From CGMWiki Vacuum method

The basic ideas of the vacuum saturation method are presented here. There are many variations on the method, and many simplifications can be made depending upon the application.

Tools and equipment

Figure 1 is a schematic drawing of the vacuum saturation method. For saturating a sand model by the vacuum method, the following tools and equipment are suggested:

A water tank capable of holding vacuum

Pneumatic and water utilities (de-ionized water is recommended)

A model container capable of holding vacuum

A source of Carbon-dioxide (CO2)

Two adjustable vacuum sources with vacuum gauges (a vacuum generator using compressed air and the Bernoulli Principle has been found to be a cheap, controllable vacuum source.)

Procedure

Several techniques are available for saturating sand in centrifuge model tests. The following is the most widely employed procedure.

Remove the dissolved air from a supply of water (preferably de-ionized water) by spraying the water into the water tank under vacuum.

Remove the air from the model container by applying vacuum to the container. An absolute pressure of about 10kPa has been found to be sufficient to ensure that the small amount of gas left in the soil would dissolve during step 5.

Relieve the vacuum in the container by slowly supplying CO2 gas. As it is almost impossible to remove all the air from the soil in step 2, the CO2 is added to displace the less soluble air in the voids.

Re-apply vacuum to the sample.

Saturate the sample with de-aired water. As explained below, there are advantages to adding water to the top of the model container. Flow can be controlled by adjusting the vacuum level in the water tank relative to the vacuum level in the container, or by using identical vacuum pressures in each container and adjusting the height of the reservoir.

The final fluid level should be higher than the top of the soil surface. Release the vacuum slowly. Monitor the fluid level and continue to add fluid as necessary.

Recent Additions to Procedure

For saturating sand in centrifuge model tests, several saturation techniques are available. One of the most widely employed method is the vacuum method described above. One commonly arising issue with this method is the time it takes for saturation. The rate of saturation is highly variable and a solution to this problem was in need to be found. The following recommendations are made from Saturation Investigations:

Increase Saturation tube diameter size.

Insert ventilation tubes in the top portion of the interface between different soil.

Discontinue use of the Aquarium Grade Monterrey Sand, instead using the finer #0-30 Monterrey sand to surround the geotextile filter.

Use Dish Soap as a solute in the solution that will be saturating the sample.

Precautions

Working with pressure and vacuum in large volumes can be extremely dangerous. When designing a vacuum saturation method, relief valves and caution may be used to ensure that safe procedures are being used.

Of all the procedures in centrifuge model building, the vacuum saturation method is perhaps the most dangerous step as far as potential disturbance to the sample. Some typical problems are outlined here.

The soil may not be 100% saturated after adding fluid. Releasing vacuum will force fluid into any drier voids. If you release the vacuum too quickly or if air is allowed to contact the surface of the soil then air may be allowed into the void space rather than fluid.

The use of inadequately de-aired water. If the vacuum in the model container is greater than the vacuum in the water tank, or if air is allowed into the water during transfer (e.g. a leak in the water line), then the dissolved air will be pulled out of solution in the sample container. If the water is in the soil, the expansion of air bubbles will bulk the soil, destroying the sample. Dripping the water into a reservoir inside the model container eliminates this risk by exposing the water to the container vacuum before it enters the soil.

Increasing the magnitude of vacuum in the model after water has entered the soil. This could cause bubbles to come out of solution and, possibly, bulk up and loosen the soil.

Applying vacuum from above too fast. This can boil the sand due to upward gradient of air pressure.

Relieving vacuum too fast. Adding CO2 too rapidly can blow holes in the surface of the sand.

In some situations the water and/or CO2 is added directly to the bottom of the soil container. In this case, extreme care should be taken not to boil the soil due to gas or water flow.

When the soil container is under vacuum, the container may tend to deform inward under atmospheric pressure. Repeated application of vacuum, or use of a container that lacks the lateral stiffness, can result in significant disturbance to the soil sample.

Pressure relief should be included to avoid dangerously over-pressurizing the soil container with CO2.

Saturating the sample from the top will cause Gravity Fingers which is more prone to saturation voids than saturating from the bottom of the sample.

Verification of saturation

A high degree of saturation can be confirmed by measuring p-wave velocity after saturation. The p-wave velocity of the soil/water mixture will quickly approach the p-wave velocity of water alone (~1500m/s) as the degree of saturation approaches 100%. See for example Gazetas (1991) As the relationship between p- wave velocity and degree of saturation is very steep, it is not necessary to measure the p-wave velocity with great accuracy. Simply ensuring the velocity is high (~1000m/s) should be sufficient. For a p-wave velocity test, all that is needed is an impact wave source, two accelerometers oriented vertically in the soil at different elevations directly beneath the wave source, and a high speed data acquisition system. The procedure consists of striking the surface of the soil (a vertical hammer hit onto a block of wood resting on the ground surface) and measuring the travel time of the p-wave pulse between the two accelerometers.

References

Gazetas (1991) [Ch. 15 in Foundation Engineering Handbook, Fang ed, Van Nostrand Reinhold, New York.].

Figures

Open

Figure 1: Vacuum saturation method example setup