A method for online prediction of crystallinity and density of ethylene high pressure polymerization products

Abstract

The present invention provides a method for online prediction of crystallinity and density of ethylene high-pressure polymerization products. The prediction method comprises the following steps: a) constructing mass conservation, energy conservation, and momentum conservation equations of an ethylene high-pressure polymerization reactor based on real-time reaction conditions; b) calculating the microstructure of all polymer molecules in a selected simulation volume through Monte Carlo simulation based on the reactor conservation equations and reaction conditions; c) encoding the microstructure of the polymer molecules into a topological information recording method recognizable by a molecular dynamics simulation tool LAMMPS; d) randomly selecting polymer molecular chains for molecular dynamics simulation from all polymer molecules, inputting the topological information of their microstructures into LAMMPS, and online predicting the crystallinity and density of the polymer. The method for predicting the crystallinity and density of ethylene high-pressure polymerization products can not only accurately predict the crystallinity and density of the polymer from the first principles, but also can associate the real-time high-pressure polymerization reaction conditions to achieve online prediction of crystallinity and density.

Description

Technical Field

[0001] The invention belongs to the field of high-pressure free radical polymerization, and in particular relates to a method for online prediction of crystallinity and density of ethylene high-pressure polymerization products. Background Art

[0002] The polymer product obtained by free radical polymerization of ethylene under high temperature and high pressure conditions is called low-density polyethylene, also known as high-pressure polyethylene (LDPE). LDPE is the lightest variety of polyethylene resin, with a density of 0.910-0.925 g / cm 3 This is because LDPE has a high degree of branching and a low degree of crystallinity, so it has a lower density than HDPE and LLDPE. Therefore, it has good toughness and processing properties and is widely used in film materials and packaging materials.

[0003] Crystallinity and density are important performance parameters of LDPE. In addition to experimental measurement methods, LDPE density prediction models are more common. For example, the empirical correlation method described in Buchelli, A., et al., Modeling fouling effects in LDPE tubular polymerization reactors. 1. Fouling thickness determination. Industrial and Engineering Chemistry Research, 2005, 44 (5), 1474-1479 predicts the density of polyethylene by temperature and pressure, but this method cannot correlate the density with the microstructure of the polymer. There is also a model that correlates density with short chain branching content, which is described in Rokudai, M.; Okada, T., Characterization of Low-Density Polyethylenes and Relationships among Melt Index, Density, and Molecular Structural Parameters. Nihon Reoroji Gakkaishi, 1980, 8 (4), 154-160. However, this method is still an empirical model, and the prediction accuracy is significantly affected by the model parameters in the empirical correlation, and its applicability is limited. There is also a more complex method for calculating polymer density, namely the PC-SAFT model, described in Gross, J.; Sadowski, G., Perturbed-chain SAFT: An equation of state based on a perturbation theory for chain molecules. Industrial & Engineering Chemistry Research, 2001, 40 (4), 1244-1260. Although this model is a state equation model with results close to the real system and successful application, it is limited to density prediction and cannot predict the crystallinity of the polymer. Therefore, there is still a lack of a method that can simultaneously predict the crystallinity and density of LDPE from the first principles.

[0004] The microstructure of a polymer has an important influence on its properties such as crystallinity and density, and the microstructure is directly controlled by the polymerization reaction conditions. If the relationship between polymerization conditions-microstructure-product performance can be established, the polymer microstructure can be controlled by online regulation of the polymerization conditions to produce products that meet the desired polymer performance. CN102666608B proposes to calculate the density of the current polymer in the high-pressure reactor with the help of a model, and feed the density value back to the controller as a controlled variable, and then regulate the operating parameters of the high-pressure polymerization equipment to achieve the switching of polymer product brands with different densities. The density prediction model in this method is only related to kinetics, thermodynamics and material balance, and the density prediction is still based on empirical formulas. In addition, due to the loss of a large amount of molecular structure information, performance prediction cannot be associated with the detailed microstructure of the polymer, and crystallinity cannot be predicted, so it is difficult to achieve precise control of product performance by polymerization conditions. Summary of the invention

[0005] Based on this, it is necessary to provide a method for online prediction of crystallinity and density of ethylene high-pressure polymerization products to address the above problems. This method can not only accurately predict the crystallinity and density of polymers from the first principles, but also correlate with the real-time high-pressure polymerization reaction conditions to achieve online prediction of crystallinity and density.

[0006] A method for online prediction of crystallinity and density of ethylene high pressure polymerization products comprises the following steps:

[0007] a) constructing the mass conservation, energy conservation and momentum conservation equations for the ethylene high pressure polymerization reaction system based on real-time reaction conditions;

[0008] b) calculating the microstructure of all polymer molecules within the selected simulation volume through Monte Carlo simulation based on the conservation equations and reaction conditions constructed in step a);

[0009] c) Encoding the microstructure of polymer molecules into a topological information recording method that can be recognized by the molecular dynamics simulation tool LAMMPS;

[0010] d) Randomly select polymer chains for molecular dynamics simulation from all polymer molecules, input the topological information of their microstructure into LAMMPS, and predict the crystallinity and density of the polymer products online.

[0011] In the steps, the reaction conditions include reactor size, polymerization temperature, polymerization pressure, ethylene feed amount, initiator feed amount, and telogen feed amount. If the reactor uses a cooling water jacket for heat exchange, the temperature and flow rate of the circulating water used for cooling are also included in the reaction conditions. It is common for tubular reactors to use cooling water jackets for heat removal, and a small number of kettle reactors also have cooling water jackets.

[0012] In the steps, the polymerization temperature and polymerization pressure used as reaction conditions are temperature distribution data and pressure distribution data along the axial direction of the reactor. In a high-pressure polymerization reactor, the temperature and pressure are often not constant, but have a certain distribution along the axial distance of the reactor. The temperature distribution usually presents an increasing type or a multi-stage first increase and then decrease alternating type, and the pressure distribution usually presents a decreasing type.

[0013] In the steps, conservation of mass refers to conservation of matter in the reactor system, and the species involved include ethylene monomer, initiator and telogen injected into the reactor at any point, and the points where these species are injected into the reactor include the reactor inlet, outlet and side wall feed port; conservation of energy refers to conservation of heat in the reactor system, including heat exchange of materials inside the reactor, heat exchange between materials inside the reactor and the reactor wall, and heat exchange between the reactor wall and cooling water in the cooling water jacket; conservation of momentum refers to conservation of pressure in the reactor system.

[0014] In the steps, the Monte Carlo simulation method can be a classical Monte Carlo method, or a series of novel Monte Carlo methods that have been mutated and still retain the random sampling function. Monte Carlo simulation is a computational method known in the art for simulating the detailed molecular structure of a polymer. The classical Monte Carlo method is generally based on the Gillespie equation, which describes that the probability of each reaction occurring at any simulation time is proportional to its reaction rate, as described in, for example, Gillespie, DT, A general method for numerically simulating the stochastic time evolution of coupled chemical reactions. Journal of Computational Physics, 1976, 22 (4), 403-434. Since the classical Monte Carlo method takes a long time to calculate, there are some novel Monte Carlo methods that have been mutated and still retain the random sampling function in the art for the purpose of algorithm acceleration, saving computing resources, etc., as described in, for example, Gillespie, DT, Approximate accelerated stochastic simulation of chemically reacting systems. 2001, 115 (4), 1716-1733.

[0015] In the step, the polymer microscopic molecular structure information obtained by Monte Carlo simulation includes the main chain length of all polymers in the selected simulation volume, the number, site, and length of long side chains, and the number, site, and length of short side chains. The selected simulation volume is 1×10 -20 ~1×10 -18 m 3 The simulation volume is an important model parameter that affects the accuracy and timeliness of Monte Carlo simulation results. If the simulation volume is too small, the frequency of different reaction types in the sample cannot accurately describe the actual reaction process, which will interfere with the randomness of the sampling results and distort the Monte Carlo simulation results; on the contrary, if the simulation volume is too large, it will seriously increase the amount of calculation and prolong the simulation time.

[0016] In the steps, the recording method for encoding the topological information of the polymer molecular microstructure belongs to a universal language system based on computer graph theory, which can be recognized by the input interface of the molecular dynamics simulation tool LAMMPS.

[0017] In the step, the polymer molecular chains for molecular dynamics simulation are randomly selected from all polymer molecules, specifically: polymer molecular chains for molecular dynamics simulation are selected from all polymer molecules in an equal logarithmic interval manner, so that the molecular weight of the selected polymer is evenly distributed in the logarithmic space, and the number of the selected polymer molecular chains ranges from 10 to 50. The molecular weight (weight average molecular weight) of ethylene high pressure polymerization products is usually between 10 3 -10 6 The range of molecular weight is used as the standard for random screening in order to obtain representative polymer chains. The number of polymer chains screened for molecular dynamics simulation should not be too small, otherwise the simulation results will deviate greatly from the actual situation; but the number of molecular chains should not be too large, otherwise it will seriously increase the computational burden of molecular dynamics simulation.

[0018] In the steps, the molecular dynamics simulation uses the LAMMPS toolbox, adopts a coarse-grained molecular dynamics simulation method, and the simulation force field is the OPLS-UA force field.

[0019] In the steps, the temperature set for the online prediction of the crystallinity and density of the polymer is 320-370K and the pressure is 0.1MPa. The polymerization reactor is in a high temperature and high pressure state, and the polymer product will not crystallize. Usually, the pressure of the polymer drops to near normal pressure during the post-processing process after leaving the reactor, and the temperature gradually decreases. During this process, the polymer begins to crystallize and fix. It is generally believed that the initial crystallization temperature of regular polymers is about 370-380K, while low-density polyethylene has more side chains and a relatively low crystallization temperature. Considering the time-consuming problem of the molecular dynamics simulation process, if the simulation temperature is set too high, it is difficult for the polymer to crystallize and the simulation time will be significantly extended. Prior experience shows that the crystallization process of the polymer can be simulated in a limited time at 350K. Further lowering the temperature can quickly calculate the crystallization state, but it deviates from the actual crystallization process, which will cause a large deviation in the prediction results. Therefore, when performing online predictions of crystallinity and density, the pressure of the simulation scene set can be set to normal pressure, and the temperature is set to 320-370K, which is more appropriate.

[0020] In the method for online prediction of crystallinity and density of ethylene high-pressure polymerization products provided by the present invention, the crystallinity and density are calculated by inputting the topological information of the polymer molecular structure into the molecular dynamics simulation tool LAMMPS. The method is based on the first principle to make predictions, thereby ensuring the accuracy and universality of the crystallinity and density prediction method; the polymer molecular structure information used is obtained from Monte Carlo simulation, and the Monte Carlo simulation is a conservation equation constructed based on real-time reaction conditions, thereby ensuring the timeliness of the prediction results and realizing the online prediction of crystallinity and density. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 Schematic diagram of a method for online prediction of crystallinity and density of ethylene high pressure polymerization products proposed in an embodiment of the present invention. DETAILED DESCRIPTION

[0022] The following combination Figure 1 The method for online prediction of crystallinity and density of ethylene high pressure polymerization product provided by the present invention is described in detail.

[0023] like Figure 1 As shown, a method for online prediction of crystallinity and density of ethylene high pressure polymerization products comprises the following steps:

[0024] S1, constructs the mass conservation, energy conservation, and momentum conservation equations for the ethylene high-pressure polymerization reaction system based on real-time reaction conditions;

[0025] S2, based on the conservation equations of the reaction system and the reaction conditions, the microstructure of all polymer molecules within the selected simulation volume is calculated through Monte Carlo simulation;

[0026] S3, encoding the microstructure of polymer molecules into a topological information recording method that can be recognized by the molecular dynamics simulation tool LAMMPS;

[0027] S4, randomly select polymer molecular chains for molecular dynamics simulation from all polymer molecules, input the topological information of their microstructure into LAMMPS, and predict the crystallinity and density of the polymer products online.

[0028] In the steps, the reaction conditions include reactor size, polymerization temperature, polymerization pressure, ethylene feed amount, initiator feed amount, and polymerization agent feed amount. If the reactor uses a cooling water jacket for heat exchange, the temperature and flow rate of the circulating water used for cooling are also included in the reaction conditions.

[0029] In the steps, the polymerization temperature and the polymerization pressure used as the reaction conditions are the temperature distribution data and the pressure distribution data along the axial direction of the reactor.

[0030] In the steps, conservation of mass refers to conservation of matter in the reactor system, and the species involved include ethylene monomer, initiator and telogen injected into the reactor at any point, and the points where these species are injected into the reactor include the reactor inlet, outlet and side wall feed port; conservation of energy refers to conservation of heat in the reactor system, including heat exchange of materials inside the reactor, heat exchange between materials inside the reactor and the reactor wall, and heat exchange between the reactor wall and cooling water in the cooling water jacket; conservation of momentum refers to conservation of pressure in the reactor system.

[0031] In the steps, the Monte Carlo simulation method may be a classical Monte Carlo method, or a series of novel Monte Carlo methods that have undergone mutations and still retain the random sampling function.

[0032] In the step, the polymer microscopic molecular structure information obtained by Monte Carlo simulation includes the main chain length of all polymers in the selected simulation volume, the number, site, and length of long side chains, and the number, site, and length of short side chains. The selected simulation volume is 1×10 -20 ~1×10 -18 m 3 , and the further preferred simulation volume is 1×10 -19 ~1×10 -18 m 3 .

[0033] In the steps, the recording method for encoding the topological information of the polymer molecular microstructure belongs to a universal language system based on computer graph theory, which can be recognized by the input interface of the molecular dynamics simulation tool LAMMPS.

[0034] In the steps, the molecular weights of the polymer molecular chains randomly screened out for molecular dynamics simulation are equilogarithmically spaced (also called equilogarithmic spacing), and the number of polymer molecular chains screened out is 10 to 50, and the further preferred number of polymer molecular chains is 20 to 30. In an optional embodiment of the present invention, it is assumed that the molecular weight of the smallest molecule among all polymer molecules is a, and the molecular weight of the largest molecule is b. Assuming that the number of molecules used for molecular dynamics simulation is x, a geometric sequence of length x is generated between [a, b], and this geometric sequence contains x numbers. According to each value in the geometric sequence, the molecules with the closest molecular weight are selected from all polymer molecules, and these molecules are used as molecules for molecular dynamics simulation.

[0035] In the steps, the molecular dynamics simulation uses the LAMMPS toolbox, adopts a coarse-grained molecular dynamics simulation method, and the simulation force field is the OPLS-UA force field.

[0036] In the step, the temperature set for online prediction of the crystallinity and density of the polymer is 320-370K, the pressure is 0.1MPa, and the further preferred temperature is 340-360K.

[0037] The present invention will be further described below by way of examples.

[0038] Example 1

[0039] according to Figure 1 The method for online prediction of crystallinity and density of ethylene high pressure polymerization products shown in the figure is used to predict crystallinity and density of products of a certain ethylene high pressure polymerization industrial reactor online. The high pressure polymerization reactor is a tubular reactor with four reactor sections and a total length of 1760m. The inlet temperature of the first reactor zone is about 437K and the inlet pressure is about 230MPa. The initiator is composed of three peroxides with different decomposition rates and is injected at the starting point of each reactor zone. Ethylene and propylene (telomerase) are injected only in the first reactor zone. The parameters of each reaction condition include reactor size, temperature distribution of the extended tube, pressure distribution of the extended tube, feed amount of each species (ethylene, initiator, telomerase) in each reactor zone, temperature and flow of circulating water for cooling, all of which are provided by real-time monitoring data of the device.

[0040] The mass conservation, energy conservation, and momentum conservation of the reactor were calculated. Monte Carlo simulation was performed based on the conservation equations. The Monte Carlo method used was the classical Monte Carlo model, and the simulation volume selected was 1×10 -18 m 3 The microstructure of all polymer molecules in the simulation volume is obtained by calculation, including the main chain, long side chain, and short side chain information.

[0041] A universal language system based on computer graph theory is used to encode the polymer molecular microstructure obtained by Monte Carlo simulation into a topological information recording method that can be recognized by the molecular dynamics simulation tool LAMMPS, which is used to input polymer molecules into LAMMPS.

[0042] 20 chains were randomly selected from all simulated polymer molecules in equal logarithmic intervals. The molecular dynamics simulation used the LAMMPS toolbox, and the topological information of the selected polymer chain microstructure was input into LAMMPS. The coarse-grained molecular dynamics simulation method was used, the simulation force field was the OPLS-UA force field, the simulation temperature was 350K, and the pressure was 0.1MPa, to predict the crystallinity and density of the polymer online. The calculation process was performed on a large supercomputer platform.

[0043] The predicted crystallinity is stable at 41.0% and the density is 0.924 kg / m 3 The total calculation time is 45 hours. The measured density of the polymerization product of this reactor provided by the factory is 0.921kg / m 3 The crystallinity is 43.1%. The relative error between the predicted value and the actual value of crystallinity is -4.9%, and the relative error of density is 0.3%.

[0044] Example 2

[0045] The difference between Example 2 and Example 1 is that 5 chains are randomly selected from all simulated polymer molecules in an equal logarithmic interval manner.

[0046] The predicted crystallinity is stable at 35.2% and the density is 0.911 kg / m 3 The total calculation time is 18 hours. The relative error between the predicted value and the actual value of crystallinity is -18.3%, and the relative error of density is 1.1%.

[0047] Example 3

[0048] The difference between Example 3 and Example 1 is that 40 chains are randomly selected from all simulated polymer molecules in an equal logarithmic interval manner.

[0049] The predicted crystallinity is stable at 45.3% and the density is 0.925 kg / m 3 The total calculation time is 103 hours. The relative error between the predicted value and the actual value of crystallinity is 5.1%, and the relative error of density is 0.3%.

[0050] The above-mentioned embodiments only express several implementation methods of the present invention, and the description is relatively specific and detailed, but it cannot be understood as limiting the scope of the present invention. For ordinary technicians in this field, several modifications and improvements can be made without departing from the concept of the present invention, which all belong to the protection scope of the present invention.

Claims

1. A method for online prediction of crystallinity and density of ethylene high pressure polymerization products, characterized in that: The steps include: a) Construct the mass conservation, energy conservation and momentum conservation equations of the ethylene high pressure polymerization reaction system based on real-time reaction conditions; b) calculating the microstructure of all polymer molecules within the selected simulation volume through Monte Carlo simulation based on the conservation equations and reaction conditions constructed in step a); c) encoding the microstructure of the polymer molecule into a topological information recording method recognizable by the molecular dynamics simulation tool LAMMPS; in the step c), the topological information recording method used to encode the microstructure of the polymer molecule belongs to a universal language system based on computer graph theory, which can be recognized by the input interface of the molecular dynamics simulation tool LAMMPS; d) Randomly select polymer chains for molecular dynamics simulation from all polymer molecules, input the topological information of their microstructure into LAMMPS, and predict the crystallinity and density of the polymer products online.

2. The method for online prediction of crystallinity and density of ethylene high pressure polymerization product according to claim 1, characterized in that: The reaction conditions include reactor size, polymerization temperature, polymerization pressure, ethylene feed rate, initiator feed rate, and telogen feed rate. If the reactor uses a cooling water jacket for heat exchange, the temperature and flow rate of the circulating water used for cooling are also included in the reaction conditions.

3. The method for online prediction of crystallinity and density of ethylene high pressure polymerization product according to claim 2, characterized in that: The polymerization temperature and the polymerization pressure used as the reaction conditions are temperature distribution data and pressure distribution data along the axial direction of the reactor.

4. The method for online prediction of crystallinity and density of ethylene high pressure polymerization product according to claim 1, characterized in that: Conservation of mass refers to the conservation of matter within the reactor system. The species involved include ethylene monomer, initiator and telogen injected into the reactor at any point. The points where these species are injected into the reactor include the reactor inlet, outlet and side wall feed port. Conservation of energy refers to the conservation of heat within the reactor system, including heat exchange between materials inside the reactor, heat exchange between materials inside the reactor and the reactor wall, and heat exchange between the reactor wall and the cooling water in the cooling water jacket. Conservation of momentum refers to the conservation of pressure within the reactor system.

5. The method for online prediction of crystallinity and density of ethylene high pressure polymerization product according to claim 1, characterized in that: The polymer microscopic molecular structure information obtained by Monte Carlo simulation includes the main chain length of all polymers in the selected simulation volume, the number, site, and length of long side chains, and the number, site, and length of short side chains. The selected simulation volume is 1×10 -20 ~1×10 -18 m 3 .

6. The method for online prediction of crystallinity and density of ethylene high pressure polymerization product according to claim 1, characterized in that: The method of randomly selecting polymer molecular chains for molecular dynamics simulation from all polymer molecules is specifically as follows: polymer molecular chains for molecular dynamics simulation are selected from all polymer molecules in an equal logarithmic interval manner so that the molecular weights of the selected polymers are evenly distributed in the logarithmic space, and the number of the selected polymer molecular chains is 10 to 50.

7. The method for online prediction of crystallinity and density of ethylene high pressure polymerization product according to claim 1, characterized in that: The molecular dynamics simulation used the LAMMPS toolbox, adopted the coarse-grained molecular dynamics simulation method, and the simulation force field was the OPLS-UA force field.

8. The method for online prediction of crystallinity and density of ethylene high pressure polymerization product according to claim 1, characterized in that: The temperature set for online prediction of polymer crystallinity and density is 320~370 K and the pressure is 0.1MPa.

Images