Polymer melt online rheology testing method and system

Abstract

The application discloses a kind of polymer melt online rheological test method and system, belong to high polymer material processing and rheological test technical field.The method is based on bypass type slit rheological online test system, part of melt in main extrusion flow channel is introduced into bypass test branch, is transported to slit test cavity by melt delivery unit, and gathers before and after pressure, test section pressure difference, temperature and rotating speed data.Theoretical flow under low-speed operating condition is corrected by multi-parameter coupling flow correction model, and actual volume flow and online rheological parameter are obtained.The slit test cavity includes test cavity front flow channel, slit test section and test cavity rear flow channel, and the length of test cavity front flow channel is greater than the length of test cavity rear flow channel, to weaken inlet disturbance.Further, online shear rate-viscosity data is input into zero shear viscosity prediction model, and zero shear viscosity prediction value is obtained.The application can improve the flow calculation accuracy and prediction stability of low shear online test.

Description

Technical Field

[0001] This invention relates to the field of polymer material processing and rheological testing technology, and in particular to a method and system for online rheological testing of polymer melts in polymer extrusion systems. Background Technology

[0002] Polymer melt rheological parameters are important characterizing parameters in processing such as extrusion, injection molding, and blow molding. Among them, zero-shear viscosity can characterize the flow characteristics of polymer melts under extremely low shear rate conditions and is closely related to molecular chain entanglement, relaxation behavior, and processing stability. Obtaining zero-shear viscosity online would have significant engineering application value for process control, material state identification, and quality assessment during processing.

[0003] Current polymer melt rheology testing largely relies on offline equipment. While offline testing can obtain relatively complete rheological profiles, the testing conditions differ from actual continuous processing, making it difficult to reflect the true state of the melt in real time at the production site. Therefore, online rheology testing is gradually gaining importance.

[0004] Current online rheological testing technologies still have the following problems: First, online testing in the low shear rate region is difficult. As the shear rate decreases, the differential pressure signal in the test chamber weakens, easily approaching the sensor's resolution limit, affecting measurement stability. Second, when using a melt pump for low-speed delivery, there is a deviation between the theoretical displacement and the actual flow rate. If the shear rate and viscosity are still directly calculated based on the theoretical displacement, the accuracy of online rheological parameter calculations will be reduced. Third, when the online measurable data only covers the low-to-medium or high-to-medium shear rate range, traditional empirical constitutive fitting is prone to parameter coupling and extrapolation instability, making it difficult to reliably identify zero-shear viscosity.

[0005] Furthermore, in slit-type online rheological testing devices, the geometry of the inlet and outlet sections of the test chamber affects the stability of the flow field within the test section. If the polymer melt does not develop sufficiently before entering the slit test section, inlet disturbances may be transmitted to the pressure measurement region, causing fluctuations in the differential pressure signal within the test section. Therefore, it is necessary to further optimize the flow channel structure before and after the slit test chamber in bypass-type online testing structures, and combine low-velocity flow correction and zero-shear viscosity prediction methods to improve the stability and reliability of online rheological test results.

[0006] Therefore, there is an urgent need to provide an integrated technical solution that takes into account the test chamber structure design, low-speed flow correction and zero-shear viscosity prediction, so as to improve the feasibility and reliability of online rheological characterization of polymer melts. Summary of the Invention

[0007] To address the aforementioned technical problems, this invention provides a method and system for online rheological testing of polymer melts.

[0008] In a first aspect, this application provides an online rheological testing method for polymer melts, employing the following technical solution: A method for online rheological testing of polymer melts is implemented based on a bypass slit rheological online testing system. The bypass slit rheological online testing system includes a bypass test branch connected to the main extrusion channel, a melt delivery unit disposed on the bypass test branch, a slit test chamber disposed downstream of the melt delivery unit, a pressure and temperature acquisition unit, and a measurement and control and data processing unit.

[0009] The slit test chamber includes a front flow channel, a slit test section, and a rear flow channel connected sequentially along the melt flow direction. The length of the front flow channel is greater than the length of the rear flow channel, allowing the polymer melt to have a longer flow development distance before entering the slit test section, thereby reducing the impact of inlet disturbances on the differential pressure signal of the test section.

[0010] The method includes the following steps: S1. A multi-parameter coupled flow correction model for the melt conveying unit under low-speed operation is established in advance. Under different melt conveying unit rotation speeds, different pressure differences between the inlet and outlet sides of the melt conveying unit, and different melt temperatures, actual volumetric flow rate data is obtained through flow calibration experiments. The model parameters of the multi-parameter coupled flow correction model are obtained by fitting the actual volumetric flow rate data as the target value. S2. During online testing, a portion of the polymer melt in the main extrusion channel is introduced into the bypass test branch, and the polymer melt is transported to the slit test chamber through the melt conveying unit. S3. At the set rotational speed N of the melt conveying unit, collect the inlet side pressure P0, outlet side pressure P3, upstream pressure P1 of the slit test section, downstream pressure P2 of the slit test section, and melt temperature T of the melt conveying unit. S4. Based on the rotational speed N of the melt conveying unit and the pressure difference ΔP between the inlet and outlet sides of the melt conveying unit. p =P3-P0 and melt temperature T, the theoretical flow rate of the melt conveying unit is corrected by the multi-parameter coupled flow correction model established in step S1 to obtain the actual volumetric flow rate Q in the bypass test branch. S5. Based on the actual volumetric flow rate Q and the pressure difference ΔP in the slit test section. s =P1-P2 and the geometric parameters of the slit test chamber, calculate the online rheological parameters of the polymer melt under the current working conditions, such as shear rate, shear stress and apparent viscosity; S6. Change the rotational speed N of the melt conveying unit and repeat steps S3 to S5 to obtain online shear rate-viscosity datasets corresponding to multiple working points. S7. Input the online shear rate-viscosity dataset into the zero shear viscosity prediction model to obtain the predicted zero shear viscosity value η0 of the polymer melt.

[0011] In a preferred embodiment, the multi-parameter coupled flow correction model in steps S1 and S4 is as follows: Where Q is the actual volumetric flow rate, N is the rotational speed of the melt conveying unit, and ΔP p The pressure difference between the inlet and outlet sides of the melt conveying unit is denoted as T, the melt temperature is denoted as a0, the baseline correction coefficient is denoted as a1 to a6, and the correction coefficients are obtained by fitting through the flow calibration experiment.

[0012] In a preferred embodiment, the flow calibration experiment in step S1 includes: under the conditions of set temperature, set melt conveying unit speed and set inlet pressure, the polymer melt is stably discharged through the bypass test branch; the mass of melt discharged per unit time is obtained by timed collection and weighing; and the melt mass is converted into actual volumetric flow rate according to the density of the polymer melt at the corresponding temperature.

[0013] In a preferred embodiment, in step S6, the rotational speed N of the melt conveying unit is changed in a stepwise manner, with each rotational speed corresponding to an online test condition point; at each online test condition point, after the pressure signal and melt temperature signal of the slit test section stabilize, the corresponding pressure data, temperature data, and melt conveying unit rotational speed data are collected.

[0014] In a preferred embodiment, the zero-shear viscosity prediction model in step S7 is a physical information neural network model associated with the rheological constitutive relation of the polymer. The physical information neural network model takes the material type and online shear rate-viscosity dataset as input, and the zero-shear viscosity η0, relaxation time λ and flow behavior index n as intermediate output parameters. Based on the intermediate output parameters, the viscosity prediction value and the zero-shear viscosity prediction value η0 are obtained.

[0015] In a preferred embodiment, the physical information neural network model applies physical boundary constraints to the intermediate output parameters, and introduces material flowability monotonicity constraints when jointly training based on multiple polymer material samples with different melt flow indices; the physical boundary constraints include: , , where z 1,i z 2,i and z 3,i σ is the unconstrained intermediate variable output by the physical information neural network model, σ is the Sigmoid function, and i represents the i-th polymer material.

[0016] The training loss function of the physical information neural network model includes a data fitting loss term, a physical monotonicity loss term, and a regularization term. The training loss function is as follows: Among them, L data L is the data fitting loss term. mono L is the physical monotonicity loss term. reg α is the regularization term, β is the weight of the physical monotonicity constraint, and β is the regularization weight.

[0017] The physical monotonicity loss term includes a zero-shear viscosity monotonicity penalty, a relaxation time monotonicity penalty, and a flow behavior index monotonicity penalty. When multiple polymer materials in the joint training are sorted from low to high according to their melt flow index, the material flowability monotonicity constraint is: when the melt flow index of the i-th material is less than the melt flow index of the (i+1)-th material, η 0,i ≥η 0,i+1 , λ i ≥λ i+1 And n i ≤n i+1 .

[0018] Secondly, this application provides an online rheological testing system for polymer melts, employing the following technical solution: An online rheological testing system for polymer melts includes a main extrusion channel, an extruder back plate, a bypass test branch, a melt conveying unit, a slit test chamber, a sensing unit, and a measurement and control and data processing unit.

[0019] The extruder rear plate is located between the extruder screen changer and the screw head. The extruder rear plate has a discharge port and a return port that communicate with the main extrusion channel. The bypass test branch includes a discharge channel and a return channel. One end of the discharge channel is connected to the discharge port, and the other end is connected to the inlet side of the slit test chamber. One end of the return channel is connected to the outlet side of the slit test chamber, and the other end is connected to the return port.

[0020] The melt delivery unit includes a melt pump disposed on the discharge channel. The slit test chamber is located downstream of the melt pump and includes a test chamber front channel, a slit test section, and a test chamber rear channel that are sequentially connected along the melt flow direction.

[0021] The length of the flow channel before the test chamber is greater than the length of the flow channel after the test chamber. Through this asymmetric flow channel length design, the polymer melt has a more developed flow state before entering the slit test section, which helps to reduce the impact of sudden velocity changes, local backflow, and pressure fluctuations in the inlet region on the pressure measurement results of the test section.

[0022] The slit test cavity adopts a split structure consisting of an upper mold core and a lower mold core. The slit test section is located between the mating surfaces of the upper mold core and the lower mold core. The front flow channel of the test cavity, the slit test section, and the rear flow channel of the test cavity are smoothly connected by a tapered transition section, a rounded corner transition section, or a curved surface transition section. The ratio of the width W to the height H of the slit test section, W / H, is greater than or equal to 10.

[0023] The sensing unit includes a first test pressure sensor and a second test pressure sensor disposed on the slit test section, at least one temperature sensor, and a pre-pump pressure acquisition unit and a post-pump pressure acquisition unit disposed on the inlet side and outlet side of the discharge melt pump, respectively.

[0024] The sensing diaphragms of the first and second test pressure sensors are flush with the inner wall of the slit test section and are positioned to avoid the disturbance zone at the entrance of the slit test section.

[0025] Within the horizontal projection plane, the angle θ1 between the central axis of the discharge port and the extruder screw axis, when viewed along the direction of melt flow out of the extruder, satisfies 0° < θ1 ≤ 90°; the angle θ2 between the central axis of the return port and the extruder screw axis, when viewed along the direction of melt flow into the extruder, satisfies 0° < θ2 ≤ 90°.

[0026] The measurement and control and data processing unit is connected to the discharge melt pump and the sensing unit respectively, and performs pressure signal acquisition, temperature signal acquisition, melt pump speed acquisition, flow correction, online rheological parameter calculation and zero shear viscosity prediction.

[0027] The beneficial effects of this invention are as follows: First, the present invention uses a bypass slit online testing structure to draw a portion of the polymer melt out from the main extrusion channel, thereby achieving online rheological testing of the polymer melt without significantly interfering with the continuous processing of the main extrusion channel.

[0028] Secondly, by setting a test cavity front flow channel in the slit test cavity with a length greater than that of the test cavity rear flow channel, the polymer melt can obtain a more fully developed flow state before entering the slit test section, which is beneficial to reduce the impact of inlet disturbance on the pressure difference measurement of the test section and improve the stability of test data.

[0029] Third, by incorporating the rotational speed of the melt conveying unit, the pressure difference between the inlet and outlet sides of the melt conveying unit, and the melt temperature into a multi-parameter coupled flow correction model, this invention can reduce the deviation between the theoretical flow rate and the actual volumetric flow rate when the melt conveying unit is running at low speed, and improve the reliability of online rheological parameter calculation.

[0030] Fourth, this invention introduces intermediate constitutive parameter output, parameter physical boundary constraints, and material flow monotonicity constraints through a physical information neural network model, which can improve the stability of zero-shear viscosity prediction under limited online shear rate-viscosity data conditions. Attached Figure Description

[0031] Figure 1 This is a schematic diagram of the structure of the online rheological testing system for polymer melts according to the present invention.

[0032] In the diagram: 1. Extruder rear plate; 2. Discharge channel; 3. Discharge melt pump; 4. Test mold; 41. Test chamber front channel; 42. Slit test chamber; 43. Sensor; 44. Test chamber rear channel; 5. Return melt pump; 6. Return channel; 7. Data acquisition device; 8. Touch screen. Detailed Implementation

[0033] The present invention will be further described below with reference to the accompanying drawings and specific embodiments. It should be understood that the following embodiments are for illustrative purposes only and are not intended to limit the scope of protection of the present invention.

[0034] like Figure 1 As shown, this embodiment provides an online rheological testing system for polymer melts, including an extruder back plate 1, a discharge channel 2, a discharge melt pump 3, a test mold 4, a return melt pump 5, a return channel 6, a data acquisition device 7, and a touch screen 8.

[0035] One end of the discharge channel 2 is connected to the discharge port of the extruder rear plate 1, and the other end is connected to the inlet side of the test die 4. The discharge melt pump 3 is installed on the discharge channel 2 and is used to transport the polymer melt drawn from the main extrusion channel to the test die 4 at a set speed. One end of the return channel 6 is connected to the outlet side of the test die 4, and the other end is connected to the return port of the extruder rear plate 1. The return melt pump 5 is installed on the return channel 6 and is used to transport the tested polymer melt back to the extruder system.

[0036] In this embodiment, both the discharge melt pump 3 and the return melt pump 5 are melt gear pumps, and are driven by variable frequency reduction motors respectively. The data acquisition device 7 can acquire the rotational speed N of the discharge melt pump 3, and obtain the inlet side pressure P0 and outlet side pressure P3 of the discharge melt pump 3 through the pump pre-pump pressure acquisition unit and the pump post-pump pressure acquisition unit respectively, thereby calculating the pressure difference ΔP between the inlet side and the outlet side of the melt conveying unit. p =P3-P0.

[0037] The test mold 4 is a slit-type online rheological test mold, comprising a front flow channel 41, a slit test cavity 42, and a rear flow channel 44 connected sequentially along the polymer melt flow direction. The slit test cavity 42 includes a slit test section, which is preferably formed by the mating surfaces of the upper and lower mold cores to facilitate the machining, disassembly, and cleaning of the slit flow channel. The front flow channel 41, the slit test cavity 42, and the rear flow channel 44 are smoothly connected by tapered transition sections, rounded transition sections, or curved surface transition sections.

[0038] The length of the front flow channel 41 of the test chamber is greater than the length of the rear flow channel 44. This structure allows the polymer melt to have a longer flow development distance before entering the slit test section, which can reduce the impact of inlet disturbance, sudden velocity changes, and local pressure fluctuations on the pressure measurement results of the slit test section. The relatively short rear flow channel 44 of the test chamber helps to shorten the flow path after the test and reduce the residence time of the melt in the test mold.

[0039] The width W to height H ratio (W / H) of the slit test section is greater than or equal to 10. In this embodiment, the slit width W is 25 mm, the slit height H is 2 mm, and the aspect ratio is 12.5. By setting this aspect ratio, the sidewall edge effect can be reduced, making the flow state within the slit test section closer to the ideal slit flow.

[0040] The sensors 43 are disposed on the test mold 4 and arranged at intervals along the flow direction of the polymer melt. The sensors 43 include a first test pressure sensor, a second test pressure sensor, and at least one temperature sensor. The first test pressure sensor is used to obtain the upstream pressure P1 of the slit test section, and the second test pressure sensor is used to obtain the downstream pressure P2 of the slit test section, thereby obtaining the pressure difference ΔP of the slit test section. s =P1-P2; The temperature sensor is used to obtain the melt temperature T.

[0041] The sensing diaphragms of the first and second test pressure sensors are flush with the inner wall of the slit test section and are positioned to avoid the disturbance area at the entrance of the slit test section, thereby reducing local flow distortion caused by protrusions or recesses at the sensor ends. In this embodiment, the spacing between adjacent sensors 43 is 50 mm.

[0042] In this embodiment, in the horizontal projection plane, the central axis of the discharge port on the extruder rear plate 1 forms an angle θ1 with the extruder screw axis when viewed along the direction of melt flow out of the extruder, and 0°<θ1≤90°; the central axis of the return port forms an angle θ2 with the extruder screw axis when viewed along the direction of melt flow into the extruder, and 0°<θ2≤90°.

[0043] The data acquisition device 7 and the touch screen 8 together constitute the measurement, control, and data processing unit. The data acquisition device 7 includes a programmable logic controller (PLC), an analog input module, and an analog output module. The analog input module is connected to the sensor 43 and the pressure acquisition units before and after the pump. The analog output module is connected to the frequency converter driving the discharge melt pump 3. The touch screen 8 or a host computer communicates with the PLC for parameter setting, operating status display, data storage, and calculation result output.

[0044] When performing online rheological testing of polymer melt using the system of this embodiment, a flow rate calibration experiment is first conducted. This is done under different discharge melt pump speeds N and different pressure differences ΔP between the inlet and outlet sides of the discharge melt pump 3. p Under different melt temperatures T, the polymer melt was stably discharged through the bypass test branch. The mass of the melt discharged per unit time was obtained by timed collection and weighing, and the actual volumetric flow rate Q was calculated based on the density of the polymer melt at the corresponding temperature.

[0045] Using the actual volumetric flow rate Q obtained from the flow calibration experiment as the target value, a multi-parameter coupled flow correction model is fitted to obtain correction coefficients a0 to a6. The multi-parameter coupled flow correction model is as follows: Where Q is the actual volumetric flow rate, N is the rotational speed of the melt conveying unit, and ΔP p Here, N represents the pressure difference between the inlet and outlet sides of the melt conveying unit, T represents the melt temperature, a0 represents the baseline correction coefficient, and a1 to a6 represent the correction coefficients obtained through flow calibration experiments. Therefore, during online testing, the real-time collected N and ΔP values ​​can be used as a reference. p The theoretical flow rate of the discharge melt pump 3 is corrected by T to obtain the actual volumetric flow rate Q in the bypass test branch.

[0046] During online testing, a portion of the polymer melt in the main extrusion channel is introduced into the discharge channel 2 through the discharge port of the extruder rear plate 1, and then transported to the test mold 4 by the discharge melt pump 3. The polymer melt flows sequentially through the test chamber front channel 41, the slit test chamber 42, and the test chamber rear channel 44, and then returns to the extruder system via the return melt pump 5 and the return channel 6.

[0047] At the set speed N of the discharge melt pump 3, the data acquisition device 7 collects the inlet side pressure P0, outlet side pressure P3, upstream pressure P1 of the slit test section, downstream pressure P2 of the slit test section, and melt temperature T of the discharge melt pump 3. Once the pressure signal and melt temperature signal of the slit test section stabilize, the data acquisition device 7 saves the pressure data, temperature data, and melt pump speed data corresponding to that operating point.

[0048] Data acquisition device 7 based on ΔP p=P3-P0 and melt temperature T are used to correct the theoretical flow rate of the discharge melt pump 3 to obtain the actual volumetric flow rate Q; then, based on ΔP s =P1-P2, the width W and height H of the slit test section, and the axial distance L between the upstream and downstream pressure measurement positions are used to obtain the online rheological parameters corresponding to the current operating point.

[0049] In this embodiment, the online rheological parameters can be obtained using conventional conversion relationships from slit rheological testing. Specifically, the wall shear stress τ w Wall shear rate γ w The apparent viscosity η is calculated according to the following relationship: The above calculations can be used to obtain the online shear rate-viscosity data corresponding to the current operating point.

[0050] Wherein, ΔP s Let P1 be the pressure difference between the upstream pressure P1 and the downstream pressure P2 of the slit test section, W be the width of the slit test section, H be the height of the slit test section, and L be the axial distance between the upstream and downstream pressure measurement positions of the slit test section. The above calculation relationships are standard parameter conversion relationships in slit rheological testing and are only used to illustrate the process of obtaining online rheological parameters in this embodiment. They are not intended as the core formulas that distinguish this invention from existing technologies.

[0051] Subsequently, the rotational speed N of the discharge melt pump 3 was changed in a stepwise manner, and the process of pressure, temperature and rotational speed acquisition, flow correction and online rheological parameter acquisition was repeated under each rotational speed condition to obtain online shear rate-viscosity datasets corresponding to multiple operating points.

[0052] The online shear rate-viscosity dataset is input into the zero-shear viscosity prediction model. This zero-shear viscosity prediction model is a physical information neural network model associated with the rheological constitutive relationship of the polymer. It first takes the material type and the online shear rate-viscosity dataset as input, outputting three intermediate parameters: η0, λ, and n. Then, based on these intermediate parameters, it obtains the predicted viscosity values ​​at each shear rate, and finally obtains the predicted zero-shear viscosity value η0 of the polymer melt.

[0053] In this embodiment, the zero-shear viscosity prediction model can adopt the Cross-Williamson constitutive relation as a specific form of rheological constitutive relation. This constitutive relation can be expressed as: Where γ is the shear rate, η is the shear viscosity, η0 is the zero-shear viscosity, λ is the relaxation time, and n is the flow behavior exponent. It should be noted that in this embodiment, this constitutive relation is used to construct the viscosity prediction layer in the physical information neural network model so that the model output is consistent with the shear-thinning behavior of the polymer melt. It is described as an optional implementation and is not intended as a general constitutive formula protected solely by this invention.

[0054] When training the physical information neural network model, physical boundary constraints are applied to the intermediate output parameters η0, λ, and n. These physical boundary constraints may include: , , where z 1,i z 2,i and z 3,i σ is the unconstrained intermediate variable output by the physical information neural network model, σ is the Sigmoid function, and i represents the i-th polymer material.

[0055] In this embodiment, when the physical information neural network model is jointly trained based on multiple polymer material samples with different melt flow indices, its loss function includes a data fitting loss term, a material flowability monotonicity loss term, and a regularization term. The material flowability monotonicity loss term is constrained as follows: when the melt flow index of the i-th material is less than the melt flow index of the (i+1)-th material, η... 0,i ≥η 0,i+1 , λ i ≥λ i+1 And n i ≤n i+1 .

[0056] In this embodiment, the training loss function of the physical information neural network model includes a data fitting loss term, a physical monotonicity loss term, and a regularization term, and its total loss function can be expressed as: Among them, L data L is the data fitting loss term. mono L is the physical monotonicity loss term. reg Here, α is the regularization term, β is the weight of the physical monotonicity constraint, and β is the regularization weight. By using the above physical boundary constraints and material flow monotonicity constraints, the parameter coupling and instability problems when extrapolating solely from finite online shear rate-viscosity data can be mitigated.

[0057] In other embodiments, the return melt pump 5 on the return channel 6 can be set or omitted according to the pressure of the main extrusion system and the return resistance; the polymer melt after testing may also not return to the main extrusion channel, but be discharged through the discharge channel. The above modifications do not affect the basic technical concept of the present invention to achieve online rheological characterization through bypass sampling, low-speed operating condition flow correction, slit pressure difference testing, and zero-shear viscosity prediction.

[0058] This invention includes, but is not limited to, the above embodiments. Any equivalent substitutions or partial improvements made under the spirit and principles of this invention shall be considered to be within the scope of protection of this invention.

Claims

1. A method for online rheological testing of polymer melts, characterized in that: The method is implemented based on a bypass-type slit rheological online testing system, which includes a bypass testing branch connected to the main extrusion channel, a melt delivery unit disposed on the bypass testing branch, a slit testing chamber disposed downstream of the melt delivery unit, a pressure and temperature acquisition unit, and a measurement, control, and data processing unit; the method includes the following steps: S1. A multi-parameter coupled flow correction model for the melt conveying unit under low-speed operation is established in advance. Under different melt conveying unit rotation speeds, different pressure differences between the inlet and outlet sides of the melt conveying unit, and different melt temperatures, actual volumetric flow rate data is obtained through flow calibration experiments. The model parameters of the multi-parameter coupled flow correction model are obtained by fitting the actual volumetric flow rate data as the target value. S2. During online testing, a portion of the polymer melt in the main extrusion channel is introduced into the bypass test branch, and the polymer melt is transported to the slit test chamber through the melt conveying unit. S3. At the set rotational speed N of the melt conveying unit, collect the inlet side pressure P0, outlet side pressure P3, upstream pressure P1 of the slit test section, downstream pressure P2 of the slit test section, and melt temperature T of the melt conveying unit. S4. Based on the rotational speed N of the melt conveying unit and the pressure difference ΔP between the inlet and outlet sides of the melt conveying unit. p =P3-P0 and melt temperature T, the theoretical flow rate of the melt conveying unit is corrected by the multi-parameter coupled flow correction model established in step S1 to obtain the actual volumetric flow rate Q in the bypass test branch. S5. Based on the actual volumetric flow rate Q and the pressure difference ΔP in the slit test section. s =P1-P2 and the geometric parameters of the slit test chamber, calculate the shear rate, shear stress and apparent viscosity of the polymer melt under the current working conditions; S6. Change the rotational speed N of the melt conveying unit and repeat steps S3 to S5 to obtain online shear rate-viscosity datasets corresponding to multiple working points. S7. Input the online shear rate-viscosity dataset into the zero shear viscosity prediction model to obtain the predicted zero shear viscosity value η0 of the polymer melt.

2. The method for online rheological testing of polymer melts according to claim 1, characterized in that: The multi-parameter coupled flow correction model in steps S1 and S4 is as follows: Where Q is the actual volumetric flow rate, N is the rotational speed of the melt conveying unit, and ΔP p The pressure difference between the inlet and outlet sides of the melt conveying unit is denoted as T, the melt temperature is denoted as a0, the baseline correction coefficient is denoted as a1 to a6, and the correction coefficients are obtained by fitting through the flow calibration experiment.

3. The method for online rheological testing of polymer melts according to claim 1, characterized in that: The flow calibration experiment in step S1 includes: under the conditions of set temperature, set melt conveying unit speed and set inlet pressure, the polymer melt is stably discharged through the bypass test branch; the mass of melt discharged per unit time is obtained by timed collection and weighing; and the melt mass is converted into actual volumetric flow rate according to the density of polymer melt at the corresponding temperature.

4. The method for online rheological testing of polymer melts according to claim 1, characterized in that: In step S6, the rotational speed N of the melt conveying unit is changed in a stepwise manner, with each rotational speed corresponding to an online test condition point. At each online test condition point, after the pressure signal and melt temperature signal of the slit test section stabilize, the corresponding pressure data, temperature data, and melt conveying unit rotational speed data are collected.

5. The method for online rheological testing of polymer melts according to claim 1, characterized in that: The zero-shear viscosity prediction model in step S7 is a physical information neural network model associated with the rheological constitutive relation of the polymer. The physical information neural network model takes the material type and online shear rate-viscosity dataset as input, and outputs the zero-shear viscosity η0, relaxation time λ and flow behavior index n as constitutive parameters. Based on the constitutive parameters, it calculates the viscosity prediction value at the corresponding shear rate, where the zero-shear viscosity η0 is the zero-shear viscosity prediction result of the polymer melt.

6. The method for online rheological testing of polymer melts according to claim 5, characterized in that: The physical information neural network model applies physical boundary constraints to the intermediate output parameters, and introduces material flowability monotonicity constraints when jointly training based on multiple polymer material samples with different melt flow indices; the physical boundary constraints include: , , where z 1,i z 2,i and z 3,i Let σ be the unconstrained intermediate variable output by the physical information neural network model, σ be the Sigmoid function, and i represent the i-th polymer material. The training loss function of the physical information neural network model includes a data fitting loss term, a physical monotonicity loss term, and a regularization term. The training loss function is as follows: , where L data L is the data fitting loss term. mono L is the physical monotonicity loss term. reg η is the regularization term, α is the physical monotonicity constraint weight, and β is the regularization weight. The physical monotonicity loss term includes a zero-shear viscosity monotonicity penalty term, a relaxation time monotonicity penalty term, and a flow behavior index monotonicity penalty term. When multiple polymer materials in the joint training are sorted from low to high according to their melt flow index, the material flowability monotonicity constraint is: when the melt flow index of the i-th material is less than the melt flow index of the (i+1)-th material, η 0,i ≥η 0,i+1 , λ i ≥λ i+1 And n i ≤n i+1 .

7. A polymer melt online rheological testing system implementing the method of any one of claims 1 to 6, characterized in that, include: Main extrusion channel, extruder back plate, bypass test branch, melt conveying unit, slit test chamber, sensing unit, and measurement and control and data processing unit; The extruder rear plate is located between the extruder screen changer and the screw head. The extruder rear plate has a discharge port and a return port communicating with the main extrusion channel. The bypass test branch includes a discharge channel and a return channel. One end of the discharge channel is connected to the discharge port, and the other end is connected to the inlet side of the slit test chamber. One end of the return channel is connected to the outlet side of the slit test chamber, and the other end is connected to the return port. The melt conveying unit includes a discharge melt pump located on the discharge channel. The slit test chamber is located downstream of the discharge melt pump and includes components along the melt flow path. The test chamber consists of a front flow channel, a slit test section, and a rear flow channel connected sequentially in the direction of movement. The sensing unit includes a first test pressure sensor and a second test pressure sensor, at least one temperature sensor, and a pre-pump pressure acquisition unit and a post-pump pressure acquisition unit respectively located on the inlet and outlet sides of the discharge melt pump. The measurement, control, and data processing unit is connected to the discharge melt pump and the sensing unit, and performs pressure signal acquisition, temperature signal acquisition, melt pump speed acquisition, flow correction, online rheological parameter calculation, and zero shear viscosity prediction.

8. The online rheological testing system for polymer melts according to claim 7, characterized in that: Within the horizontal projection plane, the angle θ1 between the central axis of the discharge port and the extruder screw axis, when viewed along the direction of melt flow out of the extruder, satisfies 0° < θ1 ≤ 90°; the angle θ2 between the central axis of the return port and the extruder screw axis, when viewed along the direction of melt flow into the extruder, satisfies 0° < θ2 ≤ 90°.

9. The online rheological testing system for polymer melts according to claim 7, characterized in that: The slit test chamber adopts a split structure consisting of an upper mold core and a lower mold core. The slit test section is located between the mating surfaces of the upper mold core and the lower mold core. The length of the front flow channel of the test chamber is greater than the length of the rear flow channel of the test chamber. The front flow channel, the slit test section, and the rear flow channel of the test chamber are smoothly connected by a tapered transition section, a rounded corner transition section, or a curved surface transition section. The ratio of the width W to the height H of the slit test section, W / H, is greater than or equal to 10. The sensing diaphragms of the first test pressure sensor and the second test pressure sensor are flush with the inner wall surface of the slit test section and are arranged in a position that avoids the disturbance area at the entrance of the slit test section.

Images