How to define contacts with erosion in Abaqus | Contact and Element deletion

[@aaroncanepa6852]

wow amazing, thx for the video, greetings from Peru

[@tecnodigitalschool]

You're welcome!

[@vonClausewitzz]

I was looking for this. Explanation is clear and easy. I think this the true approach for erosion contact. Thanks!

[@tecnodigitalschool]

Great to hear!

[@saurabhtripathi2598]

Great work...

[@tecnodigitalschool]

Thanks a lot!

[@saurabhtripathi2598]

@@tecnodigitalschool Hi, could you suggest best hardware for workstation using abaqus....

[@tecnodigitalschool]

@@saurabhtripathi2598 Heavy simulations usually are run in HPC (high performance cluster or computer center) using multiple CPUs. Going for 12, 16 or 24 cpus is already quite powerful for finite element simulations, which have some parallelization limitations compared to CFD. If you run Abaqus/Standard, therefore solving huge systems of linear equations, you can accelerate the simulation even more using a GPU (compatible with Abaqus). However, these GPUs need to be NVIDIA Quadro, which are in another level compared to the more widely known RTX.

[@dhanraj1992]

Excellent video, please keep doing such 🙂

[@tecnodigitalschool]

Thanks a a lot!

[@diegodiaz4664]

Muchas gracias por el contenido!!

[@tecnodigitalschool]

De nada! Gracias a ti!

[@aaronocampo7479]

do you know how to do this for a CEL model?

[@krystopermilan1466]

Nice explanation! Could you please share the inp files at your site to check the syntax? I am not using abaqus as preprocessor. Thanks!

[@tecnodigitalschool]

I have added the input files in the blog post. Enjoy!

[@timosaksala4797]

Nice but the erosion approach suffers from the major drawback of violating one of the first principles of physics, the conservation of mass and matter. So it is recommendable to use, instead of deletion, conversion of destroyed elements into particles having the mass of the original elements. Abaqus has the FEM-SPH for this.

[@tecnodigitalschool]

Absolutely!! The mass from the elements deleted is removed from the simulation, so in the case of the projectile it will have a significant influence if this loss of mass is not negligible. Element removal is an artifice used mainly for two reasons: 1) To represent discontinuities (cracks) in progressive failure analysis. This is very common in Continuum Damage Mechanics models. 2) To prevent convergence issues when elements are too damaged (softened) and they take huge deformations. Before the element is too distorted and the simulation crashes, we delete that element. Obviously, deleting matter is not physically representative, and that is something we have to take into account. Indeed, there are other modeling techniques to deal with this issue: - If the projectile is relatively brittle, it will break appart, but we don't want to delete those damaged elements (we need their mass). In this case, we can follow the technique proposed, using Smoothed Particle Hydrodynamics (SPH) to model the impactor or some variant of this approach. This is commonly used to model bird impact or ice impact on aerospace components. - If the projectile is extremely ductile (lead, brass...), it will not break appart, but it will take huge deformations. Using regular solid elements (lagrangian mesh) will result in the simulation crashing because of the huge element distortion. Again, we don't want to delete elements because we need their mass. The option in this case is to use an Eulerian mesh for the projectile (and Lagrangian for the target), so we will solve a CEL (Coupled Eulerian-Lagrangian) problem. In many other cases, when the deformation nor erosion of the projectile is not too severe, it is enough using a regular solid mesh. But of course, this depends on each case and requires critical thinking. Thanks for your comment Timo!

[@rachit.23.93]

Nice video sir. Can we contact via email i need some help with defining eroding surfaces in input file. Not interiror surfaces but eroding surfaces.

[@tecnodigitalschool]

Sure, send an email to [e-mail]

[@rachit.23.93]

@@tecnodigitalschool ok thank you sir.