Control method and system of dispensing type coupling equipment

Abstract

The invention provides a control method and system for dispensing type coupling equipment, and the method comprises the steps: simulating the glue discharge process of a dispensing head according to a dynamic glue discharge rate curve, and analyzing the fluid dynamic parameters of the glue discharge process to obtain the initial form of glue drops and the evolution trend of the necking process; if the wiredrawing effect occurs in the necking process of the glue drops, adjusting the slope of the glue discharge rate curve, and judging whether the wiredrawing effect still exists in the target glue drop form or not according to the adjusted glue discharge plastic curve; if the wire drawing effect still exists, the dispensing time, the dispensing pressure and the distance between the dispensing head and the working surface are adjusted by combining the rheological property of the glue; the glue outlet rate curve and the glue flow model are iteratively optimized, a target glue outlet rate curve for inhibiting the wire drawing effect and glue parameters are determined, and the glue parameters comprise viscosity, surface tension and density.

Description

Technical Field

[0001] The present invention relates to the field of information technology, and in particular to a control method and system for a dispensing coupling device. Background Art

[0002] When dispensing glue on multiple chips with a single dispensing head, the head needs to move quickly between different positions and precisely dispense glue. Due to the complex spatial relationship between the chip and the dispensing head, the dispensing head needs to dynamically adjust the dispensing rate during movement to accommodate varying dispensing requirements. However, this dynamic adjustment causes fluctuations in the dispensing rate curve, which in turn affects the formation of the glue droplet. At the moment the dispensing head stops dispensing, the viscoelasticity and surface tension of the glue cause a long, thin glue strand to form between the glue droplet and the dispensing head, known as the "stringing effect." This stringing effect not only affects dispensing accuracy but can also cause glue contamination of the chip. The rheological properties of the glue have a significant influence on the formation of the stringing effect. Glue parameters such as viscosity, elastic modulus, and relaxation time determine the speed and shape of the glue droplet necking process. When the dispensing head dispensing rate is high, the shear force on the glue increases, reducing the glue viscosity, accelerating the glue droplet necking process and making the stringing effect more pronounced. On the other hand, when the glue discharge rate is low, the viscosity of the glue is high, the necking process of the glue droplet is slow, and the drawing effect is relatively weakened. In addition, the spatial relative position relationship between the chip and the dispensing head will also affect the formation of the drawing effect. When the dispensing head needs to dispense glue at different heights or angles, the flow path and stress state of the glue will change, further exacerbating the complexity of the drawing effect. Therefore, how to suppress the glue droplet drawing effect under the dynamically changing glue discharge rate, combining the rheological properties of the glue and the spatial relative position relationship between the chip and the dispensing head, has become a key technical issue in the multi-chip dispensing process of a single dispensing head. Solving this problem requires in-depth research on the optimization of the glue discharge rate curve of the dispensing head, the control of the rheological properties of the glue, and the precise planning of the dispensing path to ensure the stability and accuracy of the dispensing process. Summary of the Invention

[0003] To solve the above technical problems, the present invention provides a control method for a dispensing coupling device, which mainly includes:

[0004] Obtain the spatial relative position relationship between the chip and the dispensing head, calculate the motion trajectory of the dispensing head on the chip, and determine the dynamic range of the glue discharge rate and the shear rate of the dispensing process based on the glue viscosity and the geometry of the dispensing path. The geometry of the dispensing path includes the degree of curvature and corner angles.

[0005] Based on the rheological properties of the glue, a mapping relationship between glue viscosity and shear rate was established to analyze the flow behavior of the glue during the dispensing process. Based on the actual needle geometry, dispensing speed, and ambient temperature during the dispensing process, the inlet pressure, flow rate, and outlet pressure were determined to construct a glue flow model.

[0006] Based on the glue flow model, the speed, direction, pressure distribution, and shear rate distribution of the glue in the dispensing needle and the dispensing head on the chip are output to determine the initial parameters of the dispensing process of the dispensing head. Combined with the positional relationship between the chip and the dispensing head, the position, speed, and actual glue output of the dispensing head within a preset time are collected to generate a dynamic dispensing rate curve.

[0007] Based on the dynamic glue discharge rate curve, the glue discharge process of the dispensing head is simulated. By analyzing the fluid dynamics parameters of the glue discharge process, the initial shape of the glue droplet and the evolution trend of the necking process are obtained;

[0008] If a stringing effect occurs during the necking process of the glue droplet, adjust the slope of the glue discharge rate curve and judge whether the target glue droplet morphology still has a stringing effect based on the adjusted glue discharge plastic curve;

[0009] If the stringing effect still exists, adjust the dispensing time, dispensing pressure and the distance between the dispensing head and the work surface based on the rheological properties of the glue;

[0010] By iteratively optimizing the glue discharge rate curve and glue flow model, the target glue discharge rate curve and glue parameters for suppressing the stringing effect are determined. The glue parameters include viscosity, surface tension and density.

[0011] Furthermore, the spatial relative position relationship between the chip and the dispensing head is obtained, and the dispensing head's motion trajectory on the chip is calculated. Based on the glue viscosity and the geometry of the dispensing path, the dynamic range of the dispensing rate and the shear rate of the dispensing head during the dispensing process are determined. The geometry of the dispensing path includes the degree of curvature and the angle of the corners. This involves: using a pre-calibrated conversion matrix between the dispensing head coordinate system and the chip coordinate system, a matrix transformation method is used to map the dispensing head coordinate system to the chip surface coordinate system, obtaining the relative position information between the dispensing head and the chip surface to obtain the dispensing starting point position coordinates and attitude angle values. The dispensing area contour is extracted from the chip surface image. Based on the dispensing starting point position coordinates, a dispensing trajectory curve equation is established. A sequence of discrete sampling point coordinates and curvature radius values at fixed intervals along the dispensing trajectory are calculated using cubic spline interpolation. For the dispensing trajectory sampling point coordinate sequence, the curvature change rate at each sampling point is calculated using a numerical differentiation method. Corner positions are identified through threshold judgment, and the sampling point density at the corners is increased, and the corner angle values are recorded. A shear rate-shear stress curve is obtained. A stress distribution equation is established based on the curvature radius and corner angle values to calculate the glue shear stress at each sampling point along the dispensing trajectory. A neural network regression method is used to calculate the dispensing head's glue delivery rate at each sampling point, combining the shear stress values and glue viscosity values at these sampling points with the curvature change rate of the dispensing trajectory. This generates a glue delivery sequence at discrete sampling points along the dispensing trajectory. Based on this delivery sequence and the sampling point coordinate sequence, a dispensing head velocity distribution curve is established. Deceleration is applied to corners and the compensated glue delivery value is calculated.

[0012] Furthermore, the spatial position data of the chip and the dispensing head are obtained, a three-dimensional coordinate system is established, and the motion trajectory of the dispensing head on the chip surface is calculated based on the three-dimensional coordinate data to obtain the geometric characteristics of the dispensing path. The dispensing path is divided into several path segments, and the dispensing rate range of each path segment is calculated. The dynamic change curve of the dispensing rate of the entire dispensing process is determined, including: obtaining the spatial position point cloud data of the dispensing head and the chip, calculating the three-axis displacement and rotation angle values of the dispensing head relative to the chip surface, and establishing the coordinate equation of the dispensing head motion trajectory based on the least squares fitting method. The projection curve on the chip surface is calculated based on the dispensing head motion trajectory coordinate equation, the trajectory image is collected to obtain the coordinate sequence of the contour edge points, and the curvature value of each point on the trajectory is calculated according to the curvature calculation formula. For the trajectory curvature value sequence, a curvature threshold and a path length threshold are set as segmentation conditions. The positions of the trajectory segment points are calculated, the coordinates of the starting and ending points of each path segment are obtained, the length and average curvature value of each trajectory segment are calculated, and a neural network regressor based on curvature characteristics is established. The input parameters include the path segment length, average curvature, and maximum curvature, and the corresponding upper and lower limits of the glue output rate are output. For the glue output rate limit of the path segment, cubic spline interpolation is used to calculate the initial glue output rate at the trajectory sampling point, and a constraint condition is set that the rate change between adjacent sampling points does not exceed a preset threshold. Based on the initial glue output rate value and the constraint condition, the variational method is used to solve the optimization equation to obtain a continuous glue output rate curve that meets the smoothness requirements. The weighted average method is used to achieve rate transition at the segment points.

[0013] Furthermore, based on the rheological properties of the glue, a mapping relationship between glue viscosity and shear rate is established. The flow behavior of the glue during the dispensing process is analyzed. Based on the actual needle geometry, dispensing speed, and ambient temperature during the dispensing process, the inlet pressure, flow rate, and outlet pressure are determined. A glue flow model is constructed, which includes: obtaining glue viscosity measurements at different shear rates, establishing a power-law function relationship between glue viscosity and shear rate, and calculating the glue's yield stress, consistency coefficient, and fluid index based on the measured data to obtain a set of rheological parameters at a reference temperature. A thermocouple array sensor is placed on the needle surface to measure the ambient temperature and glue temperature distribution during the dispensing process. Corrected rheological parameters at the actual operating temperature are calculated based on the corresponding relationship curve between the rheological parameter set and temperature. Based on the corrected rheological parameters, the shear stress distribution on the needle's inner wall is calculated based on the pressure gradient data at the needle's inlet and outlet and the glue flow rate, according to the needle's inner diameter and pipe length. This is combined with the corrected rheological parameters to obtain a velocity distribution function across the needle's cross section. The average flow velocity and volume flow rate at each needle cross section are then calculated through integration, and the pressure distribution within the needle is calculated. According to the pressure distribution, flow velocity distribution and rheological parameters, a glue flow state predictor is established, wherein the input parameters include pressure gradient, shear rate and temperature, and the output parameters are local flow velocity and viscosity.

[0014] Furthermore, based on the glue flow model, the speed, direction, pressure distribution, and shear rate distribution of the glue dispensed by the dispensing needle and the dispensing head on the chip are output, and the initial parameters of the dispensing head's dispensing process are determined. Combined with the positional relationship between the chip and the dispensing head, the dispensing head's position, speed, and actual dispensing volume within a preset time are collected to generate a dynamic dispensing rate curve. This includes: calculating the velocity and pressure field distributions within the needle based on the flow model, obtaining the shear rate distribution on the needle's cross section, and calculating the dispensing head's initial dispensing rate value based on the flow field parameters and a preset response characteristic curve. Based on the initial dispensing rate value, the acceleration time and start / stop time are calculated, and a visual sensor is used to capture a colloid trajectory image to obtain the actual dispensing path coordinates. Based on the real-time glue volume data and the actual dispensing path coordinates, the least squares method is used to fit the time-varying function of the glue volume. The actual dispensing rate sequence is obtained using a numerical differentiation method. The state variables of the Kalman filter are set to the velocity value and acceleration value, and the observation variable is set to the average glue volume within the sampling period. The filtered rate sequence is then calculated. A final rate control sequence taking path characteristics into consideration is generated according to the filtered rate sequence and the dispensing path coordinates.

[0015] Furthermore, based on the dynamic dispensing rate curve, the dispensing process of the dispensing head is simulated. By analyzing the fluid dynamic parameters of the dispensing process, the initial morphology of the dispensing droplet and the evolution trend of the necking process are determined. This method includes extracting the rate sequence during the dispensing process based on the dynamic dispensing rate curve, calculating the pressure and velocity distributions at the needle outlet, extracting the droplet profile curve sequence based on the transient process of the droplet separating from the needle, obtaining the initial morphology of the droplet, and calculating the droplet profile characteristic parameters. A deep convolutional network is established based on the droplet profile characteristic parameters. The input parameters include the profile curve sequence and the dispensing rate, and the output parameter is the predicted droplet morphology value. The network is trained using measured data to determine the evolution trend of the initial droplet morphology. Based on the predicted droplet morphology value, the surface stress distribution of the droplet is calculated by combining the surface tension coefficient measured by the capillary method and the viscosity coefficient measured by the rotational viscometer. Based on the surface stress distribution, the strain field in the necking region of the droplet is calculated using the finite element method. The fluid flow law during the necking process is solved using the fluid governing equation. The morphological evolution sequence of the droplet during its separation is calculated using the numerical integration method.

[0016] Furthermore, if a stringing effect occurs during the necking process of the droplet, the slope of the discharge rate curve is adjusted. Based on the adjusted discharge plastic curve, whether the target droplet morphology still exhibits a stringing effect is determined. This includes calculating the ratio of the neck length to the diameter based on the droplet's neck contour. If this ratio exceeds a preset threshold, the droplet is deemed to have stringed. Based on the stringing determination result, a deep neural network is used to establish a mapping relationship between stringing characteristics and discharge parameters. The input parameters include the neck length ratio and the stretching rate, and the output parameter is a correction to the slope of the rate curve. A piecewise linear interpolation method is used to correct the original discharge rate curve, reducing the slope during the necking phase. A set of fluid mechanics equations is used to calculate the stress distribution during the droplet's release process, obtaining the strain and velocity field distributions in the necking region. The length change of the droplet's neck is calculated using a numerical integration method. The neck length ratio under the corrected solution is calculated. If this ratio still exceeds the preset threshold, the process returns to parameter optimization. For corrected solutions that meet the threshold, the corresponding rate curve parameters are recorded as a sequence of discharge control parameters.

[0017] Furthermore, if the stringing effect still persists, the dispensing time, dispensing pressure, and the distance between the dispensing head and the work surface are adjusted based on the rheological properties of the glue. This includes: measuring the rheological properties of the glue at different shear rates and temperatures using a rotational viscometer, calculating the parameters of the power-law relationship between viscosity and shear rate, and acquiring stringing length data using a high-speed camera system to capture a stringing image sequence. Based on the stringing length data and rheological parameters, a temperature-compensated deep neural network is established, with input variables including ambient temperature, glue temperature, viscosity, and shear rate, and an output variable representing the stringing critical condition. A response surface optimization method is used to construct a process parameter optimization function based on the stringing critical condition, setting the dispensing time range, pressure variation range, and height adjustment range as constraints. A numerical iteration method is used to solve the optimal parameter combination. The adjusted stress field distribution is calculated using fluid mechanics equations, and the strain rate and velocity gradient in the necking region are obtained. Temperature compensation calculations are performed using real-time temperature data to obtain the compensated stress field distribution. The stringing degree is then determined to determine whether it meets a preset threshold. If not, the optimization parameters are adjusted using a gradient descent method, and the iteration process is repeated.

[0018] Furthermore, through iterative optimization of the glue rate curve and glue flow model, the target glue rate curve and glue parameters for suppressing the stringing effect were determined. The glue parameters included viscosity, surface tension, and density. This involved using a deep neural network to establish a mapping relationship between glue parameters and stringing characteristics. The input variables included viscosity, surface tension, density, and glue rate, and the output variables were string length and stress distribution. A glue parameter prediction function was obtained through training with experimental data. The mutual influence coefficients of parameter changes were calculated by combining viscometer and surface tension meter measurement data. A genetic algorithm was used to optimize the coupled relationship equation of the glue rate curve. The genetic encoding included three parameters: initial rate value, acceleration, and stability value. The string length weight coefficient was set to 0.6 and the stability weight coefficient was set to 0.4. Based on the optimized parameters, the strain distribution and deformation rate in the necking region were calculated. The string length threshold and stability index threshold were set as criteria. The comprehensive score of the optimization results was calculated to determine whether they met the preset threshold requirements. If not, the genetic encoding parameters were adjusted and the optimization process was continued. Based on the optimization results that met the requirements, the reliability of the parameter combination was verified using an orthogonal experimental method to determine the final target glue rate curve and glue parameters.

[0019] The present invention provides a control system for a dispensing coupling device, which mainly includes:

[0020] The position relationship acquisition module is used to obtain the spatial relative position relationship between the chip and the dispensing head, calculate the motion trajectory of the dispensing head on the chip, and determine the dynamic range of the glue discharge rate and the shear rate of the dispensing process of the dispensing head based on the viscosity of the glue and the geometry of the dispensing path. The geometry of the dispensing path includes the degree of curvature and corner angles.

[0021] The dispensing trajectory calculation module is used to establish a mapping relationship between glue viscosity and shear rate based on the rheological properties of the glue, analyze the flow behavior of the glue during the dispensing process, and determine the inlet pressure, flow rate, and outlet pressure based on the actual needle geometry, dispensing speed, and ambient temperature during the dispensing process to build a glue flow model;

[0022] The glue dispensing rate determination module is used to output the speed, direction, pressure distribution, and shear rate distribution of the glue in the dispensing needle and the dispensing head on the chip based on the glue flow model, determine the initial parameters of the dispensing head's dispensing process, combine the positional relationship between the chip and the dispensing head, and collect the dispensing head position, speed, and actual glue discharge within a preset time to generate a dynamic glue dispensing rate curve;

[0023] The glue flow modeling module is used to simulate the glue dispensing process of the dispensing head based on the dynamic glue dispensing rate curve. By analyzing the fluid dynamics parameters of the glue dispensing process, the initial shape of the glue droplet and the evolution trend of the necking process are obtained;

[0024] The dynamic glue discharge curve generation module is used to adjust the slope of the glue discharge rate curve if a drawing effect occurs during the glue droplet necking process, and judge whether the target glue droplet morphology still has a drawing effect based on the adjusted glue discharge plastic curve;

[0025] The glue droplet morphology simulation module is used to adjust the dispensing time, dispensing pressure, and the distance between the dispensing head and the work surface based on the rheological properties of the glue if the drawing effect still exists;

[0026] The glue dispensing parameter optimization module is used to iteratively optimize the glue dispensing rate curve and glue flow model to determine the target glue dispensing rate curve and glue parameters that suppress the drawing effect. The glue parameters include viscosity, surface tension and density.

[0027] The technical solution provided by the embodiment of the present invention may have the following beneficial effects:

[0028] The present invention discloses a control method for a dispensing coupling device. The method first obtains the spatial position relationship between the chip and the dispensing head, calculates the dispensing trajectory, and determines the range of change of the dispensing rate based on the glue viscosity and the geometric shape of the dispensing path. Then, a mapping relationship between viscosity and shear rate is established based on the rheological properties of the glue, and a glue flow model is constructed. By analyzing the fluid dynamics parameters, the morphological evolution of the glue droplets during the dispensing process is simulated. If a drawing effect occurs, the present invention iteratively adjusts the dispensing rate curve, dispensing time, pressure, distance and other parameters to optimize the glue flow model, and finally determines the target dispensing rate curve and glue parameters that suppress the drawing effect. This method can effectively improve the accuracy and quality of dispensing, reduce the scrap rate, and is suitable for precision manufacturing fields such as chip packaging. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 The figure is a flow chart of a control method of a dispensing coupling device of the present invention. DETAILED DESCRIPTION

[0030] In order to make the objectives, technical solutions and advantages of the present invention more clear, the present invention is described in detail below with reference to the accompanying drawings and specific embodiments.

[0031] like Figure 1 In this embodiment, a control method and system for a dispensing coupling device may specifically include:

[0032] S101. Obtain the spatial relative position information between the dispensing head and the chip, calculate the motion trajectory of the dispensing head and determine the dynamic range of the dispensing rate, and generate the shear rate distribution of the dispensing process by combining the glue viscosity and the path geometric characteristics. The path geometric characteristics include the path curvature and corner angles, and extract the trajectory characteristics through the collaborative work of multiple sensors to optimize the dispensing control.

[0033] S1011. Obtain the relative position data between the dispensing head and the chip surface, map the dispensing head coordinates to the chip surface coordinate system through a preset coordinate system transformation matrix, calculate the three-dimensional coordinate value and posture angle of the dispensing starting point, extract the contour line of the dispensing area from the chip surface image, construct the trajectory curve equation based on the coordinates of the dispensing starting point, use cubic spline interpolation to generate a discrete sampling point coordinate sequence every 0.2 mm, and calculate the curvature radius value of each sampling point.

[0034] S1012. For the sampling point coordinate sequence, the curvature change rate is calculated using the numerical differentiation method, and the curvature threshold is set to 0.5 to identify the corner position. When the curvature change rate exceeds the threshold, the sampling interval is reduced to 0.1 mm within 5 mm before and after the corner, and the corner angle value is recorded as 75 degrees, thereby improving the trajectory accuracy of complex path segments and ensuring the adaptability of dispensing control.

[0035] S1013. Combining the shear stress and glue viscosity data of the sampling points, a neural network regression method is used to calculate the glue discharge rate value of each sampling point, where the rate of the straight segment is set to 0.6 to 0.9 ml per second, and the rate of the corner segment is reduced to 0.3 to 0.5 ml per second. A continuous glue discharge rate initial value curve is generated through interpolation fitting, and deceleration compensation is performed on the corners to ensure the uniformity of the glue droplets.

[0036] S1014. Based on the initial value of the glue discharge rate and the coordinates of the sampling points, a motion speed distribution curve of the glue dispensing head is constructed. The speed of the straight section is set to 50 mm per second, which is gradually reduced to 15 mm per second at the corners. The glue discharge amount is increased by 15% at the corners for compensation. A smooth transition of the rate between the path segments is achieved through weighted averaging, and a dynamic glue discharge rate curve that meets the continuity requirements is generated.

[0037] S102. Establish a mapping relationship between viscosity and shear rate based on the rheological properties of the glue, analyze the flow behavior of the glue in the dispensing needle, and combine the needle geometry, dispensing speed and ambient temperature data to calculate the inlet pressure, flow rate and outlet pressure distribution, and then generate a flow model that describes the dynamic behavior of the glue, where the rheological properties include yield stress, consistency coefficient and fluid index.

[0038] S1021. Collect viscosity data of the glue at different shear rates. The shear rate range is set to 0.1 to 100 per second. Record the corresponding shear stress values. Establish a functional relationship between viscosity and shear rate through power law model fitting. Calculate the rheological parameters of the glue at a reference temperature of 25 degrees Celsius, where the yield stress is 125 Pa, the consistency coefficient is 2800 Pa·s, and the fluid index is 0.65, indicating that the glue has shear-thinning properties.

[0039] S1022. Real-time data on the temperature distribution of the environment and glue during the dispensing process was obtained. The needle inlet temperature was measured to be 25 degrees Celsius and the outlet temperature was 22 degrees Celsius. Based on the experimentally verified relationship between temperature and viscosity, the viscosity increased by approximately 3% for every 1 degree Celsius decrease in temperature. Based on this, the rheological parameters were corrected and the viscosity coefficient at the outlet was adjusted to 2890 Pa·s to reflect the changes in glue properties under actual working conditions.

[0040] S1023. Based on the geometric parameters of the needle, a dispensing needle with an inner diameter of 0.4 mm and a length of 12 mm was selected. The pressure at the inlet was 350 kPa and the pressure at the outlet was under normal pressure, respectively, measured by a micro pressure sensor. Combined with the flow rate data of 0.8 ml per minute, the shear stress distribution on the inner wall of the needle was calculated, and the wall shear stress was 280 Pa, and the pressure gradient was about 29 MPa per meter. At the same time, the average flow velocity of the cross section was solved to be 42 mm per second through the integration method. Based on the above pressure and flow velocity data, combined with the corrected rheological parameters, the Bernoulli equation was used to calculate the pressure distribution in the needle, which showed that the pressure showed an approximately linear downward trend along the axial direction.

[0041] To further describe the flow state of glue, a deep neural network predictor was constructed. Its network structure consists of 3 nodes in the input layer, including pressure gradient, shear rate and local temperature, 64 nodes in the hidden layer, and the output layer generates local flow velocity and viscosity values. The training data set contains 2,000 sets of measured working condition data. The prediction error is controlled within 5%, significantly improving the reliability of flow state prediction.

[0042] S1024. Based on the flow rate and viscosity distribution output by the deep neural network predictor, the finite difference method is used to solve the fluid control equations. The grid division spacing is set to 0.02 mm, and the complete flow field distribution inside the needle is calculated, including the velocity field, pressure field and stress field. The maximum flow velocity in the center reaches 85 mm per second, and the flow velocity at the wall is zero, forming a parabolic velocity profile. A high shear layer with a thickness of about 0.05 mm is formed near the wall, and the shear rate is as high as 150 per second. The local viscosity is reduced to 65% of the initial value due to the shear thinning effect, effectively reducing the flow resistance and optimizing the glue discharge stability.

[0043] S103. Output the speed, direction, pressure and shear rate distribution inside the dispensing needle and on the chip based on the glue flow model, determine the initial parameters of the dispensing process and combine the position relationship between the chip and the dispensing head. Use multiple sensors to collect the position, speed and actual glue output data of the dispensing head within a preset time period in real time to generate a dynamic glue output rate curve, where the initial parameters include the initial glue output rate, acceleration time and start-stop time.

[0044] A glue flow model is used to calculate the velocity and pressure fields within the needle, determining the shear rate distribution across the cross section. Combined with a pre-set response characteristic curve, the initial glue delivery rate is determined to be 0.8 ml / min. To improve control precision, a photoelectric sensor array acquires the three-dimensional position coordinates of the dispensing tip at a frequency of 200 Hz, achieving a position accuracy of 0.02 mm. A pressure sensor with a sensitivity of 0.5 kPa is placed within the needle cavity to record the dynamic changes in the glue delivery pressure in real time.

[0045] S1031. Calculate the flow field characteristics inside the needle based on the flow model. The velocity distribution is parabolic, with a maximum velocity of 75 mm / s at the center and zero velocity at the wall. The shear rate decreases gradually from 145 mm / s at the wall to zero at the center. This distribution indicates that the colloid is subjected to high shear near the tube wall. Calculate the average flow velocity and set the initial parameters based on the response characteristic curve. The acceleration time is set to 80 milliseconds and the start-stop time is set to 100 milliseconds to accommodate the mechanical response delay.

[0046] S1032. The actual path coordinates are extracted from the colloid trajectory image. The function of the change of the glue amount over time is fitted by the least squares method. The actual glue output rate sequence is generated by numerical differentiation, reflecting the process of the glue amount accelerating from zero to a steady-state value of 0.8 ml / min. The maximum overshoot is controlled within 0.15 ml / min.

[0047] S1033. For the actual glue discharge rate sequence, a Kalman filter is set to optimize data smoothness. The state variables include rate and acceleration. The observation variable is the average glue discharge amount within a 5-millisecond period. The standard deviation of the process noise is 0.05 ml / min, and the standard deviation of the measurement noise is 0.02 ml / min. After filtering, the noise of the rate sequence is significantly reduced. Subsequently, cubic spline interpolation is used to generate a continuous dynamic rate curve at intervals of 20 milliseconds. A 100-millisecond delay compensation is added at the start and stop points to eliminate the mechanical lag effect.

[0048] Combining the filtered velocity sequence with the path coordinates, a feedforward compensation method is used to adjust the dynamic glue-dispensing rate curve. At corners, the rate is reduced by 50% and the compensation time is extended by 150 milliseconds. Straight sections maintain a constant velocity, achieving control accuracy better than 2%. This method ensures the rate curve's adaptability to path characteristics through multi-sensor collaboration and algorithm optimization.

[0049] S104. Simulate the dispensing process of the dispensing head based on the dynamic dispensing rate curve, reveal the evolution trend of the initial morphology of the glue droplet and the necking process by analyzing the fluid dynamics parameters, extract the contour curve of the image sequence of the glue droplet detaching from the needle, predict the glue droplet morphology, and then calculate the surface stress distribution by combining the surface tension and viscosity data to generate the morphological evolution sequence of the glue droplet detachment process, which includes the functional relationship between the neck diameter and time, in order to optimize the dispensing process.

[0050] Based on the dynamic glue discharge rate curve, a velocity sequence was extracted and the flow field characteristics at the needle exit were calculated. The velocity distribution showed a parabolic shape, with a central velocity of 55 mm / s, zero velocity at the wall, and a pressure decreasing from 280 Pa at the center to atmospheric pressure at the edge. To simulate the formation of the glue droplet, a parameter field was constructed using a three-dimensional fluid numerical calculation method to describe the dynamic behavior of the glue at a flow rate of 0.8 ml / min. The initial glue droplet exhibited a spherical crown shape, with a height of approximately 0.8 mm and a base diameter of approximately 0.45 mm.

[0051] S1041. The contour curve sequence of the transient process of the glue droplet detaching from the needle is extracted through the edge detection algorithm, and the characteristic parameters of the glue droplet, such as height and diameter, are calculated. Observations show that the shape of the glue droplet is stable in the early stage of formation, but evolves rapidly in the necking stage. A deep convolutional network is constructed based on the contour curve sequence. The network consists of a 5-layer structure. The input layer receives 50 frames of continuous images and the corresponding glue discharge rate. The three middle layers are configured with 64, 128, and 64 convolution kernels respectively. The output layer predicts the height, diameter, and neck size of the glue droplet. The training data comes from measured samples, and the prediction accuracy reaches 95%. Compared with traditional manual analysis, this model significantly improves the efficiency and accuracy of morphological prediction, ensuring the reliability of subsequent stress calculations.

[0052] S1042. Combining the surface tension coefficient of 42 millinewtons per meter measured by the capillary method and the viscosity of 15,000 centipoise measured by the rotational viscometer, the surface stress distribution of the droplet was calculated. The Laplace pressure reached a peak of 450 Pa at the neck, and the minimum radius of surface curvature was 0.15 mm. The strain field in the necking area was calculated using the finite element method with a 0.01 mm grid. It was shown that the diameter of the neck decreased from 0.4 mm to 0.08 mm within 0.8 milliseconds and then broke. The necking speed was as high as 250 mm per second. Gravity and surface tension jointly drove the fluid to converge at both ends, forming an hourglass-shaped outline.

[0053] During the necking process, the researchers used the governing equations to solve the fluid flow patterns and establish a mapping between necking velocity and diameter. They found that when the local strain rate exceeded 1000 per second, shear thinning reduced the viscosity to 35% of its initial value, accelerating necking and reducing the risk of stringing. Based on this, they used a numerical integration method to calculate the morphological evolution sequence with a 0.01 millisecond step size, recording changes in the droplet's volume, center of gravity, and surface area, providing a complete description of the entire process from formation to detachment.

[0054] S105. If a wire drawing effect is detected during the necking process of the glue droplet, a corrected dynamic rate curve is generated by adjusting the slope of the glue discharge rate curve. The glue dispensing process is re-simulated based on the curve to obtain the target glue droplet shape, and it is determined whether the wire drawing phenomenon still exists. The contour line of the glue droplet neck is extracted to calculate the ratio of length to diameter, and the evolution of the neck shape is analyzed by numerical integration until the wire drawing effect is effectively suppressed to ensure that the glue dispensing quality meets the expected requirements.

[0055] The neck contour was extracted, and the length-to-diameter ratio was calculated. A threshold of 2.5 was set; any value exceeding this threshold was considered a sign of stringing. To quantify stringing, a deep neural network was constructed, consisting of four convolutional layers and two fully connected layers. The network inputs were the neck length ratio and stretch rate, with the number of feature maps in the intermediate layers set to 32, 64, 64, and 32, respectively. The output was a slope correction. Training data was derived from 1,500 sets of measured samples to ensure the accuracy of the mapping.

[0056] S1051. Based on the slope correction output by the neural network, the piecewise linear interpolation method is used to adjust the original glue discharge rate curve, reducing the slope of the necking stage from 0.6 ml / s² to 0.35 ml / s². At the same time, the slope is pre-increased to 0.8 ml / s² in the acceleration stage to increase the initial momentum. The corrected flow field is calculated using a set of fluid mechanics equations. The maximum strain rate in the necking area is reduced from 850 s² to 420 s², and the peak shear stress is reduced from 320 Pa to 180 Pa, thereby improving the uniformity of strain distribution and reducing the risk of wire drawing.

[0057] S1052. Based on the corrected rate curve, the numerical integration method is used to calculate the change in the length of the droplet neck with a step size of 0.2 milliseconds to generate a morphological evolution sequence. Observations show that the neck stretching speed is reduced by 45%, and the length ratio is reduced from 3.2 to 2.1, which is lower than the threshold of 2.5. If the ratio still exceeds the standard, the neural network is returned to re-optimize the parameters. Usually, it converges after 2 to 3 iterations. The final curve contains 50 control points, which fully describes the rate changes in the acceleration, stabilization and disconnection stages.

[0058] The above optimization process records the rate control parameter sequence that meets the requirements, and batch experiments verify its applicability under different glue properties. The results show that the drawing probability is reduced from 35% to below 5%, and the volume accuracy of the glue droplet remains stable.

[0059] S106. Adjust the dispensing time, pressure, and distance between the dispensing head and the working surface based on the rheological properties of the glue, determine the critical conditions for wire drawing, use the response surface optimization method to calculate the optimal combination of process parameters, and iteratively optimize until a stable wire drawing-free dispensing effect is achieved.

[0060] S1061. The rheological properties of the glue were tested. The shear rate was increased from 0.1 per second to 200 per second. The measured consistency coefficient was approximately 2800 Pa·s, the fluid index was 0.65, and the viscosity decreased by approximately 2.5% for every 1 degree Celsius increase in temperature. At the same time, analysis showed that the drawing length reached 1.8 mm at 25 degrees Celsius and dropped to 1.2 mm at 35 degrees Celsius.

[0061] Based on rheological parameters and wire drawing length data, a five-layer deep neural network was constructed. The inputs included ambient temperature, glue temperature, viscosity, and shear rate. The number of hidden layer nodes was 64, 128, 128, and 64, respectively. The critical length and strain rate of wire drawing were output. Training was based on 2,000 sets of measured data, with a prediction accuracy of 95%. The model was used to identify the influence of temperature changes on wire drawing. For critical conditions, a response surface optimization method was used to establish an optimization function. The independent variables included dispensing time of 0.1 to 0.5 seconds, pressure of 100 to 300 kPa, and height of 0.3 to 0.8 mm. The iterative solution determined the optimal combination of dispensing time of 0.25 seconds, pressure of 180 kPa, and height of 0.5 mm, effectively reducing the risk of wire drawing.

[0062] S1062. Based on the optimal parameter combination, the adjusted stress field is calculated using fluid mechanics equations. The maximum strain rate in the necking area is reduced from 850 per second to 420 per second, and the velocity gradient is reduced from 1200 per second to 680 per second. The stress distribution is more uniform. At the same time, a real-time temperature compensation mechanism is introduced. For every 1 degree Celsius increase in temperature, the pressure increases by 5 kPa and the time is shortened by 0.01 second to ensure that the stress field fluctuation is controlled within 10%. If the threshold is still not met, the parameters are adjusted using the gradient descent method with a step size of 5%. Usually, after 3 to 4 iterations, the wire drawing probability is reduced to below 5%.

[0063] S107. Determine a target glue discharge rate curve and glue parameters including viscosity, surface tension, and density for suppressing the stringing effect by iteratively optimizing the glue discharge rate curve and the glue flow model.

[0064] S1071. A deep neural network is used to establish a mapping between glue parameters and drawing characteristics. The network is designed with 4 hidden layers, and the number of nodes are 64, 128, 128, and 64, respectively. The input variables include viscosity, surface tension, density, and glue output rate. The output is the drawing length and stress distribution. The training data comes from 2000 groups of experimental samples. For example, when the input viscosity is 15,000 centipoise, the surface tension is 42 millinewtons per meter, the density is 1.2 grams per cubic centimeter, and the speed is 0.8 milliliters per minute, the predicted drawing length is about 1.8 mm. This model provides a high-precision prediction basis for parameter optimization.

[0065] Using a viscometer and a surface tension meter to measure the physical properties of the glue, they found that for every 1°C increase in temperature, viscosity decreased by approximately 2.5% and surface tension decreased by approximately 1.8%. Regression analysis established a coupling equation between the physical property parameters, and the mutual influence coefficient was calculated to quantify the dynamic relationship between viscosity and surface tension. Based on this, a genetic algorithm was used to optimize the glue droplet rate curve. The population size was set to 100, and evolution was carried out for 50 generations. The genes encoded an initial value of 0.5 ml / min, an acceleration of 1.5 ml / min squared, and a stable value of 0.8 ml / min. The fitness function calculated a comprehensive score using a weight of 0.6 for the drawing length and a weight of 0.4 for stability. The optimized rate curve significantly improved the stability of the glue droplet.

[0066] S1072. Based on the optimization parameters, the finite element method is used to solve the flow field characteristics of the droplet detachment process. The grid division accuracy is 0.02 mm. The calculation shows that the maximum strain rate in the necking area is reduced from 850 per second to 420 per second, and the deformation rate is reduced from 1200 per second to 680 per second. The uniformity of stress distribution is improved. If the comprehensive score is lower than 85 points or the drawing length exceeds 1.2 mm and the stability standard deviation exceeds 5%, the genetic parameters are adjusted and returned to the optimization step. The requirements are met after 3 to 5 iterations.

[0067] S1073. The reliability of the optimization results was verified through orthogonal experiments. A three-factor three-level experiment was designed to test the effects of viscosity, surface tension, and density respectively. Each group was repeated 5 times. The optimal combination was a viscosity of 16,000 centipoise, a surface tension of 44 millinewtons per meter, and a density of 1.22 grams per cubic centimeter. The corresponding rate curve increased from zero to 0.75 milliliters per minute within 0.2 seconds and stabilized. The coefficient of variation of the wire drawing was controlled below 8%, ensuring process robustness and dispensing quality.

[0068] The present invention provides a control system for a dispensing coupling device, which mainly includes:

[0069] The position relationship acquisition module is used to obtain the spatial relative position relationship between the chip and the dispensing head, calculate the motion trajectory of the dispensing head on the chip, and determine the dynamic range of the glue discharge rate and the shear rate of the dispensing process of the dispensing head based on the viscosity of the glue and the geometry of the dispensing path. The geometry of the dispensing path includes the degree of curvature and corner angles.

[0070] The dispensing trajectory calculation module is used to establish a mapping relationship between glue viscosity and shear rate based on the rheological properties of the glue, analyze the flow behavior of the glue during the dispensing process, and determine the inlet pressure, flow rate, and outlet pressure based on the actual needle geometry, dispensing speed, and ambient temperature during the dispensing process to build a glue flow model;

[0071] The glue dispensing rate determination module is used to output the speed, direction, pressure distribution, and shear rate distribution of the glue in the dispensing needle and the dispensing head on the chip based on the glue flow model, determine the initial parameters of the dispensing head's dispensing process, combine the positional relationship between the chip and the dispensing head, and collect the dispensing head position, speed, and actual glue discharge within a preset time to generate a dynamic glue dispensing rate curve;

[0072] The glue flow modeling module is used to simulate the glue dispensing process of the dispensing head based on the dynamic glue dispensing rate curve. By analyzing the fluid dynamics parameters of the glue dispensing process, the initial shape of the glue droplet and the evolution trend of the necking process are obtained;

[0073] The dynamic glue discharge curve generation module is used to adjust the slope of the glue discharge rate curve if a drawing effect occurs during the glue droplet necking process, and judge whether the target glue droplet morphology still has a drawing effect based on the adjusted glue discharge plastic curve;

[0074] The glue droplet morphology simulation module is used to adjust the dispensing time, dispensing pressure, and the distance between the dispensing head and the work surface based on the rheological properties of the glue if the drawing effect still exists;

[0075] The glue dispensing parameter optimization module is used to iteratively optimize the glue dispensing rate curve and glue flow model to determine the target glue dispensing rate curve and glue parameters that suppress the drawing effect. The glue parameters include viscosity, surface tension and density.

[0076] The above embodiments are intended to illustrate the technical solutions of the present invention and are not intended to limit the present invention. The present invention is described in detail with reference to the preferred embodiments only. It should be understood by those skilled in the art that the technical solutions of the present invention may be modified or replaced with equivalents without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or equivalents should be included within the scope of the claims of the present invention.

Claims

1. A method for controlling a dispensing coupling device, characterized in that: The method comprises: Obtain the spatial relative position relationship between the chip and the dispensing head, calculate the motion trajectory of the dispensing head on the chip, and determine the dynamic range of the dispensing rate and the shear rate of the dispensing process based on the viscosity of the glue and the geometry of the dispensing path; Based on the rheological properties of the glue, a mapping relationship between glue viscosity and shear rate was established to analyze the flow behavior of the glue during the dispensing process. Based on the actual needle geometry, dispensing speed, and ambient temperature during the dispensing process, the inlet pressure, flow rate, and outlet pressure were determined to construct a glue flow model. Based on the glue flow model, the speed, direction, pressure distribution and shear rate distribution of glue dispensed on the dispensing needle and the dispensing head on the chip are output, the initial parameters of the dispensing process of the dispensing head are determined, and the dynamic dispensing rate curve is generated based on the positional relationship between the chip and the dispensing head; Based on the dynamic glue discharge rate curve, the glue discharge process of the dispensing head is simulated. By analyzing the fluid dynamics parameters of the glue discharge process, the initial shape of the glue droplet and the evolution trend of the necking process are obtained; If a wire drawing effect occurs during the necking process of the glue droplet, the slope of the glue discharge rate curve is adjusted to obtain an adjusted glue discharge rate curve. Based on the adjusted glue discharge plastic curve, the dispensing process is re-simulated to obtain the target glue droplet shape and determine whether the target glue droplet shape still has a wire drawing effect. If the stringing effect still exists, the dispensing time, dispensing pressure and the distance between the dispensing head and the work surface are adjusted in combination with the rheological properties of the glue to determine the target glue discharge rate curve and glue parameters to suppress the stringing effect.

2. The method according to claim 1, characterized in that The method of obtaining the spatial relative position relationship between the chip and the dispensing head, calculating the motion trajectory of the dispensing head on the chip, and determining the dynamic variation range of the dispensing rate and the shear rate of the dispensing process of the dispensing head according to the viscosity of the glue and the geometry of the dispensing path includes: Extract the dispensing area contour line from the chip surface image, establish the dispensing trajectory curve equation according to the dispensing starting point position coordinates, and obtain the discrete sampling point coordinate sequence on the dispensing trajectory; For the coordinate sequence of sampling points on the dispensing trajectory, the curvature change rate at the sampling points is calculated, and the corner position point and corner angle value are determined according to the preset curvature threshold; The glue viscosity value is measured by a capillary viscometer and the glue discharge rate value of the glue dispensing head at the sampling point is calculated.

3. The method according to claim 2, characterized in that Also includes: The spatial position data of the chip and the dispensing head are obtained, and a three-dimensional coordinate system is established. Based on the three-dimensional coordinate data, the motion trajectory of the dispensing head on the chip surface is calculated to obtain the geometric characteristics of the dispensing path. The dispensing path is divided into several path segments, and the dispensing rate range of each path segment is calculated. The dynamic change curve of the dispensing rate of the entire dispensing process is determined, which specifically includes: Obtain the spatial position point cloud data of the dispensing head and the chip, and calculate the three-axis displacement and rotation angle values of the dispensing head relative to the chip surface; The projection curve of the dispensing head on the chip surface is calculated based on the displacement and rotation angle values to obtain the coordinate sequence of the contour edge points; Setting a curvature threshold and a path length threshold for the edge point coordinate sequence, and calculating the coordinates of the starting point and the ending point of each path segment of the projection curve; The path segment length and average curvature value are calculated according to the coordinates of the starting point and the ending point of the path segment, the glue dispensing rate limit is obtained by using a neural network regressor, and the dynamic change curve of the glue dispensing rate in the entire dispensing process is obtained by cubic spline interpolation.

4. The method according to claim 1, wherein Based on the rheological properties of the glue, a mapping relationship between glue viscosity and shear rate is established, the flow behavior of the glue during the dispensing process is analyzed, and the inlet pressure, flow rate and outlet pressure are determined according to the actual needle geometry, dispensing speed and ambient temperature of the dispensing process to construct a glue flow model, including: The rheological parameter group is calculated based on the corresponding relationship between the viscosity value of the glue at different shear rates and the shear rate; The rheological parameter group is corrected by using the temperature distribution data measured by the thermocouple array sensor to obtain the corrected rheological parameter value at the actual working temperature. The pressure gradient data and flow value measured by the micro pressure sensor and flow sensor are used to calculate the velocity distribution function on the needle cross section. A deep neural network is used to construct a glue flow model based on the velocity distribution function and the modified rheological parameter values.

5. The method according to claim 1, wherein The glue flow model outputs the speed, direction, pressure distribution, and shear rate distribution of glue dispensed on the dispensing needle and the dispensing head on the chip, determines the initial parameters of the dispensing process of the dispensing head, and generates a dynamic dispensing rate curve based on the positional relationship between the chip and the dispensing head, including: The shear rate distribution on the needle cross section is calculated based on the preset response characteristic curve and velocity field distribution, and the initial glue discharge rate value of the dispensing head is obtained through the shear rate distribution; The real-time glue quantity data and actual glue dispensing path coordinates of the dispensing head are obtained. The glue quantity variation function over time is obtained by least squares fitting, and the actual glue dispensing rate sequence is obtained by numerical differentiation method. Based on the actual glue dispensing rate sequence and the glue dispensing path coordinates, a dynamic glue dispensing rate curve is generated by using a cubic spline interpolation method.

6. The method according to claim 1, characterized in that The dispensing process of the dispensing head is simulated based on the dynamic dispensing rate curve. By analyzing the fluid dynamics parameters of the dispensing process, the initial shape of the glue droplet and the evolution trend of the necking process are obtained, including: Obtain an image sequence of the glue droplet leaving the needle, and obtain a glue droplet contour curve sequence using an edge detection method; Establishing a deep convolutional network model according to the glue droplet contour curve sequence, and outputting a glue droplet morphology prediction value by the deep convolutional network model; The predicted value of the droplet morphology is combined with the surface tension coefficient measured by the capillary method and the viscosity coefficient measured by the rotational viscometer to calculate the surface stress distribution of the droplet. The finite element method is used to calculate the strain field in the necking area of the droplet. The morphological evolution sequence of the droplet during detachment is obtained by the numerical integration method, and the evolution trend of the droplet necking process is obtained.

7. The method according to claim 1, characterized in that If a wire drawing effect occurs during the necking process of the glue droplet, the slope of the glue discharge rate curve is adjusted to obtain an adjusted glue discharge rate curve. The dispensing process is re-simulated based on the adjusted glue discharge plastic curve to obtain a target glue droplet shape, and whether the target glue droplet shape still has a wire drawing effect is determined, including: The neck contour line of the glue droplet is extracted from the image sequence, and the ratio of the neck length to the diameter is calculated based on the contour line. If the ratio exceeds the preset judgment threshold, the glue droplet is judged to have a wiredrawing phenomenon. A mapping relationship between the neck length ratio and the slope correction amount of the rate curve is established through a deep neural network. The original glue discharge rate curve is corrected using a piecewise linear interpolation method. The length change process of the glue droplet neck is calculated using a numerical integration method to obtain the neck morphology evolution sequence. Whether the target glue droplet morphology still has a wiredrawing effect is determined by whether the neck length ratio corresponding to the neck morphology evolution sequence exceeds the preset judgment threshold.

8. The method according to claim 1, characterized in that If the stringing effect still exists, adjust the dispensing time, dispensing pressure, and the distance between the dispensing head and the work surface based on the rheological properties of the glue, including: Obtain the glue drawing length data and viscosity shear rate power law relationship parameters, establish a temperature compensation model through deep neural network, and obtain the critical conditions for drawing; Aiming at the critical conditions of wire drawing, the response surface optimization method was used to construct a process parameter optimization function to obtain the optimal parameter combination of dispensing time, pressurization time, dispensing pressure and working surface distance.

9. The method according to claim 1, characterized in that The determination of the target glue discharge rate curve and glue parameters for suppressing the stringing effect includes: A prediction model of glue parameters and drawing characteristics is established, the measurement data of glue is obtained, and the coupling relationship equation between physical parameters is established. The genetic algorithm is used to optimize the glue output rate curve to determine whether the optimized parameters meet the preset threshold requirements. If the preset threshold requirements are not met, the iteration is continued to determine the target glue output rate curve and glue parameters that suppress the drawing effect.

10. A control system for a dispensing coupling device, characterized in that: The system comprises: The position relationship acquisition module is used to obtain the spatial relative position relationship between the chip and the dispensing head, calculate the motion trajectory of the dispensing head on the chip, and determine the dynamic range of the glue discharge rate and the shear rate of the dispensing process of the dispensing head based on the viscosity of the glue and the geometry of the dispensing path. The geometry of the dispensing path includes the degree of curvature and corner angles. The dispensing trajectory calculation module is used to establish a mapping relationship between glue viscosity and shear rate based on the rheological properties of the glue, analyze the flow behavior of the glue during the dispensing process, and determine the inlet pressure, flow rate, and outlet pressure based on the actual needle geometry, dispensing speed, and ambient temperature during the dispensing process to build a glue flow model; The glue dispensing rate determination module is used to output the speed, direction, pressure distribution, and shear rate distribution of the glue in the dispensing needle and the dispensing head on the chip based on the glue flow model, determine the initial parameters of the dispensing head's dispensing process, combine the positional relationship between the chip and the dispensing head, and collect the dispensing head position, speed, and actual glue discharge within a preset time to generate a dynamic glue dispensing rate curve; The glue flow modeling module is used to simulate the glue dispensing process of the dispensing head based on the dynamic glue dispensing rate curve. By analyzing the fluid dynamics parameters of the glue dispensing process, the initial shape of the glue droplet and the evolution trend of the necking process are obtained; The dynamic glue discharge curve generation module is used to adjust the slope of the glue discharge rate curve if a drawing effect occurs during the glue droplet necking process, and judge whether the target glue droplet morphology still has a drawing effect based on the adjusted glue discharge plastic curve; The glue droplet morphology simulation module is used to adjust the dispensing time, dispensing pressure, and the distance between the dispensing head and the work surface based on the rheological properties of the glue if the drawing effect still exists; The glue dispensing parameter optimization module is used to iteratively optimize the glue dispensing rate curve and glue flow model to determine the target glue dispensing rate curve and glue parameters that suppress the drawing effect. The glue parameters include viscosity, surface tension and density.

Images