JPH08168937A - Hyperfine processing supporting system and hyperfine processor - Google Patents

Abstract

PURPOSE: To drastically reduce the simulation performing time by providing a potential function part for deriving force between atoms for reproducing a mechanical property of a solid crystal and a calculating part for performing the deduction wherein the dynamic process of the atom is followed by the molecular dynamical method on the basis of the condition set in the condition setting part. CONSTITUTION: Input of the processing condition is performed in a condition setting part 11. In a potential function part 12, a parameter 21 of copper of a two-body potential function having the small calculating amount, a parameter 22 of copper of an EAM potential function having the large calculating amount, and having high precision, and a condition 23 for switching two potentials are input. The calculation of simulation of the processing is performed by a calculating unit 13 by using these conditions. The coordinates of respective atoms in respective times, output from the calculating unit 13 are sequentially held by a data holding unit 14.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention is based on an ultrafine processing support system for simulating mechanical characteristics such as cutting, grinding, deformation, polishing, friction and wear phenomena which are sliding phenomena occurring at solid interfaces, and simulation results. The present invention relates to an ultrafine processing apparatus for processing solids and evaluating characteristics.

[0002]

2. Description of the Related Art Generally, in the molecular dynamics method applied to solid state physical properties, the force received by each atom is calculated from the potential gradient. There are various potential functions used, ranging from extremely simple empirical ones to those obtained by calculating electronic states from the first principle. As for these characteristics, although an empirical potential function such as a two-body potential can obtain basic mechanical properties, it is difficult to reproduce various physical properties. On the other hand, the potential function using the first principle can accurately describe the physical properties by obtaining the electronic state peculiar to the substance. There is little empirical potential for the amount of calculation, and the amount obtained from the first principle is extremely huge. In reality, the potential function to be used is carefully selected depending on the balance between accuracy and the amount of calculation in accordance with the purpose of simulation. To simulate a mechanical phenomenon, the basic properties can be reproduced using the two-body potential.

However, it cannot be said that a material having low regularity including defects and crystal grain boundaries, a composite of different materials, or an alloy is sufficient. Atomic insertion method (EA
The potential function according to M) has recently begun to be used in various fields because of its accuracy and small amount of calculation. EAM potential is a kind of many-body potential, but Car &
Since a simplified calculation method is used instead of an accurate calculation using a density functional such as the first-principles molecular dynamics method such as the method of Pallinello et al., The amount of calculation that is comparable to the empirical potential is rather small. At present, it is considered optimal for simulating the mechanical properties of composites of different materials, alloys, and materials containing impurities and defects. However, it requires a calculation amount that is several to ten times that of the two-body potential, which is one of the main causes of hindering the expansion of the size of the machined structure model.
At present, a simulation having 100,000 to 1,000,000 atoms is carried out on a 1GFLOPS computer mainly using a two-body potential. In order to fully simulate the mechanical properties, it is necessary to deal with the crystal grain boundaries and crystals containing defects and dislocations, and at least 1,000,000 to 100
A model with a three-dimensional structure that handles 0,000 atoms is essential. For this purpose, it is necessary to expect improvement in computer capacity, but at the same time, there is a demand for the development of an algorithm that does not reduce accuracy.

[0004]

The development of an algorithm that does not reduce the accuracy is to reduce the amount of calculation without sacrificing the accuracy that appears in the result. The calculation during the simulation of the molecular dynamics method is roughly divided into calculation of the potential, calculation of the force received by each atom, and updating of the coordinates of the atom. The largest amount of calculation is the potential part, half of the total. From 80 to 90%. It is considered that reducing the calculation amount of this part contributes to the reduction of the total calculation amount, but the calculation accuracy of the potential is directly linked to the simulation result, which is a problem that cannot be solved simply.

[0005]

[Means for Solving the Problem] Two-body potential and EA
The M potential is positioned as a game-playing one in the compensatory relationship between the calculation amount and the accuracy. To fully simulate mechanical properties, it is necessary to consider grain boundaries of crystals, crystals containing defects and dislocations, and composites of different materials. For this purpose, the two-body potential is insufficient in accuracy, and it is required to use the EAM potential, resulting in an increase in the amount of calculation. In these calculations, the difference between the EAM potential and the two-body potential appears most when the regularity of the crystal structure is reduced and a large force is applied to a specific atom. At this time, in the two-body potential in which only the contributions from atoms existing in the region of the closest atom are considered, the error becomes very large with respect to the EAM potential. On the other hand, on the contrary, when the displacement from the crystal position is small and the resulting absolute value of the force is low,
The difference between the two is small and it can be considered that the error can be ignored.
Simulations of phenomena such as machining and destruction are characterized by the fact that there is a site with a large difference between these potentials and a site that can be ignored at the same time. Had to adopt. However, if the EAM potential that can ensure the accuracy is adopted only in the part with low crystallinity and the potential with low accuracy and a small amount of calculation is used in the part with high crystallinity, the accuracy is not deteriorated with respect to the overall result. , The overall calculation amount can be reduced. In addition, in the case of an atom that moves rapidly in the molecular dynamics method, it is necessary to take a finer time step during calculation. On the other hand, when the external force is small, it is possible to reproduce the thermal vibration in time steps. Therefore, by optimizing the time step size and the update frequency of the atomic coordinate position between the part that does not move rapidly and the part that does not, it is possible to reduce the calculation time while maintaining accuracy.

[0006]

[Operation] In order to maintain the accuracy of the simulation results,
The type of potential depends on the part of the processed structure model,
By selecting the step size of the simulation time, the calculation amount can be reduced without degrading the accuracy of the simulation result, and the simulation time can be shortened unless the machined structure model size is expanded or the machined structure model size is changed. Can be achieved.

[0007]

Embodiments of the present invention will be described below with reference to the drawings of processing simulation results. FIG. 1 is a block diagram of an ultrafine processing apparatus system according to an embodiment of the present invention. In FIG. 1, 11 is a condition setting unit having a function of inputting or changing the composition of solid material, setting of structure, and setting of processing conditions such as processing speed, cutting direction, cutting thickness, temperature of solid material, and 12 The potential function part that inputs two potential functions that derive the interatomic force that reproduces the mechanical properties of the solid crystal, and 13 obtain the force acting on each atom at each time by the molecular dynamics method, and after 1 time step A calculation unit having a function of deducing the dynamic process of the behavior of the atomic group by calculating the coordinates of each atom and repeating these processes, 14 is the coordinate of each atom at each time output from the calculation unit 13. Is a data holding unit having a function of sequentially holding, and 15 displays the result output by the calculation unit 13.
An output display unit such as a printer or a CRT display for confirmation, 16 is an image display unit having a function of displaying an analysis result image obtained based on the data stored in the data storage unit 14, and 17 is data storage It is a processing unit having a function of actually controlling ultrafine processing based on the data stored in the unit 14. The direction of the arrow represents the direction of information transmission.

FIG. 2 is an explanatory view showing the positional relationship between the machined structure and the structure to be machined in the present embodiment, where 18 is the structure to be machined and 19 is the structure to be machined. Here, for simplification of description, only the structure 18 to be processed is made of copper metal single crystal, and the processed structure 19 is made to be a completely rigid body without considering its internal structure. The surface cutting is performed by cutting the upper surface of 18 having a flat surface with 19.

First, the condition setting section 11 inputs processing conditions. The input here may be an input from an input device such as a keyboard, or may be the reading of data prepared in advance. In this embodiment, the data of the initial coordinates of each of the 930 atoms forming the structure to be processed 18, the processing speed, the shape of the structure to be processed 19, and the relative relationship between the structure to be processed 18 and the structure to be processed 19. The position coordinates and were read. Instead of reading the data of the initial coordinates of the atom, the shape, element, composition, crystal orientation, etc. may be input. Next, in the potential function unit 12, a copper parameter 21 of a two-body potential function with a small amount of calculation, a copper parameter 22 of an EAM potential function with a large amount of calculation, and a condition for switching these two types of potentials. 23 and are input. The site where it is better to use a highly accurate potential is an atom that moves rapidly or an irregular atomic arrangement. Since the site where the movement of the atom is strong and the site where the force that the atom receives is large are in the vicinity where the structure to be machined 18 and the structure to be machined 19 are in contact, the potential switching is simplified for simplification.
An atom existing within 2 nanometers in the direction of travel with respect to the position in front of the processed structure 19 in the direction of travel is E
The AM potential function is used, and beyond that, the two-body potential function is used.

Using the above conditions, the calculation unit 13 calculates the machining simulation. 3, 4, and 5,
In both cases, 100,000 time steps (1
The result after 00 picoseconds). 24 in these figures
Represents an atom in the structure to be processed. FIG. 3 shows the results when the potential functions are selectively used under the conditions of this embodiment. FIG. 4 shows the result obtained by using the EAM potential function with high accuracy for all the atoms and under the same conditions for other atoms.
As a result, a similar atomic arrangement is obtained as compared with FIG. FIG. 5 shows the results obtained by using the two-body potential function for all the atoms and under other conditions under the same conditions, but the results of FIG. 5 are significantly different from those of FIGS. 3, 4,
The calculation time required to obtain each of the 5 results is shown in (Table 1).

[0011]

[Table 1]

Compared to the case where the two-body potential function is used for all the atoms, it takes 3.54 times when the EAM potential function is used for all the atoms, and this embodiment takes about 1.5 times. It is a value. The execution time of this embodiment can be explained by the ratio of the number of atoms referred to when calculating the force by the two-body potential function and the EAM potential function. The comparison of the results shown here demonstrates a reduction in execution time without impairing the simulation accuracy of this example, which is a feature of the EAM potential function.

In the embodiment, two potential functions are used properly, but by using three or more potentials according to the difference in environment, it is possible to further optimize the compatibility of the execution time and the accuracy of the result. it can.

In the embodiment, the EAM potential function is used for atoms existing within 2 nanometers in the direction of travel with respect to the position of the front surface of the processed structure 19 in the direction of travel, and the two-body potential function is used for distances beyond that. Is used,
The range calculated by the EAM potential function does not have to be within 2 nanometers, and switching with the kinetic energy of each atom at each time can further improve accuracy.

The kinetic energy of each atom at each time is used to selectively use the potential function.
The same effect can be obtained by changing the unit of the step size of the simulation time using this criterion. In other words, for atoms that do not move much
The atom coordinates are updated only once per time step, and for atoms that move rapidly, the atom coordinates are updated every time step as before. As a result, with respect to atoms that do not move much, the amount of calculation is reduced to about 1/10 of that used so far, and the speed can be increased. Since only atoms with few movements are selected, the accuracy of the whole simulation is almost unchanged.

It is also possible to combine the use of the above potentials and the change in the step size of the simulation time.

[0017]

As described above, according to the present invention, at least two interatomic potentials are set at the same time, and the potentials are selectively used in the calculation unit according to the environment in which the atoms are placed, so that the calculation accuracy is not sacrificed. Moreover, the result can be obtained in a shorter time. That is, it is a feature of the present invention that the simulation execution time can be significantly reduced without sacrificing the accuracy of the simulation result. Therefore, according to the present invention, the ultra-fine machining support system that accelerates the calculation can significantly reduce the development period, material resources, and human labor required for the ultra-fine machining and the accompanying design.

[Brief description of drawings]

FIG. 1 is a configuration diagram of an ultrafine processing apparatus according to an embodiment of the present invention.

FIG. 2 is a diagram showing a relationship between a processed structure and a structure to be processed in the embodiment.

FIG. 3 is a diagram showing a display image on an image display unit in the embodiment.

FIG. 4 is a diagram showing a display image on an image display unit in an example using only many-body potentials.

FIG. 5 is a diagram showing a display image on an image display unit in an example using only two-body potentials.

[Explanation of symbols]

 11 Condition Setting Section 12 Potential Function Section 13 Calculation Section 14 Data Holding Section 15 Output Section 16 Image Display Section 17 Machining Section 18 Worked Structure 19 Machining Structure 21 Two-body Potential Function Parameter 22 EAM Potential Function Parameter 23 Two Types Of atoms that switch the potential of the atoms 24 Atoms in the structure to be processed

Claims (7)

[Claims] 1. A condition setting unit capable of setting a plurality of structure models having a three-dimensional atomic arrangement based on a solid crystal structure, and a potential function unit deriving an interatomic force for reproducing mechanical properties of a solid crystal. A calculation unit that performs a deduction that follows the dynamic process of atoms by a molecular dynamics method based on the conditions set in the condition setting unit; and a data holding unit that sequentially records and holds the results obtained by the calculation unit. A microfabrication support system for simulating at least one of cutting, grinding, deformation, polishing, friction, and wear, which is a phenomenon that occurs when the plurality of structures slide. 2. The potential function section sets at least two interatomic potentials and conditions for properly using the potentials in the calculation section according to the environment in which the atoms are placed. Ultra-fine processing support system. 3. The hyperfine machining support system according to claim 2, wherein the calculation unit selects the number of atoms to be referred to when calculating the potential value, according to the distance from the atom. 4. The hyperfine machining support system according to claim 3, wherein at least two kinds of potential functions are used, and the optimum effective radius is set independently for each potential function. 5. The hyperfine machining support system according to claim 1, wherein the condition setting unit sets the length of the time unit for calculating the motion of each atom independently for each atom. 6. An ultrafine processing support system in which the contents described in any one of claims 1 to 5 are combined. 7. An ultrafine machining comprising the ultrafine machining support system according to any one of claims 1 to 6 and a machining unit for performing ultrafine machining based on a result obtained by the ultrafine machining support system. apparatus.