Nitriding furnace control system and method based on gas flow integral differential information

Abstract

The invention belongs to the technical field, and particularly relates to a nitriding furnace control system and method based on gas flow integral differential information, and the system comprises a hardware layer, a data layer, and a control layer. The hardware layer comprises a furnace body system, a sensor group and an execution mechanism and is used for detecting signals in the furnace body and transmitting the signal data to the data layer; and the data layer comprises a data acquisition system and an analysis system. According to the method, the instantaneous volume flow is monitored, and a'flow-reaction process' quantitative relation is established; according to integral analysis on the total gas consumption amount according to actual production requirements, the material reaction progress is reversely deduced through a chemical equation, the reaction progress is accurately controlled, and theoretical support and data support are provided for manual refined reaction control; a data model is established, an integral information graph and a differential information graph of the gas flow are recorded in real time, and the process is conveniently modified after production is finished.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of nitriding process, in particular to a nitriding furnace control system and method based on gas flow integral differential information. BACKGROUND

[0002] As the core equipment for implementing the nitriding process, the control technology of the nitriding furnace has developed for many years, and various control methods coexist. The traditional nitriding furnace control mainly focuses on temperature, pressure and gas flow. Each control method provides necessary support for the implementation of the nitriding process at a certain period, but also gradually exposes some limitations that cannot meet the modern industrial high-precision and high-efficiency requirements.

[0003] In terms of temperature control, the common PID control algorithm is introduced, which automatically adjusts the power supply voltage or current of the heating element according to the deviation between the actual temperature feedback by the temperature sensor and the preset temperature. Taking a pit-type nitriding furnace as an example, by arranging thermocouples at different positions in the furnace, temperature data is collected in real time, and the PID controller dynamically adjusts the heating power according to these data, so that the uniformity of the furnace temperature is improved to a certain extent, and the stability can reach ±1℃, and the uniformity of the furnace temperature can reach ±5℃. However, for some nitriding processes with very high requirements for temperature uniformity, such as the nitriding treatment of some high-end semiconductor materials, this control precision is still insufficient. When facing complex nitriding process curves, such as containing multiple heating, holding and cooling processes and strict requirements for temperature change rate at each stage, the traditional PID control may have response lag, overshoot and other problems, which affects the consistency of product quality.

[0004] In terms of pressure control, in actual production, when the nitriding process involves mixing of multiple gases and complex reaction process, only controlling the pressure cannot completely guarantee the consistency of the reaction, because the change of gas composition may lead to differences in reaction rate and product quality under the same pressure.

[0005] In terms of gas flow control, in some complex nitriding processes, single reliance on mass flow controller is still insufficient. For example, in some processes that require accurate control of cumulative gas flow to ensure that the reaction is fully carried out or to accurately control the thickness of the nitriding layer, the traditional system lacks effective cumulative flow verification mechanism. Once the sensor or control system fails, it is difficult to detect data anomalies in time, which may lead to the scrap of the entire batch of products.

[0006] The conventional system only monitors the instantaneous volume flow, and the quantitative relationship between the flow and the reaction progress is not established; the reaction progress is a black box, and the integral analysis of the total gas consumption is lacking, and the reaction progress cannot be deduced by the chemical equation; the existing system relies on the temperature curve and the time setting to judge the process stage, ignores the influence of the activity difference of different batches of raw materials and the equipment aging on the reaction starting point, and often leads to "under-nitriding". SUMMARY

[0007] The purpose of the present application is to provide a nitrogenation furnace control system and control method based on gas flow integral differential information to solve the problems raised in the background art.

[0008] To achieve the above purpose, the present application provides the following technical solutions:

[0009] The nitrogenation furnace control system based on gas flow integral differential information comprises:

[0010] Hardware layer, data layer, control layer;

[0011] The hardware layer comprises: furnace body system, sensor group, actuator, for detecting the signal inside the furnace body, and transmitting the signal data to the data layer;

[0012] The data layer comprises: data acquisition system, analysis system, for collecting the signals inside the furnace body collected by the hardware layer, and analyzing and processing the signals, using integral / differential calculation method to process the data to obtain the differential result, and inputting the differential result into the preset mathematical model to obtain the progress of the current reaction, and giving the control rule;

[0013] The control layer comprises: PID control unit, man-machine interaction system, for dynamically adjusting the heating power, gas mass flow controller and vacuum valve opening degree by the integral result analyzed by the data layer and the flow compensated by the standard condition, and realizing man-machine interaction.

[0014] Preferably, the furnace body system comprises: front and rear door type furnace body, provided with heat preservation cylinder, temperature equalizing cylinder, silicon-carbon rod heating body and material plate in the furnace body; the sensor group comprises: gas mass flow sensor, temperature sensor, pressure sensor provided in the furnace body, and water temperature sensor for detecting cooling water; the actuator comprises: pneumatic valve for gas on-off, heating / cooling module controlled by PID, vacuum unit adopting mechanical pump and Roots pump.

[0015] Preferably, the data acquisition system comprises: data acquisition system integrated analog module, communication module, signal amplifier and filter; the data acquisition system supports local storage and cloud uploading.

[0016] Preferably, the analysis system comprises an integral / differential calculation module, a pressure and temperature compensation module, a "flow integral-reaction progress-mass prediction" mathematical model and a model correction module, a "flow integral-nitriding layer growth" mathematical model and a model correction module built in the analysis system industrial computer, and the analysis system supports three integral methods running in parallel; the integral / differential calculation module supports multi-sensor fusion integration, and when multiple branch gases are mixed, the total flow is obtained by integrating the flow of each sensor respectively; the data acquisition system supports positive / negative flow independent integration, and accurately measures the gas circulation amount of the exhaust gas recovery system.

[0017] Preferably, the man-machine interaction system comprises a local display screen displaying real-time integral graph and reaction rate graph (dP / dt), and supports parameter setting and fault alarm through a remote APP.

[0018] Preferably, the integral / differential calculation method of the data layer comprises three methods, the first method is to collect instantaneous flow signals through the analog module and the communication module of the PLC, and to perform real-time integration by using programming instructions; the second method is to collect data through the data acquisition system, to read real-time data of the sensor through SCADA, configuration software and self-defined program, and to realize integral calculation in the software; and the third method is to use the MCU built in the intelligent sensor to perform real-time integration on analog / digital signals, and to directly output the cumulative flow through the display screen or the communication interface, and to obtain the integral result by taking the average value of the integral calculation values of the three methods.

[0019] A nitriding furnace control method of gas flow integral / differential information, comprising the following steps:

[0020] (1) The sensor group of the hardware layer collects gas flow information data inside the furnace body, and transmits the data to the data layer;

[0021] (2) The data layer performs integral / differential calculation on the data by three methods, and obtains the integral result by averaging the integral / differential calculation results of the three methods;

[0022] (3) The integral result obtained in step (2) is substituted into the "flow integral-reaction progress-mass prediction" mathematical model and the model correction module, and the "flow integral-nitriding layer growth" mathematical model and the model correction module, to obtain the weight of the reaction product and the reaction progress, and to dynamically adjust the heating power, the gas mass flow controller and the opening degree of the vacuum valve.

[0023] Preferably, in step (3), the "flow integral-reaction progress-mass prediction" mathematical model and the model correction module, and the "flow integral-nitriding layer growth" mathematical model and the model correction module are configured with a correction module, which is used to correct the original mathematical model when a side reaction occurs in the reaction, so as to timely intervene to reduce the side reaction.

[0024] Compared with the prior art, the present application has the following advantages:

[0025] The present application can monitor the instantaneous volume flow, establish a quantitative relationship between the flow and the reaction progress, break the "black box" of the reaction progress, and provide theoretical support and data support for artificial fine control of the reaction by integral analysis of the total gas consumption, reaction progress calculation through chemical equations, and accurate control of the reaction progress. BRIEF DESCRIPTION OF DRAWINGS

[0026] Figure 1 The present application provides a structural diagram;

[0027] Figure 2 The present application provides a reaction rate of Example 1. DETAILED DESCRIPTION

[0028] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work fall within the scope of the present application.

[0029] The present application provides a technical solution:

[0030] The nitrogenation furnace control system based on gas flow integral and differential information comprises:

[0031] The hardware layer, the data layer, and the control layer;

[0032] The hardware layer comprises a furnace body system, a sensor group, and an actuator, which are used to detect signals inside the furnace body and transmit the signal data to the data layer;

[0033] The data layer comprises a data acquisition system and an analysis system, which are used to collect the signals inside the furnace body collected by the hardware layer, analyze and process the signals, process the data by using the integral / differential calculation method to obtain differential results, input the differential results into a preset mathematical model to obtain the progress of the current reaction, and give a control rule;

[0034] The control layer comprises a PID control unit and a man-machine interaction system, which are used to dynamically adjust the heating power, the gas mass flow controller, and the vacuum valve opening degree according to the integral results obtained by the data layer and the flow compensated by the standard condition, and realize man-machine interaction.

[0035] The furnace body system comprises a front-rear opening door type furnace body, a heat preservation cylinder, a temperature equalizing cylinder, a silicon-carbon rod heating body and a material plate are arranged in the furnace body; the sensor group comprises a gas mass flow sensor, a temperature sensor, a pressure sensor and a water temperature sensor for detecting cooling water arranged in the furnace body; the actuator comprises a pneumatic valve for gas on-off, a heating / cooling module adopting PID control and a vacuum unit adopting a mechanical pump and a Roots pump.

[0036] The furnace body adopts a front-rear split type stainless steel furnace body, an alumina fiber heat preservation layer (thickness 50mm) is laid in the inner layer, a graphite temperature equalizing cylinder is arranged in the middle, 4 groups of silicon-carbon rod heating bodies (model: G3 / 600-12, rated power 5kW) are installed on the bottom surface, and a quartz glass observation window (transmittance ≥90%) is hung on the top. The furnace components include three layers of material plates (silicon carbide material, spacing 100mm) for carrying workpieces, a thermocouple (K type, accuracy ±0.5℃) is arranged below the material plate to monitor the surface temperature of the workpiece in real time.

[0037] The silicon-carbon rod heating body is connected to the heating module of the temperature control system through a solid-state relay, and the relay control signal is 0-10V analog quantity (output by the temperature controller B2).

[0038] A Coriolis mass flowmeter (E+H Promass Q300, accuracy 0.1 level, output 4-20mA analog quantity+RS485 digital signal) is arranged at the gas inlet end, and three gas mass flow controllers (Brooks 5850E, range 0-500 SLM, control accuracy ±1% FS) are connected in parallel, and N2, NH3 and Ar gases are respectively introduced.

[0039] A pressure switch (OMRON ZSE40, threshold 0.5kPa) and a pneumatic butterfly valve (bore diameter DN50, response time <1s) are installed at the gas outlet end, and the valve opening degree is adjusted by the control signal (4-20mA) of the analysis system.

[0040] The data acquisition system comprises an analog quantity module, a communication module, a signal amplifier and a filter; the data acquisition system supports local storage and cloud uploading.

[0041] The analysis system includes an integral / differential calculation module, a pressure and temperature compensation module, a "flow integral-reaction progress-mass prediction" mathematical model and a model correction module, a "flow integral-nitriding layer growth" mathematical model and a model correction module, and the analysis system supports three integral methods running in parallel; the integral / differential calculation module supports multi-sensor fusion integration, and when multiple branch gases are mixed, the total flow is obtained by integrating the flow of each sensor respectively; the data acquisition system supports positive / negative flow independent integration, and accurately measures the gas circulation amount of the exhaust gas recovery system.

[0042] The man-machine interaction system includes a local display screen for displaying real-time integral graphs and reaction rate graphs (dP / dt), and supports parameter setting and fault alarm through a remote APP.

[0043] The integral / differential calculation method of the data layer includes three methods, the first method is to collect instantaneous flow signals through the analog module and the communication module of the PLC, and to perform real-time integration by using programming instructions; the second method is to collect data through the data acquisition system, to read real-time data of the sensor through SCADA, configuration software and self-defined programs, and to realize integral calculation in the software; the third method is to use the MCU built-in in the intelligent sensor to perform real-time integration on analog / digital signals, and to directly output the cumulative flow through the display screen or the communication interface, and to obtain the integral result by taking the average value of the three integral calculation values.

[0044] The data acquisition system core is a Siemens S7-1200 PLC integrated with an SM 1231 AI module (16-bit resolution) and a CM 1241 RS485 communication module, which is connected to an industrial computer (IPC, Advantech UNO-3082, i5 processor, 8GB memory) through a Profinet bus.

[0045] The analysis system runs on the IPC, installs WinCC V7.5 configuration software, and is built-in with a storage card (64GB, supporting 72-hour power-off data retention), which reads the integral values of the PLC and the built-in integral values of the flowmeter through the OPC UA protocol.

[0046] PLC hardware integration: use the "ADD" instruction in the ladder diagram to periodically accumulate the flowmeter analog signal (converted to kg / min) (sampling period 1s);

[0047] WinCC software integration: through C script, read the flowmeter digital signal once every 100ms, and calculate the cumulative value by using the trapezoidal integration method;

[0048] The Coriolis flowmeter calculates the cumulative flow by the internal MCU in real time, supports MODBUS RTU protocol reading, returns the current cumulative value (unit: kg) each time, and stores the data before power failure through EEPROM.

[0049] A nitrogen furnace control method of gas flow integral differential information, comprising the following steps:

[0050] (1) The sensor group of the hardware layer collects the gas flow information data in the furnace body and transmits the data to the data layer;

[0051] (2) The data layer performs integral / differential calculation on the data in three methods, and obtains the integral result by averaging the integral / differential calculation results of the three methods;

[0052] (3) The integral result obtained in step (2) is substituted into the "flow integral-reaction progress-mass prediction" mathematical model and model correction module, and the "flow integral-nitriding layer growth" mathematical model and model correction module to obtain the weight of the reaction product and the reaction progress, and dynamically adjust the heating power, gas mass flow controller and vacuum valve opening.

[0053] That is, after production starts, the data collection system collects temperature, pressure and other data information in real time, and three integral methods start to calculate gas integral information at the same time; then the information is transmitted to the analysis system, which first compensates the gas integral information according to temperature, pressure and other data information, then accumulates the integral according to time, compares the three data; if the error exceeds 1%, an alarm will be given; output real-time data, and perform model operation, optimize the model according to the built-in algorithm, and output the model calculation data; then according to the preset program, the control layer controls the process or manually adjusts the process.

[0054] In step (3), the "flow integral-reaction progress-mass prediction" mathematical model and model correction module, and the "flow integral-nitriding layer growth" mathematical model and model correction module are configured with a correction module, which is used to correct the original mathematical model when a side reaction occurs in the reaction, and to intervene to reduce the side reaction in time.

[0055] First, receive the real-time gas flow data and real-time collected data from the data collection system, including the temperature and pressure in the furnace, then compensate and integrate the data, compare the data of the three integral methods, and output to the display screen when the error is less than 1%, and at the same time import the built-in mathematical model for model verification and optimization, which can be manually intervened according to the built-in process of the system or manually.

[0056] The integral graph is drawn by loading three curves (PLC / WinCC / sensor integral) by WinCC trend control. The time range of the horizontal axis is optional (1 h / 4 h / 24 h), and the vertical axis supports double scales (volume flow m³ and mass flow kg). Clicking the curve displays the temperature, pressure value, and gas flow (with an accuracy of 1 decimal place) at the corresponding time point. Clicking any time period displays the corresponding gas flow integral value. The pressure signal is subjected to a moving average filter (window size 5 sampling points). When dP / dt> preset value, it is determined that it is in the "accelerating reaction period", triggering the orange highlight of the interface, and the reaction rate curve is drawn and output.

[0057] After the process is completed, the system automatically generates a report containing the integral / differential curve, temperature and pressure data, and reaction stage division, which is stored in the local database.

[0058] Through parallel integration of hardware (PLC), software (WinCC), and sensors (built-in MCU), the data reliability is improved by 60% compared to single PLC integration, and the error is reduced from ±2% to ±0.3%. The differential (dP / dt) curve is introduced into the nitriding furnace control, and compared with the traditional temperature curve monitoring scheme, the process intervention delay is shortened from 30 min to 2 min.

[0059] Experimental Example 1:

[0060] Process target: Silicon powder nitriding process, target: silicon powder nitriding rate greater than 95%, alpha phase content greater than 90%.

[0061] 1. Preparation:

[0062] Raw material preparation: High-purity silicon powder is selected as the raw material. The particle size of the silicon powder is 200 mesh, and the purity reaches 99.9%, ensuring stable quality and low impurity content of the silicon powder to ensure the smooth progress of the nitriding reaction.

[0063] Equipment inspection and calibration: The nitriding furnace control system is comprehensively inspected and calibrated. The working state and accuracy of the Coriolis mass flowmeter (model: E+H Promass Q500, accuracy 0.1 level), pressure sensor (accuracy ±0.01 kPa), temperature sensor (K-type thermocouple, accuracy ±0.5°C), and other equipment are checked. The flow integral system is calibrated by passing standard gas (nitrogen) to ensure that the deviation of the three integral methods (PLC hardware integral, WinCC software integral, and sensor built-in MCU integral) is within ±0.5%.

[0064] 2. Process parameter setting:

[0065] Temperature profile: The initial temperature was set to room temperature, and the furnace temperature was raised to 600°C at a rate of 10°C / min, then kept at this temperature for 3 hours, the purpose was to remove the moisture and impurities on the surface of the silicon powder. Then the temperature was raised to 1300°C at a rate of 5°C / min, and kept at this temperature, when dP / dt was 0, and after 60 min, the heating was stopped. This stage is the main reaction stage of silicon powder nitriding. Finally, the temperature was lowered to 100°C at a rate of 8°C / min, and then the furnace was discharged.

[0066] Gas flow and ratio: Nitrogen gas was introduced as the nitriding gas, and the initial flow rate was set to 300 SLM (standard liters per minute). During the temperature rise to 300°C and the holding stage, the nitrogen flow rate was maintained unchanged; during the temperature rise to 1300°C and the holding stage, the nitrogen flow rate was dynamically adjusted according to the reaction progress. At the same time, in order to promote the nitriding reaction, a small amount of hydrogen gas was introduced, and the hydrogen flow rate was set to 15 SLM. The gas ratio was nitrogen: hydrogen = 20:1.

[0067] 3. Nitriding process operation:

[0068] Charging and sealing: The prepared silicon powder was evenly charged into the graphite crucible of the nitriding furnace, the furnace door was closed and sealed, and the furnace was ensured to be in good sealing state to prevent gas leakage affecting the nitriding effect.

[0069] Temperature rise and purging: Start the heating system, and raise the furnace temperature to 600°C at the set temperature rise rate. During the temperature rise, nitrogen gas is introduced for purging to discharge the air in the furnace, so that the oxygen content in the furnace is reduced to below 1 ppm. At this time, the analysis system monitors the cumulative flow and pressure change of nitrogen gas in real time to ensure that the purging effect meets the expectations.

[0070] Main reaction stage: When the furnace temperature reaches 1300°C, the main reaction stage begins. The Coriolis mass flowmeter monitors the flow rates of nitrogen and hydrogen in real time and transmits the data to the analysis system. The analysis system integrates the gas flow in real time according to three integral methods, and combines with temperature, pressure and other data for comprehensive analysis. In the early stage of the reaction, as the reaction between silicon powder and nitrogen gas proceeds, the nitrogen flow rate gradually decreases, and the analysis system monitors the reaction progress in real time through the reaction rate graph (dP / dt). The reaction rate graph of Example 1 is shown in Figure 1. Figure 2 When the reaction rate graph shows that dP / dt begins to decrease, it indicates that the reaction rate slows down, and the analysis system automatically adjusts the nitrogen flow rate to increase to 350 SLM to maintain the continuous progress of the reaction. At the same time, the system calculates the cumulative consumption of nitrogen in real time, and infers the nitriding degree of silicon powder according to the chemical reaction formula.

[0071] Cooling and discharging: when dP / dt is 0 and after 60 min, stop heating, start the cooling system, and reduce the temperature in the furnace to 100 DEG C at the set cooling rate. During the cooling process, continue to pass a small amount of nitrogen for protection to prevent the silicon powder from being oxidized. When the temperature in the furnace is reduced to 100 DEG C, open the furnace door and take out the nitrided silicon powder product.

[0072] 4. Result detection and analysis:

[0073] Quality detection of the nitrided product: the quality of the nitrided silicon powder product is detected, the crystal structure of the product is analyzed by X-ray diffractometer (XRD), and it is determined that the product is silicon nitride (Si3N4). The microstructure of the product is observed by scanning electron microscope (SEM), and it is found that the size of the silicon nitride particles is uniform and the crystallinity is good.

[0074] Nitriding rate calculation: the content of nitrogen in the product is determined by chemical analysis method, and the nitriding rate of the silicon powder is calculated. The detection shows that the nitriding rate of the silicon powder in this time reaches 99.5%, which indicates that the nitriding effect is good, and the content of alpha phase is 92%, which indicates that the nitriding quality is good.

[0075] Compared with the traditional silicon powder nitriding process, the nitriding rate of the silicon powder is increased by more than 20% by using the nitriding furnace control system based on the integral information of gas flow in the application, and the product quality is more stable. At the same time, since the system can dynamically adjust the gas flow and temperature according to the reaction process, the utilization rate of nitrogen is increased by more than 20%, and the production cost is effectively reduced.

[0076] 5. Mathematical model construction

[0077] The core reaction of silicon powder nitriding is the reaction of silicon and nitrogen to generate silicon nitride, which is a strong exothermic reaction. The main reactions include the following two reactions:

[0078] Main reaction 1: direct nitriding to generate alpha-Si3N4

[0079] 3Si(s)+2N2(g)→Si3N4(s,α) ΔH=−744.6kJ / mol

[0080] Main reaction 2: direct nitriding to generate beta-Si3N4

[0081] 3Si(s)+2N2(g)→Si3N4(s,β) ΔH=−742.2kJ / mol

[0082] In the process of silicon powder nitriding, due to impurities in the raw materials, oxygen in the reaction environment and high temperature conditions and other factors, a variety of side reactions will be produced, mainly including:

[0083] Side reaction 1: oxidation reaction of silicon

[0084] Si(s) + O2(g) → SiO2(s) ΔH = -910.9 kJ / mol

[0085] Side Reaction 2: Reaction of Silicon with Silicon Dioxide

[0086] Si(s) + SiO2(s) → 2SiO(g) ΔH = +638.4 kJ / mol

[0087] Side Reaction 3: Formation of Silicon Oxynitride

[0088] 2SiO(g) + N2(g) → Si2N2O(s) + 0.5O2(g) ΔH = -6.6 kJ / mol (1500 K)

[0089] Side Reaction 4: Oxidation of Silicon Nitride

[0090] Si3N4(s) + 3O2(g) → 3SiO2(s) + 2N2(g) ΔH = -1088.1 kJ / mol

[0091] Side Reaction 5: Volatilization of Silicon and Gas-Phase Reactions

[0092] Si(l) → Si(g)

[0093] 3Si(g) + 2N2(g) → Si3N4(s)

[0094] When the reaction temperature exceeds the melting point of silicon (1414°C), the silicon powder will melt, and part of the silicon will volatilize to form a gas phase, which will then react with nitrogen gas in the gas phase to form silicon nitride.

[0095] Side Reaction 6: Impurity-Involved Reactions

[0096] Impurities in the silicon powder (such as Fe, Al, Ca, etc.) may participate in the reaction to form corresponding nitrides, oxides, or silicates, for example:

[0097] Fe(s) + N2(g) → Fe4N(s)

[0098] FeO(s) + SiO2(s) → FeSiO3(l)

[0099] Although these impurity reactions account for a small proportion of the overall reaction, they may have a significant impact on the reaction kinetics and product characteristics.

[0100] The shrinking core model is a commonly used model to describe gas-solid reaction kinetics, particularly suitable for cases where the solid reactant is encapsulated by a product layer. In the nitridation of silicon powder, as the reaction progresses, a layer of silicon nitride forms on the surface of the silicon particles, and the unreacted silicon core gradually shrinks, meeting the applicable conditions of the shrinking core model. The shrinking core model is based on the following assumptions: the silicon particles are spherical and have uniform particle sizes; the reaction starts from the surface of the particles and progresses towards the center; the product layer is uniform and dense, and the diffusion of nitrogen through the product layer is the rate-limiting step; the reaction temperature is constant, and temperature gradients are not considered. Under these assumptions, the conversion of silicon powder can be expressed as a function of time.

[0101] For the nitridation of silicon powder, the mathematical expression of the shrinking core model can be expressed as:

[0102]

[0103] where X is the conversion of silicon, D is the effective diffusion coefficient of nitrogen in the product layer, k is the surface reaction rate constant, R0 is the initial silicon particle radius, and t is the reaction time.

[0104] This equation takes into account both diffusion and surface reaction resistances. When the temperature is low, the surface reaction rate is slow, and the reaction is mainly controlled by the surface reaction; when the temperature is high, the surface reaction rate is fast, and the reaction is mainly controlled by diffusion.

[0105] By experimentally determining the relationship between the conversion of silicon and time at different temperatures, the model parameters D and k can be fitted using the nonlinear least squares method. These parameters usually satisfy the Arrhenius equation:

[0106] D = D0 exp(−Ed / RT)

[0107] k = k0 exp(−Ek / RT)

[0108] where D0 and k0 are the pre-exponential factors, Ed and Ek are the diffusion activation energy and reaction activation energy, respectively, R is the gas constant, and T is the absolute temperature.

[0109] The activation energy for the direct nitridation of silicon powder is in the range of 114-534 kJ / mol, and the diffusion activation energy is in the range of 106-148 kJ / mol, with specific values depending on the characteristics of the raw materials and reaction conditions.

[0110] Although the shrinking core model is simple and practical, it ignores some complex factors such as particle size distribution, non-uniformity of the product layer, and phase changes during the reaction. In order to more accurately describe the nitridation process of silicon powder, a more complex multi-phase reaction model can be established.

[0111] The particle size of actual silicon powder is usually not uniform, but has a certain distribution. A model considering particle size distribution can be expressed as:

[0112]

[0113] where f(R) is the particle size distribution function and r(R, t) is the reaction rate of particles with size R.

[0114] During the nitridation of silicon powder, various phase transformations occur, such as the transformation of α-Si3N4 to β-Si3N4. These phase transformations affect the reaction kinetics and the properties of the product. A model that takes into account phase transformations can be represented as:

[0115]

[0116]

[0117] where Xα and Xβ are the conversion rates of α-Si3N4 and β-Si3N4, respectively, kα and kβ are the formation rate constants of the α and β phases, respectively, and ktrans is the transformation rate constant from the α phase to the β phase.

[0118] The nitridation reaction of silicon powder exhibits clear stages, and a single reaction mechanism cannot be used to describe the entire reaction process. Based on thermogravimetric analysis and differential scanning calorimetry, the reaction can be divided into multiple stages, with different kinetic models used for each stage.

[0119] The direct nitridation reaction of silicon powder can be divided into two main stages:

[0120] 1000-1100°C: Interface chemical reaction control stage;

[0121] 1200-1300°C: Mixed control stage of interface chemical reaction and internal diffusion.

[0122] To more comprehensively describe the nitridation process of silicon powder, a comprehensive model that takes into account multiple side reactions needs to be established. This model is usually based on the principle of mass conservation, establishing differential equations for each component to describe its change over time.

[0123] For a reaction system containing multiple components, the mass balance equation can be established in the following form:

[0124]

[0125] where ni is the amount of substance of component i, νij is the stoichiometric coefficient of component i in reaction j (positive for products and negative for reactants), rj is the rate of reaction j, and V is the reaction volume.

[0126] A comprehensive model that takes into account the main reaction and multiple side reactions needs to establish kinetic equations for each reaction. For example, for a system containing the main reaction, oxidation reaction, and silicon nitride oxidation reaction, the following kinetic equations can be established:

[0127]

[0128]

[0129]

[0130]

[0131]

[0132]

[0133] where r1 is the main reaction rate, r2 is the oxidation reaction rate, r3 is the rate of silicon reacting with silicon dioxide, and r4 is the rate of silicon nitride oxidation reaction. The rate of each reaction can be expressed as:

[0134]

[0135]

[0136] where ki is the rate constant of reaction i, c(gas) is the concentration of the relevant gas, and n(solid) is the amount of substance of the relevant solid. The reaction rate constant is usually described by the Arrhenius equation:

[0137]

[0138] where Ai is the pre-exponential factor, Ei is the activation energy, R is the gas constant, and T is the absolute temperature.

[0139] The activation energies of different reactions vary greatly, for example, the activation energy of the nitridation reaction is usually in the range of 300-500 kJ / mol, while the activation energy of the oxidation reaction is lower, usually in the range of 100-200 kJ / mol.

[0140] Impurities in silicon powder (such as Fe, Al, Ca, etc.) have a significant impact on the nitridation reaction, mainly through the following ways:

[0141] Catalysis: promotes the progress of certain reactions;

[0142] Formation of liquid phase: reduces the reaction temperature and changes the reaction path;

[0143] Influence on product structure: changes the crystal form and microstructure of silicon nitride.

[0144] The catalytic effect of impurities can be described by changing the reaction rate constant. For example, for the reaction catalyzed by iron, it can be expressed as:

[0145] ​where [Fe] is the concentration of iron and a is the catalytic coefficient. When the impurity forms a liquid phase with silicon or silicon dioxide, the reaction mechanism changes. The presence of a liquid phase can accelerate the diffusion of nitrogen, promoting the reaction. This influence can be described by introducing a correction term for the effective diffusion coefficient:

[0146]

[0147] where liquid fraction is the volume fraction of the liquid phase and β is the correction coefficient.

[0148] Impurities have a significant influence on the transformation of a-Si3N4 to β-Si3N4. For example, studies have shown that the addition of metals with relatively strong electron donors (such as Ca, Mg, Y, etc.) is conducive to the formation of the α phase, while metals with relatively weak electron donors (such as Fe, Cu, Ag, Al, etc.) are not conducive to the formation of the α phase.

[0149] This influence can be described by changing the phase transition rate constant:

[0150]

[0151] where ktrans,0 is the phase transition rate constant of the pure system, Etrans is the phase transition activation energy, γ is the impurity influence coefficient, and [impurity] is the impurity concentration.

[0152] Comparative Example 1:

[0153] Process target: Silicon powder nitriding process, target: silicon powder nitriding rate greater than 95%, α phase content greater than 90%.

[0154] 1. Preparation:

[0155] Raw material preparation: High-purity silicon powder is selected as the raw material, with a particle size of 200 mesh and a purity of 99.9%, ensuring stable quality and low impurity content of the silicon powder to ensure the smooth progress of the nitriding reaction.

[0156] Equipment inspection and calibration: The control system of the nitriding furnace is comprehensively inspected and calibrated.

[0157] 2. Process parameter setting:

[0158] Temperature curve: Set the initial temperature to room temperature, increase the furnace temperature to 600°C at a rate of 10°C / min, then maintain the temperature for 3 hours to remove moisture and impurities on the surface of the silicon powder. Then increase the temperature to 1300°C at a rate of 5°C / min, and after maintaining the temperature for 30 hours, stop heating. This stage is the main reaction stage of silicon powder nitriding. Finally, reduce the temperature to 100°C at a rate of 8°C / min and discharge the furnace.

[0159] Gas flow and ratio: Nitrogen gas is introduced as nitriding gas, with an initial flow rate of 300 SLM (standard liters per minute). During the temperature rise to 300°C, the nitrogen flow rate is maintained constant; during the temperature rise to 1300°C, the nitrogen flow rate is dynamically adjusted according to the reaction progress. At the same time, a small amount of hydrogen gas is introduced to promote the nitriding reaction, with a hydrogen flow rate of 15 SLM. The gas ratio is nitrogen: hydrogen = 20:1.

[0160] 3. Nitriding process operation:

[0161] Charging and sealing: The prepared silicon powder is evenly charged into the graphite crucible of the nitriding furnace, the furnace door is closed and sealed to ensure that the furnace is in good sealing condition to prevent gas leakage affecting the nitriding effect.

[0162] Temperature rise and purging: Start the heating system and raise the temperature in the furnace to 600°C at the set temperature rise rate. During the temperature rise, nitrogen gas is introduced for purging to remove air in the furnace and reduce the oxygen content to below 1 ppm, with a duration of 30 min.

[0163] Main reaction stage: When the temperature in the furnace reaches 1300°C, the main reaction stage begins, with a holding time of 30 h, and continuous gas supply to maintain the pressure in the furnace at 100 KPa.

[0164] Cooling and furnace discharge: Start the cooling system and reduce the temperature in the furnace to 100°C at the set cooling rate. During the cooling process, a small amount of nitrogen gas is continuously introduced for protection to prevent the nitriding product from being oxidized. When the temperature in the furnace drops to 100°C, open the furnace door and take out the nitrided silicon powder product.

[0165] 4. Result detection and analysis:

[0166] Nitriding product quality detection: The nitrided silicon powder product is subjected to quality detection, X-ray diffraction (XRD) analysis of the crystal structure of the product to determine that the product is silicon nitride (Si3N4). Scanning electron microscopy (SEM) is used to observe the microstructure of the product, and it is found that the silicon nitride particles are uniform in size and have good crystallinity.

[0167] Nitriding rate calculation: The nitrogen content in the product is determined by chemical analysis method, and the nitriding rate of the silicon powder is calculated. The detection shows that the nitriding rate of the silicon powder in this experiment is only 75%, and the α phase content is 68%. At the same time, the gas consumption is 20% more than that in Example 1,

[0168] The process time is 30% longer than that in Example 1.

[0169] Experimental Example 2:

[0170] Process target: Gear nitriding process, target nitriding layer thickness: 60±5 μm.

[0171] Temperature profile: room temperature → 900℃ (heating rate 5℃ / min) → 1080℃ (holding for 4h) → 100℃ (cooling rate 10℃ / min);

[0172] Gas ratio: N2:NH3=3:1, total flow set to 200 SLM (standard conditions).

[0173] The operation process is as follows:

[0174] 1. Initialization:

[0175] The operator logs into the system (permission level: operator) and imports the gear nitriding process formula in the "process parameter" module;

[0176] The system automatically calibrates the sensor: open the calibration valve, introduce standard nitrogen (flow rate 100 SLM), and confirm that the three integral deviation is <0.5%.

[0177] 2. Heating stage:

[0178] The silicon-carbon rod is heated to 900℃, while Ar gas (50 SLM) is introduced to purge the air in the furnace, and the pressure sensor monitors the furnace pressure >10KPa when the exhaust valve is closed;

[0179] The analysis system calculates the cumulative flow of Ar gas in real time, and when the theoretical value (50 SLM×30min=15m³) is reached, it automatically switches to nitriding gas (N2+NH3).

[0180] 3. Nitriding stage:

[0181] The Coriolis flowmeter real-time feedback NH3 consumption mass, the analysis system calculates the N2 amount generated by reaction 2NH3→3H2+N2, combined with the integral difference to infer the gear surface nitrogen atom adsorption amount;

[0182] When the reaction rate graph shows dP / dt <0.15KPa / min for 2min, the system automatically increases NH3 flow by 5%; when 2min <0.1KPa / min, the system automatically increases NH3 flow by 8%, maintaining the reaction activity; At the same time, according to the pre-set model, the heating system is controlled by PDI.

[0183] 4. Cooling and ending:

[0184] Turn off the heating power, start the cooling water system (water temperature 25℃, flow rate 5L / min), and through the vacuum unit to reduce the furnace pressure to-0.09MPa to accelerate cooling;

[0185] After the process is completed, the system automatically generates a report, including integral / differential curves, temperature and pressure data, reaction stage division, and is stored in the local database (storage period 3 years).

[0186] During the whole nitriding process, the Coriolis mass flowmeter accurately and real-time feedbacks the consumption mass of NH3. Its output 4-20 mA analog signal, first through the RC passive filter composed of R = 1.2KΩ, C = 8nF, which can effectively filter out noise signals with a frequency higher than 199Hz. Subsequently, the signal enters the AD620 instrument amplifier, which amplifies the signal amplitude by 120 times, and then inputs to the Siemens SM 1231 AI 16-bit resolution PLC analog module for hardware integration operation. On the software level, the analysis system accurately infers the adsorption amount of nitrogen atoms on the gear surface according to the chemical reaction formula 2NH3→3H2+N2, combined with the difference between the operation results of the three ways of PLC hardware integration, WinCC software integration and sensor built-in MCU integration. Once the dP / dt in the reaction rate diagram is less than 0.15 KPa / min for 2 min, the system automatically increases the NH3 flow by 5%; when it is less than 0.1 KPa / min for 2 min, the software control algorithm automatically increases the NH3 flow by 8%, to ensure the continuous and efficient reaction.

[0187] After the completion of this gear nitriding process, the gear is comprehensively detected. According to the measurement of professional instruments, the thickness of the nitriding layer is within the target range of 60±1μm. Through the hardness testing equipment detection, the hardness of the nitriding layer is maintained between HV600-HV700, and the product qualified rate is 100%. This series of data fully shows that the system has successfully achieved the goal of high-precision gear nitriding process.

[0188] 5. Mathematical model construction

[0189] Ammonia decomposes at high temperatures to produce active nitrogen atoms, which react with iron to form a nitrided layer. The main reactions include:

[0190] Main reaction 1: Fe4N formation (ε phase, low temperature stable)

[0191] 8Fe(s) + 2NH3(g) → 2Fe4N(s) + 3H2(g) ΔH ≈ -30 kJ / mol

[0192] Main reaction 2: Fe2N formation (γ' phase, medium temperature stable)

[0193] 4Fe(s) + 2NH3(g) → 2Fe2N(s) + 3H2(g) ΔH ≈ -15 kJ / mol

[0194] Main reaction 3: Ammonia decomposition (provides active nitrogen)

[0195] 2NH3(g) ⇌ N2(g) + 3H2(g) ΔH ≈ +92 kJ / mol (proceeds in the forward direction at high temperatures, H2 can inhibit excessive nitriding)

[0196] Side reactions initiated by impurities O2 / H2O:

[0197] Side reaction 1: oxidation of iron

[0198] 2Fe(s) + O2(g) → 2FeO(s) ΔH ≈ −544 kJ / mol

[0201] 4Fe(s) + 3O2(g) → 2Fe2O3(s) ΔH ≈ −1648 kJ / mol

[0202] Side reaction 2: oxidation with participation of water vapor

[0203] Fe(s) + H2O(g) → FeO(s) + H2(g) ΔH ≈ −24 kJ / mol

[0204] Carbon-related side reactions (influence of carbon in steel):

[0205] Side reaction 3: oxidation of carbon (formation of CO / CO2)

[0206] C(s) + O2(g) → CO2(g) ΔH ≈ −394 kJ / mol

[0207] 2C(s) + O2(g) → 2CO(g) ΔH ≈ −221 kJ / mol

[0208] Side reaction 4: carbonitriding (formation of cyanide)

[0209] Fe3C(s) + NH3(g) → 3Fe(s) + HCN(g) + H2(g) ΔH ≈ +100 kJ / mol

[0210] (HCN is a toxic byproduct, carbon content must be controlled)

[0211] Oxidation of nitrided layer (at high temperatures):

[0212] Side reaction 5: oxidation of Fe4N

[0213] 4Fe4N(s) + 9O2(g) → 8Fe2O3(s) + 2N2(g) ΔH ≈ −3400 kJ / mol

[0214] Kinetic rate model construction:

[0215] Rate constants (Arrhenius form) The rate constants ki of each reaction satisfy:

[0216]

[0217] where Ai is the pre-exponential factor and Ei is the activation energy (e.g. E1 for main reaction 1 is ~ 120 kJ / mol and E2 for side reaction 1 is ~ 80 kJ / mol).

[0218] Main reaction 1 (Fe4N formation) rate r1:

[0219]

[0220] (p(NH3) is the ammonia partial pressure, is the surface coverage, reflecting the hindering of the reaction by the nitrided layer)

[0221] Main reaction 3 (NH3 decomposition) rate r3:

[0222]

[0223] (reversible reaction, k3 and k-3 are the forward and reverse reaction rate constants, respectively)

[0224] Side reaction 1 (FeO formation) rate r2:

[0225]

[0226] (correlated with the O2 partial pressure and the surface activity)

[0227] Building the mass conservation and diffusion equation model

[0228] Iron (Fe) consumption rate (iron consumed by main reactions and oxidation side reactions):

[0229]

[0230] (r5' is the Fe consumption rate by Fe4N oxidation in side reaction 5)

[0231] Building the growth model of the nitrided layer thickness

[0232] Fe4N layer thickness d growth is diffusion controlled (nitrogen diffusion in the iron matrix):

[0233]

[0234] (DN is the nitrogen diffusion coefficient and cN, surface is the surface nitrogen concentration)

[0235] Ammonia (NH3) consumption:

[0236]

[0237] (V is the reaction volume, NH3 is consumed by main reactions 1, 2, and also by the decomposition reaction)

[0238] Oxygen (O2) consumption:

[0239]

[0240] (r2' is the Fe2O3 formation rate, r5 is the Fe4N oxidation rate consuming O2)

[0241] The diffusion of nitrogen in iron follows Fick's second law:

[0242]

[0243] where c(x,t) is the nitrogen concentration at position x and time t, and DN is the diffusion coefficient of nitrogen. The initial and boundary conditions for this equation are typically:

[0244] Initial condition: c(x,0) = c0 (uniform initial nitrogen concentration in the gear)

[0245] Boundary condition: c(0,t) = cs (constant surface nitrogen concentration)

[0246] (constant deep nitrogen concentration)

[0247] Under these conditions, the analytical solution of the equation is in the form of an error function:

[0248] (where erf is the error function)

[0249] Comparative Example 2:

[0250] Process target: gear nitriding process, target nitrided layer thickness: 60±5 μm.

[0251] Temperature curve: room temperature → 900°C (heating rate 5°C / min) → 1080°C (holding for 4 h) → 100°C (cooling rate 10°C / min);

[0252] Gas ratio: N2:NH3=3:1, total flow rate set to 200 SLM (standard condition).

[0253] The operation process is as follows:

[0254] 1. Initialization:

[0255] The operator logs in to the system (permission level: operator) and imports the gear nitriding process formula in the "process parameter" module;

[0256] 2. Heating stage:

[0257] The silicon-carbon rod is heated to 900°C, while Ar gas (50 SLM) is introduced to purge the air in the furnace. The vortex flow meter transmits real-time data, and the pressure sensor monitors the furnace pressure >10 kPa to close the exhaust valve; after 30 min of aeration, the nitriding gas (N2+NH3) is turned on.

[0258] 3. Holding nitriding stage:

[0259] Hold for 4h, during which the pressure is less than 10KPa, and the NH3 flow is increased to 10KPa;

[0260] 4. Cooling and ending:

[0261] Turn off the heating power, start the cooling water system (water temperature 25℃, flow 5L / min), and accelerate the cooling by reducing the pressure in the furnace to -0.09MPa through the vacuum unit.

[0262] The same size gear is subjected to the nitriding process in the same working environment, and the target parameters are set in accordance with the above embodiment. However, in the actual operation process, the vortex flowmeter is extremely sensitive to the fluctuations of temperature and pressure of the working condition, and cannot accurately measure the actual flow of the gas, resulting in a large deviation in the amount of nitriding gas introduced into the furnace. In the heating stage, when the temperature approaches the target holding temperature, the fixed threshold control cannot flexibly adjust the heating power according to the actual reaction condition, resulting in a large overshoot of the temperature. In the holding stage, even if the reaction rate slows down significantly, the system cannot automatically increase the gas flow according to the reaction progress to maintain the reaction activity.

[0263] After the nitriding process is completed, the gear is detected, and it is found that the thickness of the nitriding layer is extremely uneven. The thickness of some areas is only 45μm, while the thickness of some areas exceeds 70μm. The hardness of the nitriding layer is extremely discrete, ranging from HV320 to HV650. According to statistics, the qualified rate of this batch of products is only 65%, which is in sharp contrast to Example 2.

[0264] Finally, it should be noted that the above content is only used to illustrate the technical solutions of the present application, and is not a limitation on the protection scope of the present application. Simple modifications or equivalent replacements of the technical solutions of the present application made by ordinary skilled in the art do not deviate from the essence and scope of the technical solutions of the present application.

Claims

1. A nitriding furnace control system based on gas flow integral differential information, characterized by: Comprise: Hardware layer, data layer, control layer; The hardware layer comprises: furnace body system, sensor group, actuator, for detecting the signal inside the furnace body, and transmitting the signal data to the data layer; The data layer comprises: data acquisition system, analysis system, for collecting the internal signal of the furnace body collected by the hardware layer, and analyzing and processing the signal, using integral / differential calculation method to process the data to obtain differential result, and inputting the differential result into the preset mathematical model to obtain the progress of the current reaction, and giving control rule; The control layer comprises: PID control unit, man-machine interaction system, for dynamically adjusting the heating power, gas mass flow controller and vacuum valve opening degree by the integral result obtained by the data layer analysis and the flow compensated by the standard condition, and realizing man-machine interaction.

2. The nitrogen furnace control system based on gas flow integral differential information according to claim 1, wherein: The furnace body system comprises: a front and rear door type furnace body, a heat preservation cylinder, a temperature equalizing cylinder, a silicon-carbon rod heating body and a material plate are arranged in the furnace body; the sensor group comprises: a gas mass flow sensor, a temperature sensor, a pressure sensor and a water temperature sensor for detecting cooling water arranged in the furnace body; the actuator comprises: a pneumatic valve for gas on-off, a heating / cooling module controlled by PID, and a vacuum unit adopting a mechanical pump and a Roots pump.

3. The nitrogenation furnace control system based on gas flow integrated differential information according to claim 1, characterized by: The data acquisition system comprises: a data acquisition system integrated analog module, a communication module, a signal amplifier and a filter; the data acquisition system supports local storage and cloud uploading.

4. The nitrogenation furnace control system based on gas flow integrated differential information according to claim 1, characterized by: The analysis system comprises: an integral / differential calculation module, a pressure and temperature compensation module, a "flow integral-reaction progress-mass prediction" mathematical model and a model correction module, and a "flow integral-nitriding layer growth" mathematical model and a model correction module built in the analysis system industrial computer; the analysis system supports three integral methods running in parallel; the integral / differential calculation module supports multi-sensor fusion integral; when multiple branches of gas are mixed, the total flow is obtained by integrating the flow of each sensor respectively; the data acquisition system supports positive / negative flow independent integral, and accurately measures the gas circulation amount of the exhaust gas recovery system.

5. The nitrogenation furnace control system based on gas flow integrated differential information of claim 1, wherein: The man-machine interaction system comprises a local display screen for displaying real-time integral graph and reaction rate graph (dP / dt), and supports parameter setting and fault alarm through remote APP.

6. The nitrogenation furnace control system based on gas flow integrated differential information according to claim 4, characterized by: The integral / differential calculation method of the data layer comprises three methods, the first method is to collect instantaneous flow signal through the analog module and communication module of PLC, and to perform real-time integral by using programming instruction; the second method is to collect data through the data acquisition system, to read real-time data of the sensor through SCADA, configuration software and self-defined program, and to realize integral calculation in software; the third method is to use the MCU built in the intelligent sensor to perform integral on analog / digital signal in real time, and to directly output cumulative flow through display screen or communication interface, and to obtain integral result by taking the average value of the three integral calculation values.

7. A control method of a nitriding furnace for gas flow integral differential information, characterized by, The method comprises the following steps: (1) The sensor group of the hardware layer collects gas flow information data inside the furnace body, and transmits the data to the data layer; (2) The data layer carries out integral / differential calculation of three methods for data, and obtains integral results by averaging integral / differential calculation results of the three methods; (3) The integral results obtained in step (2) are input into a "flow integral-reaction progress-mass prediction" mathematical model and a model correction module, a "flow integral-nitrided layer growth" mathematical model and a model correction module, to obtain the weight of the reaction product and the reaction progress, and to dynamically adjust the heating power, the gas mass flow controller and the vacuum valve opening degree.

8. The control method of a nitriding furnace for gas flow integrated differential information according to claim 7, characterized by, In step (3), the "flow integral-reaction progress-mass prediction" mathematical model and the model correction module, and the "flow integral-nitrided layer growth" mathematical model and the model correction module are configured with a correction module, which is used to correct the original mathematical model when a side reaction occurs in the reaction, and to timely intervene to reduce the side reaction.

Images