Academy of Finland has awarded Aalto & Geological Survey Finland consortium financing for the Geomeasure project. We will investigate the Finnish seabed with GTK research vessel Geomari and find new ways to interpret the results of the tests, in particular free fall penetrometer tests. To do that, we will take samples from the seabed and correlate results of experimental testing with the penetrometer data. Also, the tests will be numerically modelled with MPM. We also plan to investigate the behaviour of the deposits under cyclic loading, typical for wind turbines.
Our review paper on Material Point Method is now available here. The link will expire in a month, but I hope that by that time we will be able to provide a copy that could be used instead.
Please have a look, and let me thank all the co-authors: professors Martin Berzins, Will Coombs, Jim Guilkey, Matthias Möller and Kenichi Soga, as well as the doctoral students (some of whom graduated by now): , Tito Adibaskoro, Seyedmohammadjavad Seyedan, Roel Tielen and Quoc Anh Tran (all the names in alphabetical order). Thank you again!
The E-Unsat 2020 has been excellently organised and a resound success, despite being fully online. Even though I must say I did miss the chats with friends and informal, but often illuminating discussions with the attendees, I still have seen many great presentations and learnt much.
I was also presenting the paper authored by Dr Abed and me on validation of gas transport couplings in our THMC coupled Finite Element Method code Thebes (presentation here, and paper here). We have much more to do in relation to gas transport – my doctoral student Abhishek Gupta is responsible for that, and we hope to deliver ambitious simulations of gas transport of a generic nuclear waste repository in a year or two.
I was delighted to deliver the Young Invited Lecture on Granular Material Point Method, something we are working on with Seyed Mohammad Javad Seyedan. We will aim to publish full paper based on the presentation, but for now, the presentation is here.
Many more exciting things has happened and is happening with the research in my group – I will try to update you on that soon!
Hi, I have been very busy recently, leading to lack of updates for a very long time. During past few months I delivered a plenary lecture during the 32nd Nordic Seminar on Computational Mechanics in Oulu (slides), Quoc Anh Tran graduated with a doctoral degree (thesis), and we have a new journal paper published, “A convected particle least square interpolation material point method” in IJNME. That one is not currently freely available, so in case you need a copy, contact me at my Aalto email.
So, what the future holds? SMJ Seyedan should graduate relatively soon with another doctoral degree related to MPM, Tito Adibaskoro is slowly progressing towards the same goal. To continue the research I applied for research related to MPM both at national level (Academy of Finland proposal) and international one (Marie-Curie ITN network), and let’s keep fingers crossed.
On the THMC modelling of clays, a new student has started in January, more update on that later!
Much is happening at Aalto and I have been late with much what should be posted. Just to update things quickly, recently SMJ Seyedan has published a paper detailing a small strain enhancement for constitutive models for soils. The paper is here (open access). The idea is to have a thermodynamically consistent enhancement in the initial part of the stress-strain curve, which can be applied to any model using constant shear modulus. We tested it for Mohr-Coulomb and Modified Cam Clay, but it should work for other constitutive models too. Unfortunately, we have no such neat way to adjust the bulk modulus yet, but hopefully in the future… The paper comes with a Matlab code, so it is easy to try it out!
On another front, QA Tran has published a paper related to the null space errors in the Material Point Method. The paper is here and is also open access. The paper discusses null space errors and show how to filter them out efficiently. Such filtering is recommended when we have a problem with large deformations, in which number of material points in some grid cells increases above the initial value. That creates a problem when the data is tranferred from the nodes back to the material points – we have less nodes than the points and as such the transfer uses extra assumptions leading to errors. For example, having 4 nodes transferring data to 9 points, we need to create 9 values at the points from 4 values at the node – that is impossible without extra assumptions which may lead to errors.
The same paper also shows that use of a generalised-alpha integration scheme in the Material Point Method helps to damp out high frequency vibrations, without affecting the total energy and total momentum balance much. Very useful when we need a solution without high frequency noise – and much better than using an overall damping coefficient which alters the solution significantly.
S. Seyedan and W. T. Sołowski, “Enhancing Constitutive Models for Soils: Adding the Capability to Model Nonlinear Small Strain in Shear,” Advances in Civil Engineering, vol. 2019, Article ID 6016350, 11 pages, 2019. https://doi.org/10.1155/2019/6016350.
Tran, Q‐A, Sołowski, W. Temporal and null‐space filter for the material point method. Int J Numer Methods Eng. 2019; 1– 33. https://doi.org/10.1002/nme.6138
We have attended the 2019 MPM conference in Cambridge in numbers. I had a presentation on Uintah software and a keynote, while other members of the research group presented their research. You can find the proceedings here: http://mpm2019.eu/papers
In particular there are some presentations available:
- Keynote “Accuracy, errors and convergence of MPM”
- Uintah software presentation (100 MB, external link)
- MPM Simulation of fine particle migration process within unsaturated soils presented by Dr Xiaoqin Lei
- Continuum modelling of the granular flows in gaseous states using material point method, presented by SeyedMohammadJavad Seyedan
And the conference papers:
- MPM Simulation of fine particle migration process within unsaturated soils (Lei & Sołowski)
- Continuum modelling of the granular flows in gaseous states using material point method (Seyedan & Sołowski)
- An improved moving least squares method for the Material Point Method (Tran, Berzins & Sołowski)
- Moving least squares reconstruction for B-spline Material Point Method (Tran, Wobbes, et al.)
On the way back, I have been given opportunity to present my research at Imperial College London. In particular, you can download the abridged presentation related to our research with Dr Abed in project THEBES: THMC Behaviour of the Swelling Clay Barriers
Quoc Anh Tran, doctoral student working in the Academy of Finland project on landslide modelling, managed to get a journal paper published in Computers & Geotechnics. The paper shows number of improvements in landslides modelling with the Material Point Method. Those include strain rate dependent, structure dependent Mohr-Coulomb model for undrained analysis. We also used a simple non-local formulation for the shear-band thickness (Pinyol et al. 2017), which reduces mesh dependence. After validation, the paper shows some – I dare to say – rather nice simulation of a Saint-Monique landslide.
The paper is open access, download it from here:
We have recently been in Trondheim, the truly charming city with Nidaros cathedral, famous for its organs. The reason was the Workshop, where Quoc Anh Tran had a very nice presentation on MPM simulation of sensitivity test.