Dear Reader,

The week has whizzed through so fast that we are already at the end and here I am writing another newsletter!

Do you feel the year is moving rather fast? Maybe because it has been a busy start of the year for me, I feel that the year is running along rather quickly.

Welcome to today's Newsletter! For this week, I will be reflecting on the following:

1. Comprehensive modelling of UD composites
2. Behind the Scenes at CM Videos: I reached 3000 subscribers!
3. Quote of the week: Cryptocurrencies and Athens: Lessons from history

Technical Reflections

Comprehensive modelling of UD composites: What must you do?

My first research into composite materials was based on unidirectional composites. This was as an undergraduate student and I still remember working on hand-layup unidirectional composites. Thereafter, I took up a PhD study on unidirectional composites and started modelling it.

This week, I have been reflecting of the essential steps one has to take whilst modelling unidirectional composites. Incidentally, from my research, questions about unidirectional composites continue to be the dominant question people are asking me about composites modelling. It seems ability to model UD composites will enable you explore the bidirectional, multilayer and textile composites. I am making a three-part video on comprehensive modelling of UD composites and has already published the first one. Please watch out on the CM Videos channel for the other two, when I launch them.

So what must you do if you want to undertake a comprehensive modellling of UD composites.

Step 1: Focus on the realiable representation of the Microstructure

Realistic and comprehensive modelling of UD composites must focus on microstructure. This will involve representing correctly the fibre, matrix and interface media. If there are voids or other agreegates that are incorporated into the virtual domain, this has to be represented as well. If you are modelling multi-fibre composite, then you have to explore the use of Monte Carlo methods to generate representative RVE for the composite. I made a video about this focus on geometry design and above is the video for your reference, if you have not seen it already.

Step 2: Loading conditions imposed on the model has to be representative

The nature of loads that is used in the model is another important feature for modelling of UD composites. This is because working with a microscale representation, traditional macroscale domain loading might not be appropriate. Therefore, here are my suggestions for the load cases:

• Always model a uniaxial and shear load case simulations as shown in the figure above
• For the shear case, give priority to pure shear representation rather than a simple shear. The former gives more accurate results especially in terms of constitutive behaviour of the composite.
• Due to heterogeneous nature of the microstructure, a periodic boundary condition (PBC) is recommended always. If you are looking to limit the effect of boundary conditions on the model predictions, then a PBC is always preferred.
• Think about about the history outputs in the model, in which case you might have to apply your loads using a reference point with a linear constraint (*Equation) as shown in the figure above.

Step 3: Choose the right simulations and analysis steps to drive the modelling

The objective of the UD modelling using a microscale model is to be able to extract a set of effective properties from the simulation. To do this, you have to choose the right simulations and analysis steps to model. In this regard, I will suggest the following:

• Carry out an extensive set of simulations so that you are able, at the end to generate the whole set of effective properties for such a composite. The figure above shows the sort of uniaixal and shear simulations you can undertake.
• The choice between static or dynamic analysis; implicit or explicit analysis; damage or elastic simulations, will always depend on what outcome you are interested in generating.
• It is always good to start with a simple elastic analysis for all constituents so that in the end you can compare your numerical predictions with analytical micromechanics lamination theory to build confidence in your simulations.

Step 4: Always extract stress-strain diagrams and effective properties.

The objective should always be to extract stress-strain plots. This is why during setting up of the model, you have to focus on what history outputs to track so that you can use those to generate the stress-strain plots. Typically, I will do the following:

1. Create a reference point (uniaxial) or two reference points (shear) to use for applying loads on the model. It should be a point located away from the virtual domain of the composite.
2. Make the reference point a nodal set in the Assembly module of your simulation.
3. Create a linear constraint (*EQUATION) link between the reference point nodal setand nodal set associated with the surfaces of the model.
4. Apply the load on the reference nodes.
5. Then ask ABAQUS to give you a history output associated with the set of the reference point. This can be reaction forces (RF1, RF2, RF3) and displacements (U1, U2, U3).
6. Post-simulations, generate plots of reaction forces and displacements of the reference nodes. Convert force and displacements data (from simulation) by applying simple mechanics formula of stress = force divided by area and strain = displacement divided by edge length of RVE.
7. Plot the stress-strain data to generate the graph and from there your effective properties.

Conclusion

The above will be how I will go about a comprehensive simulation of mechanics of a unidirectional composite. If you want to incorporate failure, then you have to think about the underlying constitutive behaviour of the constituents.

Behind the Scenes at CM Videos

The YouTube channel reached 3000 subscribers!

I am happy to celebrate another milestone for the YouTube Channel. We reached the 3000 subscriber milestones. Again, this is amazing stuff for such a specialist niche area as ABAQUS and computational modelling. It is good to also see that it took me nearly a year to reach 1000 subscribers and about 8 months to reach 2500 subscribers and this latest achievement of reaching 3000 subscribers was achieved in just 4 months. It speaks to the growth of the channel and the fact that you all continue to support the vision, watch the videos and engage with the content.

I want to once again, express my since gratitude to you all for always being there and making this YouTube community a really vibrant and rewarding experience for me and for CM Videos Insiders like yourselves.

Let us now start working towards 3500 subscribers. If we keep the momentum, I will be writing about this again in the June Newsletter. Fingers crossed!

Quote of the Week

Cryptocurrencies and Ancient Athens

This week, I got in the mail this book: Record 2022 which is a yearly publication from Jesus College Oxford - my former college at Oxford University. It was good to read what is going on in the college and see names of new graduates. I greatly enjoyed the opinion pieces that were in the book. The quote for the week is taking from an opinion piece by Vili Lehdonvirta, who wrote on Cryptocurrencies and Ancient Athens. It was a fasinating piece.

In this highly revealing opinion piece, Vili contrasted the challenge of cryptocurrency (invention and evolution) with the challenge of ancient Athens where landlords were enslaving many through exorbitant rents. To change this, a self-directed, publicly regulated framework was invented by Solon who designed a better system requiring competing landowners to rely on trustworthy officers of state to police the rental market with fairness to renters enforced. In particular, Solon's system integrated a mechanism to make the trustworthiness criteria of such officers not to be influenced by the antics of men. He did this using a machine he called kleroterion which is basically an allotment machine. The machine randomly selected or rejected individuals that served as check trustworthy officers to check the activities of the landlords.

This system is similar to the crytocurrencies mechanism invented by one Satoshi Nakamoto who in response to the financial crisis of 2008 wanted to create a system that is decentralized and self-regulated. This gave birth to the blockchain and the cryptocurrencies. It had to incorporate mining of a crytographic code by computer administrators on the blockchain to help in self-regulation and independent appointment of administrators to check the ledgers associated with the blockchain.

This really fascinated me as I read it this week because history does always seem to repeat itself. Students of history are better positioned to respond to events in our times because the reference to history always provides a framework in which current events are assessed and responded to.

This brings us to the end of this week's newsletter. Thank you anyone who managed to read to the end. Let me know in a reply to this message any feedback you want me to be aware of.

Have a lovely weekend and we will catch up again next weekend.

Cheers!

​