Soil mechanics – Discover Dredging

Interpore 2026: The Cutting Of Clay in A Herschel-Bulkley Model

Proudly holding my poster for the Interpore conference

This week, I’ll be in Nantes at the Interpore 2026 conference1. This is an academic conference on everything porous, ranging from catalysts, to food, to foams, to biology and even soil mechanics. And that is a topic where my research fits in. I submitted an abstract for a poster on the work done by our intern Prasanna Ramadurai2 on the applicability of Herschel-Bulkley fluid modelling for the simulation of the cutting of clay. The accompanying presentation can be found here.

The elemental physics of clay deformation with the rate process theory

As a refresher, clay is a funny substance. The constituent particles of various minerals that form small plates, that are electrostatically charged. Due to these charges, they tend to cling together. When you deform the mass, there develops a shear plane. The particles move over each other with alternating repelling and attracting electric fields. This discrete path is reminiscent of how in particle physics moving steps are only possible when there is enough energy to push the particle past an activation threshold. This was initially postulated by Boltzmann3 and subsequently formalised by Arrenius and Glasstone in the rate process theory. Miedema further applied this to the deformation of clay4. Depending on the amount of water present, it can behave like the clay can behave like thick water or soft rock. Both captured by the same equations. Eventually the resulting shear stress curve is very similar to a Herschel-Bulkley fluid.

Shear stress model by Miedema and comparable fluid models

The work of Prasanna focused on exploring a workflow to simulate the deformation of clay using Ansys Fluent for CFD. This package does not support the deformation model as described by Miedema. But, as the resulting behaviour should be similar to the Herschel-Bulkley model, the H-B viscosity could be used. As previously described here, Prasanna managed to find the appropriate settings and setup to achieve credible results.

Result of the experiments and the simulation compared

As Fabian Kruis5 has previously done experiments in the soil bin test rig, we do have reference data from actual measurements. Fabian has recorded the deformation and analysed the internal movements with PIVlab. The vector field from PIVlab is very similar to the vector field calculated by Ansys Fluent with the Herschel-Bulkley viscosity model. However, translating the deformation to stresses and ultimately to the cutting forces on the blade is still to be improved. The results from Ansys overestimate the measured forces.

Next to the poster, I also prepared a presentation. This presentation can be accessed through the conference portal, or directly from here. Off course, when you are at the conference, you can approach me there. Or through the contact details her on this website.

Resulting shear plane angles from PIVlab and Ansys

References

Internship Prasanna Ramadurai: CFD Modelling Clay as a Fluid

Prasanna presenting his work at Damen Dredging Equipment

Prasanna Ramadurai has been doing an internship with us at Damen Dredging Equipment for my PhD project on the cutting of clay1. As an old fashioned analytical and experimental dinosaur, I have been working in my comfort zone. However, when submitting articles the response from the reviewers has been: ‘How about validating your results with a numerical simulation?’ And that is exactly what Prasanna has been doing for me these months. Now he has presented his work and I can use the results in my own research.

Relation of PI and CI to the adhesion range according to Atterberg

Clay is a strange material. It is neither a solid, nor a fluid. Depending on the amount of water in the material, it can behave like concrete or like water. The scale on which this can be described is defined by the Atterberg limits2. Well known soil parameters as Plasticity Index and Consistency Index are derived from those Atterberg limits. Atterberg himself defined the following limits:

ID	Limit name	Criteria
1	Upper liquid limit	Starts to show signs of a viscous fluid
2	Lower liquid limit	Normal Casagrande test or Fall cone test
3	Adhesion limit	When no clay sticks to a nickel spatula
4	Upper plastic limit	Can be moulded
5	Lower plastic limit	Normal rolling test for plastic limit*
6	Cohesion limit	When pieces of clay do not stick to each other anymore
7	Shrink limit	Normal shrink limit, constant volume for water content
*Atterberg proposed to roll on a paper surface, whereas ISO 17892 proposes a glass plate

The consistency limits as originally proposed by Atterberg

The resistance of a material to deformation can be expressed as the resulting stress due to a strain rate. When there is immediate stress for even the slightest movement and constant after reaching a yield stress, this is typical of a solid. On the other hand, when a material starts to move immediately and the resistance to deformation increases with the strain rate, it is a fluid. And clay is just the typical material that exhibits both phenomena.

Shear stress models depending on strain rate

In rheology, the factor which shear stress is related to the increased strain rate is called viscosity3. However, due to the internal friction in clay, the stress follows the vertical axis and consequently, the viscosity becomes infinite. Prasanna squeezed out the capabilities of the CFD program using some clever mathematical tricks of a Herschel-Bulkley fluid model to get the simulation to behave. And the results are promising enough to follow up in a separate study.

Compare CFD simulation and PIV experiments

For supervising Prasanna, I am very grateful for the assistance of Suman Sapkota for his knowledge of computational fluid dynamics. Together with my knowledge of clay, Prasanna gained a very special set of skills in this area. Prasanna will be back at the TU Delft to continue his master’s graduation project. I can recommend him for having him in your team.

References

WODCON 2025: Rolling Out A New Clay Test

Fully covered cutter head in sticky clay

Here at the WODCON 2025 in San Diego1, the theme is ‘Dredging Towards a More Resilient Future’. One of the challenges we encounter, is that even the resources of good construction sand run out. We either have to repurpose sand already dredged or find and alternative construction material. One such an overlooked material is clay. A lot of effort is put into understanding the behaviour of clay in infrastructure applications. The Dutch Centre for Legislation and Infrastructure (CROW)2 has provided recommendations on the applicability of clay for various types of construction. However, the clay has to be dredged and for the adherence potential of clay, there is another recommendation issued by World Association for Waterborne Transport Infrastructure (PIANC)3. Both do use the Plasticity Index and the Consistency Index as criteria to classify the clay. Interestingly, the clay type that is regarded as suitable for construction by the CROW, is also classified by the PIANC as to have the highest adherence potential and thus gives the most problems in dredging.

Another problem with the criteria by PIANC is that they tend to be unreliable. Something they already acknowledge in the supplied explanation to the diagram. When following the literature that led to the recommendation, it turns out the original application was not dredging but tunnel boring4. Where the problem was not so much the clogging of the cutter shield, but the collection of clay in the suction chamber. And even plotting the data used for this assessment shows a large variability. Apparently there is more to the problem of adherence than just the PI and CI. Which might be obvious when considering the original Atterberg Limits. The PI and CI are related to the plastic range of clay, whereas Atterberg already defined a range where adhesion is more relevant.

A situation similar to the cutter clogging is the covering of a drill bit in the oil industry. There, they encounter a phenomenon called ‘Bit Balling’5. It is extremely difficult to assess the bit balling potential from a physical model related to the soil parameters alone. As a solution they developed the ‘Rolling Bar Test’6. A defined amount of clay sample is put into a cylinder with the needed amount of water. Finally a rod is inserted in the sample cylinder. The whole contraption is placed onto a roller set and turned for a set of times. Each time the amount of clay sticking to the rod is measured and plotted in a graph. Eventually, most clay types will loosen their grip on the rod. But some are sticking to the rod indefinitely. Those are the clays that are also likely to show bit balling in the actual process.

Bit balling and procedure of a rolling bar test (data: Mettah, 2011)

As we know that we can’t fight the adhesion of clay, we may as well improvise, adapt and overcome the problem. Since already my graduation, I am working with clay. In that case, it was an auger, that needs the adhesion to the back shield to propagate the clay in the auger. When we were asked by a contractor to improvise a tool that could tackle this sticky clay, we developed a disc bottom cutter head that used the adhesion to move the clay over the blade to a scoop behind the blades. This worked so smoothly, that the satisfied customer bought a second. Eventually he finished to job in time and in budget7.

Category: Soil mechanics

Interpore 2026: The Cutting Of Clay in A Herschel-Bulkley Model

References

See also

Internship Prasanna Ramadurai: CFD Modelling Clay as a Fluid

References

See also

WODCON 2025: Rolling Out A New Clay Test

References

See also