CN1950768A - Method and apparatus for improving and controlling the curing of natural and synthetic moldable compounds - Google Patents

Abstract

A method for curing a moldable compound under multiple curing conditions by: (1) obtaining a time-dependent data stream of dielectric or impedance values from a plurality of sensors distributed within a curing mold, wherein for each sensor, the moldable compound is a dielectric; (2) determining impedance-related measurements from the data streams of the plurality of sensors; (3) using the impedance-related measurements from the plurality of sensors to determine predictive and/or corrective curing actions for enhancing the curing process; and (4) controlling mass production curing of parts to obtain cured parts having one or more desired characteristics.

Description

Method and apparatus for improving and controlling the curing of natural and synthetic moldable compounds

technical field

The present invention relates to methods and apparatus for monitoring and controlling the curing and setting of natural and synthetic moldable compounds. Typical of such moldable compounds are polymers.

Background technique

To date, methods of applying fixed process parameters to the processing of moldable polymers during curing and/or solidification have resulted in: reduced yields due to overly conservative cure times; and inability to accommodate cure due to fixed process parameters. and/or inherent variability in the solidification process resulting in poor product uniformity.

Attempts have been made to analyze the dielectric properties to determine the state of cure of the cured part. In particular, the following references, which are hereby incorporated by reference in their entirety, disclose various techniques that may be relevant for analyzing the state of cure and/or the process of setting:

US Patent Documents

4,344,142 Filed Aug. 6, 1975 Inventors Diehr, II et al.

4,373,092 Filed December 29, 1980 Inventor Zsolnay

4,399,100 Submitted on December 29, 1980 Inventors Zsolnay et al

4,423,371 Filed September 3, 1981 Inventors Senturia et al

4,496,697 Filed August 24, 1982 Inventors Zsolnay et al

4,510,103 Filed September 19, 1983 Inventors Yamaguchi et al.

4,551,807 Filed August 17, 1983 Inventors Hinrichs et al

4,723,908 Filed May 1, 1985 Inventor Kranbuehl

4,777,431 Filed June 27, 1986 Inventors Day et al

4,773,021 Filed September 20, 1988 Inventors Harris et al

4,868,769 Filed September 19, 1989 Inventors Persson et al

5,032,525 Filed March 31, 1988 Inventors Lee et al

5,219,498 Filed Nov. 12, 1991 Inventor Keller et al.

5,317,252 Filed September 15, 1992 Inventor Kranbuehl

5,486,319 Filed December 29, 1993 Inventors Stone et al

5,528,155 Filed June 18, 1996 Inventors King et al

5,872,447 Filed September 10, 1997 Inventor Hager, III

other publications

■ A comparative study of step curing and continuous curing methods, 1994, D.Khastgir, Indian Institute of Technology

■ AC Impedance Spectroscopy of Carbon Black-Moldablecomposites, 1999, K.Rajeshwar, University of Texas at Arlington

■ Anelastic and Dielectric Effects in Potymeric Solids, 1967, N.G.McCrum, B.E.Read, and G.Williams

Curing and solidification techniques for making components have provided, for example, some relationship between the dielectric (herein also referred to as "impedance") properties of a polymeric resin and the curing and/or freezing of such resin. However, prior art related to polymeric moldable curing and/or solidification has not fully addressed the need for direct electrical or impedance measurements during the production process, especially in the high friction and high pressure environments of injection molding or other molding types. practical aspect. Additionally, the prior art does not satisfactorily disclose the use of obtained electrical data to achieve closed-loop control of, for example, the curing and/or setting process of polymeric moldable compounds for a wide variety of molding methods and conditions. In particular, the prior art does not provide a solution or disclosure for effectively reducing defective parts during molding of large parts by curing and/or solidifying polymers. More specifically, the prior art fails to disclose efficient techniques for using electrical (i.e., impedance) data samples from the in-mold curing or/and solidification process, where such data samples are received simultaneously from multiple in-mold sensors that Distributed in such a way that it is possible to assess the state of curing and/or solidification of different parts of the large part.

The prior art also does not show how to compensate for: (a) batch-to-batch and intra-batch variations in polymerized moldable curing compound, and (b) differences in material thickness during the curing and/or setting process. Additionally, the prior art does not compensate for the additional variables introduced into the process by the nature of the equipment, tooling, and thermal history of polymerizing the moldable curable compound.

Additionally, the prior art uses dielectric or impedance measuring sensors employing opposed and parallel electrodes of precise area and spacing, and wherein the electrodes are in direct contact with the moldable compound. Although the electrodes and sensors provide a means for measuring impedance properties during curing and/or setting, their use may not be practical in a part production environment. For example, many moldable assemblies are produced using part molding techniques that allow the sensor to withstand pressures up to 30,000 psi and temperatures up to 425°F, as well as high friction environments (for example, due to moldable flow of plastic compound over the sensor). Finally, such prior art sensors must also be able to withstand mold cleaning by typical cleaning methods such as _CO2 and plastic bead impact.

Accordingly, the deficiencies described above are addressed by the curing methods and systems disclosed hereinafter. Additionally, standard practice for plastic molders due to the need for a cure time safety margin (i.e., cure time beyond what is generally considered to be expected) and/or due to inherent variability in the curing process of moldable compounds, There is therefore a need for a real-time feedback cure control system that reduces the safety margin of the molded article and at the same time prevents scrap from increasing and disrupting part production. The curing method and system disclosed hereinafter also addresses the need for a real-time feedback curing control system.

Definitions and Terminology

In the following disclosure a number of technical terms and abbreviations are used. Therefore, for convenience, many of these terms and abbreviations are described in this section. Therefore, it is recommended to refer to this section for a description of the terms used here.

Confidence Interval: A numerical range within which a specific number indicates the likelihood of an event or condition occurring, eg, a probability range such as 0.8 to 1.0.

Curing: As used below, this term refers to: (a) the chemical change of a polymer undergoing crosslinking from a softened structure to a ridge structure, and (b) the hardening of a malleable polymer so that it solidifies into a rigid structure without chemical changes.

Exponential magnitude coefficient: The magnitude coefficient (A) defined by the best exponential fit to a set of raw data, where the fitted curve (y) is described by the following equation:

y=Ae ^-αt , where t is time.

Exponential decay: The decay coefficient (α) defined by the best exponential fit to a set of raw data, where the fitted curve (y) is described by the following equation:

y=Ae ^-αt , where t is time.

Impedance Data Stream: For each sensor that operates during the curing and/or setting process to detect changes in impedance within the cured part, a time series of values are obtained during part curing, where these values indicate, for example, the Impedance measurements of the corresponding capacitor circuit (CC) provided by the molding compound. In particular, each capacitor circuit is operably configured such that the moldable compound becomes the dielectric for the corresponding capacitor circuit. Additionally, the impedance dataflow segments referred to herein are more fully described in US Patent Application No. 10/800,079, filed March 11, 2004, which is incorporated herein by reference in its entirety.

Metallic material with low CTE: A material with a low coefficient of thermal expansion.

Moldable Compounds: This term refers to (a) polymeric moldable compounds, (b) styrenic monomer compounds (SMC), (c) phenolic materials and (d) thermosetting plastics such as phenolic, urea, melamine, melamine- Phenolics, epoxy resins, unsaturated polyesters; note that the terms "polymeric moldable compound", SMC, and "phenolic material" are described in this Definitions and Terms section.

Phenolic Material: A moldable plastic material formed by the reaction of formaldehyde [HCHO] and phenol [C6H5OH], but almost any reactive phenol or aldehyde can be used. The material can be heavily reinforced or "filled" with fiberglass or other materials. Phenolics are well known for their high impact strength, excellent wear characteristics and dimensional stability over a wide temperature range. Phenolics can be thermoset molded. Commercially used phenols are phenol, cresol [CH3C6H4OH], xylenol [(CH3)2C6H3OH], p-tert-butylphenol [C4H9C6H4OH], p-phenylphenol [C6H5C6H4OH], bisphenol [(C6H4OH)2] and Resorcinol [C6H4(OH)2]. The aldehydes used were formaldehyde and furfural [C4H3OCHO]. In the uncured and semi-cured state, phenolic resins are used as adhesives, casting resins, casting compounds and laminating resins. As a molding powder, phenolic resins can be used for electrical applications.

Process Curve: Impedance data derived from a corresponding impedance data stream (this term is described above) in which the values have been "smoothed" (or otherwise processed) so that the curve can be more easily determined Slope and other mathematical curve properties. Examples of operations for smoothing impedance data streams are:

(a) For each of the one or more predetermined time segments of the impedance data stream, determine the linear least squares best fit over the segment terms, and the slope of the resulting line (i.e., m in the equation y=mx+b ). Then, the process curve becomes the cohesive sequence of the determined lines;

(b) for each of the one or more predetermined time segments of the impedance data flow, determining a best-fit third-order polynomial that models the impedance data flow term in the segment; and

(c) for each of the one or more predetermined time segments of impedance data flow, determine an exponential best fit modeling the impedance data flow items in the segment, wherein at least combined attenuation coefficients, and in some embodiments both attenuation and amplitude coefficients are determined (ie, both α and A in the following equation y=Ae ^−ax are determined).

The process profile is then a concatenation of the profiles generated for each of the one or more time periods. Note that the process curve can be derived from a single time period that is essentially the entire curing time of the part. Typically, however, such time periods may be limited to a percentage of cure time in the range of 10% to 35%.

Of course, other smoothing operators are also within the scope of the novel curing systems disclosed here, such as splines (eg, whose shape can be controlled by "control points" computed from the values of the impedance data stream).

R-squared ( ^R2 ): R-squared (also known as the coefficient of determination) is a statistical measure of the reduction in the total change in the dependent variable due to the corresponding independent variable. An R-squared value close to 1.0 indicates that the corresponding data model explains nearly all of the variability in each variable.

Moldable polymer (equivalently, "polymeric moldable compound"): This term means:

(a) Typical matrix moldable polymers

(b) phenolic compounds, and

(c) Phenolic compounds in combination with other materials including (but not limited to)

Inorganic filled phenolic, glass filled phenolic, with (but not limited to) cotton, PTFE (i.e. poly

Tetrafluoroethylene), wood flour, and graphite-filled cellulose phenolic.

In addition, those skilled in the art will understand that included in the term "moldable polymer" are: ethylene-methacrylic acid copolymer, nylon, polycarbonate, polychlorotrifluoroethylene, ethylene-acrylic acid copolymer , polyether ether ketone, polyethylene naphthalate, polyethylene terephthalate, polymethyl methacrylate, polyoxymethylene, polyvinyl chloride, polyvinylidene chloride, polyvinylidene fluoride , polyvinyl fluoride, styrene-acrylonitrile, polyethylene, acrylonitrile-butadiene-styrene, acrylic-styrene-acrylonitrile, polyamide-imide, polybutylene terephthalate, poly Carbonate/acrylonitrile-butadiene-styrene, polycarbonate/polybutylene terephthalate, polyetherimide, polyethersulfone, polyimide, polyphenylene ether, polyphenylene sulfide , polysulfone, styrene-maleic anhydride, thermoplastic elastomers, polypropylene, polystyrene, thermoplastic olefins, polytetrafluoroethylene, and mixtures thereof; ie, thermoplastic materials.

SMC: Styrenic monomer compound, such as a polyester resin that typically contains about 35 wt% styrene monomer, but can vary from about 0% to 50%.

Tonnage (Tonnage): The compression force of the molder, usually expressed in tons.

Tool steel: Steel suitable for making injection and compression molds, such as AISI type A2 tool steel.

Topological features of impedance-related data: identifiable and unique features within the process curve, such as curve slopes, peaks (e.g. local maxima), valleys (e.g. local minima), flat parts (e.g. substantially zero slope), inflection points, The rate of change of the slope of the curve, etc.

Vacuum Port: A port designed in a mold to reduce the pressure in the mold cavity to subatmospheric pressure.

Contents of the invention

The present disclosure relates to methods and systems for controlling curing (this term is also described above) and forming molded parts from a moldable compound (this term is described above in the Definitions and Terms section). In particular, molded parts resulting from the curing methods and systems disclosed herein:

(a) may be both more fully formed and more fully cured (but not overcured);

(b) have consistent and repeatable part properties, including reduced (or absence) blistering and/or porosity;

(c) have desired properties such as: compressive strength, adhesion to dissimilar materials, dimensional consistency, etc.; and

(d) Cure with a reduced average cure time per part (eg, up to about 38%).

Additionally, the curing methods and systems disclosed herein:

● Robust and repeatable in traditional component production environments, such as those producing high-volume automotive SMC molded components;

Automatically adjust cure time to compensate for mold temperature fluctuations while producing consistent parts;

- Showing the effect of in-mold introduction and placement of moldable compound on part cure rate using multiple impedance detection sensors provided in the mold (as described below);

●Identifies moldable compound flow anomalies during part cure and provides feedback to effectively alter cure process variables to improve the molding process; and

• Identify changes in the cured moldable compound and provide a mechanism to continuously improve part molding.

The curing methods and systems disclosed herein include novel features for monitoring and controlling the flow of moldable compound in a mold, and the polymerization or crosslinking of the moldable compound. In addition, such monitoring and control can be performed in real time, ie during curing (eg thermosetting) of the part, so that the number of defective parts produced can be reduced.

Additionally, the curing methods and systems disclosed herein can eliminate much of the curing time safety margin currently used to ensure that parts, especially relatively large parts, are properly formed and cured.

The present disclosure generally relates to a novel curing method and system in which a plurality of sensors are operatively distributed within a mold for detecting the state of forming a part and the state of curing at spaced apart portions of the part. Thus, output from multiple sensors can be used to determine if and/or when the mold is properly filled with moldable compound, and if and/or when the part is both properly cured and at substantially the same level throughout the part. Uniform cure rate.

For each of the multiple sensors operating during the curing process, a data stream of impedance values (here denoted "impedance data stream") is obtained, where these values indicate the Impedance measurements obtained with corresponding capacitor circuits (CC) are provided. In particular, each capacitor circuit is operably configured such that the moldable compound becomes the dielectric of the corresponding capacitor circuit. For each impedance data stream, there may be a corresponding "process curve" (as described in the definitions and terminology section above). Additionally, such a process profile may be represented as a time-series plot of impedance measurements for a corresponding impedance data stream. NOTE: Such a process curve may, but need not be, be a perfectly smooth representation of impedance data flow; however, those skilled in the art will understand that the process curve may be a connection of smooth curve segments (e.g., successive first derivatives) . Such a flow of impedance data and its corresponding process profile provide an indication or "signature" of how the part was formed and cured in its mold. In particular, various geometric features of process curves (eg, slope, local maxima, local minima, inflection points, etc.) have been identified as predictors of well formed, properly cured parts as well as various part defects. Accordingly, the methods and systems disclosed herein use characteristics (e.g., shape and/or geometric profile characteristics such as slope and/or area under the profile) obtained from a plurality of sensor profiles to determine the correct and/or incorrect cure state. NOTE: This impedance data stream may represent a time series of one or more of the following impedance types of impedance values: Impedance (Z) (i.e., a measure of the total resistance to current flow in an AC circuit, which consists of two components, the ohmic resistance and reactance, and is usually expressed in complex notation as Z=R+iX, where R is the ohmic resistance and X is the reactance), phase angle (Φ), resistance (R), reactance (X), conductance (G) and / or capacitance (C).

Additionally, for each sensor, there may be multiple impedance data streams (and their associated process curves) generated. For example, for a given sensor, the impedance data stream may be derived from the signal response output by exciting the corresponding capacitor circuit as a result of one of a plurality of predetermined different signal frequencies input to the capacitor circuit. Thus, each process profile can be derived from a corresponding individual signal frequency input to a capacitor circuit with a sensor, and the corresponding shape (or other calculated characteristic) of the resulting process profile can be used to monitor, control, and/or predict the performance of the part curing process. result.

In some embodiments disclosed herein, various time-series capacitor circuit output data components (e.g., impedance (Z), phase angle (Φ), resistance (R), reactance (X), conductance (G) or capacitor (C)) to monitor and control the part curing process. Thus, a process profile generated from these various data components can provide a unique shape (or other feature), the characteristics of which can be used to monitor and control the curing process. For example, the characteristic may include a process curve local maximum, or local minimum, curve slope, slope, identification of a process curve portion having substantially zero slope, an inflection point, an area under a process curve portion, and the like.

Process profiles obtained from a plurality of such in-mold sensors may be compared or evaluated individually or in groups to detect changes and/or abnormalities in the curing state of portions of parts, particularly relatively large parts such as automotive dashboards. Thus, localized anomalies in part curing may be detected by evaluating characteristics of a process profile derived from the output of adjacent sensors, or by comparing such process profile characteristics with corresponding characteristics of process profiles from other sensors. For this comparison, the following items can be compared:

(a) for a particular elapsed time for part cure, the average slope of a portion of the process curve from one sensor monitoring part cure versus the average slope of the corresponding part of the process curve from the other sensor monitoring part cure;

(b) the maximum value of the process curve from one sensor monitoring part cure versus the process curve from the other sensor monitoring part cure;

(c) When the part is ejected, the value of the process curve from one sensor monitoring part cure versus the value from the process curve from another sensor monitoring part cure.

This localized anomaly in part curing may be due to, for example, reduced flow of moldable compound to a portion of the mold, and/or a deviation in the cure rate of a portion of the cured part from the rest of the part (e.g., due to part thickness variations, uneven heat distribution in the mold, etc.). Additionally, a more global assessment of component curing may also be determined by, for example, evaluating the degree of agreement (or lack thereof) between corresponding process profiles (eg, at the same impedance frequency) of multiple sensors outputting impedance data. For example, when the process curves have a similar shape but are offset from each other in time, this may indicate that different parts of the part may be at different cure rates and, for example, cure time adjustments and/or adjustments to parts of the mold may be required. The temperature is adjusted so that the entire part (or subsequent parts) cures properly.

Another aspect of the present curing method and system is to use the initial portion of the impedance data stream generated by multiple sensors (inside the mold) to adjust the curing conditions for subsequent molded parts (e.g., from the same mold) such that during curing The moldable compound to be cured prior to starting substantially completely fills the mold. In particular, molder tonnage and press closure rate can be adjusted to modify the cure rate, and more particularly, the onset of substantial crosslinking in the cured moldable compound.

In at least some embodiments of the curing methods and systems disclosed herein, prior to batch curing of parts of a particular part type, a testing or sampling phase may be performed to determine the individual compositions or batches of moldable compound used for that part The curing properties of the samples. For example, the sample can be cured: (a) with different arrangements of moldable compound provided in the mold, (b) with different curing temperatures, (c) for different lengths of time, (d) with different molding tonnage. The resulting test parts and their corresponding process curves can be evaluated to determine adjustments to the curing process such that, for example, an undercured sample part from a particular batch of moldable compound (and with a specific introduction into the mold) can be extended. cure time and/or increase mold temperature. Thus, by comparing the process curve derived from this sample test with the corresponding impedance data obtained during curing of the part in a production run of that part type (e.g. where thousands of part instances can be produced), the formability of the part can be determined And curing is correct or incorrect. Additionally, if the curing and/or forming of the part is incorrect, adjustments may be made to the curing process so that the resulting part is more likely to be acceptable. In particular, this adjustment may result in the subsequent portion of the process profile (from multiple sensors) better matching that of a correctly formed and cured part. Thus, while this process profile may differ between individual parts in magnitude and/or relative timing of various profile characteristics (e.g. due to part thickness, thermal history, mold temperature and heat transfer rate, cure level, and various other factors), but for each individual part, the degree of consistency in the shape of the part's process curve and the degree of clustering of that curve can be used to predict whether the resulting part will be properly formed and cured.

Another aspect of the present disclosure is that in various embodiments of the curing system and for certain moldable compounds, the corresponding shape of one or more of the above-described process curves may exhibit a "maximum value" at a given time and/or or "minimum values", which can also be used to infer information useful in monitoring, controlling and/or predicting correct part cure times.

Another aspect of the present disclosure is that, in various embodiments and for certain moldable compounds, there is provided one or more (preferably a plurality) "evaluators" (also referred to herein as "program agents"). (programmatic agent)" or "conditions") are used to output values related to the curing time of the part. The estimator may be, for example, the corresponding slope or integrated area under one or more process curves. The output from one or more evaluators can be correlated with known cure times for samples of moldable compound to determine the predictive validity of the evaluators. An estimator that exhibits a high degree of correlation with the physically measured properties of the resulting part can be used to infer information useful in monitoring, controlling and/or predicting the correct cure time for subsequently cured parts, such as mass-produced parts. In at least one embodiment, the outputs from two or more (e.g., four) evaluators that provide the highest degree of correlation with the measured curing characteristic are combined (e.g., linearly) to produce an even better Predictor for predicting part cure time.

Another aspect of the present curing systems and methods is that embodiments thereof may include signal processing and other software and hardware components for deriving a process profile and corresponding characteristics of the profile (e.g., maximum and/or minimum values), and utilizing the The profile profile determines in real-time the most suitable cure time for the batch of parts being produced. In particular, utilizing expert systems such as independent intelligent agents, artificial neural networks, and computing architectures, and hybrid computing systems that provide statistically based decision-making systems, such as Salford Systems, 8880 Rio San Diego Dr., Ste.1045, CART in San Diego, Calif. 92108.

Additionally, it is an aspect of the present curing systems and methods that part cure times can be determined to achieve desired properties in the resulting cured part such as tensile strength, compressive strength, dynamic hardness, dimensional consistency, reduction of blisters/voids, and/or Or elimination, and adhesion to dissimilar materials.

Additional descriptions of advantages, benefits and patentable aspects of the present disclosure will become apparent from the drawings and descriptions below. All novel aspects of the disclosure, whether or not explicitly mentioned in this Summary section, alone or in combination with other aspects of the disclosure, are considered the subject of patent protection. Accordingly, such novel aspects disclosed below and/or in the accompanying drawings, which may be omitted or incompletely described from this Summary of the Invention, are hereby fully incorporated by reference into this Summary of the Invention. In particular, all claims in the claims section below are hereby fully incorporated by reference into this summary section.

Description of drawings

Figure 1 illustrates the components of one embodiment of the part curing method and system disclosed herein.

2 shows a more detailed embodiment of an impedance sensor circuit 62 that: (i) supplies an electrical signal to the sensor 17, (ii) determines an impedance indicative of the impedance of the moldable compound 16 at or near the sensor 17. Measurement.

FIG. 3A shows an exploded view of one embodiment of one of a plurality of sensors that may be used to obtain impedance measurements of cured moldable compound 16 .

Figure 3B shows an additional view of the sensor 17 of Figure 3A.

FIG. 4 shows an electrical (impedance) sensor 17 provided in the mold 18 .

FIG. 5 shows another view of the impedance sensor circuit 62 .

6 is a flowchart of the high level steps performed to determine initial curing parameters to begin curing of a part, and to determine adjustments to curing parameters while curing a part and/or for curing subsequent parts.

FIG. 7 is a flowchart providing more detailed steps of step 1014 of the flowchart of FIG. 6 .

FIG. 8 is a flow diagram illustrating an embodiment of high-level steps performed, for example, during a production run in which multiple parts expected to be substantially identical are sequentially cured one after the other.

Figure 9 shows a graph of a typical SMC (polyester, styrene monomer) impedance dataflow curve (and corresponding process curve) with time shown in seconds on the x-axis and relative conductance on the y-axis. Also shown are various points at which a curing event occurs (or is expected to occur).

FIG. 10 shows data points (and corresponding process curve 1404 ) for curing impedance data flow values for curing a part fabricated from SMC as moldable compound 16 . In particular, the x-axis represents the percentage of time in which the part is expected to cure without the use of curing system 20 .

Figure 11 shows a portion of a mold cavity 28 for curing automotive body panels, with three sensors 17 (ie sensors 17a, 17b and 17c) shown in the side walls of the mold cavity.

Figure 12 below shows a typical graph of impedance data flow (one of sensors 17) produced by curing the automobile body panel of Figure 11 at normal temperature (300°F), at 285°F, and at 315°F.

FIG. 13 shows a graph of a graph of typical impedance data flow (and their corresponding process curves) for the SMC charge placement shown in FIG. 11 prior to using curing system 20 to monitor and adjust part curing.

FIG. 14 shows a graph of typical impedance data flow (and their corresponding process curves) after reseating the SMC charge in the mold cavity 28 for an automotive body panel cured in the mold cavity 28 of FIG. 11 .

Detailed ways

FIG. 1 illustrates an embodiment of a curing system 20 disclosed herein, wherein the curing system includes the following high-level components:

(A) a curing device 45 for curing components therein;

(B) a control system 39 for controlling the curing apparatus 45 and adjusting curing parameters to reduce defects in parts cured in the mold 18; and

(C) A cure setup subsystem 104 for determining an initial set of curing parameters for a particular part, and for determining adjustments that may be used to correct curing conditions that may result in defective parts.

The curing device 45 includes the following high-level components:

(A.1) The mold 18 in which the moldable compound 16 is cured into the desired part has a mold cavity 24 for receiving the moldable compound 16 to be cured into the desired part. Distributed within the body of the mold 18 are a plurality of sensors 17 (only one shown in FIG. 1 ) for detecting the impedance characteristics of various parts of the moldable compound 16 as it cures;

(A.2) An assembly of a plurality of capacitors 68 (only one of which is shown), wherein each capacitor 68 is formed by: (i) being placed directly adjacent to the cured moldable compound 16 (and normally in contact with) one impedance sensor 17; NOTE: The moldable compound 16 is the capacitor dielectric for each such capacitor 68 .

Control system 39 includes a computing system such as computer 34 (or network of computers) on which processes for controlling curing of components are executed. In particular, the following components are provided by the computer 34 (or can be accessed by the computer 34 through, for example, a communication network such as the Internet or a local area network):

(B.1) data acquisition card 35, for

(i) generating a sinusoidal excitation voltage input to the sensor measurement unit 60, wherein the sensor measurement unit 60 provides this voltage to the sensor 17 in order to obtain a resulting signal indicative of impedance data flow from the sensor; and

(ii) Reading and digitizing the impedance signal output from each sensor 17 (and more specifically, from the amplifier 36 of Fig. 2, described further below).

(B.2) The digital signal generator 41 is used for determining and outputting signal characteristics (such as frequency and voltage), wherein the signal characteristics are used to control the signal output of the data acquisition card 35, and specifically control the signal output to the sensor measuring unit 60. signal output.

(B.3) The digital signal demodulation component 42 is used for demodulating the impedance indication signal received from the sensor 17 and the amplifier 36 through the data acquisition card 35 .

(B.4) A curing data capture database 23 for storing the impedance data stream obtained from the sensor 17, and linking this data stream with information identifying the moldable compound 16 used for part production, and for obtaining the impedance data stream therefrom. Various ambient curing parameters (eg, curing temperature, curing time, mold tonnage, etc.) of the part that resist data flow are correlated.

(B.5) Cure Analysis Subsystem 26 for analyzing output from the currently cured part obtained from the cure data capture database 23, and identifying any adjustments to the curing process that might reduce the formation of defects in the part (eg adjust part cure time), or determine adjustments that can be used to cure subsequent part instances. NOTE: The cure analysis subsystem 26 may receive initial part cure parameters from the cure control database 27, as well as cure adjustment parameters (eg, cure parameter adjustments during part cure, such as mold 18 temperature adjustments, etc.). Additionally, cure analysis subsystem 26 may receive or derive data output by sensor 17 identifying certain characteristics of the impedance data stream for use in making adjustments to the part curing process and/or identifying when part curing is to be terminated. The flowchart of FIG. 8 and its accompanying description below provide further disclosure of the operations performed by cure analysis subsystem 26 . Additionally, cure analysis subsystem 26 may include expert systems and/or "smart" system architectures for identifying topographic (eg, shape) characteristics or mathematical properties of process profiles, and/or patterns present in process profiles. Accordingly, embodiments of a solidified analysis subsystem may include one or more of the following: utilizing, for example, independent intelligent agents, expert systems with fuzzy logic, one or more artificial neural networks, and computing architectures, and providing statistically based decision-making Determine the hybrid computing system of the system, such as CART of Salford Systems, 8880 Rio San Diego Dr., Ste. 1045, San Diego, Calif. 92108. Examples of graphical representations of impedance data flow are shown in FIGS. 9 , 10 and 12 to 14 .

(B.6) A curing controller 43 for controlling the curing of components in the curing device 45 . Specifically, controller 43 communicates with cure analysis subsystem 26 for determining adjustments to the current cure (eg, cure time) of a part and/or for identifying adjustments for curing subsequent parts in the same mold 18 . In performing such adjustments, the controller 43 outputs curing commands or instructions to the curing device 45 via the command line 28 to the output device 38 . Apparatus 38 may in turn translate (if necessary) the command or instruction into a corresponding command or instruction capable of being executed by components of curing apparatus 45 (such as the embosser and mold heating system), and then provide such translated command or instruction via line 44 to Curing device 45. For example, such translated commands or instructions may be: (i) increase the mold 18 curing temperature in the vicinity of a particular one of the sensors 17 for subsequent parts cured in the mold 18, (ii) decrease the tonnage applied to the mold 18, (iii) ) prolong the curing time of the current part, and/or (iv) open the mold 18 .

(B.7) Digital input device 37 for receiving an indication (via line 40 ) when a curing cycle has been initiated, or when moldable compound 16 is provided to mold 18 . Digital input device 37 outputs corresponding notifications to curing controller 43 via line 29 . Examples of such means 37 are well known in the art.

(B.8) A sensor measurement unit 60 for generating an electrical signal input to the sensor 17 and for receiving an impedance indication signal (ie, an impedance data stream) from the sensor 17 . Note that there may be a unique sensor measurement unit 60 for each sensor 17 .

(B.9) Amplifiers 36 (at least one per sensor measurement unit 60 ) for amplifying the real-time (ie, during part curing) impedance data signal corresponding to the impedance of capacitor 68 .

With regard to the cure setup subsystem 104, this subsystem is used by a cure user to interactively determine initial cure parameter settings for curing a subsequent series of parts in a particular mold 18, and in some embodiments, Possible adjustments that can be made during curing. Cure setup subsystem 104 may perform one or more of the following tasks:

(C.1) Allows the user to access the cure information archive 31 (FIG. 1) to obtain historical cure information related to the moldable compound 16 to be used in the current cure process, such as:

(i) the typical curing temperature (and range thereof) for the moldable compound,

(ii) a typical cure profile (and variations thereof) for the moldable compound 16,

(iii) variation in curing temperature according to the composition of the moldable compound 16;

(C.2) allow the user to access information about the curing device 45 to be used, e.g. the preferred temperature setting for the curing device 45 may typically be 1 degree higher than another replica or model of the curing device 45;

(C.3) Allows the user to interact with the subsystem 104 to perform the steps of FIGS. 6 and 7 described hereinafter, where different batches of the specified moldable compound 16 may be used to cure the sample part, and different curing parameters (e.g. curing time, mold temperature, tonnage, etc.) to cure them. Alternatively, if there is sufficient historical solidification information in archive 31, then sample parts need not be produced and evaluated. In any case, the cure setup subsystem 104 outputs a set of resulting initial cure parameter settings, (any) cure adjustments to be performed during part cure (or subsequent adjustments for curing additional parts), and part cure termination conditions to Solidification control database 27 .

Data Acquisition and Control Hardware

Data acquisition and control hardware (e.g., digital signal generator 41 and data acquisition card 35 of the embodiment of FIG. ) to each impedance sensor 17. Specifically, if more than one signal frequency is input to each sensor 17, the signal frequencies are serially multiplexed into each sensor such that the signals from Sensor impedance response for each input frequency (and for each sensor). The input frequency or frequencies may be in the range of 10 Hz to 5 GHz, and corresponding conductance and/or capacitance measurements are determined from the sensor responses. Therefore, the conductance and capacitance readings (equivalently, the process curve) are specific to the moldable compound 16 under cure because the bipolar composition of the compound will produce a dielectric response pattern that is specific to the moldable compound. Additionally, the conductance and capacitance readings may be specific to the curing device 45 used.

sensor 17

An exemplary embodiment of one of the impedance sensors 17 is shown in more detail in FIG. 2 . In particular, each sensor 17 comprises a main electrode 10 serving as a capacitor plate of a corresponding capacitor 68 . An additional guard or shield electrode 11 surrounds the main electrode 10 and acts as a shield against excessive fringing of the electric field to the surface of an adjacent die 18, in which each sensor 17 is typically buried mounted. on the above surface. Those skilled in the art will understand that guard electrode 11 , which is energized along with electrode 10 (according to the signal received from sensor measurement unit 60 ), helps prevent the electric field induced at main electrode 10 of sensor 17 from forming a fringing flux or becoming non-linear. Electrodes 10 and 11 are coated by a thin (e.g., approximately 0.001 to 0.05 inch) ceramic coating 13 (FIG. 2) such as alumina ceramic or another stable dielectric insulator (e.g., over a curing process temperature range such as 300°F to 425°F Dielectrically stable) and separate from the moldable compound 16. Electrodes 10 and 11 may be composed of a low CTE metallic material such as stainless steel, titanium, a nickel-cobalt-iron alloy called Kovar(R) (which is owned by CRS Holding Inc.) embedded in a layered ceramic circuit (not shown). , a subsidiary of Carpenter Technology Corp. of Wyomissing, Pa.), nickel steel, tool steel, tungsten, super alloys, and soft magnetic alloys. Any other flat or semi-flat conductive surface in contact with the cured moldable compound (e.g., the opposing mold surface) can act as the opposing plate (i.e., the ground plate 64) of the capacitor 68 and act as a capacitive contact with the main electrode 10. coupled third electrode. Also note that the opposite plate 64 is grounded to electrical ground 25 to provide a common signal reference point. Thus, when a complex current (as described in the Definitions and Terminology section) is driven through resistor 19 (FIG. 2) to ground 25 due to the opposing plates being grounded, this current passes as Dielectric moldable compound 16. Then, the complex voltage across the resistor 19 is measured with a high precision amplifier 36 . The resulting signal is then input to the data acquisition card 35 and subsequently demodulated by the demodulation component 42 into complex impedance components (eg, conductance and capacitance) of the resulting signal.

Figures 3A and 3B show an embodiment of the sensor 17 (with the sensor housing 12 removed from the rightmost portion of Figure 3B). This embodiment comprises a nested structure of A2 tool steel components, including a sensor housing 12, a main electrode 10 and a guard electrode 11, wherein the electrodes are potted with cyanate ester along the length of the electrode 10 (i.e. in the direction of the axis 15) The material coating 76 is separated, as well as radially (from the shaft 15 ), by a thin insulating ceramic coating 13 and 13a such as alumina ceramic or other stable dielectric insulator. Ceramic coatings 13 and 13a may be applied with a thermal spraying process (ie detonation gun, plasma, or high velocity ceramic (HVOF) spraying process, which are well known to those skilled in the art). The ceramic coating 13 at the sensor surface also: (a) transfers the compressive load created by the curing process to the sensor 17, and (b) separates the electrodes 10 and 11 from the cured moldable compound 16. The coaxial cable 80 is connected to the sensor 17 by an MCX connector 14 screwed into the guard electrode 11, such as the MCX connector 14, part number 133-833-401, supplied by 299 Johnson Ave S.W., Suite 100, Waseca Manufactured by Johnson's Components of MN 56093. The center conductor 84 mates with a pin that is either integral with the electrode 10 or press fit into the electrode 10 . In some embodiments of sensor 17, main electrode 10, guard electrode 11, and housing 12, and alumina ceramic surface 13 may be fused together and electrically separated by glass or glass doped with alumina ceramic. Additionally, in some embodiments of sensor 17 (eg, FIGS. 3A and 3B ), main electrode 10, guard electrode 11, and housing 12 may be coated with 2 to 4 microns thick diamond or diamond-like material, as described by Anatech Ltd of Casidium supplied by Springfield, VA, which is then press-fit together such that the diamond or diamond-like coating provides electrical isolation between the three components (i.e. electrodes 10, 11 and housing 12), and also provides a moldable Electrical isolation between compound 16 and ceramic surface 13 ( FIG. 3A ) of sensor 17 .

Figure 4 shows how an embodiment of the sensor 17 may be positioned within the body of the mold 18 such that the ceramic surface 13 facing the mold cavity 24 provides the shape and surface texture of the part to be molded therein. Specifically, the sensor 17 may be embedded mounted in the mold 18 such that the sensor makes electrical contact with the part molded from the moldable compound 16 .

Since multiple sensors 17 may be provided in the mold 18 there will typically be at least one impedance data stream from each sensor 17 .

Sensor measuring unit 60

Each sensor measurement unit 60 ( FIGS. 1 , 2 and 5 ) provides a non-bridged circuit comprising a simple voltage divider ( FIG. 2 ) which in turn comprises a resistor 19 . Each sensor measuring unit 60 is operatively connected to one of a plurality of capacitors 68 formed by one of the sensors 17 and the moldable compound 16, wherein the sensor measuring unit both provides current to each of the capacitors and detects electrical signals from the pair. The capacitor responds to the resulting impedance value. Note that the combination of sensor measurement unit 60 and each capacitor 68 forms impedance sensor circuit 62 . The current provided to each impedance sensor circuit 62 is driven (via the opposing capacitor plates 64 described above) through the cured moldable compound 16 to the corresponding electrical ground 25 of the mold 18 . For each sensor circuit 62, a load resistor 19 (typically having a resistance of about 200 kilohms, but could range anywhere from 1 kiloohm to several megohms, such as 10 megohms) is placed in communication with the sensor circuit. The current of the sensor 17 flows in one line. The resulting voltage V2 on circuit line 20 ( FIG. 5 ) output by amplifier 36 measures the voltage across resistor 19 . The amount of attenuation and phase shift caused by the flow of complex current through capacitor 68 is determined by simultaneously measuring the voltage applied at location 21 (this applied voltage is also referred to as "excitation voltage" and also referred to as "V0"). FIG. 5 illustrates an example of a sensor measurement circuit 60 where an applied (excitation) voltage (e.g., V0=sinωt) at location 21 is placed at one terminal of amplifier 36, and this potential drives a complex current I ^* through load resistor 19 (R), then finally through the corresponding capacitor 68 formed by the sensor 17 , the moldable compound 16 , and the electrical ground 25 attached to the mold 18 .

The following description assumes a voltage amplitude of 1 volt for excitation V0 at circuit location 21 . However, if this voltage is inconsistent, all subsequent analyzes remain the same, because for the inconsistent case, the constant "k" in the equation below is defined as the difference between the negative voltage at circuit position 22 (V1) and the positive voltage at circuit position 21 (V0 ) ratio.

The excitation voltage (V0=sinωt) at position 21 drives a complex current (I ^* ) through resistor 19 to ground 25 . Specifically, voltage V0 is a digitally generated sine wave generated by a high speed data acquisition card 35, such as a PCI-MIO-16E4 card manufactured by National Instruments, Austin, TX. The data acquisition card 35 generates high quality sinusoidal signals at frequencies ranging from 10 Hz to 10 kHz as specified by, for example, an operator or user. However, other data acquisition cards 35 can also be used to generate similar or different ranges of frequencies, such as the PCI-MIO-16E1 data acquisition card manufactured by National Instruments, Austin, TX. which can generate and monitor frequencies from 10 Hz to 1.25 MHz . An embodiment of a data acquisition card 35 may also provide simultaneous data sampling, such as a card specifically designed to carefully maintain phase relationships between channels, such as the PCI-6110 card manufactured by National Instruments, Austin, TX.

When the excitation voltage V0 is applied at circuit position 21, a voltage drop occurs across the load resistor 19, leaving an attenuated and phase-shifted signal at circuit position 22 (i.e. V1=ksin(ωt+θ)=k< θ, where "<" is used to indicate the polar representation of complex numbers, and represents the term "at the phase angle of"). The moldable compound 16 between the sensor 17 and the electrical ground 25 provides a complex impedance of magnitude Z at a phase angle Φ, where the phase angle Φ is a characteristic of the cured moldable compound 16 and does not differ from that defined between V0 and V1 The phase angle difference of the phase angle θ is confused.

Computing Z and Φ is accomplished by simultaneously digitally capturing the excitation signal V0 (e.g., V0=sin(ωt)) and amplifier 36 outputting a voltage V2 on circuit line 20, where V2=sin(ωt)−ksin(ωt +θ). Alternatively, in another embodiment, the same data can be obtained by directly capturing the sine waves V0(sin(ωt)) and V1(ksin(ωt+θ)) instead of V2(sin(ωt) -ksin(ωt+θ)) is obtained. NOTE: A high speed data acquisition card 35 can be used to digitize the signal V0 at position 21 and the signal V2 at position 20, thereby maintaining a digital representation of the waveform for further digital signal processing. Note: the values of Z and Φ obtained from the sensor measurement unit 60 and the respective voltages from which the values of Z and Φ are derived (such as V0 and V2, or alternatively, V0, V1 and V2) will hereinafter be referred to as "impedance Signal Data".

Subsequently, once provided with the digital hold signals of V0 and V2, measurements of quantities k (attenuation) and θ (phase shift) are done by standard demodulation practices, as will be understood by those skilled in the art.

Once k and Θ have been measured, the determination of Z and Φ is done by analyzing the circuit depicted in Figure 5 as follows.

iI ^* ＝(V0-V1)/R

ii.Z=V1/I ^*

iii. Substitution, since V1=k<θ and V0=1

iv. Impedance (Z) ^* ＝R(k＜θ)/(1-k＜θ)＝Z＜Φ

v. As can be seen from the equations immediately above, the magnitude Z and phase angle can be easily derived from known values of R, k and θ.

vi. Convert the number of poles to a complex number to separate the real and imaginary parts, ie series resistance and reactance.

vii. Series reactance (Xs) = Z sinΦ = 1/wC, where w = 2πf

viii. Series resistance (Rs) = Z cosΦ

ix. Series capacitance (Cs) = 1/wXs

x. Series conductance (Gs) = 1/Rs

xi. Instead of a series model, impedance can also be modeled as a parallel combination of reactance (Xp) and resistance (Rp), as understood by those skilled in the art.

xii. The parallel capacitor (Cp) can be calculated from the series reactance and resistance as follows:

Cp＝-Xs/[w(Rs ² +Xs ² )]

xiii. The parallel resistance (Rp) can be calculated as follows: Rp＝-Xs/wCp Rs

xiv. The parallel reactance (Xp) can be calculated as follows: Xp=-1/wC

xv. Parallel conductance (Gp) = 1/Rp.

In various embodiments of the curing system 20, any time series of data pairs: (Z and Φ), (Rp and Xp), (Gp and Cp), (Xs and Rs) or (Gs and Cs) may be used to represent The result is solidified data (also known as impedance data flow).

In this disclosure, regardless of the type of circuit model used (such as the series model or the parallel model described above), the analysis of capacitance (C), conductance (G), reactance (X) or resistance (R) is generally done. quote. The impedance analysis performed by curing system 20 is the same regardless of which circuit model is used. That is, general references to C, G, R, and X apply equally to parallel or series data.

Additionally, it is also worth noting that the component cure monitoring, control, and adjustment capabilities of curing system 20 generally do not require sensor measurement unit 60 to be a non-bridging circuit. In particular, the curing system 20 disclosed in FIGS. 6 , 7 and 8 and described hereinafter does not require such a non-bridging circuit. Instead, the sensor measurement unit 60 may be a Wheatstone bridge or a substantially functional equivalent. Process for Curing Moldable Compound 16

The flowcharts of Figures 6, 7 and 8 illustrate the high level steps carried out in order to cure a part in which a plurality of sensors 17 are distributed within the body of the mold 18 for determining the rate of cure of the various parts of the part being molded within the mold. Figures 6 and 7 are flowcharts for determining initial cure values to be used, for example, to cure a part during production. In particular, FIGS. 6 and 7 may be viewed as techniques for calibrating the part curing process for a given mold and the type of moldable compound 16 to be provided to the mold. In step 1002 , the user interacts with cure setup subsystem 104 ( FIG. 1 ) to enter the type of part to be produced by curing apparatus 45 . The input may include: (i) identification of the moldable compound 16 to be provided in the mold 18, (ii) identification of the part configuration (e.g. identification of the mold 18 to be used, location of the in-mold sensor 17, identification of the part number of the part, and/or a description of the part to be produced such as part size and/or shape). In one embodiment, the input may include electronic computer-aided design files that provide a three-dimensional data model of the part and/or mold cavity 24 to be cured. NOTE: When such a three-dimensional data model is provided, it is possible to perform a moldable compound flow analysis of how the moldable compound flows into the mold 18 . In step 1004, (if not provided in step 1002) locations are determined for providing sensor 17 (or other capacitive detector) within mold 18 such that the sensor can generate impedance data flow during curing of the part. The locations are generally provided at a plurality of substantially different regions within the mold cavity 24 . For example, the sensors 17 may be placed in the mold 18 such that they are: (a) spaced apart at or adjacent to the ends of the mold cavity 24 (e.g., at least 2/3 of the largest dimension of the part), (b) placed in the mold Locations in cavity 24 for obtaining impedance data for regions of components having substantially different thicknesses (e.g. capacitor 68 dielectric thicknesses differing by about 25% or more), (c) placed adjacent to significant bends in mold cavity 24 (e.g., a bend greater than 30 degrees), (d) placed at a location of the mold cavity 24 relatively far from where the moldable compound 16 is initially introduced (e.g., placed or entered) into the mold cavity 24 (e.g., a sensor 17 is placed at a distance of at least 2/3 of the maximum distance the moldable compound must flow from where it is introduced into the mold cavity during curing), and/or (e) is placed in or adjacent to the mold cavity therein 24 where the flow of the moldable compound 16 is a bottleneck. In step 1008, a plurality of samples of the moldable compound 16 are cured in the mold 18 used to form the test part, wherein the samples are cured at various curing parameters for establishing different curing environments. The test parts are then evaluated to determine part quality and the type and extent of part defects, if any. Variations in curing parameters can be as follows:

(a) Variation in mold 18 tonnage; this variation may be within ±5% of the expected typical tonnage for properly forming and curing a part in mold 18.

(b) Variation in mold 18 curing temperature; this variation may be within ±10% of the expected typical temperature for properly forming and curing a part in mold 18. Note, however, that curing apparatus 45 may allow different portions of mold cavity 24 to have different temperatures. Thus, the sample can be cured in different parts of the mold cavity 24 at different temperatures. NOTE: It is not necessary to consider all theoretical combinations of temperature combinations, as typically mold operators and others skilled in the art will have sufficient expertise to recognize the relatively small amount of cure temperature variation to be tested. For example, where the mold cavity 24 is relatively thin or narrow in a particular direction (e.g., less than 20% of the largest mold cavity 24 dimension in a particular direction), it may be possible at -15% of a particular predetermined curing temperature to + Curing temperature was tested in the range of 10%. Alternatively, for relatively thick or wide portions of the mold cavity 24, the curing temperature may be tested within a range of -5% of a particular predetermined curing temperature to +15% of that temperature. In addition, when the moldable compound 16 needs to flow along a relatively extended flow path to fully form the part, it may be tested along one or more of these in the range of -5% of a specific predetermined curing temperature to +15% of this temperature. The middle temperature of the flow path.

(c) Variation in cure time in mold 18; for example, the variation may be within ±10% of the expected typical time for properly forming and curing a part in mold 18.

(d) Variation in the rate at which tonnage is applied to mold 18; for example, the variation may be within ±10% of the expected typical tonnage application rate for properly forming and curing a part in mold 18.

NOTE: Expected cure times can be determined by a number of techniques, including the use of one or a combination of: (i) cure operator expertise, (ii) cure data captured from the cure of similar parts (e.g. from the same or similar available molding compound 16, cured in a mold cavity 24 of similar shape and size, and cured using the same curing equipment 45 components such as ton presses, temperature sensors and regulators, etc.), (iii) computational simulation of the curing process , (iv) "intelligent" systems such as expert systems, having heuristic rules encoded in them, where the rules represent solidified expert domain knowledge, and/or (v) trial and error. Also, note: if, for example, it becomes apparent that curing equipment 45 is curing parts significantly differently than previously experienced (e.g., one or more curing equipment components may have been replaced, causing the curing equipment to behave differently than previous part production runs), The expected curing time may be adjusted or changed during step 1008 .

In addition to variations in the curing environment, variations in the moldable compound 16 can also be tested. For example, samples from different batches can be tested (as used herein, the term "batch" means a certain number of moldable compounds 16 from which parts are produced, where a batch is assumed to be within the moldable compound's substantially homogeneous in composition). In particular, samples from different suppliers, produced at different times or using different facilities may be tested. Additionally, samples from batches with known or unknown compositional variations can be tested.

Thus, to perform step 1008, a matrix of possible combinations of environments and variations ("batches") of moldable compound 16 may be determined, and from this matrix a particular combination may be selected for testing in this step. In one embodiment, the selection process can be automated by (i) computational simulation of the curing process, and/or (ii) an "intelligent" system, such as an expert system, with heuristic rules encoded therein, wherein the rules Represents solidified expert domain knowledge.

For each sample tested in step 1008, a stream of impedance data is obtained from the plurality of sensors 17 in the mold 18 and stored for subsequent analysis as described below.

In step 1010, for each batch of moldable compound 16 tested in step 1008, the collected impedance data streams are statistically analyzed to determine one or more impedance data stream characteristics (e.g., specific characteristics of the curing cycle). the slope value of the corresponding process curve at the part, or to identify when a local maximum or minimum is reached, etc.), the characteristic: (i) effectively correlates with the characteristics of a properly formed and cured part, and/or (ii ) are effectively associated with undesirable features of poorly formed or defective parts. Specifically, the following steps can be performed in this analysis:

(a) one or more impedance data flow characteristics (e.g., slope) of the process profile at one or more times during the part curing process (e.g., when the mold 18 is opened),

(b) If the desired component characteristic (e.g. tensile strength) can be measured numerically, determine a statistical correlation (e.g. linear regression) between:

(1) The impedance process curve characteristic, for example, the slope at one or more times during the part curing process, such as the time immediately before opening the mold 18, or the first slope value range and the second slope times at transitions between value ranges, and

(2) Measurement of part characteristics of the resulting part.

(c) Assuming such association is statistically significant (eg, ^R2 > 0.6), the impedance process profile characteristics can be used to determine whether desired characteristics are obtained in subsequently produced parts (eg, additional test parts).

For example, if one of the desired part characteristics has a pass/fail criterion associated with that part, and if one or more characteristics (slope values) of the process profile can be programmed with a first (slope) range indicating that the part passes and a second set having a second (slope) range indicating that the part does not pass, the characteristic (slope value) of the process profile can be monitored throughout at least a portion of the part cure cycle to determine when (or if) each The slope of the process curve for each sensor 17 transitions from the second (slope) range of no pass to the first (slope) range of pass. For example, assuming that the only desired part feature is a fully formed part that is substantially void-free, and within a predetermined maximum allocated part cure time, when the slope of the process curve from all sensors 17 transitions from the predetermined no-pass range to the predetermined pass range An appropriately significant statistical association (linear or otherwise) for this desired characteristic occurs when . Thus, when this maximum allotted time passes (thus, the resulting part can be identified as defective), or the impedance data streams from all sensors have slopes within the first range (thus, the resulting part can be identified as conforming ), the curing controller 43 may output a signal (via line 28, FIG. 1 ) to the curing device 45 to open the mold 18.

Returning now to the steps of FIG. 6 , and in particular to step 1014, once the associations are determined, these associations can be used to derive operations and/or conditions (e.g., embodied in program agents such as daemons, and/or or executable expressions), which in turn can be accessed to monitor and control the curing process before and during production part runs. In particular, the program agent and/or executable conditions may be:

(i) for setting the curing parameters to typical or "normal" values for the type of part to be produced (e.g. for the moldable compound 16 to be used, and for the desired characteristics in the resulting cured part);

(ii) for evaluating part molding and curing while the part is in its mold 18, and adjusting curing parameters (when undesired characteristics of resistive data flow are detected) so that the part is more likely to be defect-free; and

(iii) Used to determine the end of cure of the part.

As described below, the conditions and/or operations (eg, their stored program agents and/or conditions) can be accessed to monitor and/or affect subsequent part curing processes. Additionally, the procedural agents and/or conditions may be closed form equations (eg, linear regression equations), iterative procedures, or "IF THEN" rules as may be instantiated in an expert system rule base. Those skilled in the art will appreciate that in some embodiments, the agent may be implemented as a daemon.

FIG. 7 shows a more detailed description of an embodiment of step 1014 . In step 1104 of this figure, as in step 1008 of FIG. 6, information identifying the first batch having the sample tested is determined (this information is denoted here as "B"). Then in step 1108, information identifying the first sample tested from lot B (the sample identification information denoted herein as "S") is determined. In step 1112, it is determined whether the sample S is molded and cured into a part of proper quality. If so, then (step 1116 ) at least one in-mold curing condition associated with termination of the sample S in its mold 18 is determined. Note that a typical cure termination condition for a properly cured part is an extended portion of the process curve for each sensor 17 with substantially zero slope (e.g., in the range of -0.1 to +0.1) near the end of the part cure cycle, e.g. This extension lasts at least about 5% to 10% of the expected cure time. Then, in step 1120, it is determined whether another sample from batch B is to be examined. If another such sample is tested, step 1108 is performed again to obtain (as S) information identifying the next sample of batch B.

Conversely, if it is determined in the performance of step 1112 that the sample did not yield a part of suitable quality, then in step 1124 it is determined whether there is a correlation between the characteristics of the impedance data streams of the plurality of sensors 17 and at least one defect of the part. There are various techniques that can be used to perform this determination. In one technique, the resulting sample parts and corresponding plurality of impedance data streams (or their corresponding process curves) are manually inspected by those skilled in the curing art to identify such correlations. In particular, the impedance-related properties of a group of two or more sensors 17 may vary sufficiently to identify it manually. In another technique, the detection of part defects and their association with variations between process profiles may be statistically identified by cure setup subsystem 104 . For example, each test part may be evaluated and identified as one or more of the following: (i) a non-defective part, (ii) a defective part with a void, (iii) a defective part with a void, (iv) Under-cured parts, (v) over-cured parts, (vi) not well-formed parts, and/or (vii) defective parts due to not having desired properties, e.g. the desired range is one of the following: Tensile Strength , compressive strength, dynamic hardness, dimensional consistency. Subsequently, for each of the identifications (ii) through (vii), the process curve for each test part with that identification may be evaluated to determine abnormal characteristics that are not present in the parts identified as non-defective. In particular, for each identified defect type, and for each test part with this defect type, a difference or variation between process profiles (from different sensors 17) can be determined for that part (this characteristic is referred to as "part Intra-differential properties"). For example, the within-part differential characteristic can be one or more of the following: (1) a difference in the slope of the process curve over a specified time frame in the cure process (e.g., the last third of the cure time), (2) a difference in the slope of the process curve Differences in maximum and/or minimum values, and (3) differences in cure time values for corresponding process profile characteristics (eg, differences in maximum and/or minimum values between process profiles of different sensors 17, etc.). Then, for each part defect type, one or a combination of the intra-part differential properties can be sufficiently associated with the part defect type such that the intra-part differential properties can be used to: (i) alter the part curing environment prior to a part production run, Thereby reducing the type of part defect during production, and/or (ii) changing curing parameters during part curing (eg shortening or extending the in-mold time used to cure the part) during part curing where intra-part differential properties occur. An example of this variation between the impedance data streams (or their process curves) of different sensors 17 corresponding to or associated with component defects is shown in Table A below.

Form A Unusual intra-part variation between impedance data from different sensors possible associated component defects The process curves for two or more sensors 17 are similar in shape but are offset from each other in cure time by an excessive amount, such as at least 10% of the total expected cure time; e.g., for one of the process curves for sensor P ₀ For each impedance value Z ₀ , there is a corresponding impedance value Z _i for each of the other sensors P _i , where Z _i is both within 1% of the maximum change in expected impedance value from value Z ₀ and occurs at Z ₀ within 1 second of . voids in parts The process curves of two or more sensors 17 are similar in shape but offset from each other by an excessive amount in impedance magnitude, for example the maximum value of one process curve is at least 50% greater than the maximum value of the other process curve. voids in parts For two or more sensors 17, the maximum slope of their corresponding process curves between the onset of curing and its maximum value differs by more than 25% voids in parts For two or more sensors 17, the maximum slopes differ by more than 25%, where for each process curve each maximum slope is the maximum value between the maximum value of the process curve and the point of maximum rate of change of the process curve. Localized areas of the part may be undercured, for example, the part may be porous, ie the part has portions where there is gas trapped within a large number of small air pores (bubbles). Alternatively/additionally, the part may develop blisters, i.e. after the part has been demolded,

Localized swelling or raised air bubbles on a face.

In another embodiment for determining (in step 1124) whether there is a correlation between the characteristics of the impedance data streams of the plurality of sensors 17 and at least one defect in the part, the intra-part differential characteristics may be different from previous Comparison of within-part differential properties obtained in sample test (and/or part production run) collections, wherein the resulting sample parts are molded from similar moldable compounds 16 and such resulting parts differ in shape and size above similar. In particular, embodiments of the curing methods and systems disclosed herein can collect large repositories of resistive flow data and/or intra-part differential properties over time. For example, for each of a plurality of different types of parts previously produced (e.g., through previous sample testing and/or part production runs), the intra-part differential characteristics and/or impedance data streams for each produced part may be The collection is archived with information associated with:

(i) part type characteristics (such as the moldable compound used16, which may include consistency or variation in compound composition, part shape, part size variation, maximum and minimum part extent, and other indications of part size and/or volume ),

(ii) the number and relative locations of the plurality of sensors 17 used to obtain impedance data streams (e.g., the maximum distance around any sensor excluding another sensor, the maximum distance between any two sensors, and/or whether the sensors are adequate located within the mold to effectively assess the indication of part curing conditions),

(iii) curing parameters used to cure the part (e.g., curing parameters at each sensor, tonnage applied during part curing, rate of tonnage applied, in-mold curing time, curing equipment 45 used, mold 18 heat transfer rate, etc.), and

(iv) Resulting component characteristics (eg, detected component defects, component strength properties, component elastomeric properties, thermal conductivity of the component, etc.).

Thus, statistical and intelligent processing can be used to compare the characteristics of a new part to be produced with previously produced parts having similar characteristics in order to not only identify possible impedance data streams that may be associated with various potential part defects feature, and also assists in properly positioning the sensor 17 within the mold 18.

In another technique for determining (in step 1124) whether there is a correlation between the intra-part variability characteristic of the impedance data flow and at least one defect in the part, a computational simulation or model can be performed to determine a simulated version of the impedance data flow and whether the corresponding intra-part variance characteristics are likely to be associated with the part defects actually obtained from sample S.

If, in step 1124, it is determined that there is a sufficient predictive association with at least one part defect (e.g., at least about ^R2 > 0.6), then in step 1128, one or more cure process adjustments that will reduce the likelihood of this defect are determined . Such operations may be determined by one skilled in the curing art through statistical analysis and/or by simulating or modeling how certain operations are likely to affect the part forming and curing process. Then, in step 1132, the determined adjusted code is associated with the impedance data stream of sample S and stored. Note: Also associated with and stored with it is the identification of the moldable compound 16 and the batch identification of the moldable compound 16 . Of course, if there are additional such associations between: (i) one or more additional in-part differential characteristics, and (ii) a defect in the part, steps 1124 through 1132 may be repeated until no more are detected to such a connection.

Regardless of the outcome of step 1124, step 1136 is performed in which it is determined whether there is a single sensor 17 (hereinafter referred to as a "marked sensor") whose impedance data stream has a valid correlation with the likelihood of creating a defect in the part At least one characteristic of , wherein the correlation is substantially limited to the output from this single sensor (ie, a correlation detected here would not have been detected in step 1124). If an association is identified in step 1136, then (in step 1140) one or more curing process adjustments intended to reduce the likelihood of the occurrence of defects are determined, wherein such adjustments preferably substantially only affect A limited range of curing components (eg, such effects are preferably limited to the range of components that do not substantially change the curing parameters at any other sensors 17). An example of such a limited effect is to change the mold cavity 24 temperature only in the area including the identified sensor 17 . Another example of such limited influence may occur in the following situation: moldable compound 16 is injected into the mold 18 near the identified sensor 17, and the identified sensor 17 provides anomalous impedance data; or an increase in injection pressure close to the identified sensor 17 .

Then, in step 1144, the adjusted code determined in step 1040 is associated with the impedance data stream from sample S and stored. Note: Also associated with and stored with it is the identification of the moldable compound 16 and the batch identification of the moldable compound 16 . Of course, if there are additional such associations (for sample S) between: (i) impedance data stream from another identified sensor 17, and (ii) a defect in the part, the steps may be repeated 1136 to 1144 until such associations are no longer detected.

Following step 1144, step 1120 is encountered for determining whether there is another sample from the current batch B to be analyzed. An affirmative result of this step will cause step 1108 to be performed again. However, if there are no more samples from batch B, then in step 1148 the following is determined:

(i) a set of initial cure parameters to be used to begin curing the part from the moldable compound 16 having impedance properties similar to (or identical to) the samples from Lot B;

(ii) at least one part cure termination program agent and/or condition (for example such an agent or condition may be "after 4 minutes of curing at 340 degrees, open mold 18"); and

(iii) A set of one or more program agents and/or conditions for correcting abnormal impedance measurements obtained from curing of the part in mold 18 (eg, extending or shortening in-mold curing time).

Step 1148 may be viewed as a step of combining or synthesizing the results obtained from steps 1128 and 1140 such that such adjustments and cure termination criteria are based on multiple samples from Lot B. Note, however: In alternative embodiments, steps 1128 and 1140 may only identify abnormal impedance values (or process profile characteristics) without determining cure parameter adjustments. In this later embodiment, step 1148 determines curing parameter settings, for example, by first classifying each abnormal impedance data flow characteristic, where each class identifies a single abnormal curing condition, and then (i) for each Such a class determines a set of one or more composite process profiles, and then (ii) determines initial cure parameters, and one or more program agents and/or conditions (such as a predetermined cure temperature change or a predetermined cure time change) such that Use composite process profiles to terminate or adjust part cure. NOTE: The following are representative examples of program agents and/or conditions that may be determined by cure analysis subsystem 26 and subsequently used to evaluate impedance data streams (or process profiles) from multiple sensors 17:

(a) determining a maximum impedance value for each of one or more predetermined segments in the process profile;

(b) for each of one or more predetermined segments in the process profile, the time at which the maximum impedance value is determined;

(c) determining a minimum impedance value for each of one or more predetermined segments in the process profile;

(d) for each of the one or more predetermined segments in the process profile, the time at which the minimum impedance value is determined; and

(e) Determining, for each of the one or more predetermined segments in the process profile, an integrated area under a plot of the segment impedance value versus time.

Note, however, that other impedance-related measurements are also contemplated for use in various embodiments of the curing methods and systems disclosed herein, such as: (1) one or more process curve points identified by various derivation conditions (e.g. inflection point, etc.), (2) one or more coefficients of a polynomial fit to the segment of the impedance data stream, (3) the centroid (or its coordinates) of the area under the plot of the process curve segment , and/or (4) one or more coefficients of the higher order derivative of the process curve fit to the impedance data flow segment. Also included within the scope of the curing systems and methods disclosed herein are program agents and/or conditions that are not easily described geometrically, such as predicted curing output by artificial neural networks, fuzzy logic systems, or heuristic-based evaluators. time.

Then, in step 1152, it is determined whether there is another batch in which samples were tested and in which their corresponding impedance data streams (or process curves) were examined for association with component defects. If there are additional such batches, step 1104 and subsequent steps are performed again. However, if there are no such additional batches, then the flowchart of FIG. 7 ends.

NOTE: When performing the steps of Figure 7, if the sample moldable compounds 16 are not differentiated by batch, then all samples may be considered to be from a single batch. Thus, the result of a single execution of step 1148 is a composite collection of curing parameters for each part manufactured by curing apparatus 45 and moldable compound 16 .

For the moldable compounds 16 disclosed herein, the following table shows representative examples of various impedance data flow (or process profile) characteristics that can be determined to be indicative of and/or associated with particular part characteristics, and When such properties indicate an action (e.g., termination of cure of the part or a defective part being molded), the rightmost column indicates what action should be taken in order to: (i) adjust and continue curing of the current part, ( ii) terminating curing of the current part, and/or (iii) identifying adjustments to initial curing parameters for subsequent parts.

Form B Impedance Data Flow/Process Curve Characteristics and Its Determination Possible Causes of Curved Behavior Resulting part characteristics or part production characteristics adjustment or cure termination An extended "flat" portion is detected for each process profile, (e.g., a flat portion that extends at least about 5% to 10% of the expected cure time); The best fit 3rd order polynomial (or more generally, the best fit nth order polynomial, n ≥ 3) for modeling the impedance data is determined, and the first derivative of this polynomial (i.e. material is cured fully cured part Terminate part curing

Slope) yields a substantially zero slope (e.g., in the range of -0.1 to +0.1) near the end of the curing cycle. An unintended extended "flat" portion of one of the process curves, (e.g., a flat portion that extends at least about 5% to 10% of the expected part cure time); The best-fit third-order polynomial (or more generally, the best-fit nth-order polynomial, n ≥ 3) for modeling the impedance data is determined, and the first derivative (i.e., slope) of the generated polynomial yields essentially A slope of zero (e.g. in the range -0.1 to +0.1). After full tonnage, moldable compound 16 flows Voids in parts Increase the tonnage, and/or increase the amount of time at "full tonnage" before starting the cure timer, and/or for subsequent parts, decrease the mold closing speed Impedance of process curve in at least one pass void in the widget Increase the tonnage or note

Values that have not increased by at least 10% from the value at the start of curing, then the derivative of the best fit 3rd order polynomial (or more generally, best fit nth order polynomial, n ≥ 3) of the impedance data for the last 20 seconds Decreases (i.e. does not rise to a peak) before reaching essentially zero slope (e.g. in the range -0.1 to +0.1) Around sensilla 17, the moldable compound does not enter the solidified phase Gap Inject pressure, and/or reduce mold temperature, and/or (if possible, or for subsequent parts) add cure inhibitor to moldable compound 16. Terminate part curing if process profile characteristics persist. A fluctuation of at least one process profile during the expected first third of the part cure time, wherein such fluctuations in the process profile produce multiple pairs of local maxima followed by local minima, each pair in its member with process curve (or corresponding impedance data flow) between the maximum After full tonnage, moldable compound 16 flows Voids in parts Increase the tonnage, and/or, increase the amount of time at "full tonnage" before starting the cure timer, and/or, for subsequent parts, decrease the mold closing speed

A variance of at least 5% of the variation. For example, the mean square error between the impedance data and the best-fit 3rd order polynomial (or more generally, the best-fit nth order polynomial, n≥3) of the impedance data for the last 20 seconds is greater than the predetermined impedance value. At least one fluctuation of the process profile (such fluctuations as described above) during the time expected to be the last third of the part cure time, e.g., a best-fit 3rd order polynomial of the impedance data to the last 20 seconds of impedance data (or more generally, the mean squared error between the best fit polynomials of degree n, n≥3) is greater than the predetermined impedance value. Accumulation of gaseous by-products in the mold cavity 24 Voids and/or porosity in parts The gaseous by-products are released from the mold cavity 24 by increasing ventilation, venting (e.g., deflation) of the mold 18. The extension of one of the process curves around its peak Too much inhibitor added to moldable Added before demolding the part Increase mold temperature by 18 (at least in

A long "flat" portion (e.g., a flat portion that extends at least about 5% of the expected cure time); such a flat portion can be determined by generating an impedance data modeled about 20 seconds centered on the peak. A best fit polynomial of order 3 (or more generally, a polynomial of order n best, n ≥ 3), and then determine that the first derivative (i.e. slope) of this polynomial yields essentially A slope of zero (e.g. in the range -0.1 to +0.1). In compound 16 curing time or the portion of the mold cavity 24 close to the sensor 17 that provides a process profile with an extended "flat" portion, and/or (if possible, or for subsequent parts) to reduce cure inhibitors in the moldable compound 16 . According to generate e.g. best fit 3rd order polynomial (or more generally, best fit nth order Different areas on the part are completing their cure at different times Increased curing time before demoulding the part Increase the temperature of the local area, and/or (if possible, or for subsequent parts) change the moldable compound16

polynomial, n≥3), and the first derivative of the polynomial (i.e., the slope) measures a substantially zero slope (e.g., in the range of -0.1 to 0.1) for at least 15 seconds, from at least one of the sensors 17 The process curve is flattened, whereas the process curve (or more precisely, the generated nth order polynomial) from at least one other sensor 17 does not reach a substantially zero slope. The way of introducing into the mold cavity 24 At least one process profile from one of the sensors 17s ₁ is flattened for at least 15 seconds while the process profile from at least one other sensor 17s ₂ is not flattened (i.e. expected cure time in the last 1/3 of that time Solidification does not occur substantially simultaneously throughout the part. Increased cure time before demolding the part (e.g. increasing in-mold time until all process profiles are flattened), and/or voids in the part with higher mold Increase the temperature of the localized area, and/or (if possible, or for subsequent parts) alter the introduction of the moldable compound 16 into the mold cavity 24

To within 1/4, and at least 15 seconds before the process curve of s ₂ similarly flattens, the process curve from s ₁ has acquired a slope in the range -0.1 to 0.1. temperature

Examples of at least some of the correspondences in Table B are illustrated in the examples provided in the Examples section below.

As noted above, the steps of FIGS. 6 and 7 may be performed through collaborative interaction between cure setup subsystem 104 ( FIG. 1 ) and one or more curing operators (or other curing specialists). Curing parts during production runs

FIG. 8 is a flowchart of one embodiment of the steps performed by curing system 20 when curing a part, particularly during mass production of parts. The steps of this figure are typically performed after the flow charts of FIGS. 6 and 7 have been performed. Thus, it is assumed that the cure control database 27 has been properly populated with data indicative of initial cure parameter settings, cure adjustments and cure termination conditions. In step 1204 (if necessary), the curing operator interacts with the operator interface 32 (FIG. 1) to input information about the parts to be produced from the curing apparatus 45 to the control system 39 (and specifically to the curing controller 43). type. Such input may include: (i) identification of the moldable compound 16 to be provided into the mold 18, (ii) identification of the part configuration (e.g., identification of the mold 18 to be used, location of the in-mold sensor 17, Identification of the part number of the part to be produced, and/or a description of the part to be produced, such as part size and/or shape). In one embodiment, such input may be in the form of an electronic computer-aided design file providing a three-dimensional data model of the part and/or mold cavity 24 to be cured. In step 1208, it is determined whether the curing parameters can be set according to the moldable compound lot identification. If not, then in step 1212 the initial curing parameters are set to the composite settings as discussed above with reference to step 1148 in FIG. 7 . Alternatively, if the initial curing parameters can be set according to the batch, then in step 1216, a selection is made for batch (B), wherein the sample test of batch (B) (for example, according to FIGS. 6 and 7 ) has Or likely to have the impedance data flow characteristic most similar to the impedance data flow characteristic of the current batch from which the part is to be produced. Note: Step 1216 may be performed by curing (in the curing device 45) a small quantity of parts from the current batch and comparing the resulting impedance process curve with the corresponding process curve stored in the cure control database 27 (for available Each previous test batch of molding compound) to determine the closest match. Additionally or alternatively, such matching may be performed by comparing the composition of the current batch to the composition of a previous test batch. Thus, once Lot B is selected, in step 1220 the initial curing parameters for Lot B are set as the initial curing parameters for the current lot. It should be noted, however, that it is also within the scope of the present curing methods and systems to interpolate (or otherwise combine) such initial curing parameters from the initial curing parameters of multiple previous test batches. For example, when previously tested batches _B1 and _B2 both appear to be likely candidates for selection, initial cure parameter values from these two batches can be obtained by, for example, corresponding initial cures from batches _B1 and _B2 A mean, median or weighted sum of parameter values (where the weights may be combined according to the perceived proximity of each of these batches to the current batch).

Regardless of the outcome of decision step 1208 , step 1224 is executed in which a signal is received by cure controller 43 indicating that the part is being cured by curing apparatus 45 . Subsequently, in step 1228, the sensor measurement unit 60 (e.g., one for each sensor 17) begins providing the initial impedance data stream portion of at least one impedance data stream from each of the plurality of sensors 17 to the computer 34 (FIG. 1 ). . More specifically, each sensor 17 provides the impedance signal to its corresponding sensor measurement unit 60, and the sensor measurement unit 60 provides the corresponding impedance signal to the data acquisition card 35, and the data acquisition card 35 provides its corresponding output to the solution. The demodulation component 42, and the demodulation component 42 outputs the corresponding obtained impedance data stream to the curing data capture database 23. In addition, the demodulation component 42 notifies the curing controller 43 that there is an impedance data stream to be evaluated (or more precisely , its part). The cure controller 43 then notifies the cure analysis subsystem 26 that there is (a portion of) the impedance data stream to be evaluated. The cure analysis subsystem 26 then retrieves the impedance data streams from the cure data capture database 23 and performs a data smoothing operation on each impedance data stream (portion thereof) to obtain data for the corresponding process profile. Subsequently, the cure analysis subsystem 23 evaluates the process profile data for determining the status of the part cure process. The cure analysis subsystem 26 may evaluate such a process profile using a number of program agents and/or various executable conditions as previously identified in Tables A and B in order to determine when to cure a part: (1) properly formed and Curing, (2) well formed and fully cured, and (3) inspection of process profile data for parts that do not indicate well formed, properly cured. Consequently, the cure analysis subsystem 26 will eventually issue a notification to the cure controller 43 that (i) the part being cured is expected to be well formed and cured correctly, or alternatively, (ii) at least one anomaly in the process profile has been detected , which may indicate that the part was formed and/or cured incorrectly.

Thus, in step 1232, upon receipt of such a status notification from the cure analysis subsystem 26, the cure controller 43 determines whether the notification indicates that a well formed and properly cured part has been produced, or that the part was improperly formed and cured. / or improperly cured. Thus, if it is determined that the part is correctly formed and cured, then in step 1236 cure controller 43 outputs a command or instruction (via line 28 ) for instructing curing apparatus 45 to open mold 18 and release the part therein. Then, in step 1240, the cure controller 43 waits for input (via line 29) indicating that a new part is being cured in the mold 18, or that there are currently no more parts to be cured. Thus, upon receipt of such input, cure controller 43 determines (step 1244) whether to stop the curing process within computer 34 (i.e., perform step 1248) or to stop the curing process due to receipt of input on line 29 indicating that another part is to be cured. Continue the curing process. Note: In the latter of these two alternatives, step 1208 and subsequent steps are performed again.

Alternatively, if in step 1232 it is determined that the part may be improperly formed and/or cured, then in step 1252 cure controller 43 alerts the curing operator via operator interface 32 that the current part may be defective. Then, in step 1256 , the curing controller 43 determines whether at least one operation can be identified to reduce the detected anomaly in the impedance data received from the plurality of sensors 17 . NOTE: Such determinations by cure controller 43 may be made using input from cure analysis subsystem 26 . In particular, along with notification from cure analysis subsystem 26 indicating that the current part may be improperly formed and/or cured, the cure analysis subsystem may also provide an indication of one or more corrective adjustments to perform curing on the current part . Table B above provides a representative example of some corrective adjustments that may be performed. If cure analysis subsystem 26 provides identification of one or more such corrective adjustments, then in step 1260 cure controller 43 selects (or more generally identifies) one or more corresponding commands or instructions to send to the cure A device 45 (via line 28 and at least one input device 37) for performing corrective adjustments. NOTE: In at least some embodiments, cure controller 43 may select commands or instructions for all such adjustments identified by cure analysis subsystem 26 . However, it is within the scope of embodiments of curing system 20 that curing analysis subsystem 26 may provide sequencing of such corrective adjustments such that curing controller 43 may issue such commands or instructions in a particular order. Subsequently, in step 1264 , the curing device 45 executes the received command or instruction for adjusting a curing parameter (eg, an adjustment to curing time), and then encounters step 1128 again.

In at least some embodiments of curing system 20, curing analysis subsystem 26 may have a number of possible corrective adjustment alternatives that may be performed. Additionally, impedance data indicative of incorrect part forming and/or curing may be localized to a particular range at or near one or more (but not all) of the sensors 17 . For example, with three sensors 17 installed in the mold 18, if all three sensors 17 indicate that the part is in the last third of its cure cycle, and a particular one of the sensors shows fluctuations in its process curve, then This condition may be an indication of gaseous by-products trapped in the component at or near that particular sensor, which may eventually lead to porosity in the portion of the component close to that particular sensor. Thus, in order to select (or rank) such corrective adjustment alternatives, cure analysis subsystem 26 may prioritize the alternatives that have the least impact on the overall cured part. In at least some curing environments, this means that it is preferable to select corrective adjustments that affect components substantially only in the vicinity of the sensor or sensors outputting anomalous impedance data. For example, in the above example where a localized component anomaly (i.e. trapped gaseous by-product) is identified, the following corresponding localized corrective adjustments may be preferred over other alternatives: (i) by cleaning the Any blockages to locally reduce gaseous by-products, or (ii) (for subsequent part production) create additional vacuum ports near that particular sensor. In particular, the above alternatives may be preferred over a wider range of part correction adjustments: globally reducing gaseous by-products by venting the entire mold 18 to correct the current part, or (for subsequent part production) at Venting the entire mold near the start of the cure cycle will increase the part cure cycle time.

Examples and Case Studies

Figure 9 shows a typical SMC (polyester, styrene monomer) impedance data flow with time in seconds on the x-axis and relative conductance on the y-axis. Also shown are various points at which curing events occur (or are expected to occur). FIG. 9 shows data points (and corresponding smoothed process curve 1404 ) for cured impedance data flow values for a part made of SMC as moldable compound 16 . Specifically, the x-axis represents the percentage of time that the part is expected to be cured without using the curing system 20 . NOTE: With (i) the embosser closing the mold 18, (ii) the SMC coming into contact with the sensor 17 from which the impedance data stream is obtained, and (iii) this sensor being electrically coupled to the corresponding grounded capacitor plate 64, the process curve is initially when rising. The process curve 1404 continues to rise as the moldable compound 16 begins to soften, whereby ions and molecular entities in the moldable compound are more able to move within the electric field of the sensor. As the moldable compound 16 reaches the freeze point (ie, the time of highest crosslinking rate), the process curve 1404 exhibits a "peak" (at approximately point 1408). After peak 1408, the impedance value drops rapidly as the polyester and styrene react and the cross-linking restricts the movement of ions and molecular entities within the sensor's electric field. The process curve 1404 then "winds up" at flat line conditions (approximately at point 1412) as the remainder of the styrene-styrene reaction occurs. Thus, assuming that the cure analysis subsystem 26 recognizes such a flat-line condition relatively quickly, and detects that the corresponding process profiles of each of the plurality of sensors 17 enter such a flat-line condition at substantially the same time, it can be achieved by using the cure system 20 to save about 20% of curing time. To perform this analysis of process profiles 1404, cure analysis subsystem 26 may first identify peaks 1408 for each process profile. Once this point is identified, the cure analysis subsystem 26 then iteratively calculates the slope of successive portions of each process profile 1404 until a series of slope values approaching zero are determined for each process profile (such slopes are indicative of detected transition to the flat line condition). NOTE: The appropriate slope (or sequential set of slopes) to end curing of a particular part can be determined empirically by evaluating or observing one or more of the resulting part characteristics. For example, blistering before or after part post-bake is often used on SMC parts to identify the point of sufficient cure, as will be understood by those skilled in the art.

10 shows a graph of a typical impedance data flow (and its corresponding process curve 1504) for an instrument panel of a light truck manufactured by SMC, where the x-axis time is expressed as a percentage of the curing rate before using the curing system 20. . The impedance data flow graph in FIG. 10 differs slightly from that in FIG. 9 because approximately 30% of the way through the part's expected cure time (i.e., is shown starting at point 1508), the The pressure clamp pressure changes. Cure system 20 is configured such that cure analysis subsystem 26 identifies such a flattening in the slope of process curve 1504 despite the increase in impedance value starting at point 1508 . By analyzing sample test parts (eg, according to the procedure in Figure 6), the appropriate slope to end part cure is empirically determined by identifying when blistering begins to occur. Program agents using this appropriate slope setting are provided to the cure analysis subsystem 26 for use in determining earlier termination of part cure than previously used. The cure analysis subsystem 26 uses the process curve point, typically at 1512, to open the molder. Embodiments of curing system 20 configured in this manner reduced part curing time by an average of 18% over more than two months of substantially continuous operation.

Figure 11 shows a portion of a mold cavity 24 for curing automotive body panels, with three sensors 17 (ie sensors 17a, 17b and 17c) shown in the side walls of the mold cavity. Also shown is the location where multiple layers of SMC (as moldable compound 16 ) are provided in the mold cavity 24 for curing into the charge pattern of the automotive body panel component. Normal curing temperatures for this part (prior to using the curing methods and systems disclosed herein) are known to be 300 degrees Fahrenheit for the lower portion 1604 of the mold (i.e., the area approximately below the dotted line) and 300 degrees Fahrenheit for the upper portion 1608 of the mold ( That is, the area approximately above the dotted line) is 310 degrees Fahrenheit. Normal cure time (before using the curing methods and systems disclosed herein) is 105 seconds per part. In order to determine the effect on the flow of impedance data from the sensors 17a-17c caused by changes in mold cavity 28 temperature, the temperature was intentionally varied by ±15 degrees Fahrenheit from the normal 300 degrees Fahrenheit. Impedance data streams were collected for multiple part cures at multiple temperatures ranging from 285 to 315 degrees Fahrenheit. Figure 12 below shows a typical plot of the impedance data flow resulting from curing at normal temperature (300 degrees Fahrenheit), 285 degrees Fahrenheit, and 315 degrees Fahrenheit. Subsequent evaluation of the test parts showed that adequate cure at rated conditions was achieved at approximately 65 seconds. Additionally, Figure 12 shows that the impedance data flow shifts to the right as the curing temperature decreases. This shift reflects slower melting and reaction rates as the temperature is lowered, as expected. However, for each temperature, the slope of the corresponding process curve approaches zero when the part (or portion thereof) is properly cured. Thus, by incorporating this information into program agents and/or conditions accessible by cure analysis subsystem 26, embodiments of cure system 20 are able to reduce the average part cure time to 64 seconds.

13 shows a graph of typical impedance data flow (and their corresponding process curves) for the three sensors 17a, 17b, and 17c shown in FIG. The material layout is shown in Figure 11. Note: Different parts of the part fully cure at substantially different times as indicated by the different times for each process curve to reach a process curve slope near zero. In fact, tests indicate that the moldable compound 16 (SMC) solidifies more rapidly near the sensor 17c. Thus, during execution of the steps in Figure 6, an additional step is performed in an attempt to provide initial cure conditions that will bring the cure rates of the different parts of the part closer to each other. Specifically, the charges in FIG. 11 are rearranged such that at least some of the charges are closer to sensor 17c (ie, about 6 inches closer). FIG. 14 shows a graph of a typical impedance data flow (and its corresponding process curve) after repositioning the SMC charge in the mold cavity 28 . Furthermore, it can also be seen in Figure 14 that the freezing points are closer to each other for these three sensors.

While various embodiments of the invention have been described in detail, it will be apparent to those skilled in the art that modifications and adaptations can be made to these embodiments. However, it should be clearly understood that modifications and adaptations are within the scope of the invention as set forth in the appended claims.

Claims (17)

1. A method for curing multiple instances of a part, comprising: providing a plurality of sensors in a mold for curing said part instances, wherein each sensor is operable to generate a signal related to an impedance within a current one of said part instances cured in said mold; for each sensor, receiving a corresponding time series of impedance measurements for measuring said signal generated by said sensor while said current part instance is curing; determining, for a first one of said sensors, at least first impedance-related data obtained from a corresponding time series of said impedance measurements; For a second one of said sensors, determining second impedance-related data from a corresponding time series of said impedance measurements; identifying an association between: (a) a relationship between said first and second impedance-related data, and (b) status information indicative of at least one characteristic of said current component instance; When the state information indicates a defect in the current component instance, obtaining hardened data indicating at least one corrective action for the current component instance or one of subsequent ones of the component instance; and At least one instruction obtained from the curing data is transmitted to a curing equipment component for curing the part instance in the mold, the transmitting causing the curing equipment component to change curing conditions in accordance with the corrective operation. 2. The method of claim 1, wherein the first sensor provides first impedance data for a first portion of the component instance and the second sensor provides a second impedance for a second portion of the component instance data wherein: (a) the first portion and the second portion differ in thickness by about 25 percent or more; (b) at least one of the first and second sensors is adjacent to the component instance where (c) the first and second sensors are spaced apart by at least two-thirds of the largest dimension of each component instance, and (d) the first and second sensors are spaced apart by a moldable compound At least two-thirds of the maximum distance it must flow from where it is introduced into the mold during curing. 3. The method of claim 1, wherein at least one of the first and second sensors comprises two electrodes insulated from each other. 4. The method of claim 1, wherein the corresponding time series of the impedance measurements of the first and second sensors are determined using at least one non-bridging circuit. 5. The method of claim 1, wherein the corresponding time series of the impedance measurements of the first and second sensors are determined using at least one bridge circuit. 6. The method of claim 1, wherein said impedance measurement comprises one of: (i) impedance (Z), which is a measure of the total resistance to current flow in an AC circuit, (ii) phase angle, ( iii) resistance, (iv) reactance, (v) conductance, and (vi) capacitance. 7. The method of claim 1, wherein the part instance is cured from one of: (a) a polymeric moldable compound, (b) a styrene monomer compound, (c) a phenolic material, and (d) thermoset plastic. 8. The method of claim 1 , wherein the first impedance-related data comprises one of: (a) a portion of the time series of impedance measurements from the first sensor having a slope that is substantially zero identification of, (b) information related to whether the value of the time series of impedance measurements of the first sensor has increased by at least 10% from a value substantially at the beginning of curing the current part instance, and (c) Information relating to whether the values of the time series of impedance measurements produce a plurality of fluctuations of at least 5% of the maximum variation of the time series of impedance measurements. 9. The method of claim 8, wherein the second impedance-related data comprises a change from the first impedance-related data, the change being associated with a defect in the current component instance. 10. The method of claim 1, wherein the determination of the correlation is based on a predetermined relationship between the identification of the defect and a plurality of time series of impedance measurements for each of a plurality of previously cured instances of at least one part type. relevant. 11. The method of claim 1, wherein the status information includes an identification of the at least one characteristic that is one of: (i) the desired range of one of the following: tensile strength, compressive strength, dynamic hardness, and dimensional consistency, and (ii) an indication of one or more undercured conditions. 12. The method of claim 1, wherein the defect identifies the current part instance as one of: (i) a defective part with a void, (ii) a defective part with a void, (iii) Under-cured parts, (iv) over-cured parts, (v) not well-formed parts, and (vi) due to not having one of the desired properties such as tensile strength, compressive strength, dynamic hardness, dimensional consistency as desired Defective parts caused by range. 13. The method of claim 1, wherein corrective operations include one of the following: (a) changing the cure time of the current part instance or one of subsequent part instances in the part instance, (b) changing the cure temperature of the current part instance or one of the subsequent part instances, (c) change the tonnage of the current part instance or one of the subsequent part instances, (d) clean the vacuum port , and (e) add one or more vacuum ports. 14. The method of claim 1, wherein said transmitting step comprises generating said at least one instruction in a computer, wherein said computer performs the process of determining a preference between said at least one corrective operation and a different corrective operation. step. 15. The method of claim 14, wherein said preference is dependent on the range of said components affected by each of said at least one corrective operation and said different corrective operation. 16. An apparatus for curing multiple instances of a part in a mold, wherein said mold has a plurality of sensors, each sensor for generating a Impedance-related signals, including: for each sensor, a corresponding sensor measurement unit for providing one or more electrical signals to said sensor and for obtaining corresponding impedance related data from said sensor while said current part instance is curing; The first one or more components that determine: (a) first impedance-related data obtained from corresponding impedance-related data from a first of said sensors; and (b) second impedance-related data from corresponding impedance-related data of a second of said sensors; wherein said first and second impedance data are each the result of a predetermined calculation comprising determining at least one of: slope, local maximum and local minimum, substantially flat curve range, area, The rate of change of the slope and the point of inflection; A second one or more components for identifying an association between: (a) a difference between said first and second impedance-related data, and (b) at least status information for a feature; a controller for obtaining information indicative of at least one corrective operation for correcting a defect in said current component instance or one of subsequent ones of said component instances, wherein said information follows said Changes in status information; and A third one or more components for communicating at least one instruction issued by said controller to a curing equipment component for curing said part instance in said mold, said communicating causing said curing equipment component to follow The corrective operation alters curing conditions, wherein the at least one instruction is obtained using the information indicative of the at least one corrective operation. 17. The apparatus of claim 16, wherein the second one or more components include a data repository storing a plurality of associations for associating: (a) data from two or more a difference between the impedance-related data of each sensor in the set of a plurality of said sensors, and (b) at least one characteristic indicative of a part instance previously cured using at least one of said mold and said curing apparatus assembly status information.

Images