Actual test: Stress dependent behavior of sand slope failure.

Most recent publications:

Allersma H.G.B., 2001, Some applications of imaging technologies in experimental geotechnics. Bauingenieur, Springer VDI Verlag, no:76, pp.204-207.

Allersma H.G.B., 2001, Optical analysis of stress and strain around a penetrating probe in a granular medium. 4th International Conf. on Micromechanics of granular material, Sendai, May 21-25, pp.85-88

Allersma, H.G.B., A.A. Kirstein, R.B.J. Brinkgreve, 2001, Centrifuge and numerical modelling of suction pile behavior under various loading conditions. XVth Int. Conf. for Soil Mech. And Foundation Eng., August 27-31, Istanbul, pp.1327-1330.

Lee, S.W., Y.N. Lee, H.G.B. Allersma, 2001, A study on replacement behaviour of soft soil by centrifuge modelling tests. The Journ. of Korean Geotechnical Eng. Society, Vol.16, No.6, pp. 141-151. (in Korean language).

Allersma, H.G.B., A.Rohe, 2001, Centrifuge tests on the failure mechanism of dykes. Proc. 14th Southeast Asian Geotechnical Conference, Hongkong, 10-14 Dec 2001, pp.1077-1081.

Personalia

Name:

Dr. Henderikus G.B. Allersma

Position:

Associate Professor

Address:

Delft University of Technology
Faculty of Civil Engineering
Geotechnical Laboratory
Stevinweg 1
2628 CN Delft
The Netherlands

Telephone:

+31-15 278 3326

Fax:

+31-15 278 3328

Email:

H. Allersma@CT.TUDelft.nl

Specialization

Development and application of measuring and test techniques in experimental geotechnics

centrifuge modelling; optical stress strain analysis; computer soil testing

Start Centrifuge TUDelft by Dialling Telephone Number

The stress dependent behaviour of soil can be modelled in a correct way in a centrifuge. It is therefore possible to cause failure of a vertical cut or a sheet pile wall in a small clay sample. Thanks to the cohesion, failure would never occur in a normal gravity field. If, however, the model is placed in a centrifuge the self weight of the material can be increased so that the shear stresses in a small model can be brought to the same level as the cohesion stresses. Also, sand behaves stress dependent because the interparticle friction is larger at a low stress level. This is caused by the fact that at higher interparticle loads some local crushing takes place at the particle surface, resulting in a lower internal friction angle. In this figure load displacement diagrams of a circular footing on Dune sand at different gravities are shown. The load is devided by the actual acceleration. It appeared that the bearing loads are relative lower at larger gravities. It means that small scale tests at 1g result in too optimistic values. On the other hand it can be seen that the influence of the gravity is not so strong at higher stress levels.

A small geocentrifuge (for movie click here) has been designed and built at the Geotechnical Laboratory of the University of Delft. The radius is 1.20m and the space for samples and equipment is 400x400x500mm. Samples with a weight of 300N can be accelerated up to 160g. Due to the small size the cost of operation is low and a large number of test can be performed in a short period. In general the weight of the sample boxes and in flight testing equipment is so low that they can be transported by hand by one person.

Automated preparation devices assure a good reproducibility of the clay diagram, (photo)and sand (photo) samples. It appears possible to make an air free clay slurry under normal atmospheric conditions. An additional centrifuge has been built to prepare normaly consolidated clay samples. The computer controlled sand preparation device produces sand samples with a standard deviation in porosity of less than 0.2%. Due to the accuracy slight differences in tests procedures can be made visible.

Several accesoires are available to perform advanced in flight tests, such as i.e.
- sand pouring machine (photo, diagram)
- two dimensional loading system (photo, diagram)
- gas supply system (photo, diagram)
- water circulation system (system 1, system 2)
- vane apparatus (photo)
- pile driving hammer (diagram)
- in flight excavation (diagram))
- wave simulator (diagram)
- simulation suction pile installation (photo, diagram)
- simulation pollution behaviour (photo)
click here for more details

The devices are controlled by a PC compatible single board computer which is located in the spinning part of the centrifuge (486, 66MHz processor, 16 Kbyte RAM, 32 Mbyte solid state disk, 1 Gbyte HD just in the center). At this moment (2000) much more powerfull single board computers are available. However, for our control applications the 66MHz system is fast enough.
The computer is accessible in a normal way via slip rings and line driver units. The communication with the on-board mechanical equipment is realized by commercial available AT-slot interface boards. One board contains 8 analog inputs, one voltage controlled output and 32 digital input/output lines (see diagram). Because the measuring and control signals have not to be transported via slip rings the noise is reduced to a very low level. The interface board can be accessed very easily in a BASIC program, so that the user can write his own software to define a particular test procedure.
The limited space for sensors is neutralized by using digital image processing techniques to digitize automatically and on line surface displacements of clay and sand samples. The displacement of objects can be measured also by this technique.

In recent years several research projects have been carried out, such as for example: stability of dikes, widening of embankments (for movie click here) (prototype height 6m), sliding behavior of spudcans (prototype diameter 14m), stability of embankments during water infiltration (for movie click here), simulation of gas blowouts (for movie click here), simulation of land subsidence (prototype depth 70m) (for movie click here), buckling behavior of large diameter piles (prototype diameter 2.6m; lenght 35m), behavior suction foundations (prototype dimensions diameter=9m, height=10m) during installation (for movie click here), etc., etc.


Thanks to its convenient and interactive operational modes, the centrifuge has proven to be very suitable for educational purposes. On the other hand, the geocentrifuge is powerfull enough to investigate several types of practical problems.
Because of the relative low cost of operation, the number of tests in a project (30-80) are not limited by financial aspects. For this reason the centrifuge is extremely well-suited for mechanism studies. In alsmost all the projects new and unexpected phenomena are observed.
Up to now it was good possible to combine the research work of graduate students with projects which are financed by companies. In recent years more than 20 students (also from abroad) have been involved in centrifuge projects. Thanks to the financial support of companies little money was needed from the University to build the equipment. The research projects have also been resulted in several publications.

Students who follow the course Experimental Geotechnics are performing a short research project. They are free in the choice of the subject. Before they are doing the tests they are making some calculations and they should write a proposal. In many cases they build simple equipment by them selves. Finally they have to write a report containing background tests, test results, evaluation and conclusions.

Short description actual and recent projects:

(see publications for more details)

Simulation of gas blowouts (SHELL EXPRO UK))

- In this test program the behaviour of a soil layer was observed during an external gas blowout. Homogeneous sand samples and samples of sand with clay layers were used. In the centrifuge a blowout was simulated by injecting air at the bottom side of the sample. The main goal of the research was to find a relation between gas flow and crater depth. This is a typical example of a problem which can only be solved by experiments, because there is not any theory available to describe this complicated process.
Since the centrifuge has no rotating joints for gas high pressure cylinders are used to store 2000 liters of air on board. The required gas flow can be supplied to the test box by a computer controlled air supply system.
A relation could be found between the energy of the gas and the crater depth. Furthermore several other interesting phenomena could be observed. It appeared that cratering occurs only when the gas flow is started suddenly. However, if there is a large diameter shallow footing at the surface, then also a crater is formed at a gradual increase of the gas flow. Initially it was believed that a clay layer limits the depth of the crater. The centrifuge tests show, however, that the clay layer causes a crater because the gas is accumulated in the first instance under the clay and finally break through suddenly. This process shows a similarity with liquefaction during earth quakes.

Failure of dikes by water infiltration (Rijkswaterstaat)

- The behaviour of dikes are investigated during wave overtopping. A typical configuration of the dike is a sand body covered by a clay layer with a thickness of ca. 0.5 meter. During wave overtopping water is infiltrated into the dike. The rising ground water table causes finally failure. The water infiltration in flight could be achieved in a simple and cheap way by means of an air lift.
In contradiction with the general opinion failure is not following the shear band theory of Bishop or Fellenius. It appeared, however, that the rising ground water table is tilting the clay layer. Since the friction with the underground is reduced to zero the clay cracks due to its own weight. After cracking the sand body erodes gradually. A field test has shown that this mechanism is also active in reality.
Due to the centrifuge tests the field test has been observed more critically, so that the real failure mechanism could be visualized.
In a later test program in the centrifuge the stream lines are made visible, so that the effect of drains could be visualized. It appeared that a drain improves the stability significantly. As an answer on critical notes that a real dike is not homogeneous, clay lenses are inserted into the sand. It appeared that clay lenses has not always a negative effect on the stability.

Sliding behaviour of spudcans on sand (SHELL Int. Explotation & Prod.)

- For jack-up units founded offshore upon sand by means of spudcans, sliding of the windward leg is often the governing foundation criterion for a site specific integrety assessment. At this moment centrifuge tests is the only possibility to get more insight in the sliding behaviour of spudcans. A typical output of a test series is shown here.
Several interesting phenomena could be observed in the centrifuge tests. It appeared that more than full embedment penetration was needed to reach the maximum bearing capacity. A spicular tip decreases the bearing capacity somewhat, where a rough spudcan has a higher bearing capacity than a smooth footing. For details see this graph
Test on the sliding capacity have shown that the first sliding check (H = V tg a) as described in the Recommanded Practice for Site Specific Assessment of Mobile Jack-Up Units is applicable to restricted loading conditions and may therefore be too optimistic. Hence, it is concluded that this check should not be used when assessing the integrety of the windward leg foundation. Much better fit is obtained by more advanced theoretical models, e.g. based on Brinch Hansen's theory.

Simulation of the trap-door mechanism (no support)

- The trap-door is a representative deformation mechanism for several phenomena. It can be related to tunneling, land subsidence and geological phenomena. In the Netherlands land subsidence as a result of gas production is measured in the province of Groningen.
A simple test technique has been developed to simulate the trap-door mechanism in the centrifuge. Several configurations are tested, with homogeneous and layered soils. By comparing 1g tests and 100g tests it could be shown very clear that the stress level has a great influence on the deformation mechanism. It appeared that a clay layer has little influence on the deformation of the sand.
By inserting a rocky layer the cracking behaviour could be examined. Sudden deformations could be observed, when the rocky layer is undermined. This can be a mechanism which causes light earth shocks during gas production.

Simulation of flowing hoppers (no support)

- A number of orientating tests have been carried out to visualize the flow in hoppers at higher g levels. It could be visualized that the flow pattern is strongly dependent on the stress level. If there is thunnel flow or mass flow is not only dependent of the hopper shape but also influenced by the relative density of the material and the stress level.

Buckling behaviour large diameter piles (SHELL Int. Exploration & Prod.)

- In practise it was observed in some cases that hollow cylinder piles are deformed significantly during driving. The deformation starts at the tip (peanut shape) and can be extended over a lenght of more than 20 meter. Due to this deformation a serious problem arises, for example, if an insert pile has to be installed. The aim of this project is to investigate which mechanism is effecting buckling. Centrifuge tests as well as 1g test in a callibration cell are carried out. For centrifuge testing a pile driving hammer has been developed which can operate at 150g. After driving the pile can be loaded with the computer controlled in flight loading system.
The tests have shown clearly that an initial injury is required to deform the pile during driving. It was found that the buckling has no negative effect on the

The behavior of new embankments on soft soil (no support)

- When a new embankment is constructed on soft soil a part of the dumped material will be penetrated into the subsoil. Some orientating tetst have been performed in the centrifuge to simulate this process. Some stages during the sand suppletion is shown in the figures 1, 2, 3, 4. The deformations can be measured more in detail by image processing.
It is believed that centrifuge tests can yield valuable information about the mechanism of this process. The stiffness of the clay can be varied and the effect of stabilizing techniques e.g. sand layer or geotextile can be examined. In this movie an embankment is made in flight, where the clay surface was first stabilized by a thin layer of sand.

The behavior of embankments during widening (Rijkswaterstaat)

- Due to increasing traffic several roads, founded on soft soil, have to be widened. During widening of the sand embankment, cracks can occur in the asphalt pavement. An important question is if cracking can be reduced by changing the procedure of widening. A computer controlled sand pouring machine has been developed to enable manipulation of the widening procedure in flight in a centrifuge test.
The main objective of this research program is to compare the results of centrifuge tests with calculation methods. The advantage of centrifuge models is, that a relative simple configuration can be tested in the first instance. A good comparison is possible only if the soil parameters are well known. Therefore a lot of attention is paid to investigate methods in order to determine soil parameters at 1g and in soil layers during flight.
The tests have shown that cracks occur due to a gradient in the vertical displacement.

Making of embankments on soft soil by dumping method (Hyundai Inst. of Constr. Technology, Korea)

- The aim of this project is to study the behaviour of soft sub soil during dumping of granular material for making embankments. Several aspects of which is believed that they will influence the process will be examined. Important are the relations between shape of the embankment and dumping strategy, particle size and the use of geotextile. Also different clay types are used, under which a clay from the shore of Korea. Click here (detail) to see an example of an embankment made in flight. In this test the embankment is poured layer by layer. In order to digitize the shape of the embankment automatically a computer controlled device has been developed. An example of a scanning pattern is shown here. It is already clear that the strategy of dumping is of influence on the shape of the embankment.

The behavior of suction cans in layered soil (Clyde Petroleum Expl. b.v., Fugro Engineers b.v.)

- Suctions cans are used offshore as anchoring elements for floating platforms or as foundation element for several different applications. A large can (i.e. diameter 9m, height 10m) is placed up side down at the sea floor. By pumping water out of the can, water is flowing through the soil, so that the soil stress around the tip of the can is reduced. Due to the reduction of the soil stress and the low pressure inside the can the can is penetrating into the soil. The aim of this project is to investigate the possibility of installation in layered soil (sand, clay, sand). It was in particular uncertain what will happen in case a can penetrates from a clay layer into a dense sand.
Centrifuge tests have shown up to now that a clay layer did not hamper the penetration of a suction pile. Recently real scale suction cans are installed in layered soil, similar to the centrifuge samples. It appeared that also in this case it was possible to install the cans without heave of the clay layer.

Simulation of suction can installation in sand (no support)

- In this project it is analysed how suction cans behave during installation in homogeneous sand. The installation of suction cans is based on a vertical load caused by pressure difference and friction reduction thanks to groundwater flow, driven by the pressure difference. Load displacement diagrams during installation can be derived, so that the different driving and resistance components can be separated. Cans with different heights and diameters can be used, were also the wall thickness is varied. Furthermore the bearing capacity during static and dynamic loading will be analysed.
A test technique has been developed to simulate the installation of a can by suction at 150g in the centrifuge. By performing the test in a pressurized container (photo, diagram), also the water depth could be simulated in a correct way.
During a test the significant parameters are measured. It was found that there is a linear relation between the pressure difference, the wall thickness, the lenght and the diameter of the suction can.
If the diameter become too small upheave occur. If the can is penetrated mechanically a 7 times higher load is required than penetration by suction. The can tends to a vertical position during suction installation, even when the initial position is not straight.

Horizontal bearing capacity of suction piles (Suction Pile Techn., v.Hattum en Blankevoort)

- In this research project centrifuge tests are performed to examine the horizontal bearing capacity of suction piles. Parameters which have been tested are: H/D ratio, loading angle, attachment height cable, density sand, surface roughness pile. About 40 centrifuge tests have been performed to examine the influence of the different parameters.
The test results have been compared with a three dimensional FE calculation. It was found that there was a good agreement with FE calculation and tests. Calculations as well as tests show that attachment of the cable at 2/5 from the bottom yields the maximum bearing capacity. This position depends somewhat on the angle of the cable. Increasing the diameter of the pile is much more effective than using strips. A spectacular result was that the horizontal bearing capacity increases significantly if 1/3 of the top was removed after installation. This was the case in sand as well as in clay.

Cyclic bearing capacity of suction piles (Suction Pile Techn., v.Hattum en Blankevoort)

- In this research project centrifuge tests will be performed to investigate the loading speed and time dependent behaviour of vertically loeded suction piles. It is believed that suction piles can yield a significant higher pull out load in the case of fast loading (i.e. due to wave action). Another unknown point is in how far long term (cyclic) loading is influencing the bearing capacity.

Suction caisons for stabilization of the Jumbo offshore installation ship (Kahn Scheepvaart b.v.)

- In this research project centrifuge (Fig.1, Fig.2) and 1g tests will be performed to investigate the vertical pull out resistance of suction caisons. Caisons of different shapes are used in this case. The caisons will be used as anchors to stabilize temporarily super heavy lift ships.
Points of interest are:
= influence of compartiments on bearing load and installation
= influence of permanent suction
= influence of wall shape

Installation of suction piles by percussion method (MATech)

- In this research project a new installation technique of suction piles has been investigated. In this technique the trapped water will not be removed by continuous pumping. Instead of that a vessel will be brought to under pressure. By opening a valve suddenly the trapped water will be removed with large capacity under maximum pressure difference. It is believed that this technique will be valuable in the case that the sea floor is not flat, in course sand and in the case of slender piles.
In the first instance 1G tests have been performed on model piles with dimension D=65mm H=80mm. A computer controlled test setup has been developed to generate the short pressure pulses and to measure the significant paramaters such as pressure during pulse and displacement. The tests show the tendensy that percussion technique has advantages if the conditions are less ideal. Some conclusions are:
upheave reduces with shorter pulse time (example in course sand)
penetration in uneaven sand surface only possible by percussion method
penetration in course sand only possible using percussion technique

The positive test results have led to the decision to perform a test program on this subject in the geotechnical centrifuge. Thanks to the 1G tests some experience was obtained with the ins and outs, so that it was more easy to build the in flight test setup. The test is performed in a pressurized container in order to simulate a large water depth. Parameters which can be measured are: pressure during pulse, displacement, water consumption. The first orientating tests were looking promissing. In this figure an example is shown of the shape of the pulse at different stages of the installation. The pulse time is approximately 50 ms.

The sliding behavior of spudcans in layered soil (SHELL EXPRO UK)

- Spudcans are used in offshore for the foundation of jack up platforms. If the spudcans are founded on sand the penetration depth is not so large as in clay. In order to validate the design rules a test program has been carried out to examine the resistance against sliding. Horizontal loads are applied to the spudcans during strong wave actions. In this project the attention is focused on the sliding behavior of spudcans in layered soils (sand on clay). An item of research is to determine the thickness of the sand layer at which is the transition between a sand type spudcan and a clay type spudcan. The load due to wave action can be simulated by a two dimensional loading system. The test are performed with a model spudcan with a diameter of 60mm, at an acceleration of 150g.
It was found, for example, that the sliding capacity increases at decreasing height of the sand layer.

Effect of repeated loading in the upheaval buckling behaviour of pipelines (SHELL Int. Exploration & Prod.)

- A representative segment of a buried pipeline was subjected to cyclic loading (simulation of buckling due to thermal effects) in order to examine the upheaval behaviour. In this figure the displacement of the pipe segment is plotted during cycles with increasing maximum load. It was found that the cyclic loading did not decrease the pull out load of the pipe segment significantly.

Upheaval resistance of trenched pipelines covered with a loose packed backfill (SHELL EXPRO UK)

- A representative segment of a buried pipeline in loose packed soil was subjected to static loading in order to determine the upheaval resistance. The tests have been performed on a soil from the site (offshore East of Scotland), so that predictions could be made for a particular design. Several aspects have been examined in detail, such as density of the soil, clay content and loading speed. Pressure sensors were installed in order to visualize the change in pore pressure during loading. The advantage of performing small scale centrifuge tests on this problem is, that it is rather easy to test methods of which are believed that they improve the upheaval resistance. It appeared that the measured data fit quit well in the calculations used for designing buried pipes.

Effect of waves on the behaviour of pipes at the sea floor.(no support).

- Some orientating tests have been performed to investigate the influence of waves on the behaviour of buried and not covered pipes at the sea floor. A very simple device has been developped to simulate wave action at g-levels up to 150 times earth gravity. In this movi the operation of the device can be watched. Up to now it was found that the waves cause densification of a sand bed. Also a tendency could be observed that heavy pipes are sinking in a sand bed if waves are generated. By using a somewhat more smart system also running waves can be simulated.

Simulation of earthquakes

- Some orientating tests have been performed to see if the small centrifuge can be used to simulate earthquakes. A container with soil was mounted on a platform with wheels. At 100 g the platform was moved by a pneumatic actuator. Aftre a short displacement the platform bumps against a ridged wall, which causes a shock wave in the sample. To demonstrate the effect of the bump a clay dike was built at a sand layer. In this picture the deformation of the dike after two bumps is visualised by subtracting the images.

Collapse ship embankment

- In order to make predictions about the safety of foundation of bridges knowledge is required about the collapse between a ship and the wall. Some orientation tests are carried out in the centrifuge to show that this phenomenon can be investigated by this test technique. Some photographs and some prilimary results are presented here.

Sub-gouge effects of scraping ice bergs on the sea floor

Failure mechanism of dikes (Rijkswaterstaat)

- In order to get more insight in the stability of dikes at high water levels, an in-situ method is in development to test a dike against failure. Before such a method can be developed more insight is required in the failure mechanism of several types of dikes. In this test program some dikes types shall be schematized and the stability at high water levels shall be investigated. In this movie it is shown how a dike with half thickness can be brought to failure by increasing the water level to the maximum height (microbes are used to mark the clay surface).

Centrifuge tests on failure of dikes by overburden pore water pressure (Delft Cluster)

- A centrifuge test program will be carried out to investigate the behaviour of dikes during rising of the water level. It is assumed in this case that the clay dike is founded on a permeable layer of sand. The increasing water level is supposed to increase the pore water pressure in the sand. The pore pressure causes an uplift force at the bottom of the profile of the clay dike. Thanks to friction reduction failure can be triggered.
In order to have a good control over the pore water pressure the pressure under the clay dike can be adjusted by a computer controlled pressure system. In the test setup shown in this diagram the pore water pressure can be increased uniformly in the sand layer underneath the embankment. In the diagram shown here it is possible to create a pressure gradient in the sand layer underneath the clay layer. The pump is activated by the on-board computer, using the output of one of the pressure sensors.
A photograph of a typical test result is presented here. At low pore pressure a shear band is formed starting at the toe of the dike. Reduction of the friction by increasing pore pressure causes deeper shear bands.
In this movie it can be shown that both mechanisms happen almost at the same moment. In this movie a test is shown where the slope is stabilized by a berm. Furthermore an imperfection is made in the clay behind the dike. A similar mechanism is now realized as assumed in some calculations. A good fit was found between tests and numerical calculations.
This detail shows that a complex shear band pattern can be formed.
A next step is to find methods which improves the stability of dikes. Some possibilities are shown here. The advantage of centrifuge modelling is that you are free to test any design. Even if the design is not practical or too expensive. Trials can led to new better ideas. It was found that a sheeting pile was not very effective. Most effective up to now were horizontal nails half way the slope.

Deformation measurement during field test on failure of dikes by overburden pore water pressure (Delft Cluster)

Sand in the centrifuge (Dutch National Science Foundation)

- In this research project all phenomena around the behavior of sand at higher g-levels are tested. In particular attention is paid in how far the sand bed preparation, model dimensions, particle diameter, etc. influences the behavior of footings.
It was found that the bearing capacity is extremely sensitive for changes in i.e. the relative density and lost of fine material. These phenomena show that sample preparation is one of the most important aspect in centrifuge modelling. This research program was very helpful in evaluating the reliability of the test results on spudcans.
(The research is described in a thesis, which is available on request).

Failure mechanism of sheet pile walls (National Institute of Industrial Safety, Ministry of Labour, Japanese )

-In cooperation with the Ministry of Labour of Japan (National Institute of Industrial Safety) a test program is being carried out to investigate the failure mechanism of sheet pile walls during collapse. Several configurations are tested, where the failure mechanism is made more clear by means of image processing.
Click here to see a typical test result of an excavation between two sheet pile walls. The support of the wall was removed during flight. Click here to get an on line impression of the centrifuge test.
In this picture or movie an example is shown of an excavation of a sheet pile wall which was stabilized by another sheet pile wall at some distance.

In flight excavation of soil (no support)

- A low budget device has been developed to perform in flight excavation in tests on sheet pile walls in clay or sand. In this movie the operation of the device in clay can be observed. The operation in a test on a sheet pile wall without anchor in sand is shown here.
It can be seen that a sheet pile wall with anchor behaves different.
More details about the device will be mailed to you on request.

Detection pollution transport in centrifuge test (EC)

- In cooperation with other European geotechnical centrifuge groups a project (NECER) has been started to develop measuring and test techniques in order to simulate and monitor the transport of pollution in soil. One of the advantages of centrifuge tests is that the time scale can be reduced. A process which takes several years in reality can be simulated in the centrifuge in a few hours.
As a first excersise orientating tetst have been carried out to examine the capillar rising at higher gravities. For more details click here.
In a following up a test program has been carried out to examine the infiltration of LNAPL and DNAPL in sand. The problem was schematized as a two dimensional system, where the oil was supplied by a linear source. In the following movies the infiltration of LNAPL and DNAPL is visualized. In reality the infiltration of, LNAPL takes 30 minutes, where the movie with DNAPL is real time. In thje test with LNAPL the sand was partly saturated, were a capillar transition zone was present between the saturated zone and the dry zone. Three stages during a test with LNAPL are presented in the following pictures 1; 2; 3 The influence of the capillary and saturated zone are clearly visible.
The contours (determined by image processing) of several stages are presented in this picture. In this movie it is shown how digital image processing is applied to digitize the contours of an optical tracer.


One of the measuring methods to follow this process in more detail is the use of digital image processing. The grey level distribution over the polluted spot can be used to estimate the concentration of the pollution at some point during the test. Some stages during a test are visualized in this three dimensional plot. The grey level distribution can be converted into false colours. Research is carried out to examine in how far the grey level can be used to digitize the concentration of the pollution.

Simulation cleaning process of viscous LNAPL

One of the main goals of research on pollution transport is the simulation of cleaning processes. The advantage of centrifuge tests is that processes which would take several days at normal g level can be simulated in a few hours. As a pilot test the removal of a viscous LNAPL by means of a forced ground water flow is simulated. Firstly infiltration of LNAPL in partly saturated sand was started. During this proces a hydraulic gradient was applied, so that the ground water starts to flow. It appeared that the LNAPL was removed after approximately two hours (75 days prototype). The test can be visualized in this movi. Click here to see a photograph of the test setup, after a test has been performed. The accumulation of the removed contaminent is clear visible.

Photoelastic stress measurement in soil around tunnel and piles (no support).

An optical measuring method will be applied to visualize stresses in soil (crushed glass) around a tunnel and neighbouring piles. The aim of the test is to examine changes in stress when the diameter of the tunnel is decreased. The test setup and measuring device is shown here. The first measuring data will be available soon.

Non soil mechanical problem

Question 1
When you are going to Jupiter and you have to decide if it is useful to take candlelight with you or not. The following questions can be asked:
- will the flame be longer, shorter or equal
- is there a g level at which the flame stops burning

To demonstrate the power of a centrifuge it can be said that this problem can be solved within 10 minutes by an experiment.

An additional question in a centrifuge test is:
- what is the angle between the axis of the flame and candle if the hinged platform rotates 90 deg (a or b).

Question 2
What do you think about the resistance of a chicken egg against gravity. The egg is placed on a flat surface, whithout any further support.

A technique based on digital image processing has been applied in order to measure displacements in soil samples. This method was initially developed to determine displacements of objects and deformations of soil surfaces in the small geocentrifuge of the laboratory. The system comprises a PC with a frame grabber, a video camera and a software toolkit. Because the system is PC based a high degree of interactive operation is possible. By means of users written programs objects in a captured frame can be isolated and digitized automatically.
Beside centrifuge research several other applications are found in which the method could be used to measure displacements. An example is a field test, were wave overtopping was simulated in order to examine the failure mechanism of a dike (for movie click here). The surface displacement could be detected by means of labels, which are placed at the slope of the dike.
In the laboratory the method is used to measure the diameter of triaxial samples during a test.

click here for more details).

Photo-elastic stress measurement

Stresses in a granular material (i.e. during cone penetration) can be measured by means of grains of photoelastic material. In this technique sand is replaced by crushed glass. The assembly is made transparent (no stress) by submerging the pores with a liquid which has the same refraction index as glass. Due to shear stresses (stress) glass becomes anisotropic for light. This phenomenon can be made visible by polarized light. In this movie the development of the optical phenomenon underneath a pile tip is visualised.
The stress underneath a tip is shown here in more detail
The phenomenon in the grains is made visible in more detail in a shear test with disks of perspex (no shear, shear). This movie shows some cycles in a more dynamic way. The grains are used as sensors in this case. First they translate the contact forces into a second rank tensor and secondly they make that the stress components can be measured. A computerized measuring device and calculation technique has been developed to apply this method on model test.
The stresses can be made visible in i.e. hoppers, so that the effect of filling , flow and i.e. an (inclusions can be examined.
The measuring technique is a valuable tool to analyze the stress strain behaviour of granular material in soil testing devices. In this example the stress distribution is visualized in a simple shear apparatus. It can be shown that the stress distribution is not very well defined in a direct shear cell . The visual patter in circular polarized light is combined with the measured principal stress directions. The straight pattern of the major principal stresses means that the stress distribution is very uniform in shear. By digitizing the co-ordinates of the black markers It was found that the major directions of stress and strain coincide fairly well.
The photoelastic technique can also be applied in geotechnical centrifuge research. As a pilot test the stress pattern is made visible at 150g in the granular material below a vertical loaded spudcan footing. It can e.g. be visualized how the stress pattern changes before and after sliding.
Recently a new compact measuring device has been developed to enable digitization of stress components in centrifuge models with optical sensitive material. In this technique an image processing station is used to simulate the mechanical scanner. In the first instance images are stored at every 3 deg. rotation of a polarization filter. Next the images are played back where the averaged light intensity is determined in sub frames, yielding the course of the light intensity during rotation of the polarizer in a material point. The obtained harmonic wave yields the principal stress difference (amplitude) and the principal stress direction (phase) in a material point. The measured directions can be used to derive a map of the principal stress trajectories underneath a spudcan footing, before and after sliding.
This test technique is up to now the only way to obtain field information about the stress distribution in granular material.

Computer soil testing

Other researchers involved:

A. Jacobse, Master Student, Using suction anchors for ship stabilization
A. Rohe, student, Failure mechanism of dikes in centrifuge (Delfts Cluster)
A. Bizzarri, PhD study, Influence stress path on creep of clay
W. Broere, Post Doc, Optical measurement of stress in photoelastic grains.
Y. Hayashi, Guest Researcher, Kumamoto University, Japan. Soil improvement.
R. Krabbedam, Guest Researcher, Jumbo Shipping, Suction anchors.
J.J. de Visser, memeber technical staff, model preparation and operator centrifuge.
A. Mensinga, member technical staff, electronical facilities.
F. Riem Vis, member technical staff, design equipment.

Graduate students and other researchers who have been involved in centrifuge research:

S. van Baars, L. Ravenswaay, M.G. den Braber, I.A.G. Ligtenberg
H. Stuit (PhD), W. van Niekerk, M. Borop (Denmark), J. Hedegaard (Denmark)
O. Mareschal (Belgium), M. Senders, G. Kramer, F. van Woensel (Belgium)
F. Plenevaux (Belgium), J.F. Wintgens (Belgium), S. Dresken.
Y. Toyosawa (Ministry of Labour, Japan), J. Kajabova, T. Simon (Belgium)
M.A. Dumont (France), B. Ferres (France), R. Peek, G. Esposito (Italian)
C. Dufour (Belgium), D.Maes, S.W. Lee (S. Korea), M. Pimoulle (France), O. Dupont (Belgium).

Most important publications up to 1994

Allersma, H.G.B., 1982. Determination of the stress distribution in assemblies of photoelastic particles. Experimental Mechanics 1982(9). pp.336-341.

Allersma, H.G.B., 1982. Photoelastic investigation of the stress distrubution during penetration. Proc. ESOPT2, Amsterdam (may 1982) A.A. Balkema, Rotterdam, Vol.2, pp.79-83.

Allersma, H.G.B., 1982. Photoelastic stress analysis and strain in simple shear. Proc. IUTAM Conf. On deformation and failure of granular materials, Delft, A.A. Balkema, Rotterdam, pp. 345-353.

Allersma, H.G.B., 1987. Optical analysis of stress and strain in photoelastic particle assemblies. PhD Thesis, Delft University of Technology. (available on request)

Allersma, H.G.B., 1988. Optical analysis of stress and strain around the tip of a penetration probe. First Int. Conf. on Penetration Testing, Orlande, Florida, pp.615-620.

Allersma, H.G.B., 1990. On line measurement of soil deformation in centrifuge tests by means of image processing. 9th Int. Conf. on Experimental Mechanics, Copenhagen, pp.1739-1748.

Allersma, H.G.B., 1991. Using image processing in centrifuge research. Proc. Int. Conf. Centrifuge91, Boulder ,Colorado, A.A. Balkema, Rotterdam, pp. 551-558.

Allersma, H.G.B., S. van Baars, 1992. Modelproeven in een geotechnische centrifuge. De Ingenieur, 104, 7/8, pp.8-12.

.

Publications from 1994

Allersma, H.G.B., 1994. Using image processing in analyzing stresses in photoelastic granular material. 10th Int. Conf. on Experimental Mech., Lisbon, 18-22 July, 1994.

Allersma, H.G.B., 1994. The University of Delft geotechnical centrifuge. Int. Conf. Centrifuge94,Singapore, 31 August-2 Sept. pp.47-52.

Allersma, H.G.B., H.G. Stuit, P. Holscher, 1994. Using image processing in soil mechanics. Thirteenth Int. Conference on Soil Mechanics and Foundation Engineering. Jan. 3-10, New Delhi,India, pp. 1341-1344.

Allersma, H.G.B., 1994. Measurement of the deformation of a dike by means of image processing. External Report, Geotechnical Laboratory, Delft University of Technology.

Allersma, H.G.B., I.A.G. Ligtenberg & B.A.N. Koehorst. Simulation of failure of dikes by water infiltration by waves. Int. Conf. Centrifuge94, Singapore 31 August-2 Sept. pp.289-294.

Allersma, H.G.B., A.P. Kooijman & W. van Niekerk, 1994. Simulation of cratering in a small geotechnical centrifuge. Int. Conf. Centrifuge94, Singapore 31 August-2 Sept. pp.325-330.

Allersma, H.G.B., 1995. Using image processing in centrifuge research. European Workshop: New Techniques for Centrifuge Modelling. Nantes, Jan. 5-6.

Allersma, H.G.B., 1995. Development of miniature equipment for a small geotechnical centrifuge. European Workshop: New Techniques for Centrifuge Modelling. Nantes, Jan. 5-6.

Allersma, H.G.B., L. Ravenswaay and E. Vos, 1994. Investigation of road widening on soft soils using a small geotechnical., Transportation Research Record No.1462, National Academic Press Washinghton, D.C. pp. 47-53.

Allersma, H.G.B., 1994. Development of miniature equipment for a small geotechnical centrifuge. Transportation Research Record No.1432, National Academic Press Washinghton, D.C. pp. 99-105.

Allersma, H.G.B., 1995. Simulation of the trap-door problem in an geotechnical centrifuge. Second Int. Conf. on the Mechanics of Jointed and Faulted Rock, Vienna, Austria, A.A. Balkema, Rotterdam, pp.521-526.

Allersma, H.G.B., 1995. Vijf jaar geotechnisch centrifuge-onderzoek bij de TU Delft., de Ingenieur, nr.8, mei 1995, pp.32-34.

Allersma, H.G.B., 1995. Meten en regelen bij 150g. RBelektronica, dec. 1995.

Allersma, H.G.B., Mareschal, O. 1995. Het bezwijken van dijken bij waterinfiltratie. Land+water, dec. 1995.

Bizzarri, A, H.G.B. Allersma, B.A.N Koehorst, 1995. Preliminary tests on soft clay with a biaxial apparatus. Int. Symp. On Compression and Consolidation of Clayey Soils, Hiroshima, A.A. Balkema, Rotterdam, pp.27-32.

Allersma, H.G.B., 1995. Optical measurement of stress and strain in granular material. Second Int. Conf. on Composites Engineering. New Orleans, August 21-24, pp. 17-18.

Allersma, H.G.B., 1995. Simulation of subsidence in soil layers in a geotechnical centrifuge. Fifth International Symposium on Land Subsidence, The Hague, the Netherlands, IAHS Publication No. 234 pp.117-126.

Allersma, H.G.B., 1995. Simulation of buckling of open ended cylinder piles during driving. External Report Nr.384, Geotechnical Laboratory, Delft University of Technology.

Allersma,H.G.B., O. Mareschal, 1995. Centrifuge research on the stability of dikes submitted to wave overtopping. Report nr.383, Geotechnical Laboratory, Delft University of Technology.

Stuit, H.G., 1995. Sand in the geotechnical centrifuge. PhD-Thesis, Delft University of Technology. (available on request)

Senders, M., H.G.B. Allersma, 1996. Simulation of suction can installation in a geotechnical centrifuge, Report, Geotechnical Laboratory, Delft University of Technology.

Allersma, H.B.G., 1996. Computer applications in experimental soil mechanics. Proc. Sec. Int. Conf. in Civil Eng. on Computer Applications Reasearch and Practise. University of Bahrain, College of Engineering, Vol.2, pp501-511.

Allersma, H.G.B., 1996. Using centrifuge research in analysing stability of dikes. Genie Civil-Genie Cotiers, IVemes Journees Nationales, Dinard, France, pp.289-298.

Allersma, H.G.B., 1996. Modelonderzoek naar het gedrag van dijken bij waterinfiltratie. Syllabus open sprekersdag afdeling voor Geotechniek, KIVI, Delft.

Allersma, H.G.B., 1996. Using digital image processing in a field measurement. Geotechnique 46(3), pp. 561-563.

Plenevaux, F., J.F. Wintgens and H.G.B. Allersma, 1996. Simulation of suction anchor installation in a geocentrifuge. Report no.402, Geotechnical Laboratory, Delft University of Technology.

Allersma, H.G.B., 1996. Simulation of flowing hoppers in a geotechnical centrifuge. 12th Int. Congress of Chemical and Process Eng., Praha, August 25-30, 1996.

Allersma, H.G.B., B. Hospers and M.G. den Braber, 1996. Centrifuge tests on the sliding behaviour of spudcans. 49th Canadian Geotechnical Conference. St John's September 23-25, 1996, pp. 199-206

Allersma, H.G.B., F. van Woensel, 1996. Centrifuge research on the sliding of spudcans in layered soils. Report nr. 410.

Allersma, H.G.B., 1997. Centrifuge tests on widening of embankments on soft soil. Third Geotechnical Engineering Conference, Egypt (space view) Cairo University, Jan. 5-8,1997, pp. 43-54.

Allersma, H.G.B., 1997. Toepassing digitale beeldbewerking in de geotechniek. Open sprekers dag afdeling voor Geotechniek, KIVI, 19 feb.

Allersma, H.G.B., 1997. Bloemlezing centrifuge-onderzoek TUDelft. Studiedag 'De centrifuge in de geotechniek', KIVI afdeling voor Geotechniek, NIRIA-vaksectie Civiele Techniek, 19 maart.

Allersma, H.G.B., F.J.A. Plenevaux, J,-F.P.C. Wintgens 1997. Simulation of suction pile installation in sand in a geocentrifuge. 7th Offshore and Polar Eng. Conference, Hawaii, May 25-30, pp. 761-766.

Allersma, H.G.B., 1997. Using imaging technologies in experimental geotechnics. Second Int. Conf. on Imaging Technology Techniques and Civil Eng. Applications, Davos, May 25-30, pp. 1-9.

Toyosawa, Y., N. Horii, H.G.B. Allersma, 1997. Failure mechanism of anchored sheet pile wall in centrifuge tests. Proc. 32th Japan National Conference on Geotechnical Engineering, 15-17 July, Kumamoto, Japan, Vol.2, pp.1697-1698. (in Japanese).

Allersma, H.G.B., 1997. Centrifuge tests on the influence of shape on the sliding behaviour of spudcans. 8th Int. Conf. on the Behaviour of Off-Shore Structures, 7-10 July, Delft, the Netherlands, Vol.1, pp.225-233.

Allersma, H.G.B., 1997. Optical analysis of methods to influence the stress distribution in hoppers. Int. Symposium on Mechanics of Particulate Material, June 29-July 2, Chicago, pp.431-440.

Allersma, H.G.B., 1997. Possibilities of small geotechnical centrifuges. Int. Conference on Soil Mechanics&Foundation Eng., Hamburg, Sept.6-12, 1997, pp.2015-2018.

Allersma, H.G.B., B. Hospers and J.G. den Braber, 1997. Centrifuge tests on the sliding behaviour of spudcans. Canadian Geotechnical Journal, Vol.34, Nr.5, pp.658-663

Allersma, H.G.B., 1997. Application of small beam centrifuge in geotechnical research. Autumn Geotechnical Seminar of Korean Geotechnical Society, Eng. College Korea Univ., Seoul, November 27, 1997, pp.46-58.

Allersma, H.G.B., 1997. Some applications of drains in geotechnicacs. Autumn Geotechnical Seminar of Korean Geotechnical Society, Eng. College Korea Univ., Seoul, November 27, 1997, pp.59-62.