High-fidelity pressure control of extrusion manufacturing processes using primary and auxilary pressure sensors

Abstract

A system for achieving high-definition pressure control of manufacturing operations utilizes data from one or more primary sensors, each of which is associated with a relatively high degree of accuracy based on the primary sensor's proximity to material involved in the manufacturing operation, as well as data from one or more secondary sensors, each of which is located more remote from the material involved in the manufacturing operation than the primary sensors. Data from the one or more primary sensors is evaluated to detect errors or trends indicative of primary sensor failure. Suspect or compromised primary sensor data is then disregarded or filtered, using data from the secondary sensors, historical data, or some combination thereof.

Description

HIGH-FIDELITY PRESSURE CONTROL OF EXTRUSION MANUFACTURING PROCESSES USING PRIMARY AND AUXILARY PRESSURE SENSORSFIELD OF THE DISCLOSURE

[0001] This disclosure relates generally to pressure control of manufacturing processes and, more particularly, to increasing the fidelity of pressure control by adjustment of manufacturing parameters based on (i) real-time, near-real time, historical, and / or differential pressure data provided by one or more primary sensors susceptible to reliability challenges such as electromagnetic interference, extreme temperatures, and / or abrasive media involved in the manufacturing operation, (ii) data indicative of pressure provided by one or more auxiliary sensors which are more remote from, or otherwise resistant to, such reliability challenges, and (iii) selective consideration of such real-time and offset pressure data from either the one or more primary sensors or the one or more auxiliary sensors, or some combination of them, to minimize effects of outlier readings or erroneous data.BACKGROUND

[0002] A variety of manufacturing processes utilize pressure control to manage the introduction of one or more media to a cavity or region of a manufacturing system. For instance, injection molding systems and extrusion molding systems typically employ hydraulic pressure sensors measuring the hydraulic pressure exerted on the back of an injection piston, or load cells measuring the amount of force exerted on a load cell on the back of a screw, either of which are disposed well upstream of a gate at an inlet of one or more mold cavities or dies. Such hydraulic pressure sensors or load cells serve as a proxy for pressure of molten polymeric material in the mold cavity, but are widely used and reasonably reliable. Because hydraulic pressure sensors or load cells (referred to generally herein as “machine sensors”) are isolated from the molten polymeric material entering the mold cavity or die, they can provide a control signal that is of high quality with respect to stability and consistency. Machine sensors are also generally not exposed to extremely high temperatures, abrasive materials (such as glass microbeads or fibrous resins increasingly prevalent in media used in injection molding, extrusion, and other manufacturing operations), and because they are situated in relatively low “traffic areas” of injection molding systems, are less susceptible to electromagnetic interference (EMI) than melt pressure transducer (MPT) sensors and / or die pressure sensors mounted downstream of a screw tip, or proxy sensors, such as ultrasonic sensors or strain gauge sensors, mounted to generate and transmit data indicative of pressure downstream of a screw tip. However, such machine sensors, while more remote from challenges to sensor reliabilityencountered downstream of a screw tip, are inherently susceptible to error because of their lack of proximity to the molten polymer.

[0003] A potential error source for machine sensors is if an operator were to adjust the cushion in front of a shot of molten polymeric material, as this can affect the offset used in calibrating a controller to data received from the machine sensors to produce pressure values expected from a MPT or die pressure sensor. Additionally, changes in back pressure during the recovery phase of a screw in a barrel of the injection molding system can have unexpected impacts on offset. Other sources of error for machine sensors include friction, compliance of mechanical components, and fluid compression in the media used in the manufacturing operation.

[0004] Errors from machine sensors are dynamic in the course of a shot of an injection molding system.

[0005] As an alternative to machine sensors, MPT sensors and / or die pressure sensors may be used as sensors in intimate contact with molten polymeric material downstream of a screw, such as in, or immediately upstream of, a gate or nozzle as the molten polymeric material enters a mold cavity, feed system to a plurality of mold cavities, and / or extrusion die. Alternatives to MPT sensors, such as ultrasonic sensors or strain gauge sensors, may be used to generate and transmit data indicative of pressure at a location downstream of the screw. With MPT and die pressure sensors mounted downstream of the screw, due to their immediate proximity to molten polymeric material, and proxy sensors, such as ultrasonic or strain gauge sensors targeting regions downstream of the screw, some of the potential error sources associated with machine sensors are avoided. However, MPT sensors, die pressure sensors, and proxy sensors have a tendency to drift out of calibration. They are mounted in or near what might be characterized as a relatively high traffic area of the press or molding system. Because they are in fluid communication with molten polymeric material, MPT sensors and / or die pressure sensors are inclined to leak, delaminate, and / or otherwise experience a failure of electronic components, particularly when exposed to abrasive media, such as polymer having glass beads or fibrous material. Decisions as to what sensors to employ also suffer from competing considerations, such as minimizing size while providing sensors that are sufficiently mechanically robust to withstand the harsh environment to which the sensors are exposed, such as temperatures at or above 700 °F.

[0006] MPT sensors and die pressure sensors are also delicate sensors, that may need to be used in concert with amplifiers and sensitive electronics that are susceptible to interference andsignal distortion. Nonetheless, because hydraulic pressures detected by hydraulic sensors are only exposed to about 10% of the pressures to which actual MPT and / or die pressure sensors are exposed, hydraulic pressures often need significant amplification to arrive at what might be considered a proxy for true plastic pressure. MPT sensor and / or die pressure sensors, when functioning properly, can provide high quality representations of true plastic pressure at any given instant, and therefore provide a desirable data source to control an injection molding process.

[0007] It would be desirable to provide a manufacturing process that could benefit from the accuracy of pressure sensor data from sensors in immediate proximity to molten polymeric material at the entrance of a mold cavity, feed system to multiple mold cavities, extrusion die, or otherwise downstream of a screw, while also having the reliability of auxiliary sensors, such as machine sensors, positioned more remote from the flow of molten material, and that have high quality with respect to stability and consistency.SUMMARY OF THE DISCLOSURE

[0008] Various manufacturing systems can be controlled with improved precision and efficiency by controlling manufacturing operations in real-time or near-real time with reliable, high precision sensors, referred to herein as primary sensors, in combination with generally reliable secondary sensors that are positioned and selected to be resilient and less susceptible to error. Near-real time applies to calculations or control based on a time difference of up to 1000 milliseconds.

[0009] In a molding system, for instance, primary sensors, including melt pressure transducer (MPT) sensors and / or die pressure sensors, may be mounted in a nozzle adapter or otherwise in front of (i.e., downstream of) a screw tip. Proxy sensors, such as ultrasonic sensors or strain sensors, may be used in lieu of MPT sensors. Secondary sensors may include hydraulic sensors or load cells, generally referred to herein as machine sensors, disposed upstream of the screw, remote from high temperatures and otherwise generally isolated from conditions that contribute to a relatively harsh environment for primary sensors. It should be appreciated that, in some embodiments, the machine sensors may be the primary sensors and the MPT, die pressure, or proxy sensors are the secondary sensors. These embodiments are generally referred to herein as a “reverse offset” system.

[0010] While secondary sensors are generally more robust than primary sensors from a physical durability standpoint, primary sensors can provide greater accuracy due to their greaterproximity (or target being closer) to molten polymeric material entering the mold cavity, feed system, or extrusion die. It is found that systems can be developed that utilize readings from primary and secondary sensors to test the quality of signals from primary sensors in real time, identify and selectively monitor, filter, ignore, or attenuate errant primary sensor signals. Imminent or approaching failure of primary sensors can also be detected. The difference between a secondary sensor signal and a fully functioning primary sensor signal is defined herein as the “high-definition offset vector” or the “offset vector.”

[0011] Upon detection of errant primary sensor signals, manufacturing operations can failover to apply historical offsets to the secondary sensor signals on either a temporary or continuing basis. For instance, upon detection of an errant primary sensor signal, rather than employing data from the errant primary sensor signal in a control algorithm for adjusting screw velocity to maintain or modify melt pressure, pressure data from an immediately-previous shot at a time corresponding to the time at which the errant primary sensor signal was detected could be substituted for the errant primary sensor signal. Alternately, pressure data from another historically-recorded shot could be substituted for the errant primary sensor signal.

[0012] One manner of detecting errant primary sensor signals involves analyzing signal output behavior from the primary sensor, such as by analyzing whether the signal behavior is as expected, for example by scanning for abnormal jumps, sawtooth behavior, or other signatures indicative of a spurious signal. With this first level of quality check, no secondary signal would be relied upon for comparison. According to another manner of detecting errant primary sensor signals, comparisons of readings from primary and secondary sensors, identifying a range of acceptable variances or tolerances in discrepancies, and, in real-time or near-real-time, determining whether the variances are within or outside of such predetermined limits. Even if such comparisons reveal variances that are within an acceptable range, a system can be programmed to identify one or more changes in variances over a plurality of shots, or over a particular duration of time, indicative of a signal drift or other early indication of a potentially failing primary sensor.

[0013] With sufficient collection of historical data, a manufacturing system can continue to operate for an extended period of time even with a faulty primary sensor, by substituting the historical data for the erroneous readings from the faulty primary sensor. By monitoring ambient conditions, fluctuations in media used in the manufacturing operation, and other variables that might impact the extent to which historical pressure data is likely to differ from actual data that could be collected from a fully functioning primary sensor, a system of the present disclosurecan be programmed to correct for expected deviation from historical pressure data. When operating a manufacturing system in such a manner, the primary sensor is essentially a phantom or virtual primary sensor, thereby permitting uninterrupted operation even in the absence of a fully operational primary sensor.BRIEF DESCRIPTION OF THE SEVERAL VIEWS OF THE DRAWINGS

[0014] While the specification concludes with claims particularly pointing out and distinctly claiming the subject matter that is regarded as the present invention, it is believed that the invention will be more fully understood from the following description taken in conjunction with the accompanying drawings. Some of the figures may have been simplified by the omission of selected elements for the purpose of more clearly showing other elements. Such omissions of elements in some figures are not necessarily indicative of the presence or absence of particular elements in any of the exemplary embodiments, except as may be explicitly delineated in the corresponding written description. None of the drawings are necessarily to scale. For example, the dimensions and / or relative positioning of some of the elements in the figures may be exaggerated relative to other elements to help to improve understanding of various embodiments of the present invention.BRIEF DESCRIPTION OF THE SEVERAL VIEWS OF THE DRAWING

[0015] FIG. 1 is a schematic view of an extrusion molding system provided with primary and secondary sensors;

[0016] FIG. 2 is an example graph depicting the offset between melt pressure data produced by a MPT sensor and melt pressure data produced by a machine sensor;

[0017] FIG. 3 is a flow chart illustrating examples of the manner in which a system of the present disclosure may operate to detect errors in a primary sensor and steps the system may take to compensate for such errors;

[0018] FIGS. 4A-4C are example graphs depicting failure signatures of the error detection analyses disclosed herein; and

[0019] FIG. 5 is an example method for manufacturing molded parts using a high-definition pressure signal.DETAILED DESCRIPTION

[0020] Generally speaking, aspects of the present disclosure include systems and approaches for controlling a molding machine (e.g., an extrusion molding machine) where anumber of sensors are positioned inline and upstream from the die orifice. The sensors are arranged to calculate a relative change in pressure of the molten polymer during the extrusion process, and to compare any relative changes with a reference value to determine whether corrective adjustments should be made to produce stable and consistent products. This information may be used to make real-time adjustments of melt pressure setpoints and / or driving forces on the molten polymer.

[0021] In some examples (and as will be described herein), the extrusion molding machine may incorporate a single sensor, two sensors, or more than two sensors used to increase the fidelity of pressure control by adjustment of manufacturing parameters in real-time or near realtime.

[0022] Turning to the drawings, an extrusion molding process is herein described. As illustrated in FIG. 1 , the extrusion molding machine 100 includes an extrusion unit 102. The extrusion unit 102 includes a hopper 106 adapted to accept material in the form of pellets 108 or any other suitable form. In many of these examples, the pellets 108 may be a polymer or polymer-based material such as, for example, post-consumer regrind (PGR). Other examples are possible. The extrusion unit 102 further includes a heated barrel 1 10, a screw 112, and a die 116.

[0023] The hopper 106 feeds the pellets 108 into a heated barrel 110 of the extrusion unit 102. Upon being fed into the heated barrel 1 10, the pellets 108 may be driven to the end of the heated barrel 110 towards a barrel end cap 1 10a by axial motion of the screw 1 12 while the screw 112 remains stationary with respect to the heated barrel 110. The heating of the heated barrel 110 and the compression of the pellets 108 by the rotating screw 1 12 causes the pellets 108 to melt, thereby forming a molten plastic material or polymer 114. The molten plastic material 114 is typically processed at a temperature selected within a range of about 130 °C to about 410^0 (with manufacturers of particular polymers typically providing extruders with recommended temperature ranges for given materials).

[0024] The axial motion of the screw 1 12 forces the molten plastic material 1 14 toward a die 116 to form an extrudate plastic material that will ultimately be the molded part.

[0025] An actuation unit 122 is provided that is operably coupled with the screw 112 to facilitate powered rotation thereof. In some examples, the actuation unit 122 may be in the form of a hydraulic motor. In other examples, the actuation unit 122 may be in the form of an electric motor. In these and other examples, the actuation unit 122 may additionally or alternativelyinclude a valve, a flow controller, an amplifier, or any of a variety of other suitable control devices for extrusion molding apparatuses or non-extrusion molding apparatuses.

[0026] The die 1 16 may be associated with post mold handling equipment 125. A final product 134 results after the molten plastic material 114 is extruded to form extrudate 131 and undergoes any desired operation by post mold handling equipment 125. The extrudate 131 may take an intermediate form 132 during the post mold handling equipment 125 operation. Molten thermoplastic material 114 cools significantly as it flows through the die 116, however, in some embodiments, the post mold handling equipment 125 may supply additional cooling. In some embodiments, the post mold handling equipment 125 may include a plurality of options to generate various molded parts (e.g., tube, rod, film, blown film, bags, pellets, bottles, etc.). The post mold handling equipment 125 may also include electronic components capable of generating an output signal 136 that communicates via signal line 152a condition and / or characteristic of the extrudate 131 such as, for example, thickness, temperature, line speed, etc. to control systems (e.g., a remote controller 146 and / or a native controller 140).

[0027] As previously noted, the extrusion molding machine 100 also includes a native controller 140 communicatively coupled with the extrusion molding machine 100 via connection 145. The connection 145 may be any type of wired and / or wireless communications protocol adapted to transmit and / or receive electronic signals. In these examples, the native controller 140 is in signal communication with a screw control 126 via connection 145. The native controller 140 may command the screw control 126 to control the axial motion of the screw 112 at a rate that maintains a desired molding process such that variations in material viscosity, die temperatures, melt temperatures, and other variations that may influence extrusion rate are taken into account by the native controller 140. Such adjustments may be made in real time by the native controller 140. In one embodiment, when the actuation unit 122 is a hydraulic motor, the screw control 126 can include a hydraulic valve associated with the screw 1 12. In another embodiment, when the actuation unit 122 is an electric motor, the screw control 126 can include an electric controller associated with the screw 112. In the embodiment of FIG. 1 , the native controller 140 can, for example, generate a signal that is transmitted from an output of the native controller 140 to the screw control 126 to control an axial screw position and / or velocity of the screw 112.

[0028] The native controller 140 can be an on-board controller that is original to the extrusion molding unit 102 and built along with the extrusion molding unit 102. The native controller 140 can be disposed in a number of positions with respect to the extrusion molding machine 100.As examples, the controller 140 can be integral with the extrusion molding machine 100, contained in an enclosure that is mounted on the machine, contained in a separate enclosure that is positioned adjacent or proximate to the machine, or can be positioned remote from the machine. In some embodiments, the native controller 140 can partially or fully control functions of the machine via wired and / or wired signal communications as known and / or commonly used in the art. As such, in some examples, modifications to the control architecture of the native controller 140 or removal of the native controller 140 may be time consuming, expensive and in some instances impossible.

[0029] The native controller 140 can be any of a variety of suitable controllers for controlling the molding process. In some arrangements, the native controller 140 may include a PID controller natively configured to implement a PID control algorithm. In addition to the native controller 140 controlling the axial motion of the screw 122 of the actuation unit 122 using a melt pressure sensor 129, the control variable may, for example, be a temperature or a pressure associated with the molten thermoplastic material 114 at a particular location in the extrusion molding unit 102. A controlled molten thermoplastic pressure may correspond, for example, to (1 ) an extrusion pressure detected via an extrusion pressure sensor 128 located at or near the actuation unit 122, (2) a melt pressure detected via a melt pressure sensor 129 located at or near the die 1 16, or (3) a pressure detected via a die pressure sensor 130 located proximate to an end of the die 116. In some embodiments, the melt pressure sensor 129 and / or the die pressure sensor 130 serve as a “primary sensor” and the extrusion pressure sensor 128 serves as a “secondary sensor.” The native controller 140 may generally be configured to provide a control signal to control an operation of the extrusion molding unit 102 (e.g., a signal to the screw control 126 to control the axial motion of the screw 112 based upon the sensed control variable being provided as an input to the control algorithm of the native controller 140 (e.g., based upon comparison of the sensed control variable and a setpoint defined by the native controller 140).

[0030] The extrusion pressure sensor 128 can facilitate detection (direct or indirect) of the extrusion pressure inside of the heated barrel 110 (i.e., the pressure of the heated barrel 110 at the beginning of the screw 112) by providing a feedback signal via a signal line 144 to the native controller 140. The native controller 140, in some embodiments, can detect the extrusion pressure from the feedback signal and can control (e.g., feedback control) the pressures within the extrusion molding machine 100 by controlling the screw control 126, which controls the rates of extrusion by the extrusion molding unit 102.

[0031] The melt pressure sensor 129 can facilitate detection (direct or indirect) of the actual melt pressure (e.g., the measured melt pressure) of the molten thermoplastic material 114 at or near the die 116. The melt pressure sensor 129 may or may not be in direct contact with the molten thermoplastic material 114. In some embodiments, the melt pressure sensor 129 can be a pressure transducer that transmits an electrical signal via a signal line 141 to an input of the native controller 140 in response to the melt pressure at the die 116. In some embodiments, the melt pressure sensor 129 can facilitate monitoring of any of a variety of additional or alternative characteristics of the molten thermoplastic material 114 at the die 116 that might indicate melt pressure, such as temperature, viscosity, and / or flow rate, for example. If the melt pressure sensor 129 is not located within the die 116, the native controller 140 can be set, configured, and / or programmed with logic, commands, and / or executable program instructions to provide appropriate correction factors to estimate or calculate values for the measured characteristic in, at, or near the die 1 16. It is to be appreciated that sensors other than a melt pressure sensor can be employed to measure any other characteristics of the molten thermoplastic material 114, the screw 112, the barrel 1 10, or the like that is known in the art, such as, temperature, viscosity, flow rate, strain, velocity, etc. or one or more of any other characteristics that are indicative of any of these.

[0032] The die pressure sensor 130 may facilitate detection (direct or indirect) of the melt pressure of the molten thermoplastic material 114 in, at, or near the die 116. The die pressure sensor 130 may or may not be in direct contact with the molten thermoplastic material 1 14. In some examples, the die pressure sensor 130 can be a pressure transducer that transmits an electrical signal via a signal line 151 to an input of the native controller 140 or to an input of the remote controller 146 in response to the die pressure within the die 116. In other embodiments, the die pressure sensor 130 can facilitate monitoring of any of a variety of additional or alternative characteristics of the thermoplastic material 114.

[0033] Still referring to FIG. 1 , a remote controller 146 (e.g., a controller that includes a PID controller) can be in signal communication with the native controller 140, the extrusion pressure sensor 128, the melt pressure sensor 129, and / or the die pressure sensor 130. Prior to retrofitting the remote controller 146 to the extrusion molding machine 100, the native controller 140 can be in signal communication with any one or more of the extrusion pressure sensors 128, melt pressure sensor 129, and die pressure sensor 130 in the manner described above. To retrofit (e.g., associate) the remote controller 146 to the extrusion molding machine 100, output from one or more of the extrusion pressure sensor 128, melt pressure sensor 129, and / orthe die pressure sensor 130 may be connected to the remote controller 146 (and optionally, disconnected from the native controller 140), thereby diverting sensor output from the sensor(s) to the remote controller 146 in lieu of the native controller 140. Once the retrofit is complete, the native controller 140 may no longer directly receive feedback signals from the one or more sensors disconnected from the native controller 140. Instead, the remote controller 146 receives these feedback signals, and generates and transmits a modified feedback signal to the native controller 140 that enhances the operation of the native controller 140 by affecting the manner in which the native controller 140 controls the operation of the extrusion molding machine 100. The native controller 140 and the remote controller 146 thus operate in a closed- loop type control arrangement that mimics the arrangement that existed prior to addition of the remote controller 146.

[0034] In some embodiments as described above, the extrusion pressure sensor 128, melt pressure sensor 129, and die pressure sensor 130 already exist on the extrusion molding unit 102 prior to the retrofitting and are in signal communication with the native controller 140. In such embodiments, retrofitting of the remote controller 146 to the extrusion molding apparatus 100 involves disconnecting the sensors from the native controller 140 and reconnecting the sensors to the remote controller 146. Alternatively, in some arrangements, one or more of the extrusion pressure sensor 128, the melt pressure sensor 129, and the die pressure sensor 130 may not already exist in the extrusion molding unit 102 prior to the retrofitting. In these examples, retrofitting of the remote controller 146 to the extrusion molding machine 100 may include installing one or more sensors to the extrusion molding unit 102 and connecting the installed sensor(s) to the remote controller 146.

[0035] In some embodiments, retrofitting the remote controller 146 to the extrusion molding machine 100 may include diverting (or installing and connecting) still other sensor output to the remote controller 146 instead of the native controller 140 in a manner analogous to that described above, wherein the still other sensor(s) are configured to measure a control variable to which a control strategy of the native controller 140 and / or remote controller 146 pertains (e.g., a temperature sensor, flow sensor, etc.).

[0036] Further still, the process of retrofitting the remote controller 146 to the extrusion molding apparatus 100 may include connecting an output of the setpoint of the native controller 140 to the remote controller 146. That is, whereas the native controller 140 prior to the retrofitting may provide the native controller setpoint to the control algorithm of the native controller 140, the native controller 140 after the retrofitting may additionally provide its setpointvia a signal communication to the remote controller 146. It should be understood that, in light of the fact that the setpoint defined by the native controller 140 may change over the course of the extrusion molding process (e.g., as a product of ramping or stepping-up of the control variable setpoint during control iterations), the providing of the native controller setpoint to the remote controller 146 may include providing the setpoint each time the native controller 140 defines the control variable setpoint (e.g., at each iteration of the control loop). Thus, the remote controller 146 maintains awareness of the current value of the native controller setpoint, so as to correctly provide a modified feedback signal to be provided to the native controller 140, as described herein.

[0037] The native controller 140 includes software (not illustrated) adapted to control its operation, any number of hardware elements (not illustrated; such as, for example, a non- transitory memory module and / or processors), any number of inputs, any number of outputs, and any number of connections. The software may be loaded directly onto a non-transitory memory module of the native controller 140 in the form of a non-transitory computer readable medium, or may alternatively be located in the remote controller 146 and be in communication with the native controller 140 via any number of controlling approaches. The software includes logic, commands, and / or executable program instructions which may contain logic and / or commands for controlling the extrusion molding machine 100 according to a set of setpoint parameters. The software may or may not include an operating system, an operating environment, an application environment, and / or a user interface.

[0038] The hardware uses the inputs to receive signals, data, and information from the extrusion molding machine 100 being controlled by the native controller 140. The hardware uses the outputs to send signals, data, and / or other information to the extrusion molding machine 100. In various embodiments the pathway between the native controller 140 and the extrusion molding machine 100 may be a physical connection or a non-physical communication link that works analogous to a physical connection, direct or indirect, configured in any way described herein or known in the art. In various embodiments, the native controller 140 can be configured in any additional or alternate way known in the art.

[0039] As previously indicated, the primary sensors 129, 130 are positioned downstream from the secondary sensor 128. Due to the inherent material properties of the molten plastic material 114, the material will exhibit lower compressibility at a location at or near the end of the screw 112 closest to the screw control 126 (i.e., where the secondary sensor 128 is generally positioned) as compared to a location at or near the die 1 16 (i.e., where the primary sensors129, 130 are generally positioned). Therefore, the pressure readings from the primary sensors 129,130 are more accurate than those obtained from the secondary sensor 128. However, due to the closer proximity to the melt stream, operators and handling equipment, the primary sensors 129,130 are not as durable or resistant to damage as the secondary sensor 128.

[0040] It is appreciated that in some examples, the extrusion molding machine 100 may additionally sense any changes to the viscosity of the molten polymer material 114 and adjust operational parameters in response to these changes in addition to sensing and adjusting parameters in response to changes in the density of the molten polymer material 1 14. As a non-limiting example, the extrusion molding machine 100 may measure a shear rate (e.g., a speed that the molten polymer material 1 14 is moving in relation to the amount of applied pressure) of the molten polymer material 114. Such measurements may occur via the use of a) measurement of axial screw position and / or velocity coupled with a sensor positioned at or near the die. Other suitable approaches are possible. This measured shear rate may be compared with a reference shear rate, and any differences between these values may be used to adjust an operational or control parameter to cause a subsequently measured shear rate to be equal to (or within an allowable threshold value of) the reference shear rate.

[0041] As described above, because melt pressure transducers (MPT) sensors generally provide more accurate measurement of melt pressure than a machine sensor, calculating the closed-loop error based on the output of a MPT sensor 129 is more likely to control the extrusion molding machine 100 in manner that produces an acceptable molded part (e.g., a molded part within the design tolerances). Accordingly, the MPT sensor 129 may serve as a primary sensor and be in the form of a MPT sensor. Alternatively, the primary sensor may instead take the form of an ultrasonic sensor or a proxy sensor, such as a strain gauge sensor, to indirectly measure melt pressure. Unless specifically described otherwise, any reference to the MPT sensor 129 envisions the alternate use of a proxy sensor.

[0042] It should be appreciated that while both the extrusion pressure sensor 128 and the MPT sensor 129 and / or the die pressure sensor 130 may be configured to provide sensor data indicating a melt pressure, due to their different locations and / or the differences in how their pressure data is generated, the melt pressure output by the extrusion pressure sensor 128 and the melt pressure output by the MPT sensor 129 and / or the die pressure sensor 130 are offset from one another. If the native controller 140 and / or remote controller 146 is tuned to control the extrusion molding machine 100 based on sensor data reported by the MPT sensor 129 and / or the die pressure sensor 130, the setpoint pattern does not indicate the appropriatevalues for controlling the extrusion molding machine 100 based on the melt pressure data output by the extrusion pressure sensor 128, and vice versa. Accordingly, to maintain the intended operation of the extrusion molding machine 100, the native controller 140 and / or remote controller 146 may be configured to quantify the offset (i.e., a “delta”) between the melt pressure data output by the extrusion pressure sensor 128 and the melt pressure data output by the MPT sensor 129 and / or the die pressure sensor 130. Thus, when switching between melt pressure data sources, the native controller 140 and / or remote controller 146 may apply the offset to the new melt pressure data source prior to inputting the melt pressure to the PID controller. Said another way, by applying the offset to the melt pressure output of the extrusion pressure sensor 128, the input to the PID controller approximates the melt pressure value the would be output by the MPT sensor 129 and / or the die pressure sensor 130 when properly functioning.

[0043] It should be appreciated that by quantifying the offset between the extrusion pressure sensor 128 and the MPT sensor 129 and / or the die pressure sensor 130, the native controller 140 and / or remote controller 146 is able to switch between melt pressure data sources with minimal delay (e.g., less than 10 milliseconds (ms), less than 5ms, less than 3ms, etc.). As such, the native controller 140 and / or remote controller 146 may be able to switch data sources and still produce an acceptable molded part without interrupting the extrusion process. Additionally, being able to readily switch data sources enables operators of the extrusion molding machine 100 to continue the extrusion process without stopping to repair the MPT sensor 129 and / or the die pressure sensor 130 and / or reconfigure the native controller 140 and / or remote controller 146 with a new set of setpoint values. This enables the operator to reduce extrusion molding machine downtime by scheduling repair of the MPT sensor 129 and / or the die pressure sensor 130 during a pre-planned stoppage period.

[0044] For simplicity of explanation, the following generally refers to the control functionality being implemented in the native controller 140. It should be appreciated that in other embodiments, any of the functionality described with respect to the native controller 140 may additionally or alternatively be implemented by the remote controller 146. In these embodiments, the functionality may be performed by the remote controller 146 by instructing the native controller 140 to carryout one or more operations that implement the described functionality.

[0045] With simultaneous reference to FIG. 2, illustrated is an example graph 200 depicting the offset (as represented by line 215) between melt pressure data produced by the MPTsensor 129 or die pressure sensor 130 (as represented by the line 205) and melt pressure data produced by the extrusion pressure sensor 128 (as represented by the line 210). As illustrated, the offset signal 215 that represents the offset vector remains relatively constant. However, the extrusion process exhibits high frequency changes (measured in milliseconds) for the axial motion of the screw.

[0046] Accordingly, to quantify the offset between the primary pressure signal 205 and the extrusion pressure signal 210, the native controller 140 may be configured to store an offset vector 215 that stores offset values for many different segments of the extrusion process. For example, the segments may be defined temporally (e.g., 1 second, 5 seconds, 15 seconds, 30 seconds) and / or based on a non-time dependent variable. Within the segment, the offset vector 215 may include samples for every 1-5ms of the extrusion process. Thus, in the event of a failure of the primary sensors, the offset applied to the extrusion pressure signal 210 prior to inputting the pressure data to the PID controller accounts for the shifts in the offset throughout the extrusion process. As a result, the secondary pressure signal produced by adding the offset to the extrusion pressure signal 210 more closely matches the expected the primary pressure data throughout the extrusion cycle.

[0047] It should be appreciated that due to changes in operating conditions (e.g., viscosity of the molten material, humidity, wear and tear, etc.) the offset between the melt pressures produced by the MPT sensor 129 or the die pressure sensor 130 and the extrusion pressure sensor 128 changes over time. Accordingly, the techniques disclosed herein relate to adjusting the offset vector 215 over time to maintain an accurate representation of the melt pressure data offset in case there is a need to switch to control via the secondary sensor, the extrusion molding machine still exhibits a response that closely matches the response when controlled using the primary sensor.

[0048] In one example, the native controller 140 maintains the offset vector 215 as a rolling average of the offset a predetermined number of prior segments (e.g., 10 segments, 50 segments, 100 segments, etc.). As another example, the native controller 140 detects a threshold difference in the offset during a most recent extrusion cycle and the stored offset vector (e.g., a difference of 2%, 5%, 10%, etc. in an error metric between the offset vectors). In yet another example, the native controller 140 is configured to periodically update the offset vector 215 (e.g., every 8 hours, every 100 rotations, etc.).

[0049] FIG. 3 schematically illustrates a flowchart 220 for utilizing a machine plastic pressure signal 230 (e.g., a pressure signal produced by the extrusion pressure sensor 128) and ameasured plastic pressure signal 228 (e.g., a pressure signal produced by the MPT sensor 129 or the die pressure sensor 130) to attain a high-definition pressure output 272. In some embodiments, the actions described with respect to the flowchart 220 are performed by the native controller 140 of the extrusion molding machine 100. It should be appreciated that while the description of FIG. 3 assumes that the measured plastic pressure signal 228 is the primary pressure signal, in other embodiments (referred to herein as “reverse offset embodiments”), the machine plastic pressure signal 230 is the primary pressure signal. Accordingly, any techniques described below with respect to the use of the measured plastic pressure signal 228 as the primary pressure signal envision the alternate implementation in a reverse offset embodiment that use the machine plastic pressure signal 230 as the primary pressure signal.

[0050] At the input signal processing step 250, the native controller 140 processes the machine plastic pressure signal 230 and / or the measured plastic pressure signal 228. For example, the native controller 140 may convert the voltage levels of the signals 228, 230 to corresponding melt pressure values based on a known conversion associated with the MPT sensor 129 or the die pressure sensor 130 and the extrusion pressure sensor 128, respectively. Additionally, the native controller 140 may calculate an offset between the signals 228, 230 to produce an offset vector for historization. As described above, the offset vector may include multiple samples of the offset between the signals 228, 230 throughout the execution of the extrusion process.

[0051] Additionally, in some embodiments, the native controller 140 may historize the measured plastic pressure signal 228 for use in a subsequent portion of the extrusion process should a failure be detected during a future execution of the sensor health analysis / failure signature detection stage 252. Similar to historizing the offset vector, the native controller 140 may store a plurality of samples of the measured plastic pressure signal 228 such that the native controller 140 is able to provide historized pressure data that reflects the changes in the measured plastic pressure signal 228 throughout the execution of the extrusion process. In some embodiments, the historized measured plastic pressure signal 228 is a rolling average of prior measured plastic pressure signal 228 (e.g., the measured plastic pressure signal 228 data from the prior three, five, ten, etc. extrusion segments).

[0052] At the sensor health analysis / failure signature detection stage 252, the native controller 140 analyzes the signals 228, 230 to determine whether the primary sensor (typically the MPT sensor 129 or the die sensor 130) is functioning properly. While the flowchart 220 depicts a sequential order for the signature detection analyses 254-262, in alternateembodiments, a different sequence of signature detection analyses 254-262 may be implemented. In some further alternate embodiments, two or more of the signature detection analyses 254-262 may be performed in parallel. It should be appreciated that, in some embodiments, the signature detection analysis 254-262 are performed within 1 -5 milliseconds.

[0053] In some embodiments, if the native controller 140 has previously detected a failure signature during a previous execution of the sensor health analysis / failure signature detection stage 252, the native controller 140 may skip the analyses 254-262 and proceed to a signal fault compensations step 264. In other embodiments, the native controller 140 may continue to perform the analyses 254-262 to detect whether the failure signature has stopped manifesting (and / or has stopped manifesting for a threshold duration (e.g., 100 segments, 200 segments, etc.).

[0054] Starting with the circuit health analysis 254, the native controller 140 may perform a circuit continuity check by comparing the measured plastic pressure signal 228 to an out-of- range threshold. To this end, the measured plastic pressure signal 228 exceeding the out-of- range threshold may indicate an error in the ability of the MPT 129 or the die pressure sensor 130 to physically sense a melt pressure (such as when physical components of the MPT sensor 129 or the die pressure sensor 130 are damaged by the molten material). In this scenario, the MPT sensor 129 or the die pressure sensor 130 may output a voltage signal indicative of the error (e.g., 16V).

[0055] With simultaneous reference to FIG. 4A, illustrated is an example graph 300 showing an example signature that fails the circuit health analysis 254. More particularly, the graph 300 depicts melt pressure data derived from the measured plastic pressure signal 228 (as represented by the line 305a) and melt pressure data derived from the machine plastic pressure signal 230 (as represented by the line 310a) and the offset therebetween (as represented by the line 315a). As shown in the graph 300, shortly after time .010 seconds the line 305a includes a vertical asymptote 306 indicating that measured plastic pressure signal 228 produced a voltage value that is out-of-range. Accordingly, the native controller 140 may detect a failure in circuit health by either detecting a voltage value of the measured plastic pressure signal 228 or by detecting the out-of-range threshold represented by the vertical asymptote 306 in the pressure signal derived therefrom.

[0056] Turning to the electromagnetic interference (EMI) detection analysis 256, the native controller 140 may detect the presence of EMI present in the measured plastic pressure signal 228. The data signature of the EMI may manifest itself as a local signal anomaly in the offsetvector. Accordingly, if the native controller 140 detects the presence of the local signal anomaly for at least a predetermined duration (e.g., 1 ms, 10ms, 100ms, etc.), the native controller 140 may determine that EMI is impacting the measured plastic pressure signal 228. As one example, the native controller 140 may detect the presence of the anomaly by determining that the derivative of the offset signal, or the absolute value thereof, exceeds a threshold value. In some embodiments, the native controller 140 additionally detects that the anomaly occurs for a threshold duration (e.g., 2ms, 10ms, 15ms, and so on).

[0057] With simultaneous reference to FIG. 4B, illustrated is an example graph 325 showing an example signature that fails the EMI detection analysis 256. More particularly, the graph 325 depicts melt pressure data derived from the measured plastic pressure signal 228 (as represented by the line 305b) and melt pressure data derived from the machine plastic pressure signal 230 (as represented by the line 310b) and the offset therebetween (as represented by the line 315b). As shown in the graph 325, the measured plastic signal 305b experiences periodic anomaly event(s) 307, which manifests as a period in which the offset signal 315b rapidly changes value as well. Accordingly, the native controller 140 may detect an EMI failure signature by either detecting that the duration of the local signal anomaly 307 exceeds a threshold and / or that a total number of local signal anomalies that occurred during a predefined time frame (e.g., 1 second, 5 seconds, 10 seconds, 30 seconds, and so on) exceeds a threshold.

[0058] Turning to the signal drift analysis 258, the native controller 140 may detect that the offset signal has shifted from a reference level by a threshold amount. As described above, in some embodiments the native controller 140 is configured to detect a relatively small amount of signal drift to update the stored offset vector. Accordingly, in these embodiments, the threshold amount of drift detected during the signal drift analysis 258 is larger than the threshold required to update the stored offset vector and / or is determined based upon a different reference offset vector (e.g., a difference of 10%, 15%, 25%, etc. in an error metric between the offset vectors). In some embodiments, the error metric is a difference between the average offset of the offset samples included in the offset vectors.

[0059] Turning to the semiconductor failure analysis 260, the native controller 140 may detect a failure in the physical components included in the MPT sensor 129 or the die pressure sensor 130. A common type of MPT failure that occurs is due to delamination, which manifests by introducing sawtooth shaped artifacts in the measured plastic pressure signal 228. Of course, other types of electrical component failure may occur depending upon the physical arrangementof components of the MPT sensor 129. Accordingly, the native controller 140 may be configured to analyze the measured plastic pressure signal 228 to detect a threshold number of sawtooth artifacts that manifest due to the electrical component failure.

[0060] With simultaneous reference to FIG. 4C, illustrated is an example graph 350 showing an example signature that fails the semiconductor failure analysis 260. More particularly, the graph 350 depicts melt pressure data derived from the measured plastic pressure signal 228 (as represented by the line 305c). As shown in FIG. 4G, the line 305c includes many sawtooth artifacts, degrading the quality of the measured plastic pressure signal 228.

[0061] It should be appreciated that the particular primary sensor failure conditions described with respect to tests 254-260 are merely example failure conditions and different embodiments may include additional, fewer, and / or alternate failure conditions. Accordingly, the sensor health analysis / failure signature detection stage 252 also includes a catch-all other failure signature detection analysis block 262 at which the native controller 140 is configured to detect the alternate and / or additional failure conditions.

[0062] If the native controller 140 did not detect a failure signature during the sensor health analysis / failure signature detection stage 252, or, in some embodiments, if the failure signature is no longer detected, the native controller 140 may skip the signal fault compensation step 264 and report the measured plastic pressure signal 228 as the high-definition pressure output 272. On the other hand, if the native controller 140 does detect a failure condition in any of the analyses 254-262, the native controller 140 may compensate for the failure via the signal fault compensation step 264.

[0063] During the signal fault compensation step 264, the native controller 140 applies techniques to produce a high-definition pressure output that compensates for the measured plastic pressure signal 228 being an unreliable input for closed loop control of the extrusion molding machine 100. While the flowchart 220 depicts a sequential order for the compensation techniques 266-270, the ordering represents a general preference of the compensation techniques 266-270. Accordingly, the controller 140 may only perform a single one of the compensation techniques 266-270 in any given extrusion cycle.

[0064] Starting with the compensation / filtering techniques 266, in some scenarios the failure signature may be resolved by applying signal processing techniques that improve the measured plastic pressure signal 228. For example, smoothing techniques may improve the quality of the measured plastic pressure signal 228 in the event of semiconductor electronics failure. As otherexamples, the native controller 140 may perform outlier identification and exclusion, time-based filtering and time-weighted filtering techniques, as well as other techniques known in the art.

[0065] Next, the native controller 140 may perform realtime offset compensation 268. During this step, the native controller 140 may add the offset vector to the machine plastic pressure signal 230 to produce the high-definition pressure output 272. More particularly, when executing the extrusion process, the controller 140 may identify a current time within the extrusion process and add the corresponding sample included in the offset vector to the machine plastic pressure signal 230. As a result, the native controller 140 inputs into the closed loop control process a pressure signal that appears to have been generated by the MPT 129 or the die pressure sensor 130, but is actually derived from the extrusion pressure sensor 128. It should be appreciated that in the reverse offset embodiments, the native controller 140 may instead subtract the offset vector samples from the measured plastic pressure signal 228 to produce the high-definition pressure output 272.

[0066] Additionally, the native controller 140 may perform historization recall techniques 270 to substitute historical measured plastic pressure data for the measured plastic pressure signal 228. It should be appreciated that the viscosity of the molten plastic material 1 14 changes over time. Accordingly, when relying on historized measure plastic pressure data, the native controller 140 may identify an amount of time and / or a number of cycles that have occurred since the historized measured plastic pressure data was collected and apply a correction factor that accounts for the shift in viscosity. As such, when the native controller 140 implements the historization recall techniques 270, the native controller 140 is still able to account for dynamic changes in the operating conditions.

[0067] FIG. 5 illustrates an exemplary method 500 for manufacturing products. The method 500 may be performed by operating a controller (such as the native controller 140 or remote controller 146) of a molding machine (such as an extrusion molding machine 100 or an injection molding machine) in accordance with a program stored on a non-transitory memory. The molding machine may include an extruder (such as the screw control 126) configured to control advancement of a molten material advancement device (such as the screw 112).

[0068] The molding machine may also include at least one primary pressure sensor. For example, the primary pressure sensor may be a melt pressure transducer (MPT) sensor 129 or a die pressure sensor 130 in direct contact with molten material (such as the molten plastic material 114) downstream of the molten material advancement device, or (ii) a proxy sensor that, while not being in direct contact with molten material, provides data indicative of pressureof molten material downstream of the molten material advancement device. The molding machine may also include at least one secondary pressure sensor (such as the extrusion pressure sensor 128) not in direct contact with molten material downstream of the molten material advancement device.

[0069] The method 500 may begin at block 502 when the controller receives data provided by both the at least one primary sensor and the at least one secondary sensor. In some embodiments, the controller may calculate a plurality of deltas between the data provided by the at least one secondary sensor and the data provided by the at least one primary sensor at a plurality of times throughout a segment of an extrusion process. The segments may be defined temporally or based on a non-time dependent variable. The plurality of deltas may form an offset vector representative of an offset signal (such the offset signals 215, 315).

[0070] In some embodiments, the controller stores a historical offset vector in a memory of the controller, wherein the historical offset vector is based on a plurality of deltas calculated at a plurality of times throughout a segment of the extrusion process. In these embodiments, the controller may be configured to detect that an average of the plurality of deltas calculated during the segment of the extrusion process exceeds an average of the plurality of deltas indicated in the historical offset vector by a threshold value and update the stored offset vector to be based on the plurality of deltas calculated during the segment of the extrusion process in response thereto. Additionally or alternatively, in some embodiments, the controller may be configured to detect a temporal trigger to update the stored offset vector and updating, via the controller, the stored offset vector to be based on the plurality of deltas calculated during the segment of the extrusion process in response thereto.

[0071] At block 504, the controller analyzes the data provided by the at least one primary sensor to detect whether one or more indicia of sensor health failure are present. In some embodiments, the indicia of sensor health failure indicate that the pressure data received from the at least one primary sensor is outside of an expected range of pressure values. For example, the controller may perform the analyses 254-262. Accordingly, in some embodiments, the controller detects that the plurality of deltas between the data provided by the at least one secondary sensor and the data provided by the at least one primary sensor for the segment of the extrusion process exhibits a pattern indicative of the sensor health failure.

[0072] When an indicia of sensor health failure is not present (the “NO” branch of decision 506), the controller proceeds to block 508 where the controller controls the extruder in a manner that adjusts a velocity of the molten material advancement device to achieve a desired meltpressure at the primary sensor based on pressure data received from the at least one primary sensor.

[0073] On the other hand, when an indicia of sensor health failure is present (the “YES” branch of decision 506), the controller proceeds to block 510 where the controller controls the extruder of the molten material advancement device in a manner that adjusts the velocity of the molten material advancement device based on adjusted pressure data received from the at least one secondary sensor to achieve the desired melt pressure. For example, the controller may implement the realtime offset techniques 268. As a result, in some embodiments, the controller is able to switch control techniques during the extrusion process in which the controller detects the sensor health failure.

[0074] In some alternate embodiments, the controller instead controls the extruder of the molten material advancement device in a manner that adjusts the velocity of the molten material advancement device based on historical pressure data received from the at least one primary sensor to achieve the desired melt pressure at block 510. For example, the controller may implement the historization recall techniques 270. In these embodiments, the controller may be configured to determine an amount of time and / or a number of cycles that have occurred and since the reception of the historical pressure data and based thereon, scale the historical pressure data to account for a shift in viscosity of the molten material.

[0075] In some embodiments, the controller is further configured to analyze the data provided by the at least one primary sensor to detect a resolution of the sensor health failure. In response to detecting the resolution of the sensor health failure, the controller may resume control of the extruder in a manner that adjusts a velocity of the molten material advancement device to achieve the desired melt pressure at the primary sensor based on the pressure data received from the at least one primary sensor.

[0076] While particular embodiments of the present invention have been illustrated and described, it would be obvious to those skilled in the art that various other changes and modifications can be made without departing from the spirit and scope of the invention. It is therefore intended to cover in the appended claims all such changes and modifications that are within the scope of this invention.

Claims

CLAIMSWhat is claimed is:1 . A method for manufacturing products, comprising: providing an extruder configured to control advancement of a molten material advancement device; providing at least one primary pressure sensor, the primary pressure sensor being one of (i) a melt pressure transducer sensor or a die pressure sensor in direct contact with molten material downstream of the molten material advancement device, or (ii) a proxy sensor that, while not being in direct contact with molten material, provides data indicative of pressure of molten material downstream of the molten material advancement device; providing at least one secondary pressure sensor not in direct contact with molten material downstream of the molten material advancement device; providing a controller in communication with each of the at least one secondary pressure sensor, the extruder, and the at least one primary sensor; operating the controller according to a program stored on a non-transitory computer readable medium to: receive the data provided by both the at least one primary sensor and the at least one secondary sensor, analyze the data provided by the at least one primary sensor to detect one or more indicia of sensor health failure, and when an indicia of sensor health failure is not detected, control the extruder in a manner that adjusts a velocity of the molten material advancement device to achieve a desired melt pressure at the primary sensor based on pressure data received from the at least one primary sensor, and when an indicia of sensor health failure is detected, control the extruder of the molten material advancement device in a manner that adjusts the velocity of the molten material advancement device based on adjusted pressure data received from the at least one secondary sensor to achieve the desired melt pressure.

2. The method of claim 1 , wherein operating the controller to detect the one or more indicia of sensor health failure comprises: detecting, via the controller, that the pressure data received from the at least one primary sensor is outside of an expected range of pressure values.

3. The method of claim 1 , wherein during a segment of an extrusion process of the molten material advancement device, operating the controller comprises: calculating, via the controller, a plurality of deltas between the data provided by the at least one secondary sensor and the data provided by the at least one primary sensor at a plurality of times throughout the segment of the extrusion process.

4. The method of claim 3, wherein operating the controller to detect the one or more indicia of sensor health failure comprises: detecting, via the controller, that the plurality of deltas between the data provided by the at least one secondary sensor and the data provided by the at least one primary sensor for the segment of the extrusion process exhibits a pattern indicative of the sensor health failure.

5. The method of claim 4, wherein operating the controller comprises: storing, via the controller, a historical offset vector in a memory of the controller, wherein the historical offset vector is based on a plurality of deltas calculated at a plurality of times throughout a previous segment of the extrusion process.

6. The method of claim 5, wherein operating the controller comprises: detecting, via the controller, that an average of the plurality of deltas calculated during the segment of the extrusion process exceeds an average of the plurality of deltas indicated in the historical offset vector by a threshold value; and in response to detection, updating, via the controller, the stored offset vector to be based on the plurality of deltas calculated during the segment of the extrusion process.

7. The method of claim 6, wherein operating the controller comprises: detecting, via the controller, a temporal trigger to update the stored offset vector; and in response to the detection, updating, via the controller, the stored offset vector to be based on the plurality of deltas calculated during the segment of the extrusion process.

8. The method of claim 3, wherein the segment of extrusion process is defined temporally or based on a non-time dependent variable.

9. The method of claim 1 , wherein operating the controller comprises: analyzing, via the controller, the data provided by the at least one primary sensor to detect a resolution of the sensor health failure, and in response to detecting the resolution of the sensor health failure, resuming, via the controller, control of the extruder in a manner that adjusts a velocity of the molten material advancement device to achieve the desired melt pressure at the primary sensor based on the pressure data received from the at least one primary sensor.

10. An extrusion machine comprising: an extruder configured to control advancement of a molten material advancement device; at least one secondary pressure sensor not in direct contact with molten material downstream of the molten material advancement device; at least one primary pressure sensor, the primary pressure sensor being one of (i) a melt pressure transducer sensor or a die pressure sensor in direct contact with molten material downstream of the molten material advancement device, or (ii) a proxy sensor that, while not being in direct contact with molten material, provides data indicative of pressure of molten material downstream of the molten material advancement device; a controller communicatively coupled with each of the extruder, the at least one primary pressure sensor, and the at least one secondary sensor; wherein the controller is configured to: receive the data provided by both the at least one primary sensor and the at least one secondary sensor, analyze the data provided by the at least one primary sensor to detect one or more indicia of sensor health failure, and when an indicia of sensor health failure is not detected, control the extruder in a manner that adjusts a velocity of the molten material advancement device to achieve a desired melt pressure at the primary sensor based on pressure data received from the at least one primary sensor, and when an indicia of sensor health failure is detected,control the extruder in a manner that adjusts the velocity of the molten material advancement device based on adjusted pressure data received from the at least one secondary sensor to achieve the desired melt pressure.1 1 . The extrusion machine of claim 10, wherein to detect the one or more indicia of sensor health failure, the controller is configured to: detect that the pressure data received from the at least one primary sensor is outside of an expected range of pressure values.

12. The extrusion machine of claim 10, wherein during a segment of an extrusion process, the controller is configured to: calculate a plurality of deltas between the data provided by the at least one secondary sensor and the data provided by the at least one primary sensor at a plurality of times throughout the segment of the extrusion process.

13. The extrusion machine of claim 12, wherein to detect the one or more indicia of sensor health failure, the controller is configured to: detect that the plurality of deltas between the data provided by the at least one secondary sensor and the data provided by the at least one primary sensor for the segment of the extrusion process exhibits a pattern indicative of the sensor health failure.

14. The extrusion machine of claim 13, wherein the controller is configured to: store a historical offset vector in a memory of the controller, wherein the historical offset vector is based on a plurality of deltas calculated at a plurality of times throughout a previous segment of the extrusion process.

15. The extrusion machine of claim 14, wherein the controller is configured to: detect that an average of the plurality of deltas calculated during the segment of the extrusion process exceeds an average of the plurality of deltas indicated in the historical offset vector by a threshold value; and in response to detection, update the stored offset vector to be based on the plurality of deltas calculated during the segment of the extrusion process.

16. The extrusion machine of claim 15, wherein the controller is configured to:detect a temporal trigger to update the stored offset vector; and in response to the detection, update the stored offset vector to be based on the plurality of deltas calculated during the segment of the extrusion process.

17. The extrusion machine of claim 10, wherein the segment of extrusion process is defined temporally or based on a non-time dependent variable.

18. The extrusion machine of claim 10, wherein the controller is configured to: analyze the data provided by the at least one primary sensor to detect a resolution of the sensor health failure, and in response to detecting the resolution of the sensor health failure, resume control of the extruder in a manner that adjusts a velocity of the molten material advancement device to achieve the desired melt pressure at the primary sensor based on pressure data received from the at least one primary sensor.

19. A method for manufacturing products, comprising: providing an extruder configured to control advancement of a molten material advancement device; providing at least one secondary pressure sensor, the secondary pressure sensor being one of (i) a melt pressure transducer sensor or a die pressure sensor in direct contact with molten material downstream of the molten material advancement device, or (ii) a proxy sensor that, while not being in direct contact with molten material, provides data indicative of pressure of molten material downstream of the molten material advancement device; providing at least one primary pressure sensor not in direct contact with molten material downstream of the molten material advancement device; providing a controller in communication with each of the at least one secondary pressure sensor, the extruder, and the at least one primary sensor; operating the controller according to a program stored on a non-transitory computer readable medium to: receive the data provided by both the at least one primary sensor and the at least one secondary sensor, analyze the data provided by the at least one primary sensor to detect one or more indicia of sensor health failure, and when an indicia of sensor health failure is not detected,control the extruder in a manner that adjusts a velocity of the molten material advancement device to achieve a desired melt pressure at the primary sensor based on pressure data received from the at least one primary sensor, and when an indicia of sensor health failure is detected, control the extruder of the molten material advancement device in a manner that adjusts the velocity of the molten material advancement device based on adjusted pressure data received from the at least one secondary sensor to achieve the desired melt pressure.

20. A method for manufacturing products, comprising: providing an extruder configured to control advancement of a molten material advancement device; providing at least one primary pressure sensor, the primary pressure sensor being one of a (i) melt pressure transducer sensor or a die pressure sensor in direct contact with molten material downstream of the molten material advancement device, or (ii) a proxy sensor that, while not being in direct contact with molten material, provides data indicative of pressure of molten material downstream of the molten material advancement device; providing at least one secondary pressure sensor not in direct contact with molten material downstream of the molten material advancement device; providing a controller in communication with each of the at least one secondary pressure sensor, the extruder, and the at least one primary sensor; operating the controller according to a program stored on a non-transitory computer readable medium to: receive the data provided by both the at least one primary sensor and the at least one secondary sensor, analyze the data provided by the at least one primary sensor to detect one or more indicia of sensor health failure, and when an indicia of sensor health failure is not detected, control the extruder in a manner that adjusts a velocity of the molten material advancement device to achieve a desired melt pressure at the primary sensor based on pressure data received from the at least one primary sensor, and when an indicia of sensor health failure is detected,control the extruder of the molten material advancement device in a manner that adjusts the velocity of the molten material advancement device based on historical pressure data received from the at least one primary sensor to achieve the desired melt pressure.21 . The method of claim 20, wherein controlling the extruder based on historical pressure data comprises: determining, via the controller, a number of cycles that have occurred and since the reception of the historical pressure data; and based on the number of cycles, scaling, via the controller, the historical pressure data to account for a shift in viscosity of the molten material.

Images