Walkthrough_Notebook.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "34005242",
   "metadata": {},
   "source": [
    "# Walkthrough of features created during summer project\n",
    "\n",
    "This notebook has three principle aims:\n",
    "1. To explain the scientific background and motivation behind the code written over the summer, and set the context for the result that you will be reproducing.\n",
    "2. To highlight that the code written is in many ways system agnostic; it can be generalised to many different materials, interatomic potentials and fracture systems (i.e, if you were a potential researcher this code would be easy to adapt to your specific problem).\n",
    "3. To provide the specific instructions needed to reproduce the result obtained from NCFlex.\n",
    "\n",
    "The code provided is part of the matscipy package, and therefore is held to a high standard of documentation and has unit tests which serve to test it thoroughly."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c1d57b8d",
   "metadata": {},
   "source": [
    "## An introduction to fracture\n",
    "Fracture is something that we as humans actually have a pretty intuitive grasp on. If you pull something brittle hard enough, it breaks, and (theoretically) if you were to push two clean surfaces of a broken material together, bonds should reform and it should heal. You can rationalise this using an energy argument - as you pull a material, what you're doing is storing elastic potential energy by stretching all of the atomic bonds a little. Growing a crack relieves this elastic energy - but at the cost of creating new surfaces, which are energetically unfavourable, so fracture will only occur when the crack growth can relieve enough energy to overcome this surface cost.\n",
    "\n",
    "Why do we care about modelling fracture? Well, pretty simply, everything breaks, and pretty much all fracture modelling and prediction that exists today is either completely or semi-empirical. The aim of this work is to completely do away with using experimental data in fracture models (as fracture experiments are quite rough at the best of times) and instead build a bottom-up way to model fracture which derives its information directly from first-principles. In the future, the idea is that you could predict key fracture properties such as the crack initation stress and crack path selection completely from a model without having to do any experiment at all - thus saving a huge amount of money (as fracture testing is expensive!).\n",
    "\n",
    "We can seperate fracture out into two main regimes. These are:\n",
    "1. Slow fracture, which occurs in a material very close to its breaking point, which has not yet critically failed. This is similar concept to diffusion - the crack advances by 'jumping over' energy barriers, using energy from thermal excitations in the material, and progesses very slowly. This notebook is focused on this regime.\n",
    "\n",
    "2. Fast (or dynamic) fracture, which occurs when a material is pulled to stress intensity above its fracture toughness. In this case, crack motion is very fast (usually a large fraction of the speed of sound in the material). You will be doing some dynamic fracture simulations in LAMMPS to attempt to reproduce the fracture toughness found from this notebook. (Don't worry - nearly all the code is provided)."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ed19b57f",
   "metadata": {},
   "source": [
    "## Difficulties with modelling atomistic fracture\n",
    "Modelling atomistic fracture well is really, really difficult. These difficulties can essentially be broken up into two categories - those that are interatomic potential related and those that are boundary conditions related.\n",
    "\n",
    "### Inter-atomic potential problems\n",
    "At the very tip of a crack in a brittle material, bonds are breaking, which is a fairly extreme environment for atoms to be in. To get accurate forces between atoms, you need either a highly specialised inter-atomic potential, or a first principles method like density functional theory (DFT). For many years, inter-atomic potentials simply weren't good enough, meaning that the most successful method for modelling fracture was Quantum-Mechanics/Molecular-Mechanics or QM/MM, which is a method where an inner region around the crack tip is described using ab-initio molecular dynamics with DFT, and this is coupled through a boundary to a much larger number of atoms which are described using a cheap interatomic potential. Whilst this method was certainly successful, it was limited to very short timescales by the intense cost of computing an ab-initio calcuation every timestep. \n",
    "\n",
    "Recently however, the development of machine learned interatomic potentials such as the atomic cluster expansion (ACE) type potentials have revolutionised the field of atomic modelling, including fracture, as they are capable of an accurate description of the bond breaking process whilst being cheap enough for much longer time and length scale simulations. Detailed information about these are outside the remit of this notebook, but all simulations being done today will use a carbon ACE potential, which has been shown to be good at describing both the bond breaking and surface reconstruction processes necessary for describing fracture accurately [1].\n",
    "\n",
    "### Boundary condition problems\n",
    "When we model an atomistic domain, we'd like to be modelling it as if it were small part of a much larger structure, to reflect a relevant real life situation. This requires *flexible boundary conditions* which essentially allows the boundary to update as the crack moves, mimicing being a cut-out of a much larger system. However, before now, the majority of fracture simulations used *static boundary conditions* which essentially can be thought of as considering the entire fracture system to be the size of the modelled atomistic domain - on the order of just a few nm. As you can imagine, fracture in a nano-particle is not reflective of fracture in a large body, so the results from these static boundary condition simulations are inaccurate.\n",
    "\n",
    "To model flexible boundary conditions, we essentially make an assumption, which is that past some radial distance away from the crack tip in our atomistic domain, the displacement of atoms follows that predicted from the continuum solution around a crack tip. Then, as the crack moves, we update the positions of these boundary atoms according to the continuum solution with the new updated crack tip position as the origin.\n",
    "\n",
    "A lot of the work done this summer was to ensure that this approximation holds - that is, that the displacements of atoms around a crack tip tend to the continuum solution sufficiently quickly for flexible boundary conditions to work and provide good results. This work will be highlighted here, and there are essentially two parts. The first part is all about something called the Cauchy-Born shift corrector in multi-lattices, which is relevant for diamond. The second part is about surface relaxation. \n",
    "\n",
    "### Cauchy-Born corrector shift in multi-lattices\n",
    "Diamond is a multi-lattice. Which means that it's a non-Bravais lattice, as it has two atoms in it's primitive unit cell. You can essentially think of the diamond structure as two inter-penetrating FCC lattices, which we'll call 'sublattices'. Remember a second ago when I said that flexible boundary conditions only work when 'the atomic displacements tend to the continuum solution' - well that doesn't actually work in multi-lattices. For Bravais lattices, mapping a contiuum displacement field to some atoms is totally easy - you just displace the atoms pretending they're points in a continuous space deforming according to the continuum solution (this is the 'Cauchy-Born approximation'). However, for multi-lattices, if you do this, you find that the atoms have high forces on them. This is because there is an extra degree of freedom that exists between sublattices - one sublattice can move relative to the other - which simply can not be known from the continuum solution, and has to be found by relaxing the atoms under the influence of an interatomic potential. \n",
    "\n",
    "Uh oh! This is a BIG problem. The whole point of flexible boundary conditions is that we can pretend we are modelling a small part of a much larger domain - but if in order to actually displace atoms according to the continuum solution at the boundary we have to relax every atom, it makes the method essentially useless. So, how do we fix this? Well, we need some way to predict the difference between the true atomic positions and the Cauchy-Born approximation predicted positions. This difference is called the Cauchy-Born corrector shift.\n",
    "\n",
    "This was implemented in matscipy through the use of a regression model. We first need to fit the model to be able to predict the Cauchy-Born predictor shifts, and then we can test it. This section is a walk through of how you can interface with code that I've written to do this for a simple case."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "5e059145",
   "metadata": {
    "code_folding": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "optimising lattice parameter\n",
      "       Step     Time          Energy         fmax\n",
      "*Force-consistent energies used in optimization.\n",
      "LBFGS:    0 16:00:27      -73.005280*       0.6081\n",
      "LBFGS:    1 16:00:27      -72.979340*       2.5741\n",
      "LBFGS:    2 16:00:27      -73.006763*       0.0115\n",
      "LBFGS:    3 16:00:27      -73.006764*       0.0002\n",
      "LBFGS:    4 16:00:27      -73.006764*       0.0000\n",
      "optimised lattice parameter: 3.5658038209718814\n"
     ]
    }
   ],
   "source": [
    "#import modules and set up calculator\n",
    "import numpy as np\n",
    "from ase.build import bulk\n",
    "from ase.lattice.cubic import Diamond\n",
    "from ase.optimize.precon import PreconLBFGS\n",
    "from ase.constraints import ExpCellFilter\n",
    "from matscipy.cauchy_born import CubicCauchyBorn\n",
    "from ase.calculators.lammpslib import LAMMPSlib\n",
    "from ase.optimize import LBFGS\n",
    "\n",
    "#set up calculator\n",
    "cmds = ['pair_style hybrid/overlay pace table spline 10000',\n",
    "        'pair_coeff * * pace c_v3.yace C', 'pair_coeff * * table d2.table D2 9.0']\n",
    "calc = LAMMPSlib(amendments=[\"mass 1 12.101\"],lmpcmds = cmds,log_file='test.log',keep_alive=True)\n",
    "\n",
    "#find eq lattice constant\n",
    "el              = 'C'\n",
    "a0_init         = 3.56\n",
    "print('optimising lattice parameter')\n",
    "unit_cell = Diamond(size=[1,1,1],symbol=el,latticeconstant=a0_init,pbc=(1,1,1))\n",
    "unit_cell.set_calculator(calc)\n",
    "ecf = ExpCellFilter(unit_cell)\n",
    "uc_optimise = LBFGS(ecf)\n",
    "uc_optimise.run(fmax=0.0001)\n",
    "a0 = unit_cell.get_cell()[0,0] #get the optimised lattice parameter\n",
    "print('optimised lattice parameter:',a0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "cfcd3e28",
   "metadata": {},
   "outputs": [],
   "source": [
    "#define a function which we can use to impart a strain field to a set of atoms\n",
    "def E_cart3D(x, y, z, eps=None):\n",
    "    #Here, this function is extremely simple, it just takes a strain vector (eps) and converts it into a set of\n",
    "    #matrices, which are the same for each atom (as the applied strain is uniform).\n",
    "    #In more general use, for example within the fracture code,\n",
    "    #this function would take a set of atomic positions (x,y,z) and evaluate the continuum strain field\n",
    "    #at those points, and return the strain state in matrix form for each atom.\n",
    "    E = np.zeros([len(x), 3, 3])\n",
    "    E[:, 0, 0] = eps[0]\n",
    "    E[:, 1, 1] = eps[1]\n",
    "    E[:, 2, 2] = eps[2]\n",
    "    E[:, 1, 2], E[:, 2, 1] = eps[3], eps[3]\n",
    "    E[:, 0, 2], E[:, 2, 0] = eps[4], eps[4]\n",
    "    E[:, 0, 1], E[:, 1, 0] = eps[5], eps[5]\n",
    "    return E\n",
    "\n",
    "#define a function which can be used to test a fitted regression model\n",
    "def model_prediction(dirs, eps, method, cb, E_func, coordinates='cart3D', atol=1.5e-5):\n",
    "    #this function allows us to evaluate the true shift and predicted shift for a single strain vector, in\n",
    "    #order to compare how good the regression model is\n",
    "    \n",
    "    #create the rotation matrix from the specified directions\n",
    "    A = np.zeros([3, 3])\n",
    "    for i, direction in enumerate(dirs):\n",
    "        direction = np.array(direction)\n",
    "        direction = (1/np.linalg.norm(direction))*direction\n",
    "        A[:, i] = direction\n",
    "\n",
    "    # generate rotated unitcell (according to dirs)\n",
    "    atoms = Diamond(cb.el, latticeconstant=cb.a0,\n",
    "                    size=[1, 1, 1], directions=dirs)\n",
    "    atoms.set_pbc([True, True, True])\n",
    "    atoms.calc = calc\n",
    "    atoms_copy = atoms.copy()\n",
    "    cb.set_sublattices(atoms, A)\n",
    "    \n",
    "    #get the true cauchy-born corrector shift\n",
    "    shift_diff_true = cb.eval_shift(\n",
    "        eps, atoms)\n",
    "    #get the model predicted shift\n",
    "    shifts = cb.predict_shifts(\n",
    "        A, atoms, E_func=E_func, eps=eps, method=method, coordinates=coordinates)\n",
    "    shift_diff_predicted = shifts[0, :]\n",
    "    \n",
    "    return shift_diff_true,shift_diff_predicted"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e322f37a",
   "metadata": {},
   "source": [
    "Now we've defined some functions we can use to apply a continuum strain field to a single diamond unit cell, and then measure the true Cauchy-Born corrector shift and compare it to our model prediction. It's now time to initialise and fit our model, which is done simply by creating a CubicCauchyBorn object from the matscipy package, which was created and is maintained by me - if you'd like to have a look at the code and documentation, please see https://github.com/libAtoms/matscipy/blob/master/matscipy/cauchy_born.py"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "5a34e1bb",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Performing initial regression fit for CB corrector.....\n",
      "Fit completed\n"
     ]
    }
   ],
   "source": [
    "#Initialise the Cauchy-Born shift corrector object\n",
    "cb = CubicCauchyBorn(el, a0, calc, lattice=Diamond)\n",
    "\n",
    "#Fit the Cauchy-Born corrector model\n",
    "cb.initial_regression_fit()\n",
    "\n",
    "#Now, test how good it is at predicting the cauchy-born shift correctors for\n",
    "#shear strains between -0.02 and 0.02, which is a reasonable range for fracture,\n",
    "#for a rotated diamond unit cell\n",
    "dirs = [[1, 1, 0], [-1, 1, 0], np.cross([1, 1, 0], [-1, 1, 0])]\n",
    "eps_vals = np.linspace(-0.02,0.02,100)\n",
    "true_shifts = np.zeros([len(eps_vals),3])\n",
    "predicted_shifts = np.zeros([len(eps_vals),3])\n",
    "\n",
    "for i,eps in enumerate(eps_vals):\n",
    "    eps_vec = [0,0,0,eps,-eps,eps]\n",
    "    true_shift, predicted_shift = model_prediction(dirs, eps_vec, 'regression', cb, E_cart3D)\n",
    "    #print(true_shift,predicted_shift)\n",
    "    true_shifts[i,:] = true_shift\n",
    "    predicted_shifts[i,:] = predicted_shift"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "2ded9587",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "\n",
    "fig,axs = plt.subplots(2,1)\n",
    "fig.suptitle('Cauchy-Born model predictions vs True shift')\n",
    "axs[0].scatter(predicted_shifts[:,0],true_shifts[:,0],label='predicted shift x component')\n",
    "axs[0].plot(true_shifts[:,0],true_shifts[:,0],'r--',label='true shift x component')\n",
    "axs[0].set_xlabel(r'Model predicted shifts $\\AA$')\n",
    "axs[0].set_ylabel(r'True shifts $\\AA$')\n",
    "axs[0].legend()\n",
    "axs[1].scatter(predicted_shifts[:,1],true_shifts[:,1],label='predicted shift y component')\n",
    "axs[1].plot(true_shifts[:,1],true_shifts[:,1],'r--',label='true shift y component')\n",
    "axs[1].set_xlabel(r'Model predicted shifts $\\AA$')\n",
    "axs[1].set_ylabel(r'True shifts $\\AA$')\n",
    "axs[1].legend()\n",
    "fig.set_tight_layout(True)\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "398d9717",
   "metadata": {},
   "source": [
    "As you can see, the model predictions are remarkably good - and match the true shifts very well, particularly for small shifts. In the context of fracture, this allows for a much faster rate of force decay as we move away from the tip of a crack, allowing our simulations to converge, when previously they would not. The fact that the model shows deviation at large shifts also is not really a problem - this is because the region near the crack tip is relaxed explicitly, and we only need this model to work well in the far field where strains are very small anyway. There are some figures showing this increased rate of decay in my report."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4cc5887d",
   "metadata": {},
   "source": [
    "### Surface reconstruction in fracture\n",
    "\n",
    "A major violation of the approximation mentioned above - that forces around the tip of a crack decay to 0 as they approach the boundary - are the two surfaces formed by fracture. As surface relaxation is highly non-linear, the continuum solution is unable to describe them even close to correctly, leading to giant forces that do not decay with distance from the crack tip. To solve this problem, code was written which was able to essentially map out the surface relaxation in a small unit cell of exposed surface in the crystal, and 'paste' this relaxation onto the exposed surfaces directly. This greatly reduced the surface forces. Functionally, this code is very easy to use - let's look at an example."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "43d3905a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "176259c4f12f4bb4a6c3b3c6a7e83031",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(NGLWidget(), VBox(children=(Dropdown(description='Show', options=('All', 'C'), value='All'), Dr…"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from ase.visualize import view\n",
    "#First, let's make a large supercell of a structure and view it\n",
    "surf_dir = 1\n",
    "size = [10,20,10]\n",
    "directions = [[1, 1, 0], [-1, 1, 0], [0, 0, 1]]\n",
    "\n",
    "#create the rotation matrix from the specified directions\n",
    "A = np.zeros([3, 3])\n",
    "for i, direction in enumerate(dirs):\n",
    "    direction = np.array(direction)\n",
    "    direction = (1/np.linalg.norm(direction))*direction\n",
    "    A[:, i] = direction\n",
    "\n",
    "shift_vec = np.zeros([3])\n",
    "shift_vec[surf_dir] += 0.01\n",
    "\n",
    "bulk = Diamond(directions=directions, size=size,\n",
    "               symbol=el, latticeconstant=a0, pbc=(1, 1, 1))\n",
    "\n",
    "#as this is a multilattice - we need to determine the sublattices. We can do this using a cauchy_born object\n",
    "cb.set_sublattices(bulk,A)\n",
    "\n",
    "bulk.positions += 0.01\n",
    "bulk.wrap()\n",
    "\n",
    "\n",
    "\n",
    "cell = bulk.get_cell()\n",
    "cell[surf_dir, :] *= 2  # vacuum along surface axis (surface normal)\n",
    "slab = bulk.copy()\n",
    "slab.set_cell(cell, scale_atoms=False)\n",
    "pos = slab.get_positions()\n",
    "\n",
    "sorted_surf_indices = np.argsort(pos[:, surf_dir])\n",
    "\n",
    "# apply the surface reconstruction to the top surface only\n",
    "surface_atom = np.argmax(pos[:, surf_dir])\n",
    "surface_coords = pos[surface_atom, :]+np.array([0.01, 0.01, 0.01])\n",
    "slab.set_calculator(calc)\n",
    "initial_force = slab.get_forces()\n",
    "slab_before = slab.copy()\n",
    "view(slab, viewer='ngl')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d5c405f7",
   "metadata": {},
   "source": [
    "We can have a look at the surface forces in this big slab before a relaxation is mapped and applied. As you will see, the surface force is high."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "13ba399c",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "max force before 4.414600202281976\n"
     ]
    }
   ],
   "source": [
    "#print initial large max surface force\n",
    "natoms = len(slab)\n",
    "maxfbefore = np.max(\n",
    "    np.abs(initial_force[sorted_surf_indices[int(natoms/2):natoms]]))\n",
    "print(f'max force before {maxfbefore}')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c64fb170",
   "metadata": {},
   "source": [
    "Now, let's map out the reconstruction by calling the 'map_surface' function from the SurfaceReconstruction object. This object is also part of matscipy, and was created by me - please feel free to look at the code and documentation here: https://github.com/libAtoms/matscipy/blob/master/matscipy/surface_reconstruction.py"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "11fab2e0",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "PreconLBFGS:   0  16:02:17     -357.027432       4.5539\n",
      "PreconLBFGS:   1  16:02:17     -357.393159       3.5087\n",
      "PreconLBFGS:   2  16:02:17     -357.499978       2.4410\n",
      "PreconLBFGS:   3  16:02:17     -357.549429       2.8353\n",
      "PreconLBFGS:   4  16:02:17     -357.617982       2.4494\n",
      "PreconLBFGS:   5  16:02:17     -357.701595       1.3530\n",
      "PreconLBFGS:   6  16:02:17     -357.738395       0.9105\n",
      "PreconLBFGS:   7  16:02:17     -357.754389       0.6444\n",
      "PreconLBFGS:   8  16:02:17     -357.762584       0.4998\n",
      "PreconLBFGS:   9  16:02:17     -357.766682       0.4825\n",
      "PreconLBFGS:  10  16:02:17     -357.769579       0.4162\n",
      "PreconLBFGS:  11  16:02:17     -357.772927       0.3178\n",
      "PreconLBFGS:  12  16:02:17     -357.777014       0.2187\n",
      "PreconLBFGS:  13  16:02:17     -357.780417       0.1810\n",
      "PreconLBFGS:  14  16:02:17     -357.781655       0.1100\n",
      "PreconLBFGS:  15  16:02:17     -357.782011       0.1043\n",
      "PreconLBFGS:  16  16:02:17     -357.782481       0.0744\n",
      "PreconLBFGS:  17  16:02:17     -357.783024       0.0741\n",
      "PreconLBFGS:  18  16:02:17     -357.783095       0.0653\n",
      "PreconLBFGS:  19  16:02:17     -357.783133       0.0527\n",
      "PreconLBFGS:  20  16:02:17     -357.783199       0.0450\n",
      "PreconLBFGS:  21  16:02:17     -357.783271       0.0351\n",
      "PreconLBFGS:  22  16:02:17     -357.783279       0.0351\n",
      "PreconLBFGS:  23  16:02:17     -357.783292       0.0319\n",
      "PreconLBFGS:  24  16:02:17     -357.783315       0.0263\n",
      "PreconLBFGS:  25  16:02:17     -357.783336       0.0290\n",
      "PreconLBFGS:  26  16:02:17     -357.783363       0.0075\n",
      "PreconLBFGS:  27  16:02:17     -357.783365       0.0040\n",
      "PreconLBFGS:  28  16:02:17     -357.783366       0.0034\n",
      "PreconLBFGS:  29  16:02:17     -357.783367       0.0007\n",
      "PreconLBFGS:  30  16:02:17     -357.783367       0.0009\n",
      "PreconLBFGS:  31  16:02:17     -357.783367       0.0008\n",
      "PreconLBFGS:  32  16:02:17     -357.783367       0.0007\n",
      "PreconLBFGS:  33  16:02:17     -357.783367       0.0007\n",
      "PreconLBFGS:  34  16:02:17     -357.783367       0.0006\n",
      "PreconLBFGS:  35  16:02:17     -357.783367       0.0005\n",
      "PreconLBFGS:  36  16:02:17     -357.783367       0.0005\n"
     ]
    }
   ],
   "source": [
    "from matscipy.surface_reconstruction import SurfaceReconstruction\n",
    "#Now, map the surface shift in a single unit cell very cheaply\n",
    "sr = SurfaceReconstruction(\n",
    "                    el, a0, calc, dirs, surf_dir, lattice=Diamond, multilattice=True)\n",
    "sr.map_surface(fmax=0.0005, shift=0)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8a1134ff",
   "metadata": {},
   "source": [
    "Now, we can just map this single surface unitcell reconstruction directly onto a larger surface. You will see that after mapping, the forces after have gone down"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "7fb9c49f",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Now apply this shift to the surface, and visualise it again\n",
    "sr.apply_surface_shift(slab, surface_coords, cb=cb, xlim=None,\n",
    "                       ylim=None, zlim=None, search_dir=-1, atoms_for_cb=bulk)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "3d3eb80c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "bb30f8935a544ea78a2eff1565c2f7a0",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(NGLWidget(max_frame=1), VBox(children=(Dropdown(description='Show', options=('All', 'C'), value…"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "final_force = slab.get_forces()\n",
    "view([slab_before,slab], viewer='ngl')\n",
    "#Use the slider to look at the structures before and after the surface reconstruction is mapped!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "6c76fac9",
   "metadata": {
    "code_folding": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "max force after 0.00313143039105717\n"
     ]
    }
   ],
   "source": [
    "#Get the final force and show it's much lower\n",
    "natoms = len(slab)\n",
    "maxfafter = np.max(\n",
    "    np.abs(final_force[sorted_surf_indices[int(natoms/2):natoms]]))\n",
    "print(f'max force after {maxfafter}')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4197e485",
   "metadata": {},
   "source": [
    "Clearly, the surface force is now much lower - and we've not had to relax anything except from a small unit cell of surface, meaning the computational cost is very cheap; much cheaper than relaxing the full surface. This code is fully general for any 1x1 reconstruction in any material, and is also implemented in matscipy for specific cases for symmetry breaking reconstructions including the 2x1 Pandey diamond 111 reconstruction, and the silicon 110 3x1 reconstruction."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "643a7f7f",
   "metadata": {},
   "source": [
    "# A full NCFlex simulation and fracture toughness\n",
    "\n",
    "The details of how NCFlex works are covered in the report - but in brief, you can think of it as a way to explore the domain where slow (non dynamic) fracture is possible. Before we go any further, we need to define the stress intensity factor $K$. You can think of $K$ as describing how an applied far field stress at the boundary of a material far from the crack is amplifed by the geometry of a given crack. Dynamic fracture occurs when $K$ reaches a value $K_{+}$ which is a material dependent constant, called the fracture toughness. We can find this using NCFlex by finding the largest stable $K$ value allowed in the regime where slow fracture is possible. Running a full NCFlex simulation is straightforward but time consuming and often requires a few hours. I'm very aware that you'd rather be doing something else right now, so here we'll just load a figure from previous calculations that were done."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "9641a0b6",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from IPython.display import Image\n",
    "Image(filename=\"NCFlex_figure.png\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cea7c72d",
   "metadata": {},
   "source": [
    "The different curves here represent different region I radii used in NCFlex. The idea is that the greater the region I radius, the less the finite size effect and subsequently the more accurate the simulation is, but the more expensive. From this curve, the maximum value of $K$ that the crack can be stable for ($K_{+}$) is 3.83$\\pm$0.03 MPa$\\sqrt{\\mathrm{m}}$."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9dcbfcf4",
   "metadata": {},
   "source": [
    "# Reproducing this result\n",
    "\n",
    "Now it's your turn! It's time to reproduce this result by looking at dynamic fracture. Essentially, what I'd like you to do is run a number (~5) of dynamic fracture simulations where you apply a set value of $K$ and measure the crack speed, and then plot a graph of crack speed vs $K$. What you should see is that once $K$ is larger than approximately $K_{+}$, the crack speed should massively increase. Ideally, by taking a few measurements, you should be able to pinpoint a rough value of $K$ where this occurs, and this should line up with the predicted value of $K_{+}$ presented here. \n",
    "\n",
    "The simulation you'll be running will be a 'thin-strip' type fracture configuration and it'll be at finite temperature. This means that the $K_{+}$ you'll measure will be slightly lower than that predicted here for 0 Kelvin - this is caused by a mixture of different boundary conditions and the fact that at finite temperature, you have to pull slightly less hard to overcome the energy barrier for fracture.\n",
    "\n",
    "I've written all the code to do this for you already. On 40 cores, each of the runs should take you ~5 minutes."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1e5d4d72",
   "metadata": {},
   "source": [
    "# Detailed instructions for running the LAMMPS MD simulations"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6340909e",
   "metadata": {},
   "source": [
    "In the folder this notebook is present in, there should be a seperate folder named LAMMPS_investigation. In your IDE of choice, navigate to this folder and open the python file params.py. This is the file which provides the information to set up the crack geometry for the MD. At the top of this file, there is an area where you can change the value of K to be supplied to the simulation. I'd like you to test the following values of K:\n",
    "\n",
    "K values to test:\n",
    "3.75, 3.78, 3.81, 3.84, 3.87\n",
    "\n",
    "Once you've put in a value of K, save the params.py file. Now open the run_simulation.sh script. In this script, adjust the number of processors you'd like to use to run the MD. Once that's done, save the script and make it runable with \n",
    "```\n",
    "chmod u+x run_script.sh\n",
    "```\n",
    "and then run \n",
    "```\n",
    "./run_script.sh\n",
    "```\n",
    "After about 5 minutes (on 40 cores), the last lines printed to terminal at the end of the simulation will be the crack speed. Record this for the value of K you put in, and then re-run the simulation for each value of K above. Plot a graph of crack speed against K - it should look like a line that fluctuates around 0 and then steeply jumps up once $K$>$K_{+}$. If you'd like to see what the MD is doing, run the box below in this notebook, and move the slider to see the crack grow. \n",
    "\n",
    "For the result (and it's error), look at the two $K$ values you've taken either side of the point where the crack speed jumps up. These will be your error bounds for $K_{+}$, with the actual $K_{+}$ value being the average of these values. In this case this is a good way to estimate error as one of the side effects of introducing finite temperature is a 'smearing effect' - at 0K, $K_{+}$ is a sharp transition from stationary to fast crack, but at finite temperature this transition is more gradual. If you'd like, feel free to take more measurements to try and refine your guess for $K_{+}$. The final value should be a little lower than the result found from NCFlex above, probably around 3.78 MPa$\\sqrt{\\mathrm{m}}$. This is because of epistemic uncertainty associated with the difference in boundary conditions and finite temperature."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "id": "7161c230",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "65e678ef61d24057836c982554afe72f",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(NGLWidget(max_frame=10), VBox(children=(Dropdown(description='Show', options=('All', 'C'), valu…"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#USE THIS BOX TO VISUALISE THE OUTPUT OF AN MD TRAJECTORY\n",
    "import ase\n",
    "traj = ase.io.read('LAMMPS_investigation/dump.lammpstrj',index=':')\n",
    "for t in traj:\n",
    "    t.symbols = 'C23040'\n",
    "view(traj, viewer='ngl')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2d93cb48",
   "metadata": {},
   "source": [
    "### References\n",
    "[1] J. Chem. Theory Comput. 2023, 19, 15, 5151–5167"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "myvenv",
   "language": "python",
   "name": "myvenv"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}