of Compaction on Liquefaction
Loosely deposited sand has an unstable particle arrangement and can more readily change into a denser state during shaking, so it tends to be more susceptible to liquefaction.
In contrast, well-compacted sand has particles that are already in closer contact with one another and therefore tends to have greater resistance to liquefaction than loose sand.
Experiment Comparing Loose and Compacted Sand
| Condition | Items to Observe | Direction of Discussion |
|---|---|---|
| Loose sand | Settlement, sand boils, changes in the sand surface | Particle arrangement changes more readily |
| Compacted sand | Compare the same observations | Greater resistance to liquefaction |
When comparing compaction conditions, keep the container size, water level, type of sand, model weight, shaking intensity, and other conditions the same.
Effect of Particle Size on Liquefaction
Susceptibility to liquefaction is affected not only by particle size but also by particle-size distribution, density, degree of compaction, fines content, permeability, and many other factors.
In a typical model experiment, relatively uniform sand may make the phenomenon easier to observe.
In contrast, very coarse gravelly materials or materials containing a large amount of clay may behave differently under the same model conditions.
Effect of Fine Particles
When sand contains fine particles such as silt or clay, the movement of water between particles and the state of contact between particles may change.
However, it is not appropriate to conclude simply that a higher fines content always prevents liquefaction.
Actual liquefaction assessments consider particle-size distribution, ground density, groundwater level, earthquake motion, and other factors together.
Effect of Shaking Intensity on Liquefaction
Under weak shaking, the arrangement of sand particles may not change substantially and liquefaction may not occur.
As shaking becomes stronger, particles are repeatedly disturbed and pore water pressure is more likely to increase.
However, shaking in a model experiment is a simplified representation of actual earthquake motion. The shaking intensity of the model cannot be directly equated with actual seismic intensity or ground acceleration.
Effect of Shaking Duration on Liquefaction
Even at approximately the same shaking intensity, liquefaction may not occur if the shaking duration is short, while changes may become apparent when shaking continues for a longer period.
This is because repeated shaking gradually changes the arrangement of the sand particles and the pore water pressure.
In comparative experiments, shaking intensity and shaking duration should be recorded separately.
Experiment Comparing Different Shaking Durations
| Condition | Items to Observe | Direction of Discussion |
|---|---|---|
| Short-duration shaking | Model settlement and sand boils | Liquefaction may not occur |
| Long-duration shaking | Compare the same observations | Effect of repeated shaking |
Experiment Changing the Weight of the Model Building
Even on the same liquefied ground, the amount of settlement may vary depending on the weight and base area of the model.
A heavier model applies a greater load to the ground and may therefore settle more readily when liquefaction reduces the ground’s bearing capacity.
In addition, even if two models have the same weight, differences in their base areas result in different pressures being transmitted to the ground.
Experiment Changing the Base Area of the Model Building
Even when models have approximately the same weight, a model with a smaller contact area applies a greater average pressure to the ground.
Therefore, when comparing differences in settlement, record not only the weight of the model but also its base area.
Example Experimental Conditions for a Liquefaction Experiment
| Item | Example |
|---|---|
| Container | Transparent plastic container |
| Ground material | Sand with a relatively uniform particle size |
| Water level | Sand saturated to approximately the level of the sand surface |
| Model | Small model building or similar object |
| Shaking method | Repeated shaking in a fixed direction |
| Observations | Settlement, sand boils, water level, sand surface, and uplift |
The conditions above are examples for organizing observations. Follow the laboratory manual for the actual experimental conditions.
Example Observation Results from a Liquefaction Experiment
| Observation Stage | Observation Result | Direction of Discussion |
|---|---|---|
| Before shaking | The model was stable on the sand surface | The sand-particle framework supported the load |
| After shaking began | Water seeped out from the sand surface | Possible increase in pore water pressure |
| During continued shaking | The model gradually settled | The bearing capacity of the ground decreased |
| During continued shaking | Sand and water rose in some areas | A phenomenon corresponding to a sand boil |
| After shaking | The overall sand surface became lower | The sand may have become more densely packed |
Example of How to Write the Results
Before shaking, the model building remained stable on the sand surface. When repeated shaking was applied to the container, water began to rise from part of the sand surface, after which the model building gradually settled into the sand. In some areas of the sand surface, water containing sand was observed moving upward toward the surface.
After shaking stopped, the sand surface was lower than before shaking, suggesting that the sand particles may have changed to a denser arrangement.
Indicators for Determining Whether Liquefaction Occurred
In a model experiment, liquefaction should not be identified simply because water moved. Instead, several phenomena should be considered together.
- The model building settled.
- Water rose to the ground surface.
- A phenomenon resembling a sand boil was observed.
- The sand surface settled.
- A buried lightweight object rose upward.
- The bearing capacity of the sandy ground decreased during shaking.
Possible Reasons Why the Model Did Not Settle
If the model hardly settled during a liquefaction experiment, the sandy ground may not have been sufficiently saturated.
Other possible causes include excessive compaction of the sand, weak shaking, insufficient shaking duration, or a model that was too light.
Even if settlement did not occur, do not simply describe the experiment as a failure. Instead, check systematically whether the conditions required for liquefaction were present.
When No Sand Boil Was Observed
Even if liquefaction occurs, a clearly visible sand boil does not necessarily appear at the model surface.
If water drains slowly through the sides of the container or across the entire sand layer, localized upward discharge may not be noticeable.
Therefore, it is important not to determine whether liquefaction occurred solely from the presence or absence of a sand boil.
Why Model Settlement May Vary Between Experiments
Even when the same experiment is repeated, differences in sand packing, water level, model position, and shaking intensity may cause variations in settlement.
Because the degree of sand compaction has a particularly strong influence on the results, comparative experiments should use consistent methods for placing and compacting the sand.
How to Keep the Degree of Sand Compaction Consistent
If the sand is placed in the container using a different method each time, its initial density will vary.
In comparative experiments, use the same mass of sand and keep the pouring height, number of additions, compaction procedure, and other conditions as consistent as possible.
In rigorous experiments, dry density and relative density may be measured, but school laboratory exercises may use simplified procedures depending on the purpose of the experiment.
How to Keep the Water Level Consistent
In experiments that are not intended to compare groundwater levels, it is important to add water to the same height each time.
Recording the position of the water surface relative to the sand surface, or the height of the water level measured from the bottom of the container, makes experimental conditions easier to compare.
How to Keep Shaking Conditions Consistent
When the container is shaken by hand, it is difficult to keep the intensity and number of shakes completely consistent.
If a shaking table or similar apparatus is available, reproducibility can be improved by keeping the frequency, amplitude, and duration of shaking constant.
Even when shaking manually, it is important to use a consistent duration and approximate number of shaking cycles and to record the conditions.
Differences Between a Liquefaction Model and Actual Ground
A model experiment greatly simplifies actual ground conditions in order to observe the basic mechanism of liquefaction.
In actual ground, many factors interact in complex ways, including the thickness and depth of sand layers, particle-size distribution, groundwater level, earthquake motion, geological structure, and building foundations.
Therefore, conditions that produce liquefaction in a model cannot be directly applied to evaluate the liquefaction risk of an actual site.
Considering the Limitations of the Model Experiment
The model container is extremely small compared with actual ground, and its side walls affect the movement of sand and water.
In addition, the sand particles themselves are the same size as real sand particles, so a model in which only the container is reduced in size does not have complete geometric similarity.
The shaking method also differs from actual earthquake motion. Therefore, the model experiment should be treated as a model for understanding the basic mechanism of the phenomenon.
Relationship Between Liquefaction Countermeasures and Experimental Results
Methods for reducing conditions that promote liquefaction include compacting the ground, solidifying the ground, lowering the groundwater level, and making it easier for excess water pressure to dissipate.
Comparing loose and compacted sand or changing the water level in a model experiment can help demonstrate the principles behind these countermeasures.
Liquefaction Countermeasures Using Compaction
When sandy ground is compacted, the sand particles become more densely arranged, leaving less room for further contraction during shaking.
As a result, the ground can have greater resistance to liquefaction than loose sandy ground.
Liquefaction Countermeasures Using Drainage
An increase in pore water pressure during an earthquake is an important factor in liquefaction.
Therefore, methods that make water easier to drain or lower the groundwater level can help reduce the occurrence of liquefaction.
Liquefaction Countermeasures Using Ground Improvement
Another method is to solidify the ground using cement-based materials or similar substances so that the sand particles are bound together.
In such improved ground, particles cannot move as freely as they can in loose sandy ground, increasing resistance to liquefaction.
Main Sources of Error and Observation Problems
| Cause | Effect on Results | Possible Improvement |
|---|---|---|
| Different sand packing conditions | Liquefaction susceptibility changes | Standardize the sand placement and compaction methods |
| Different water levels | The saturated zone changes | Measure and standardize the water level |
| Different shaking intensities | Settlement and time to liquefaction change | Standardize the shaking conditions |
| Different model positions | Settlement varies | Place the model in the same position |
| Different model weights | Susceptibility to settlement changes | Use the same model |
| Air bubbles remain in the sand | The sand does not become fully saturated | Ensure that the sand is thoroughly saturated with water |
Example Discussion When the Experiment Did Not Work as Expected
Even after shaking was applied, little settlement of the model building or sand boiling was observed. Air may have remained within the sand layer, preventing it from becoming fully saturated. In addition, the sand may have been strongly compacted when it was placed in the container, producing dense ground that was less susceptible to liquefaction.
Furthermore, because the container was shaken manually, the shaking intensity may have been too low to provide sufficient repeated shearing.
How to Write Improvements
When discussing improvements, do not simply write “shake more strongly” or “perform the experiment more carefully.” Instead, describe specifically how the conditions affecting liquefaction should be standardized.
How to Standardize Ground Conditions
- Use the same type and mass of sand.
- Standardize the method used to place the sand.
- Use the same number of compaction operations.
- Add water to the same level.
- Reduce the amount of trapped air in the sand layer as much as possible.
How to Standardize Shaking Conditions
- Keep the shaking duration constant.
- Keep the number of shaking cycles constant.
- Use a shaking apparatus if possible.
- Record the shaking amplitude and frequency.
- Record the time from the beginning of shaking until liquefaction-related phenomena appear.
Difference Between a Weak and a Strong Discussion
Example of a Weak Discussion
The building sank when the sand was shaken, so liquefaction occurred.
Example of a Better Discussion
When repeated shaking was applied to loose sandy ground saturated with water, water rose from the sand surface and the model building gradually settled. As the shaking caused the arrangement of the sand particles to change, pore water could not drain sufficiently and the water pressure increased. This reduced the effective stress between the sand particles, causing the ground to lose bearing capacity.
Organizing the discussion in the sequence “observed phenomenon → changes that occurred within the ground → relationship to the reduction in bearing capacity” makes the reasoning behind the interpretation clearer.
Example Discussion Comparing Groundwater Levels
In the sample with the groundwater level raised close to the sand surface, settlement of the model and upward movement of water were observed, whereas the sample with a lower groundwater level showed smaller changes.
In the sample with the higher groundwater level, the pores between sand particles were filled with water closer to the surface, so the effects of increased pore water pressure during shaking may have extended closer to the ground surface.
Example Discussion Comparing Compaction Conditions
The model settled substantially in loosely packed sand, whereas settlement was smaller in compacted sand.
In loose sand, there was more room for the particles to move into a denser arrangement, making pore water pressure more likely to increase during shaking. In contrast, the particles in the compacted sand were already densely arranged in the initial condition, giving the ground greater resistance to liquefaction.
Example Discussion of a Sand Boil
During shaking, turbid water and sand rose from part of the sand surface. This phenomenon is considered to correspond to a sand boil, in which increased pore water pressure within the sand layer causes groundwater to move toward the surface while carrying sand particles upward.
Example Discussion of Uplift
The lightweight model buried in the sand moved upward after shaking. Liquefaction reduced the resistance of the ground and caused the water-rich sand to behave in a more fluid-like manner, increasing the relative influence of buoyancy acting on the model.
Useful Expressions for Laboratory Reports
- After shaking began, water was observed rising from the sand surface.
- The model building gradually settled as repeated shaking continued.
- Sand and water rose to the ground surface, and a phenomenon corresponding to a sand boil was observed.
- After shaking, the sand surface was lower than before shaking.
- The increase in pore water pressure is considered to have reduced the effective stress between the sand particles.
- The model building is considered to have settled because the bearing capacity of the ground decreased.
- The effects of liquefaction were more apparent near the surface under conditions with a higher groundwater level.
- Settlement was smaller in compacted sand than in loose sand.
- Differences in shaking intensity and duration may have affected the results.
- The side walls of the model container may have affected the behavior of the sand and water.
- The model experiment is a simplified representation of actual ground conditions and cannot directly evaluate the liquefaction risk of an actual site.
Checklist for Discussing Liquefaction Experiments
- Can you explain why liquefaction is more likely to occur in loose sandy ground saturated with water?
- Can you explain the roles of sand particles and pore water?
- Do you understand the relationship between pore water pressure and effective stress?
- Did you relate settlement of the model building to the reduction in ground bearing capacity?
- Did you explain why sand boils occur?
- Did you explain why lightweight underground structures may rise?
- Did you consider differences in groundwater level?
- Did you consider differences in the degree of sand compaction?
- Did you record shaking intensity and duration?
- Did you discuss variations in experimental conditions?
- Do you understand the differences between the model and actual ground?
- Did you relate liquefaction countermeasures to the experimental results?
Summary
In a liquefaction experiment, shaking loose sandy ground saturated with water makes it possible to observe phenomena such as settlement of a model building, upward movement of water to the ground surface, sand boils, settlement of the sand surface, and uplift of lightweight objects.
Liquefaction occurs when shaking causes sand particles to move toward a denser arrangement while pore water cannot drain quickly enough. The resulting increase in pore water pressure reduces the effective stress carried through contacts between sand particles.
By comparing factors such as groundwater level, degree of sand compaction, particle size, shaking intensity, and shaking duration, the experiment can be used to discuss the conditions under which liquefaction is more likely to occur.
In a laboratory report, rather than simply stating that “the model sank,” it is important to connect the observations to what occurred within the ground, explain why the bearing capacity decreased, and discuss why differences in groundwater level and compaction conditions produced different results.
