Thinning method and device

Abstract

The invention relates to the technical field of crystal processing, in particular to a thinning method and device. The thinning method comprises the following steps: providing a diamond, and enabling the axial direction of the diamond to be basically horizontal; the contour coordinates (Xi, Yi and Zi) of the diamond are obtained, the X direction is perpendicular to the plane where the diamond is located, the Z direction is the vertical direction, and every two of the X direction, the Y direction and the Z direction are perpendicular; layer-by-layer laser thinning is performed in the X direction based on the contour coordinates (Xi, Yi, Zi). By constructing the three-dimensional model, the diamond is precisely cut and thinned in a layered mode, and the technical effect of improving the diamond thinning quality is achieved.

Description

Technical Field

[0001] This application relates to the field of crystal processing technology, and in particular to a thinning method and apparatus. Background Technology

[0002] Diamond thinning technology is a key process for precision machining of hard and brittle materials, and is widely used in high-end manufacturing fields such as semiconductor wafers, LED substrates, optical crystals and ceramic substrates.

[0003] In existing technologies, diamond thinning mainly includes two categories: mechanical grinding and laser thinning. Mechanical grinding uses high-hardness tools or grinding wheels for physical contact processing, which can easily cause stress damage to polycrystalline diamond. Furthermore, the high hardness of the material results in high material consumption and poor processing flexibility. Laser thinning, on the other hand, uses lasers to process the wafer layer by layer from the outer edge to the center. Although it is more efficient, pits or obvious lines often appear in the center of the wafer after processing. Existing thinning methods can damage the substrate and disrupt local flatness, ultimately making it difficult to meet the high-precision requirements of thinning quality.

[0004] Therefore, the technical problem with existing technologies is that the thinning quality is relatively low. Summary of the Invention

[0005] This application provides a thinning method and apparatus, which constructs a three-dimensional model to precisely cut and thin diamond into layers, thereby achieving the technical effect of improving the quality of diamond thinning.

[0006] On the one hand, the thinning method provided in this application adopts the following technical solution:

[0007] A thinning method, comprising:

[0008] Provide diamonds so that their axes are arranged in a basically horizontal manner;

[0009] Obtain the contour coordinates (X) of the diamond. i ,Y i Z i ), where the X direction is perpendicular to the plane where the diamond is located, the Z direction is vertical, and the X, Y, and Z directions are perpendicular to each other;

[0010] Based on contour coordinates (X i ,Y i Z i Laser thinning is performed layer by layer along the X direction.

[0011] As a preferred option, it also includes:

[0012] Determine the maximum thickness X in the X direction. max ;

[0013] Preset target thickness X t And single layer thickness Xh ;

[0014] Based on target thickness X t And single layer thickness X h Determine the number of layers N to be removed. x ;

[0015] The laser sequentially targets N in the X-axis direction. x Thinning layer by layer.

[0016] As a preferred option, it also includes:

[0017] Determine the total removal amount ΔX in the X direction. 总 for:

[0018] △X 总 =X max -X t

[0019] Remove the number of layers N x for:

[0020]

[0021] As a preferred option, it also includes:

[0022] Determine the maximum thickness X in the X direction. max Preset target thickness X t and initial single-layer thickness

[0023] When removing the (k+1)th layer, the surface roughness Ra of the kth layer is used as a reference. k Determine the corrected single-layer thickness X h (k+1);

[0024] Based on the modified single-layer thickness X h (k+1), where the laser sequentially thins the diamond layer by layer in the X direction; where k≥1.

[0025] Preferably, the determination of the corrected single-layer thickness X h (k+1) includes: corrected single-layer thickness X h (k+1) is:

[0026]

[0027] Where, η k+1 γ is the surface roughness feedback correction factor; SGN(·) is the cutting state adjustment amplitude coefficient; S is the sign function; S is the cutting state characteristic value; ΔS is the cutting state difference value; δ is the cutting state attenuation coefficient; Ra target ω represents the target surface roughness; ω is the surface roughness correction sensitivity coefficient.

[0028] Preferably, the difference in cutting conditions is ΔS:

[0029]

[0030] Among them, S k The cutting state characteristic value of the k-th layer;

[0031] μ S,k-1 The mean value of the cutting state characteristic values ​​of the (k-1)th layer;

[0032] σ S,k-1 denoted as the standard deviation of the cutting state characteristic values ​​of the (k-1)th layer.

[0033] As a preferred option, it also includes:

[0034] Based on the log-linear regression model, a nonlinear relationship is established between surface roughness Ra and the laser wavelength λ, laser power P, laser frequency f, diamond moving speed v, and cutting track spacing d: In(Ra)=β0+β1λ+β2P+β3f+β4v+β5d.

[0035] Based on the nonlinear relationship In(Ra)=β0+β1λ+β2P+β3f+β4v+β5d, β0, β1, β2, β3, β4, and β5 are determined through linear regression model fitting. This allows us to establish the nonlinear relationship between surface roughness Ra and laser wavelength λ, laser power P, laser frequency f, moving speed v, and cutting track spacing d.

[0036] Preferably, X-axis cutting compensation is also included:

[0037] When removing the (k+1)th layer, the diamond tilt angle α and the Z-axis single-layer step size Z are used. h Determine the theoretical compensation amount ΔX 理论 ;

[0038] Based on the theoretical compensation amount △X 理论 and the surface roughness Ra of the (k+1)th layer k+1 Determine the actual compensation amount △X real As a preferred option, the theoretical compensation amount ΔX 理论 for:

[0039] △X 理论 =Z h ·tanα;

[0040] The actual compensation amount △X real for:

[0041] △X real =△X 理论 ·(1+C·σ Ra,k+1 )

[0042] Where C is the material correction factor;

[0043] σ Ra,k+1 Let be the standard deviation of the surface roughness of the (k+1)th layer;

[0044] β is the laser cone angle.

[0045] On the other hand, the thinning device provided in this application adopts the following technical solution:

[0046] A thinning apparatus for performing the thinning method to thin diamond, comprising:

[0047] A support mechanism for fixing a diamond, the support mechanism being deflectable and movable to cause the diamond to deflect and move;

[0048] A laser mechanism, located above the support mechanism, is used for laser thinning of diamond.

[0049] In summary, this application includes at least one of the following beneficial technical effects:

[0050] This application improves the precision and quality of diamond thinning by constructing diamond profile coordinates and then thinning the diamond layer by layer based on these coordinates. Attached Figure Description

[0051] Figure 1 This is a flowchart illustrating the thinning method described in this application;

[0052] Figure 2 This is a coordinate schematic diagram of the thinning method described in this application;

[0053] Figure 3 This is a schematic diagram of the first processing path of the thinning method described in this application;

[0054] Figure 4 This is a schematic diagram of the second processing path of the thinning method described in this application;

[0055] Figure 5 This is a schematic flowchart of the layer-by-layer thinning method described in this application;

[0056] Figure 6 This is a schematic diagram of the diamond layer-by-layer thinning process described in this application;

[0057] Figure 7 This is a schematic flowchart of the layer-by-layer thinning method described in this application;

[0058] Figure 8 This is a schematic diagram of the layer-by-layer thinning correction described in this application;

[0059] Figure 9This is a schematic diagram of the model construction process of the thinning method described in this application;

[0060] Figure 10 This is a schematic diagram of the compensation amount flow of the thinning method described in this application;

[0061] Figure 11 This is a schematic diagram of diamond deflection during the thinning process described in this application;

[0062] Figure 12 This is a schematic diagram of the thinning device described in this application.

[0063] Explanation of reference numerals in the attached drawings: D, diamond; 100, bearing mechanism; 200, laser mechanism. Detailed Implementation

[0064] The serial numbers assigned to components in this document, such as "first" and "second," are used solely to distinguish the described objects and have no sequential or technical meaning. The terms "connection" and "linkage" used in this application, unless otherwise specified, include both direct and indirect connections (linkages). It should be understood that the terms "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "clockwise," and "counterclockwise," indicating orientations or positional relationships, are based on the orientations or positional relationships shown in the accompanying drawings and are used solely for the convenience of describing this application and simplifying the description. They do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of this application.

[0065] In this application, unless otherwise expressly specified and limited, "above" or "below" the second feature can mean that the first feature is in direct contact with the second feature, or that the first feature is in indirect contact with the second feature through an intermediate medium. Furthermore, "above," "on top of," and "over" the second feature can mean that the first feature is directly above or diagonally above the second feature, or simply that the first feature is at a higher horizontal level than the second feature. "Below," "below," and "under" the second feature can mean that the first feature is directly below or diagonally below the second feature, or simply that the first feature is at a lower horizontal level than the second feature.

[0066] This application provides a thinning method and apparatus, which achieves the technical effect of improving the thinning quality of diamond D by constructing a three-dimensional model for precise cutting and thinning.

[0067] To better understand the above technical solutions, a detailed description of the technical solutions will be provided below in conjunction with the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are merely illustrative of this application and are not intended to limit the scope of this application.

[0068] Diamond D thinning technology is a key process for precision machining of hard and brittle materials, playing an irreplaceable role in high-end manufacturing fields such as semiconductors, optics, and electronics. Currently, diamond D thinning mainly relies on two processes: mechanical grinding and laser thinning. Mechanical grinding involves physical contact between high-hardness tools or grinding wheels and the surface of diamond D, using the cutting action of abrasive particles to remove material and achieve thinning. However, due to the extremely high hardness of diamond D, significant contact stress is generated during mechanical grinding, easily causing irreversible stress damage to the crystal structure of polycrystalline diamond D. This leads to defects such as microcracks and lattice distortion within the diamond D, affecting its subsequent performance. Simultaneously, the high hardness results in extremely rapid wear of the grinding consumables, increasing processing costs and limiting the flexibility of machining complex-shaped diamond D, making it difficult to adapt to diverse processing needs.

[0069] Laser thinning processes use a laser beam to process the wafer layer by layer from the outer edge towards the center, removing material using the thermal or photochemical effects of the laser. Compared to mechanical grinding, this process is more efficient. However, existing laser thinning processes have poorly designed processing paths, leading to uneven distribution of laser energy on the wafer surface. This often results in pits or obvious processing lines in the central area of ​​the wafer after processing, compromising the local flatness of the diamond D surface. Furthermore, existing laser thinning processes lack a dynamic adjustment mechanism for the processing. Once key parameters such as laser parameters and processing thickness are set, they are not changed, making it impossible to cope with changes in material properties or fluctuations in processing conditions. Ultimately, this results in thinned diamond D failing to meet the requirements of high-precision manufacturing in terms of both thickness accuracy and surface quality.

[0070] In summary, existing diamond D thinning methods generally suffer from low thinning quality, specifically high surface roughness, insufficient thickness accuracy, and susceptibility to stress damage or processing defects. There is an urgent need for a diamond D thinning method and apparatus that can balance processing accuracy, surface quality, and processing efficiency.

[0071] This application provides a thinning method, such as Figure 1 , 2 As shown, it includes: providing diamond D, such that the axis of diamond D is arranged in a basically horizontal manner; obtaining the contour coordinates (X...) of diamond D. i ,Y i Z i ), where the X direction is perpendicular to the plane containing diamond D, the Z direction is vertical, and the X, Y, and Z directions are mutually perpendicular; based on the contour coordinates (X i ,Y i Z i Laser thinning is performed layer by layer along the X direction.

[0072] The thinning method described in this application aims to achieve preliminary layered thinning of diamond D through a reasonable processing layout, precise coordinate acquisition, and scientific cutting path design, which is beneficial to improving the thinning quality of diamond D. Figure 1 , 2 As shown, it specifically includes:

[0073] A diamond workpiece D to be thinned is provided. In this embodiment, diamond D or diamond D workpiece refers to sheet diamond. This application mainly focuses on thinning sheet diamond D; in other embodiments, ingot diamond D can also be thinned. The diamond D is fixed on a support mechanism so that the axial direction of the diamond D is basically horizontal. The axial direction refers to the main extension direction of the diamond D. The purpose of the basically horizontal arrangement is to ensure that the surface of the diamond D and the direction of the laser beam maintain a stable relative position, avoid inconsistent laser action depth caused by the tilt of the diamond D, and facilitate uniform thinning along the X-direction.

[0074] Next, obtain the contour coordinates (X) of diamond D. i ,Y i Z i The coordinate system is defined as follows: the X-axis is perpendicular to the plane where the diamond D is located, which is the thickness direction of laser thinning and also the main direction of material removal; the Z-axis is vertical and perpendicular to the horizontal plane; the Y-axis is perpendicular to both the X-axis and the Z-axis, together forming a three-dimensional rectangular coordinate system.

[0075] The contour coordinates are obtained using a laser displacement sensor for scanning measurement, which can perform a full scan of the surface of diamond D. During the scanning process, the laser beam emitted by the laser displacement sensor is directed to the surface of diamond D, and the X, Y, and Z coordinates of each point on the surface of diamond D are obtained, i.e., (X... i ,Y i Z i ).

[0076] Based on the obtained contour coordinates (X i ,Y i Z i Laser thinning is performed layer by layer along the X-axis. The relative motion between the laser and the diamond D and the cutting path planning are as follows: Figure 3 , 4 As shown, specifically: the laser mechanism is fixed, and the diamond D is moved in a coordinated manner in the X, Y and Z directions by the supporting mechanism, so that the laser beam acts on the diamond D from top to bottom, and the diamond D is thinned layer by layer along the X direction.

[0077] During each layer thinning process, the movement of diamond D is divided into two dimensions: one is the Z-axis stepping movement, and the other is the cutting path movement in the XOY plane. Among them, during the thinning process of the same layer, the Z-axis stepping movement refers to the diamond D moving by a preset step amount along the Z-axis after completing a certain depth of thinning in the XOY plane, so that the laser beam can act on the processing area of ​​the next depth. During the thinning process of each layer, diamond D remains fixed in the Z-axis, and only through the coordinated movement of the Y and Z axes, the laser beam cuts the layer (vertically) in an S-shaped or bow-shaped path in the XOY plane until the thinning of that layer is completed.

[0078] The specific movement of the S-shaped path is as follows: Diamond D starts from an initial position in the Z direction and moves from one end to the other along the Y direction, while simultaneously moving a distance corresponding to a preset single-layer thinning thickness along the X direction. During the movement, the laser beam continuously acts on the surface of Diamond D, removing material from that area. When it moves to the other end in the Y direction, Diamond D moves a small distance in the opposite direction along the X direction (to avoid overlapping cutting marks), then moves in the opposite direction along the Y direction while continuing to move along the X direction, forming an S-shaped cutting trajectory. The specific movement of the bow-shaped path is as follows: Diamond D moves in a straight line along the Y direction while simultaneously moving a preset distance along the X direction, completing the processing of one cutting path. When it moves to the end point in the Y direction, Diamond D moves a preset cutting path interval distance along the Z direction, then moves in a straight line in the opposite direction along the Y direction while continuing to move along the X direction, forming a bow-shaped cutting trajectory.

[0079] The advantages of using an S-shaped or bow-shaped path are: it ensures that the laser beam fully covers the processing area within the XOY plane, eliminating blind spots; simultaneously, this path design allows for a more uniform energy distribution on the diamond D surface, avoiding thermal damage caused by excessive laser energy concentration in localized areas, or incomplete material removal due to insufficient energy. In actual processing, the appropriate cutting path can be selected based on the shape and size of the diamond D. For regular rectangular diamond Ds, a bow-shaped path is preferred for higher processing efficiency; for irregularly shaped diamond Ds, an S-shaped path is preferred for greater adaptability. After each layer is cut, the diamond D moves along the X direction by a preset single-layer thickness distance to proceed to the next layer, until the thinning of that layer is complete.

[0080] In this way, diamond D can be uniformly thinned layer by layer along the X direction, and the thickness of diamond D can be initially controlled. At the same time, the cutting path design reduces processing defects and facilitates precise thinning.

[0081] Furthermore, clearly defining the goals and layers of the thinning process makes the layer-by-layer thinning process more controllable. For example... Figure 5 , 6 As shown, the thinning method also includes: determining the maximum thickness X in the X direction. maxPreset target thickness X t And single layer thickness X h Based on target thickness X t And single layer thickness X h Determine the number of layers N to be removed. x ; In the X-axis, the laser sequentially targets N x Thinning layer by layer.

[0082] Based on the obtained diamond D profile coordinates (X i ,Y i Z i ), Calculate the X values ​​of all contour points i The coordinates are used to determine the maximum thickness X of diamond D in the X direction. max ;X max It reflects the maximum dimension of diamond D in the thickness direction and is the main basis for calculating the total removal amount and the number of layers removed.

[0083] Based on the application requirements of diamond D, a target thickness X is preset. t Target thickness X refers to the final thickness that diamond D needs to achieve after thinning. t The settings need to be tailored to the specific application scenario to meet the requirements of high-precision applications. Simultaneously, the initial single-layer thickness X is preset. h This refers to the thickness of material that needs to be removed during each layer thinning process. Single layer thickness X h The initial settings need to take into account both processing efficiency and processing quality: if the single layer thickness is too large, although the processing efficiency is high, it may lead to excessive material removal in a single operation, resulting in large processing stress and affecting surface quality; if the single layer thickness is too small, the processing efficiency is too low and the processing cost is increased. When setting the initial settings, an appropriate initial value can be selected according to the material of diamond D, the initial thickness, and the target surface quality requirements.

[0084] Based on the preset target thickness X t And single layer thickness X h Determine the number of layers N to be removed. x That is, how many single-layer thinning operations are needed to remove diamond D from X max Reduce to X t Next, according to the determined number of layers N to be removed... x In the X-axis direction, diamond D is thinned layer by layer using a laser. By breaking down the continuous thinning process into controllable single-layer operations, the blindness of the thinning process is avoided. The thinning thickness and number of operations for each layer are quantified, which facilitates individual control of the thinning quality of each layer and reduces the risk of overall quality fluctuations caused by sudden parameter changes.

[0085] Among them, the total removal amount △X is determined in the X direction. 总 for:

[0086] △X 总 =X max -X t

[0087] Remove the number of layers N x for:

[0088]

[0089] Precise quantification of the number of layers removed: The total removal amount is the total thickness that needs to be removed from diamond D from its initial maximum thickness to the target thickness. The derivation logic of the calculation formula is: Total removal amount = Initial maximum thickness - Target thickness, therefore the expression is:

[0090] △X 总 =X max -X t

[0091] Since the thickness of each layer is X h Theoretically, the number of removal layers is the total removal amount ÷ the thickness of a single layer, i.e., ΔX 总 / X h However, in actual processing, if the total removal amount △X 总 Cannot be X of single layer thickness h Divisible by X, the remaining thickness (even if less than X) h A separate thinning operation is also required; otherwise, the final thickness of diamond D will exceed the target value. Therefore, the calculation results need to be rounded up, and the final formula for the number of layers removed is:

[0092]

[0093] in, The rounding up symbol.

[0094] Furthermore, a surface roughness feedback mechanism is introduced. By monitoring the surface roughness in real time, the thickness of subsequent single-layer processing is dynamically adjusted to form a closed-loop control, achieving a balance between processing efficiency and surface quality. Surface roughness Ra is a key indicator for measuring the quality of processed surfaces. The value of surface roughness Ra reflects the smoothness of the surface. During laser thinning, the Ra value after the previous layer is cut directly reflects the processing quality of that layer: if the Ra value is high, it indicates that the material removal of that layer is not uniform, and there may be problems such as local material residue, stress concentration, or uneven laser energy distribution. If the original single-layer thickness is used for subsequent processing, these defects may be further amplified, making it difficult to obtain a high-quality surface. Therefore, it is necessary to reduce the single-layer thickness of subsequent processing to remove material more finely, gradually correct surface defects, and reduce the Ra value of subsequent processing. Conversely, if the Ra value after the previous layer is low, it indicates that the processing state of that layer is good, the material removal is uniform, and the surface is smooth. In this case, the single-layer thickness of subsequent processing can be appropriately increased to improve processing efficiency and shorten the processing cycle while ensuring surface quality. Therefore, the dynamic adjustment strategy based on surface roughness can optimize the processing parameters in real time according to the actual state during the processing, and achieve a dynamic balance between processing quality and efficiency.

[0095] like Figure 7 , 8 As shown, the method also includes: determining the maximum thickness X in the X direction. max Preset target thickness X t and initial single-layer thickness When removing the (k+1)th layer, the surface roughness Ra of the kth layer is used as a reference. k Determine the corrected single-layer thickness X h (k+1); Based on the modified single-layer thickness X h (k+1), where the laser sequentially thins the diamond layer by layer in the X direction; where k≥1.

[0096] It should be noted that, as Figure 8 As shown, the surface roughness Ra of the (k+1)th layer refers to the roughness of the surface remaining on diamond D after the kth layer is thinned. In other words, since the thinning of diamond D involves removing material layer by layer along the X direction, a new processed surface is formed after the kth layer is processed. This surface serves as the reference surface for the (k+1)th layer, and its roughness directly affects the processing effect of the (k+1)th layer. Therefore, when removing the (k+1)th layer, the surface roughness Ra formed after the kth layer is used as the reference surface. k As a basis for adjustment. In one embodiment, the surface roughness Ra k The measurement was performed using a white light interferometer, which features high measurement accuracy and non-contact measurement, enabling precise acquisition of surface roughness data without damaging the processed surface.

[0097] Determine the maximum thickness X of diamond D in the X direction. max Simultaneously preset target thickness X t With the initial single-layer thickness This is the initial single-layer removal thickness set before the thinning operation begins, taking into account laser parameters and material properties. When performing the (k+1)th layer thinning operation, the surface roughness Ra of the kth layer needs to be referenced. k Ra k It is the surface roughness of the surface remaining on diamond D after the (k-1)th layer is thinned (i.e., the surface roughness of the substrate after the kth layer is thinned); if Ra k A higher Ra indicates that the material removal of the k-th layer is not uniform or that stress concentration exists; if Ra k A lower value indicates that the thinning effect of the k-th layer is better and the material removal is more uniform. This is based on the surface roughness Ra of the k-th layer. k The corrected single-layer thickness X is calculated. h (k+1), then according to X h (k+1) Thinning of the (k+1)th layer. This avoids the one-size-fits-all problem of "fixed single-layer thickness": when the thinning effect is poor, the single-layer thickness is reduced for fine processing; when the thinning effect is good, the single-layer thickness is maintained (or increased) to improve efficiency, thus achieving dynamic control that prioritizes quality and adapts to efficiency.

[0098] Determine the corrected single-layer thickness X h (k+1) includes: corrected single-layer thickness X h (k+1) is:

[0099]

[0100] Where, η k+1 γ is the surface roughness feedback correction factor; γ is the cutting state adjustment amplitude coefficient; SGN(·) is the sign function; S is the cutting state characteristic value; γ is the cutting state adjustment amplitude coefficient; δ is the cutting state attenuation coefficient; Ra target ω represents the target surface roughness; ω is the surface roughness correction sensitivity coefficient.

[0101] It should be noted that, The initial single-layer thickness during the thinning operation can be set manually; η k+1γ is the surface roughness feedback correction factor, serving as a coefficient based on the previous roughness adjustment; γ is the cutting state adjustment amplitude coefficient, reflecting the degree of influence of the control cutting state difference on the thickness; SGN(·) is the sign function, used to determine the direction of the cutting state difference: it takes 1 when △S>0 and -1 when △S<0; △S is the cutting state difference value, used to quantify the cutting state deviation between the current layer and the previous layer; δ is the cutting state attenuation coefficient, used to weaken the influence of large deviations and avoid excessive adjustment amplitude.

[0102] η k+1 Its function is to adjust the thickness based on the roughness deviation of the previous layer, η k+1 for:

[0103]

[0104] Among them, Ra target The target surface roughness can be adjusted based on the actual process.

[0105] ω is the surface roughness correction sensitivity coefficient.

[0106] When Ra k >Ra target This indicates that the quality of the previous layer is not consistent, and the thickness of a single layer needs to be reduced. Therefore, η k+1 <1, making X h (k+1) decreases; when Ra k ≤Ra target At that time, the thinning quality meets the standard, maintaining the initial thickness to improve thinning efficiency, therefore η k+1 =1.

[0107] Correcting the single-layer thickness X h In the formula for calculating (k+1), [1+γ·SGN(△S)·e -δ|△S| [] is the adjustment item for the cutting state. SGN(△S) aligns the adjustment direction with the deviation direction of the cutting state. A positive deviation increases the thickness, while a negative deviation decreases the thickness; e -δ|△S| The adjustment range is made more gradual when the deviation is large, avoiding quality fluctuations caused by sudden parameter changes; at the same time, by combining the feedback of surface roughness and cutting status, the thickness of a single layer can be precisely and dynamically adjusted, ensuring that the surface quality meets the target without excessively sacrificing processing efficiency.

[0108] Furthermore, the difference in cutting conditions is ΔS:

[0109]

[0110] Among them, S k The cutting state characteristic value of the k-th layer is a parameter used to reflect the laser action state of that layer, such as the uniformity of laser energy distribution and the stability of diamond D movement.

[0111] μ S,k-1 The average value of the cutting state characteristic value of the (k-1)th layer is used to reflect the overall level of the cutting state of the previous layer;

[0112] σ S,k-1 It represents the standard deviation of the cutting state characteristic values ​​of the (k-1)th layer, reflecting the degree of fluctuation in the cutting state of the previous layer.

[0113] Where k≥1; when k=1, the 0th layer is the initial surface of diamond D.

[0114] △S is the relative deviation between the cutting state of the current layer and the previous layer: the absolute deviation is obtained by subtracting the mean of the previous layer from the current value, and then divided by the standard deviation of the previous layer to eliminate the difference in magnitude, so that △S can objectively reflect the degree of change in the cutting state (for example, if S k If the value is much greater than the average of the previous layer, and the fluctuation of the previous layer is small, then ΔS will increase significantly. The core function of ΔS is to provide feedback on the cutting state for correcting the thickness of a single layer, so that the thickness adjustment not only conforms to the surface quality, but also adapts to the stability of the cutting process, further improving the controllability of the thinning process.

[0115] More specifically, the physical meaning of the cutting state difference value ΔS is that its value is the standardized deviation of the cutting state characteristic value of the k-th layer relative to the cutting state characteristic value of the (k-1)-th layer, which is calculated by dividing by the standard deviation σ of the (k-1)-th layer. S,k-1 This eliminates the influence of the absolute level of the cutting state between different processing layers, making ΔS more comparable. The magnitude and sign of ΔS reflect the changes in the cutting state: when ΔS>0, it indicates that the cutting state characteristic value of the k-th layer is higher than the average level of the (k-1)-th layer, and the cutting intensity increases, possibly due to factors such as increased laser power and decreased moving speed. In this case, it is necessary to pay attention to whether it will lead to an increase in surface roughness. When ΔS=0, it indicates that the cutting state characteristic value of the k-th layer is consistent with the average level of the (k-1)-th layer, and the cutting state is stable, maintaining the current processing parameters. When ΔS<0, it indicates that the cutting state characteristic value of the k-th layer is lower than the average level of the (k-1)-th layer, and the cutting intensity decreases, possibly due to factors such as decreased laser power and increased moving speed. In this case, it is necessary to pay attention to whether the material removal efficiency meets the requirements.

[0116] The larger the absolute value of ΔS, the more significant the difference in cutting state between layer k and layer (k-1), requiring a larger adjustment to processing parameters (such as layer thickness and laser parameters). Conversely, the smaller the absolute value of ΔS, the more stable the cutting state, requiring a smaller adjustment. For example, if ΔS = 1.5, it means the characteristic value of the cutting state of layer k is 1.5 standard deviations higher than the mean of layer (k-1), indicating a significant increase in cutting intensity, which may lead to increased surface roughness. In this case, it is necessary to reduce the layer thickness or lower the laser power. If ΔS = -0.5, it means the cutting intensity is slightly reduced, and the layer thickness or laser power can be appropriately increased to ensure processing efficiency. By calculating S and ΔS, changes in the cutting state during processing can be monitored in real time, providing correction parameters for calculating the layer thickness and making the adjustment of the layer thickness more precise.

[0117] It should be noted that the cutting state characteristic value S is a comprehensive index used to quantitatively describe the cutting state during laser thinning. The magnitude of the cutting state characteristic value S reflects the stability of the cutting process, energy utilization efficiency, and material removal effect. During laser thinning, the cutting state is affected by various factors, such as laser parameters (wavelength, power, frequency), diamond movement speed, cutting path spacing, and material properties. Changes in these factors can cause fluctuations in the cutting state, thereby affecting the processing quality. By defining the cutting state characteristic value S, the complex cutting state can be transformed into a quantifiable index, facilitating the monitoring and analysis of the changing trends of the cutting process. The core physical meaning of the cutting state characteristic value S is that: a higher value indicates greater cutting intensity and faster material removal speed, but may be accompanied by greater processing stress and surface roughness; a lower value indicates lower cutting intensity and slower material removal speed, but may result in better surface quality. Therefore, the changing trend of the cutting state characteristic value S can provide an important basis for adjusting cutting parameters.

[0118] In one embodiment, the cutting state characteristic value S is calculated based on multi-dimensional signals acquired in real time during the processing. This is obtained through feature extraction and weighted fusion. During each layer thinning process, the following three types of signals are acquired simultaneously to comprehensively reflect the cutting state: Laser reflection signal: The reflected laser power after the laser acts on the diamond D surface is acquired using a laser power monitor. Changes in the reflected power reflect changes in the absorption rate of the laser on the diamond D surface, thus reflecting the progress of the material removal process. For example, when the material removal rate is high, the surface roughness may increase, and the reflected power may fluctuate; when the material removal is uniform, the reflected power is relatively stable. Plasma signal: The plasma spectral signal generated during laser processing is acquired using a spectral analyzer. The intensity and spectral distribution of the plasma are related to the absorption of laser energy and the degree of ionization of the material, reflecting the intensity of the interaction between the laser and the material. Higher plasma signal intensity indicates higher laser energy utilization efficiency and greater cutting intensity. Diamond vibration signal: The vibration acceleration signal of diamond D during the processing is collected using a piezoelectric accelerometer. The amplitude and frequency distribution of the vibration signal reflect the stress changes and the uniformity of material removal during the processing. The larger the vibration amplitude, the greater the processing stress, which may indicate uneven material removal or stress concentration.

[0119] The three types of raw signals were preprocessed to remove noise interference and extract effective features; filtering was performed using a low-pass filter; normalization was performed to normalize the filtered signals so that the numerical range of each type of signal was unified between [0,1], which facilitates subsequent weighted fusion; laser reflection signal feature parameters, plasma signal feature parameters, and vibration signal feature parameters were extracted from the preprocessed signals, and the above three feature parameters were fused into a cutting state feature value S by weighted summation; in other words, for the cutting state feature value S of the k-th layer k The calculation process is as follows: During the processing of the k-th layer, the above three types of signals are collected in real time. After preprocessing and feature extraction, multiple S values ​​are calculated using a weighted fusion formula (with a collection frequency of 10kHz and a processing time of 10s, 100,000 S values ​​can be obtained). The average value of these S values ​​is taken as the cutting state feature value S of the k-th layer. k Ensure S k It can reflect the overall cutting state of the machining of this layer; similarly, the cutting state characteristic value S of the (k-1)th layer can reflect the overall cutting state of the machining of this layer. k-1 It is the average of multiple S values ​​calculated during the processing of the (k-1)th layer.

[0120] Furthermore, such as Figure 9As shown, in order to achieve the optimized design of laser parameters, a nonlinear relationship model between surface roughness Ra and laser wavelength λ, laser power P, laser frequency f, diamond moving speed v, and cutting track spacing d was established through orthogonal experiments and log-linear regression analysis, providing a basis for the selection and adjustment of processing parameters.

[0121] Surface roughness Ra is a core indicator for evaluating the quality of thinning processes, and laser parameters directly affect the interaction between the laser and diamond (D), thus determining the surface roughness. Different combinations of laser parameters produce different processing effects. For example, excessive laser power may lead to over-melting of the material, increasing surface roughness; excessively high moving speed may result in incomplete material removal, also affecting surface quality. Therefore, establishing a quantitative relationship model between Ra and laser parameters allows for the derivation of the optimal laser parameter combination based on a preset Ra target value, maximizing processing efficiency while ensuring surface quality, and providing theoretical support for parameter adjustments during the processing.

[0122] like Figure 9 As shown, the method also includes: based on a log-linear regression model, establishing a nonlinear relationship between surface roughness Ra and the wavelength λ of the laser, the power P of the laser, the frequency f of the laser, the moving speed v of the diamond, and the cutting track spacing d: In(Ra)=β0+β1λ+β2P+β3f+β4v+β5d.

[0123] Based on the nonlinear relationship In(Ra)=β0+β1λ+β2P+β3f+β4v+β5d, β0, β1, β2, β3, β4, and β5 are determined through linear regression model fitting. This allows us to establish the nonlinear relationship between surface roughness Ra and laser wavelength λ, laser power P, laser frequency f, moving speed v, and cutting track spacing d.

[0124] Specifically, the laser wavelength λ, laser power P, laser frequency f, diamond movement speed v, and cutting path spacing d are selected as the core influencing factors, each with four levels (covering the conventional processing range), and L is used. 16 (4 5 Sixteen groups of experiments were designed using an orthogonal array to avoid the high cost of full factorial experiments. The same batch of diamonds and a uniform cutting path were used in the experiments to ensure data comparability. After the experiments were completed, the Ra value of each group was measured using a white light interferometer, and the average value of five feature points was taken as the result. Then, the natural logarithm lnRa was taken to complete the data preprocessing and prepare for linear regression.

[0125] Establish a log-linear regression model:

[0126] In(Ra)=β0+β1λ+β2P+β3f+β4v+β5d

[0127] The least squares method was used to solve for the regression coefficients β0, β1, β2, β3, β4, and β5, yielding the regression equation. The model fit was verified to be good enough (R0). 2 The F-test result was significant (F>3.33), indicating that the model can effectively explain the variation in Ra and has good reliability.

[0128] Taking the natural logarithm of the fitted equation, we obtain the nonlinear relationship between Ra and the laser parameters:

[0129]

[0130] In this way, the optimal parameter combination can be deduced from the target Ra, or the parameter adjustment range can be calculated when Ra deviates from the target during processing, reducing trial and error costs and improving processing efficiency and surface quality stability.

[0131] Furthermore, such as Figure 10 , 11 As shown, the influence of the laser cone angle β is introduced (the laser beam is not an ideal parallel light, but has a certain cone angle β, which will cause edge deviation of the laser's depth of action), making the compensation calculation more consistent with the actual processing scenario; the thinning method also includes X-axis cutting compensation:

[0132] When removing the (k+1)th layer, the diamond tilt angle α and the Z-axis single-layer step size Z are used. h Determine the theoretical compensation amount ΔX 理论 ;

[0133] Based on the theoretical compensation amount △X 理论 and the surface roughness Ra of the (k+1)th layer k+1 Determine the actual compensation amount △X real .

[0134] Theoretical compensation amount △X 理论 for:

[0135] △X 理论 =Z h ·tanα;

[0136] Actual compensation amount △X real for:

[0137] △X real =△X 理论 ·(1+C·σ Ra,k+1 )

[0138] Where C is the material correction factor;

[0139] σ Ra,k+1 Let be the standard deviation of the surface roughness of the (k+1)th layer;

[0140] α is the diamond tilt angle, which satisfies: β is the laser cone angle.

[0141] In actual processing, the diamond tilt angle α and surface roughness can cause thickness errors in the X direction, which need to be corrected through cutting compensation. First, the diamond tilt angle α is measured, based on the Z-axis single-layer step amount Z. h According to the formula △X 理论 =Z h • tanα is used to calculate the theoretical compensation amount, which offsets the projection error caused by tilt; combined with the surface roughness Ra of the (k+1)th layer. k+1 Standard deviation σ Ra,k+1 By introducing a material correction factor C (0.1-0.3), the actual compensation amount ΔX is obtained. real for:

[0142] △X real =△X 理论 ·(1+C·σ Ra,k+1 )

[0143] During compensation, it is superimposed on the preset single-layer thickness to ensure that the amount of removal in each layer is uniform and to improve the thickness control accuracy.

[0144] This application also provides a thinning device, such as Figure 12 As shown, the above-described thinning method can be used to thin diamond D, including: a support mechanism 100 and a laser mechanism 200. The support mechanism 100 is used to fix diamond D and can be deflected and moved; the laser mechanism 200 is located above the support mechanism 100 and is used to perform laser thinning on diamond D.

[0145] The support mechanism 100 adopts vacuum adsorption clamping to avoid stress damage caused by mechanical clamping. It is driven by a linear motor to realize translation in the X / Y / Z directions and deflection around the Y / Z axis. The integrated measurement interface can collect data such as contour coordinates and roughness in real time. The laser mechanism 200 uses a pulsed laser with adjustable focusing function to support precise adjustment of laser parameters. The control system links the support mechanism 100 and the laser mechanism 200 to automatically execute the processing path, dynamically adjust parameters and cut compensation, ensuring efficient implementation of the thinning method and meeting the requirements of high-precision processing.

[0146] Although preferred embodiments of this application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of this application.

[0147] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and variations.

Claims

1. A thinning method characterized by, Comprise: providing a diamond (D), arranging the axial direction of the diamond (D) to be substantially horizontal; Obtaining profile coordinates (X i ,Y i ,Z i ) of a diamond (D), wherein the X direction is perpendicular to the plane in which the diamond (D) lies, the Z direction is the vertical direction, and the X direction, the Y direction and the Z direction are perpendicular to each other; Based on the contour coordinates (X i ,Y i ,Z i ), layer-by-layer laser thinning is performed in the X direction.

2. The thinning method of claim 1, wherein Further comprise: determining the maximum thickness X in the X direction max ; Pre-set target thickness X t and single layer thickness X h ; Based on target thickness X t and single layer thickness X h , determine number of layers N x to remove The laser sequentially thins the layers in the X direction x one layer at a time.

3. A thinning method according to claim 2, wherein Further comprise: determining the total removal amount ΔX in the X direction 总 is: ΔX 总 = X max - X t Removing the number of layers N x is:

4. The thinning method of claim 2, wherein Further comprise: determining the maximum thickness X in the X direction max , the preset target thickness X t and the initial single layer thickness For the removal of the (k+1)th layer, based on the surface roughness Ra of the kth layer k , the modified single layer thickness X h (k+1) is determined; based on the corrected single layer thickness X h (k+1), sequentially thinning the diamond (D) layer by layer in the X direction; wherein k≥1.

5. A thinning method according to claim 4, wherein The determination of the corrected single layer thickness X h (k+1) includes the correction of the single layer thickness X h (k+1) is: wherein η k+1 is a surface roughness feedback correction factor; γ is a cutting state adjustment amplitude coefficient; SGN(·) is a sign function; S is a cutting state characteristic value; ΔS is a cutting state difference value; δ is a cutting state attenuation coefficient; Ra target is a target surface roughness; ω is a surface roughness correction sensitivity coefficient.

6. A thinning method according to claim 5, wherein The cutting state difference value is ΔS: wherein S k is the cutting state characteristic value of the kth layer; μ S,k-1 is the mean value of the cutting state characteristic values for the k-1 layer; σ S,k-1 is the standard deviation of the cutting state characteristic values of the k-1th layer.

7. A thinning method according to any one of claims 1 to 6, characterized in that, Further comprise: Based on the logarithmic linear regression model, a nonlinear relationship In(Ra)=β0+β1λ+β2P+β3f+β4v+β5d is established between the surface roughness Ra and the wavelength λ of the laser, the power P of the laser, the frequency f of the laser, the moving speed v of the diamond, and the cutting groove interval d; Based on the nonlinear relationship In(Ra) = β0 + β1λ + β2P + β3f + β4v + β5d, by linear regression model fitting, β0, β1, β2, β3, β4, β5 are determined to determine the nonlinear relationship of surface roughness Ra and laser wavelength λ, laser power P, laser frequency f, moving speed v and cutting path interval d 8. The thinning method of claim 3, wherein, Further comprise cutting compensation in the X direction: When the k+1 layer is removed, based on the diamond tilt angle a and the Z direction single layer step amount Z h , the theoretical compensation amount AX 理论 is determined. based on the theoretical compensation amount ΔX 理论 and the surface roughness Ra of the k+1 layer k+1 , the actual compensation amount ΔX real is determined.

9. The thinning method of claim 8, wherein, Theoretical compensation amount ΔX 理论 is: ΔX 理论 = Z h • tan α; The actual compensation amount ΔX real is: ΔX real = ΔX 理论 • (1 + C • σ Ra,k+1 ) Wherein, C is the material correction coefficient; σ Ra,k+1 the standard deviation of the surface roughness for the (k+1)th layer; β is the laser cone angle.

10. A thinning device characterized by comprising: The thinning method as claimed in any one of claims 1-9 can be performed on the diamond (D) to thin the diamond (D), comprising: A bearing mechanism (100) for fixing the diamond (D), the bearing mechanism (100) being deflectable and movable to drive the diamond (D) to deflect and move; A laser mechanism (200) located above the bearing mechanism (100) for laser thinning the diamond (D).

Images