WO2024118692A1 - Universal process control for injection molding - Google Patents

Abstract

Methods to fill a mold cavity of an injection molding apparatus with molten thermoplastic material during an injection molding process may include: identifying parameters for the process, associated with the mold cavity, molten thermoplastic material, or type of injection molding apparatus; providing an expected process condition at a pre-determined time during the process based on the parameters; measuring an actual process condition of the apparatus at the pre-determined time; based on a difference between the expected and actual process conditions, providing a predetermined curve of a target process condition over time for a remainder of the process; measuring an experienced process condition of the apparatus over time for the remainder of the process; iteratively comparing the experienced process condition to the target process condition at a given time on the predetermined curve to detect a variance; and automatically adjusting the process so that the variance remains within a threshold.

Description

UNIVERSAL PROCESS CONTROL FOR INJECTION MOLDING

CROSS-REFERENCE TO RELATED APPLICATIONS

[0001] The present disclosure claims priority to U.S. Provisional App. No. 63/429,067, filed November 30, 2022 and entitled “Universal Process Control for Injection Molding,” the entirety of which is incorporated by reference herein.

FIELD OF THE DISCLOSURE

[0002] The present disclosure generally relates to injection molding and, more particularly, to a universal process control for injection molding.

BACKGROUND

[0003] The most overwhelming challenges that the plastics industry are trying to resolve today pertain to sustainability, supply chain shortages, and the lack of skilled labor. Within the injection molding sector, manufacturers perform a full validation whenever a mold is moved to a different injection molding machine (IMM) or there is a material change. These full validations are labor intensive, expensive, and use a lot of material. Moreover, these methods may or may not utilize scientific molding principles. There has been a demand for a standard “part process” development method to transfer a mold between IMMs that was more efficient and scientific and to embrace variation in all resins.

SUMMARY

[0004] As provided herein, iMFLUX’s Auto-Viscosity Adjust (AV A) technology has made transferring a mold between IMMs easier with their low, constant pressure injection molding process. This technology enables the injection molding process the ability to independently adjust parameters in real time. This disclosure provided herein focuses on developing part process with iMFLUX and its features that are independent of resin and machine. As the present disclosure provides, iMFLUX was able to develop a part process for two molds that were independent of IMM.

BRIEF DESCRIPTION OF THE DRAWINGS

[0005] FIG. 1 is an exemplary flow chart illustrating conventional methods for validating the same mold in two different machines with two different resins;

[0006] FIG. 2 is an exemplary flow chart illustrating a universal process control method for validating the same mold in two different machines with two different resins; [0007] FIG. 3 illustrates an MIV transfer verification process, from a conceptual test plan for the VMP;

[0008] FIG. 4 illustrates a Haitian Zahafir injection molding machine;

[0009] FIG. 5 illustrates a JSW injection molding machine;

[0010] FIG. 6 illustrates a Mitutoyo Crysta-Apex S 500 series coordinate-measuring machine;

[0011] FIG. 7 illustrates a Dynisco LMI5500 melt indexer;

[0012] FIG. 8 illustrates a melt indexer sample collection process;

[0013] FIG. 9 illustrates a process capability report for inner barrel part height by material; and

[0014] FIG. 10 illustrates a process capability report for inner barrel part perimeter by material.

DETAILED DESCRIPTION

[0015] In 2015, a consortium of leading global Medical Device OEMs began collaborating on a new validation strategy that had the capability to reshape industry standards for mold transfer validation requirements. This group’s mission was to execute a proof-of-concept study inspired from a whitepaper written in 2000 by Rod J. Groleau, founder of the RJG Association and RJG Technologies, titled Location Independent PPAP Streamlined for Global Manufacturing.

Groleau, who constructed and defined the term systematic molding in the late 70s, suggested the use of a lean systematic “part process” model as an alternative to machine validation. Groleau’s whitepaper defines a “part process” validation as a validation that shifts the focus from the response of the machine to the part(s) in the mold and whether this part is being produced consistently.

[0016] In other words, this method optimizes in response of the “plastic’s point of view.” Since scientific molding process parameters are derived from plastic variables, not IMM variables, an injection mold has the capability to be moved between different IMMs despite them having different tonnages, intensification ratios, screw sizes, etc.

[0017] This idea was and is difficult for molders who use traditional, nonscientific methods to comprehend. The fact that a tool can move in-between such different presses, even hydraulic to electric, without a lengthy and careful requalification is difficult to accept. It appears too easy. It should be more complicated to be more credible. [0018] The integrity of scientific, or systematic, molding is carried most predominantly by RJG, General Polymers, John Bozzelli, and other molding experts. Each of these ambassadors strongly recommends the use of cavity pressure sensors when molding because these transducers are the only way to truly know the condition of the melt, and therefore assume part consistency in molds.

[0019] For the consortium to develop this suggested part process validation model, a machine-independent variables (MIV) method was created where actual plastic conditions and parameters that characterize a part are recorded from one IMM to then be converted to have matching specified outputs or results on a different IMM. Examples of these recorded parameters were fill time, actual melt temperature, volumetric shot size, and hold pressure. Since these variables were part-specific and not machine-specific, like the traditional validation approach, they were independent of the four distinctly different IMMs in this study. It was determined that this system could replicate optimized plastic parts data in different IMMs deemed as capable with negligible part variation as seen from its dimensional results. This study satisfied visual and dimensional Ppk requirements of 1 .5 or higher on the other three IMMs.

[0020] The result of this lean systematic method research was a well-documented case study to expose the medical device industry to the “Part Process Development/Validation” strategy as a robust alternate methodology to traditional validation concepts.. The ultimate paradigm shift this team faced after was to convince the average molder to focus on the plastic part process outputs rather than machine set points. It is important to note that the initial full IQ, OQ, and PQ part validations are still required, and this method is not applicable in the case of material changes.

[0021] Where scientific molding left off in 2017 is a notable case-study, but manufacturers are still tied to their material suppliers. This is where iMFLUX’s technology has found opportunity. Not only does their singular hold phase technology naturally let parts dictate the process by allowing velocity to be variable to accommodate for any parts’ geometry, but it is also naturally the most ideal option for processing from the “plastic’s point of view” when including their other features.

[0022] IMFLUX’s AVA technology allows for highly variable resins like sustainable resins to be processed without user adjustments. These alternative resins have varying pellet sizes and molecular weights making them difficult to process conventionally. Post-consumer regrind (PCR), ocean bottle grade plastic, bio-filled plastics and scrap are examples of sustainable resins that have this history of being challenging to process. However, they are competitive alternatives to virgin resin being that they are abundant, more affordable, and craved by consumers.

[0023] AVA is the component that allows for resin to be independent of mold qualification when using cavity pressure sensors. This technology allows the IMM to behave like a giant rheometer to regulate and maintain stable flow rates into the mold. As most processors know, if the flow rate varies, viscosity varies; if viscosity varies, parts will vary. It allows the screw velocity to be variable to shear the resin as needed to achieve an input cavity pressure at an input fill time. This quality assists with repeatable part weights and dimensions.

[0024] The Medical Device OEMs’ independent variables mentioned before were fill time, actual melt temperature, volumetric shot size, hold pressure. iMFLUX’s independent variables, derived from its built-in intelligence, are fill time, process factor A (PFA), and PFA time. AVA technology solves for Melt Pressure as an output. In real-time, this feature analyzes the delta between the target fill time and the actual time being achieved and accommodates for variation in the resin. This feature achieves a stable process when Target Time, or fill time, is equivalent to Actual Time. When experimenting with several resins, by enabling this feature and inputting the same Target Time for each resin, AVA communicates with our melt pressure transducer (MPT) to solve for the Melt Pressure it needs to keep making the same part dimensionally. For the first time, processors are seeing IMMs solve for the parameters it needs to make the same part regardless of the material or IMM type.

[0025] For comparison, FIG. 1 is an exemplary flow chart illustrating conventional methods for validating the same mold in two different machines with two different resins, while FIG. 2 is an exemplary flow chart illustrating a universal process control method for validating the same mold in two different machines with two different resins, as provided herein. Although FIG. 2 illustrates an example method for two machines and two resins, the universal process control method can be applied for more than two machines and/or more than two resins.

Materials and Methods

Materials

[0026] Three polypropylene (PP) post-consumer resins (PCRs), a PP post-industrial regrind (PIR), and a virgin PP were used for this study due to their wide melt flow index (MFI) range. The following table describes each resin’s name and type as well as its reported MFI.

[0027] Table 1 . Material descriptions and characterization.

[0028] These resins will be referred to as virgin, PCR10, PCR20, PCR40, and PIR for the remainder of this paper.

Injection Molding Process: Machines and Molds

[0029] A 3600 kN Haitian Zhafir (as shown at FIG. 4) and a 3440 kN JSW (as shown at FIG.

5) were injection molding machines (IMMs) that were used to process each material in two different molds.

[0030] Each of these molds were designed to produce parts for the assembly of a deodorant container. The parts are referred to as Inner Barrels and Outer Barrels in this paper. The cavitation for each mold was four, and they both had two end of fill cavity pressure sensors to record the response of each leg of the Design of Experiments (DOEs) and validations.

Process Window Development Method: Largest Empty Corner Rectangle (LECR)

[0031] This study begins with a range-finding method referred to as LECR. The process range is determined by the fastest and slowest fill time at the shortest and longest Step Times. In conventional molding terms, Step Time is equivalent to injection time and hold times combined. Typically ranging from -2 to 0, PFA is a factor multiplied by cavity pressure and added/subtracted from melt pressure to achieve a response after the part achieves 95-99% fill. DOE Method

[0032] Minitab® is the statistical software used to generate this research’s response surface DOE. A DOE is a standard, statistical method used to study and justify variation within plastic parts under conditions that are assumed to be the source of that variation. This method outputs a combination of parameters to test in several trials and measure how this affects plastics parts’ dimensions. Through a DOE matrix continuous factors and categorical factors can be isolated.

[0033] Continuous factors are derived IMM inputs which can also be considered machine learned inputs. Examples are PFA Time, PFA, and AVA Target Time, also referred to as fill time. PFA Time is Step Time minus Fill Time. In conventional terms, it can be interpreted as hold time. Categorical factors have a countable number of categories or distinct groups. Examples are the use of several IMMs or various materials which were hypothesized to not be crucial in this study.

[0034] This study had 65 randomized trials on each IMM tested with five differing resins. On each IMM, sixty-five randomized DOE legs were collected per mold. For each leg, five kilograms of material were used and a resin sample was collected for MFI analysis. Three shots were collected at the beginning, middle, and end of each DOE leg to be weighed and dimensionally measured.

[0035] This DOE trial data is then analyzed in a Response Optimizer to determine the most ideal settings across IMMs and tested resins. The Response Optimizer provides a single set of validation inputs to be tested for all IMM and Material combinations.

Dimensional Measurement Method

[0036] A coordinate-measuring machine (CMM) is a device that measures the geometry of physical samples by sensing discrete points on the surface of the parts with a probe. A Mitutoyo Crysta-Apex S 500 series CMM (as shown at FIG. 6) was used to measure the dimensions for the inner and outer barrel parts.

[0037] A Gage R&R is developed to ensure that any variation in dimensional measurements comes from the parts and not the process and measurement method.

[0038] The height and perimeter were the dimensions measured on the CMM for the inner barrel parts.

MFI Method [0039] A melt indexer measures a polymer’s resistance to flow, also known as viscosity, at a set temperature, under a force, for a duration of time. To determine the MFI of the resin for each leg, Dynisco’s LMI5500 melt indexer (as shown at FIG. 7) was used.

[0040] A stopper was placed at the bottom capillary to ensure no pellets escaped the heated barrel. Approximately six grams of resin was maneuvered into the barrel. A piston was used to pack the pellets down. A 2.06-kg weight was placed on top of the piston and left to set for seven minutes. The stopper was then removed, and excess resin was trimmed before sample collection began. FIG. 8 provides a depiction of this process.

[0041] A sample is then extruded for a set amount of time before being trimmed and weighted on a scale. The weight was input into the melt indexer to generate an MFI value in g/1 Omin. This test was done using a sample from each DOE.

Validation Method

[0042] The collection and evaluation of DOE data should be able to justify validation settings. A validation can ensure a process is capable of consistently delivering quality parts with a set of the inputs that were studied. Minitab® is the statistical software used to generate all validations in this study. A single validation was established for each mold. The validation metrology results were analyzed for their mean and standard deviation on the different IMMs and resins. An overall capability analysis is drawn as well.

Results

Validation Results: Capability Analysis by Material

[0043] For both inner and outer barrel molds, 25 samples were collected of each material at the suggested validation parameters of Minitab®’s response optimizer on each IMM. The validation capability report shown at FIG. 9 is for overall height of the inner barrel parts by material based on both IMMs.

[0044] These charts are showing the capability of each resin with the data from both IMMs. There is a bi-modal distribution in KAL20. Aside from this trend the data between the IMMs was indistinguishable. Each material displays a Cpk value over 1 .33, meaning there would be no defects with these validated parameters. There is a slight mean shift between the resins that iMFLUX plans to address with 2nd level intelligence.

[0045] The validation capability report shown at FIG. 10 is for the perimeter of the inner barrel parts by material based on both IMMs. [0046] These charts are showing the capability of each resin with the data from both IMMs. There is a bi-modal distribution in FT200WV, KAL10, and KAL40. Aside from this trend the data between the IMMs was indistinguishable. Each material displays a Cpk value over 1 .33, meaning there would be no defects with these validated parameters. There is a slight mean shift between the resins that iMFLUX plans to address with 2nd level intelligence.

Discussions

[0047] In the molding industry where tools move so readily from machine to machine and the demand for greener resins continue to escalate, the ability to quickly establish a process that produces good-quality parts is vital from a material expense, sustainability, and labor efficiency standpoint. iMFLUX’s Universal Process allows for the IMM type to be independent of the mold qualification process when using the following three independent factors: AVA Fill Time, PFA, and PFA Time. This technology’s feedback loop in its autonomous system

[0048] The economics of adopting this approach could potentially not only save tens to hundreds of thousands of dollars for each mold moved, but the speed-to-market advantages and operations flexibility would be simply invaluable. This case study has demonstrated that iMFLUX’s part capability is robust when using these factors. Molds have the ability transfer in and out of facilities without redoing a mold qualification while ensuring the integrity of parts are maintained.

[0049] The patent claims at the end of this patent application are not intended to be construed under 35 U.S.C. § 1 12(f) unless traditional means-plus-function language is expressly recited, such as “means for” or “step for” language being explicitly recited in the claim(s). The systems and methods described herein are directed to an improvement to computer functionality, and improve the functioning of conventional computers.

Claims

What is Claimed is:

1 . A method for applying universal processing conditions to fill a mold cavity of an injection molding apparatus with molten thermoplastic material during an injection molding process, the method comprising: identifying one or more parameters for the injection molding process, each of the one or more parameters associated only with one of the mold cavity, the molten thermoplastic material, or a type of injection molding apparatus; based on the one or more parameters, providing an expected process condition at a predetermined time during the injection molding process; measuring an actual process condition of the injection molding apparatus at the predetermined time during the injection molding process; based on a difference between the expected process condition and the actual process condition, providing a predetermined curve of a target process condition over time for a remainder of the injection molding process; measuring an experienced process condition of the injection molding apparatus over time for the remainder of the injection molding process; iteratively comparing the experienced process condition at a given time in the remainder of the injection molding process to the target process condition at a corresponding given time on the predetermined curve to detect a variance; and when the variance exceeds a threshold, automatically adjusting the injection molding process so that the variance is within a threshold.

2. The method of claim 1 , wherein the one or more parameters identified are one or more of a group consisting of the melt flow index of the molten thermoplastic material, the cavitation geometry of the mold cavity, and the type of injection molding apparatus.

3. The method of claim 2, wherein only one parameter is identified.

4. The method of claim 1 , wherein the one or more parameters identified are one or more of a group consisting of auto viscosity adjust (AV ) trigger time, process factor A (PFA), and PFA time.

5. The method of claim 1 , wherein the expected process condition is an expected melt pressure or an expected flow front pressure, the actual process condition is an actual melt pressure or an actual flow front pressure, and the difference is a pressure difference between an expected melt pressure and actual melt pressure or an expected flow front pressure and an actual flow front pressure.

6. The method of claim 1 , wherein the target process condition is a target melt pressure or a target flow front pressure, the experienced process condition is an experienced melt pressure or an experienced flow front pressure, and the variance is a pressure difference between a target melt pressure and an experienced melt pressure or a target flow front pressure and an experienced flow front pressure.

7. The method of claim 1 , wherein the expected process condition is an expected cushion of the molten thermoplastic material located between a front of an injection molding screw or ram to an end of a barrel of the injection molding apparatus, the actual process condition is an actual cushion of the molten thermoplastic material measured between the front of the injection molding screw or ram to the end of the barrel of the injection molding apparatus, and the difference is a cushion difference between the expected cushion and the actual cushion.

8. The method of claim 1 , wherein the target process condition is a target cushion of the molten thermoplastic material located between a front of an injection molding screw or ram to an end of a barrel of the injection molding apparatus, the experienced process condition is an experienced cushion of the molten thermoplastic material measured between the front of the injection molding screw or ram to the end of the barrel of the injection molding apparatus, and the variance is a cushion difference between the expected cushion and the actual cushion.

9. The method of claim 8, wherein automatically adjusting the injection molding process so that the variance is within the threshold includes adjusting a pressure to change the experienced cushion and not adjusting a shot size of injection thermoplastic material.

10. A mold including a mold cavity, the mold qualified for use in an injection molding apparatus configured to execute a method of applying universal processing conditions to fill the mold cavity with injection molten thermoplastic material during an injection molding process, the method comprising: identifying one or more parameters for the injection molding process, each of the one or more parameters associated only with the mold cavity, the molten thermoplastic material, or the type of injection molding apparatus; based on the one or more parameters, providing an expected process condition at a predetermined time during the injection molding process; measuring an actual process condition of the injection molding apparatus at the predetermined time during the injection molding process; based on a difference between the expected process condition and the actual process condition, providing a predetermined curve of a target process condition over time for a remainder of the injection molding process; measuring an experienced process condition of the injection molding apparatus over time for the remainder of the injection molding process; iteratively comparing the experienced process condition at a given time in the remainder of the injection molding process to the target process condition at a corresponding given time on the predetermined curve to detect a variance; and when the variance exceeds a threshold, automatically adjusting the injection molding process so that the variance is within a threshold.

11 . The mold of claim 10, wherein the one or more parameters identified are one or more of a group consisting of the melt flow index of the molten thermoplastic material, the cavitation geometry of the mold cavity, and the type of injection molding apparatus.

12. The mold of claim 11 , wherein only one parameter is identified.

13. The mold of claim 10, wherein the one or more parameters identified are one or more of a group consisting of auto viscosity adjust (AVA) trigger time, process factor A (PFA), and PFA time.

14. The mold of claim 10, wherein the expected process condition is an expected melt pressure or an expected flow front pressure, the actual process condition is an actual melt pressure or an actual flow front pressure, and the difference is a pressure difference between an expected melt pressure and actual melt pressure or an expected flow front pressure and an actual flow front pressure.

15. The mold of claim 10, wherein the target process condition is a target melt pressure or a target flow front pressure, the experienced process condition is an experienced melt pressure or an experienced flow front pressure, and the variance is a pressure difference between a target melt pressure and an experienced melt pressure or a target flow front pressure and an experienced flow front pressure.

16. The mold of claim 10, wherein the expected process condition is an expected cushion of the molten thermoplastic material located between a front of an injection molding screw or ram to an end of a barrel of the injection molding apparatus, the actual process condition is an actual cushion of the molten thermoplastic material measured between the front of the injection molding screw or ram to the end of the barrel of the injection molding apparatus, and the difference is a cushion difference between the expected cushion and the actual cushion.

17. The mold of claim 10, wherein the target process condition is a target cushion of the molten thermoplastic material located between a front of an injection molding screw or ram to an end of a barrel of the injection molding apparatus, the experienced process condition is an experienced cushion of the molten thermoplastic material measured between the front of the injection molding screw or ram to the end of the barrel of the injection molding apparatus, and the variance is a cushion difference between the expected cushion and the actual cushion.

18. The mold of claim 10, wherein automatically adjusting the injection molding process so that the variance is within the threshold includes adjusting a pressure to change the experienced cushion and not adjusting a shot size of injection thermoplastic material.