CA2931412C

CA2931412C - Manufacturing process control systems and methods

Info

Publication number: CA2931412C
Application number: CA2931412A
Authority: CA
Inventors: Ashley Stone
Original assignee: Meyer Edo
Current assignee: Meyer Edo
Priority date: 2016-05-27
Filing date: 2016-05-27
Publication date: 2023-06-13
Anticipated expiration: 2036-05-27
Also published as: CA2931412A1

Abstract

A method of controlling a manufacturing process having a machine to form a material in to a component. The method comprises the steps of establishing an initial set of operating parameters for the machine, producing an initial component from the machine, inspecting the component to determine its acceptability relative to a desired component, determining a variation in the operating parameters to improve the acceptability of the component, effecting changes in the operating parameters and inspecting subsequent components to determine their acceptability.

Description

CA Patent Application Agent Ref: 76798/00016

2

3 TECHNICAL FIELD

4 [0001] The present invention relates to a process control for a manufacturing process, such as a molding systems and methods for controlling such a process. The present invention is 6 exemplified by application to a molding system, more specifically, to a hot-runner system, a melt 7 flow modular assembly, an injection molding method, an injection mold, hot runner, and a valve 8 gate device but is applicable to other manufacturing processes.

BACKGROUND
11 [0002] Injection molding (British English: moulding) is a manufacturing process for 12 producing parts from both thermoplastic and thermosetting plastic or other materials, including 13 metals, glasses, elastomers and confections. Material is fed into a heated barrel, mixed, and 14 forced into a mold cavity where it cools and hardens to the configuration of the cavity. After a product is designed, usually by an industrial designer or an engineer, molds are made by a mold 16 maker (or toolmaker) from metal, usually either steel or aluminum, and precision-machined to 17 form the features of the desired part. Injection molding is widely used for manufacturing a 18 variety of parts, from the smallest component to entire body panels of cars.
19 [0003] Injection molding utilizes a ram or screw-type plunger to force molten plastic material into a mold cavity; this produces a solid or open-ended shape that has conformed to the 21 contour of the mold. It is most commonly used to process both thermoplastic and thermosetting 22 polymers, with the former being considerably more prolific in terms of annual material volumes 23 processed.
24 [0004] Thermoplastics are prevalent due to characteristics which make them highly suitable for injection molding, such as the ease with which they may be recycled, their versatility 26 allowing them to be used in a wide variety of applications, and their ability to soften and flow 27 upon heating.
28 [0005] Injection molding consists of high-pressure injection of the molten plastics 29 material, referred to as plastic melt, into a mold, which shapes or forms the polymer into a desired shape. Molds can be of a single cavity or multiple cavities. In multiple cavity molds, 22928473.1 CA Patent Application Agent Ref: 76798/00016 1 each cavity can be identical and form the same parts or can differ and produce multiple different 2 geometries during a single cycle. Molds are generally made from tool steels, but stainless steels 3 and aluminum molds are suitable for certain applications.
4 [0006] Aluminum molds typically are ill-suited for high-volume production or parts with narrow dimensional tolerances, as they have inferior mechanical properties and are more prone 6 to wear, damage, and deformation during injection and clamping cycles;
however, they are more 7 cost-effective in low volume applications as mold fabrication costs and time are considerably 8 reduced. Many steel molds are designed to process well over a million parts during their lifetime 9 and can cost hundreds of thousands of dollars to fabricate.
[0007] When thermoplastics are molded, typically pelletized raw material is fed through 11 a hopper into a heated barrel with a feed screw. Upon entrance to the barrel, the thermal energy 12 increases and the Van der Waals forces that resist the relative flow of individual chains are 13 weakened as a result of increased space between molecules at higher thermal energy states. This 14 reduces its viscosity, which enables the polymer to flow under the influence of the driving force of the injection unit. The feed screw, typically an Archimedean screw, delivers the raw material 16 forward, mixes and homogenizes the thermal and viscous distributions of the polymer, and 17 reduces the required heating time by mechanical shearing of the material and adding a significant 18 amount of frictional heating to the polymer. The material is fed forward through a check valve 19 and collects at the front of the screw into a volume known as a shot.
The shot is the volume of material, which is used to fill the mold cavity, compensate for shrinkage, and provide a cushion 21 (approximately 10% of the total shot volume which remains in the barrel and prevents the screw 22 from bottoming out) to transfer pressure from the screw to the mold cavity. When enough 23 material has gathered, the material is forced at high pressure and velocity through a gate and into 24 the part-forming cavity by moving the screw along its axis. To prevent spikes in pressure, the process normally utilizes a transfer position corresponding to a 95-98% full cavity where the 26 screw shifts from a constant velocity to a constant pressure control.
Often injection times are 27 well under one second and the cooling time of the part in excess of four seconds. Once the screw 28 reaches the transfer position the packing pressure is applied, which completes mold filling and 29 compensates for thermal shrinkage, which is quite high for thermoplastics relative to many other materials. In thermal gating the packing pressure is applied until the material located in the mold 31 gate (cavity entrance) solidifies. The gate is normally the first place to solidify through its entire 22928473.1 CA Patent Application Agent Ref. 76798/00016 1 thickness due to its small size. Once the gate solidifies, no more material can enter the cavity;
2 accordingly, the screw returns and acquires material for the next cycle while the material within 3 the mold cools so that it can be ejected and be dimensionally stable.
This cooling duration is 4 dramatically reduced using cooling lines to circulate water or oil from a thermolator or preferably using an organic refrigerant. Once the required temperature has been achieved, the 6 mold opens and an array of pins, ejectors, etc. is driven forward to remove the article from the 7 mold, referred to a "de-molding". Then, the mold closes and the process is repeated. The thermal 8 gating, where the closing of the gate is accomplished by solidified plastic, is possible for small 9 flow requirements with a melt flow channel with small gates. For high plastic flow rates, a valve gated hot runner is used in which mechanical valves control the flow of melt from a common 11 supply, or hot runner, to the mold. A modulating assembly modulates the melt flow. Faster cycle 12 time may be attained because no gate cooling is required to shut off the melt flow, and no gate 13 re-heating is required to open the gate to the melt flow.
14 [0008] A parting line, sprue, gate marks, valve pin marks, and ejector pin marks are usually present on the final part, often even after prolonged cooling time.
None of these features 16 are typically desired, but are unavoidable due to the nature of the process. Gate marks occur at 17 the gate that joins the melt-delivery channels (sprue and runner) to the part-forming cavity.
18 Parting line and ejector pin marks result from minute misalignments. The wear, gaseous vents, 19 clearances for adjacent parts in relative motion, and/or dimensional differences of the mating surfaces contacting the injected polymer also create marks on the molded surface of the part. The 21 dimensional differences can be attributed to non-uniform, pressure-induced deformation during 22 injection, machining tolerances, and non-uniform thermal expansion and contraction of mold 23 components, which experience rapid cycling during the injection, packing, cooling, and ejection 24 phases of the process. Mold components are often designed with materials of various coefficients of thermal expansion. These factors cannot be simultaneously accounted for, without 26 astronomical increases in the cost of design, fabrication, processing, and the part quality 27 monitoring. The skillful mold and part designers, will position these aesthetic detriments in 28 hidden areas, if feasible.
29 [0009] Inevitably, to eliminate the gate marks, it is necessary to improve gate performance. The gate marks and gate residuals are called vestige. The molding quality is directly apportioned to 31 shape, configuration, degradation of the vestige and vestige height and shape. The gate quality is 22928473.1 CA Patent Application Agent Ref: 76798/00016 1 nowadays-major issue in injection molding art, particularly for food and beverage packaging.
2 [00010] It is well-known in the field of injection molding art that some structure must be 3 placed in the mold gate, at a particular time in the molding cycle, to inhibit the flow of molten 4 material into the cavity of a mold, so that the molded part may be cooled, and subsequently opened to remove the molded parts. This must be done without creating drool of the molten 6 material in the molding surface. This drool would create undesirable marks on the next moldings, 7 and this is largely un-acceptable.
8 [00011] As noted above, there are essentially two broad categories of melt flow 9 modulating assemblies, or flow inhibiting techniques known in the field of injection molds, namely, thermal gating in which the gate at the exit of the nozzle is rapidly cooled at the 11 completion of the injection operation to form a solid or semi-solid plug of the material being 12 injected into the gate; and valve gating in which a mechanical means is employed to inhibit the 13 flow of material being injected into the mold cavity.
14 [00012] Each category has its own advantages and disadvantages relative to the other.
Numerous systems using thermal gating are known in the art of the hot-runners.
16 [00013] Valve gating systems are generally of one of two types, namely inline and lateral 17 systems of gate closing. A wide variety of systems of each type have been developed. Referring 18 now to the inline gating choices, there is in the art of the injection molding, mainly three types of 19 valve gate closing choices: axial pin motion, rotary pin motion with shutoff, and a rotary pin with dynamic melt flow control without positive shutoff.
21 [00014] Many valve mechanisms used in the injection molding industry are constructed in 22 such a way as to move a valve pin assembly in an axial direction along the nozzle melt channel 23 from fully open to fully closed position. This is the predominant structure when comparing based 24 on the motion of the valve pin.
[00015] An example of this is found in U.S. Patent No. 4,268,240, U.S.
Patent No.
26 6,086,357, U.S. Patent publication No. 2011/0293761 Al, U.S. Patent No.
8,047,836 B2, U.S.
27 Patent No. 7,600,995 B2, or for example, U.S. Patent No. 7,044,728 B2 and U.S. Patent 28 publication No. 2005/0100625 Al where the plastic is transferred from a hot-runner manifold to 29 a nozzle. This type of the melt delivery goes around the pin and then rejoins the melts from each side of the pin, and reconstitutes the tubular flow just below the valve pin tip. Therefore, this 31 kind of the valve pin motion, being axial, causes melt flow, arriving laterally at the pin, to be 22928473.1 CA Patent Application Agent Ref: 76798/00016 1 divided by the valve pin or stem.
2 [00016] The flow is rejoined again into a single path as it passes in to the mold cavity, 3 resulting in moldings with undesirable weld lines created by the once divided polymer volumes, 4 visibly affecting quality of the products. These weld lines can adversely affect both the aesthetic and performance qualities of the final molded product, and it is significantly advantageous to 6 avoid their creation when molding certain products.
7 [00017] Some alternatives to prevent melt separation have been proposed, e.g. the valve 8 pin may be shielded, as in U.S. Patent No. 4,412,807 which shows an apparatus in which the 9 plastic flow channel in the nozzle is kept separate from the valve pin in an effort to avoid dividing the melt stream. The channel is a crescent shaped cross section, which is known to be 11 less than ideal for encouraging plastic flow, especially in the opposing sharp corners.
12 Furthermore, when the valve pin is in the open position to let plastic material to pass into the 13 mold cavity, it creates a stagnant area of poor plastic flow directly adjacent the front face of the 14 pin. These areas of poor plastic flow can result in material degradation, which can adversely affect the performance and physical properties of the molded product.
16 [00018] U.S. Patent No. 4,925,384 shows a similar design that permits the plastic to come 17 into contact with the valve pin but restricts it from passing around the pin to form a weld line.
18 This patent describes an approach that does not cause pronounced division of the melt flow. This 19 design also suffers from a melt channel with sluggish flow areas and requires difficult and expensive machining processes to produce the nozzle housing, having an unusual melt channel 21 cross section.
22 [00019] Alternatively, valve gates may be structured to rotate the pin and close or open the 23 gate that way.
24 [00020] U.S. Patent No. 3,873,656 shows a valve having taps, which rotate to open or close. This is similar to the approach described above. It is not compact or easy to manufacture 26 and has sharp edges, susceptible to damage, where it mates with the sprue channels.
27 [00021] A rotating nozzle is shown in U.K. Patent No. 872,101.
The entire injection unit 28 nozzle rotates on an axis parallel to the flow of plastic as opposed to the perpendicular or angular 29 rotation axis of the two patents mentioned previously. The nozzle front portion remains in forced contact with the delivery bushing, to prevent plastic leakage between the two.
The construction 31 shown is very bulky, consuming a substantial amount of space.

5 22928473.1 CA Patent Application Agent Ref: 76798/00016 1 [00022] Further example of the attempt to reduce weld line and part marks is disclosed in 2 the U.S. Patent No. 5,499,916 where the stem rotates with limited contact with the melt flow but 3 does not allow melt separation.
4 [00023] A further example to improve melt flow delivery, and melt flow temperature uniformity as well as hot runner balancing is attempted in the application of the rotating pin is

6 disclosed in U.S. Patent Publication No. 2007/0065538 Al. The valve pin is operatively

7 connected to a motor that has fast acceleration and deceleration rates.
The valve pin is made in

8 the form of an Archimedean screw or screw pump so that the pin is positioned within the melt

9 flow assembly in the hot-runner nozzle. By rotating and pumping melt flow in the direction of the melt flow, the valve-pin "pump" reduces pressure drop within the melt flow assembly, 11 supposable creating favorable melt delivery and melt conditioning.
12 [00024] However, when rotation is in the direction to retard melt flow of the molten 13 material traveling in the direction of the cavity, higher-pressure drop is created in the melt flow 14 assembly of the hot-runner and therefore, balancing the pressure and flow to ensure that drop-to-drop uniformity is maintained. Besides the positive effect on the uniformity of the melt, that is 16 critical for food packaging and medical parts, this valve gate molding system can effectively 17 produce acceptable quality gate vestige mark and at the same time ensure that the closing of the 18 gate is accomplished by rotation of the pin screw "pump" within melt in the melt flow assembly 19 and at the same time improve temperature uniformity of the melt in the hot-runner nozzle.
[00025] In each of the systems described above, and in inline systems, generally, a valve 21 pin aligned with the gate is moved parallel to the direction of movement of molten material 22 (generally referred to as "melt") through the gate, between a position wherein the pin extends 23 into the gate to block flow through the gate, and a position wherein the pin is retracted from the 24 gate permitting flow there-through into the mold cavity. In order to be aligned with the gate, the valve pin is located inside the injection nozzle and is at least partially within the flow path of the 26 melt.
27 [00026] For these and other reasons, inline valve gating suffers from a variety of 28 problems.
29 [00027] One common problem is wear of the valve pin due to contact with the nozzle and/or gate, which can lead to leaks or failure of the valve.
31 [00028] Another common problem is the conversion of the melt from the tubular flow CA Patent Application Agent Ref: 76798/00016 1 entering the nozzle to an annular (or other non-continuous) flow, which is caused by the valve 2 pin or other related components being within the melt flow. Such a non-continuous flow can 3 result in weld or knot lines in the molded product produced as the melt flow recombines within 4 the gate or mold cavity, and this can result in weakened or unacceptable molded products. This is particularly a problem when molding preforms for water containers where good part appearance 6 and gate quality are an essential for successful sales of bottled water or other clear liquids.
7 [00029] The water bottles are made in a two-stage process and, require in a first stage to 8 produce a preform, and in a second-stage the preform is air inflated against a cavity of the mold 9 in a shape of the bottle. The bottle preforms are made from the polyethylene terephthalate, abbreviated as PET. The PET preform molding process, in particular, requires tubular melt flow, 11 and having the pin inside the melt flow, does not help improve melt flow in the molding process.
12 [00030] During the injection process, the molten plastic material is injected into the mold 13 cavity under very high pressure, often above 15,000 PSI. Once injected, in a short injection time, 14 often less than one second, molten plastic enters cooling and solidification phase lasting 2 to 30 seconds. During this process, a definite time is selected, within hold time in the process, when to 16 close the valve gate. Closing valve gate means that gate volume should be filled by pin tip 17 volume so as to block plastic flow through the gate.
18 [00031] The axial movement of the valve pin assembly accomplishes this.
19 [00032] In principle, the valve pin conical surface and gate conical surface each have a complementary sealing surface. When these surfaces are brought together the flow of the 21 material through the gate stops.
22 [00033] Usually, as it is well recognized in the injection molding art, the gate closing is 23 initiated just about when the cavity is filled, and the injection time hold interval is about to end.
24 After the valve pin is fully forward and in a predominantly closed position, no more plastic melt is possible to enter the cavity. Mold cooling helps to remove heat from the molded part and helps 26 to solidify the part and cool it so that can be handled in post molding cooling process. The post 27 molding process, by itself is the complex process when molding PET
preforms or any food 28 packaging containers like K-cups. The post molding process often requires specialty equipment 29 and additional complexities.
[00034] The valve pin stays closed until the mold is fully open and perhaps even just 31 before mold fully closed position after ejection of the molded part. Of course, timing when to 22928473.1 CA Patent Application Agent Ref: 76798/00016 1 start opening the valve gate is largely dependent on the valve pin driving apparatus.
2 [00035] Fast acting valve gate systems allow for more flexibility and better timing and 3 control of the gate mechanism. Currently air piston operated valve gates require closing time up 4 to one second due to lack of proportionality between air pressure and axial force. Once air piston start moving it only stops at the hard stop at the end of the stroke.
Similarly, servo motor driven 6 pins, introduce nonlinearity largely because gear box and nonlinearities in magnetic structures of 7 the current servo motors and drives.
8 [00036] As noted above, there are various options for the valve gate pin configuration and 9 ways for opening and closing the gate. There are, however, only a few options for powering the valve pins. It is known in the art of injection molding and hot runners to provide an electrical or 11 fluid actuator to power the pin of the valve gate.
12 [00037] The electric motors, air motors or hydraulic motors mostly power the rotary pins.
13 For axially moving valve pin, typically, the actuators are the pneumatic or hydraulic type. The 14 moving air piston type actuator is predominantly used today to power axially moving valve gate pin due to its simplicity and compactness. All other motors, including servo motors and drives 16 require conversion of rotary to linear motion via transmission elements or gearbox.
17 [00038] The disadvantage of using an air piston cylinder to power the valve gate 18 assembly, besides extensive drilling of the substantial number of air channels, is that the 19 pneumatic piston actuator may require specialized valves and air hoses to deliver and control the compressed air. The pressure of the air supply in each location is different and is very difficult to 21 ensure consistent high air pressure at each valve pin location. Even when air pressure is 22 available, often in range 75 PSI (pounds per square inch) to 120 PSI, flow rate, cleanliness and 23 capacity of the air compressors may not be always adequate. Often just differences in hose length 24 will change the mold performance due to different air supply pressure seen by each valve gate.
Just the fact that the piston seal stiction alone, in the multi cavity mold, may be different at 26 different operating temperatures is material and illustrates the level of randomness involved in 27 these systems. The mechanical tolerance, location in the mold, air supply line arrangements, and 28 environmental contamination, maybe enough to result in less than an optimal valve gate opening 29 or closing time. These and other variations result in differences in part quality and quality of the gate vestige. These and other variations are not desirable.
31 [00039] Yet another disadvantage of the air operated valve gates is that the pin can only be 22928473.1 CA Patent Application Agent Ref: 76798/00016 1 positioned at the fully open position or at the fully closed position, and cannot be positioned 2 between these two positions, unless additional pistons or complexities are installed. Moreover, as 3 the compressed air temperature varies during the day, this inhibits molding good parts without 4 continuous process adjustments and monitoring. A further disadvantage of the air piston operated valve gates is that they are relatively easy to get contaminated by the PET
dust or air impurities 6 and then get slow to move and not very accurate in the closing position of the pin.
7 [00040] A further disadvantage is that the air exhaust contaminates freshly molded parts, 8 and parts for medical and food packaging industry are very sensitive to parts cleanliness.
9 [00041] The most important disadvantage of the air operated valve gate systems is that air is exhausted to the environment and large volume of air is used for these operations.
11 Compressing and delivering air to the molding system is very expensive and compressed air is 12 delivered with overall compressor efficiency less than 40%. That means only a portion of the 13 electrical energy used for compressing and delivering compressed air is used and converted into 14 a useful motion of the valve pin.
[00042] Hydraulic pistons are often used for large valve gated assemblies and relatively 16 high axial force requirements, but using hydraulic oil and mist in the vicinity of freshly molded 17 medical or food packaging parts, is not acceptable.
18 [00043] Electric motors with rotary motion are being used for generating axial motion.
19 [00044] The motors and gear transmission assemblies are very large in volume and mostly not suitable for applications with a higher number of cavities.
21 [00045] In some applications, like food packaging and medical molding industry, the use 22 of the electric actuators for the valve gates is demanded due to their cleanliness. Air and 23 hydraulics just generate too much of the air contaminant dispersion to be acceptable in clean 24 environments like medical moldings and food packaging.
[00046] Electrical actuators are becoming more compact and being now available in a 26 variety of the configurations, which allows them to be used as actuators for the valve gate 27 assemblies in injection molding systems.
28 [00047] One example of such an electrically operated valve gate pin is disclosed in the 29 U.S. Patent Publication No. 2005/0100625 Al. In that patent, a valve gate assembly for regulating a flow of molten material into a mold is operated by the electric motor. The electric 31 motor operates via a mechanical transmission to move the valve pin, and infinitely positions the 22928473.1 CA Patent Application Agent Ref: 76798/00016 1 valve pin between the fully closed position and the fully open position by using a position 2 feedback device in a closed loop servo control mode. Various electric motors are proposed for 3 this application, but servo controls of this nature are largely impractical for high cavitation molds 4 where 96 to 144 individual PET preforms may be molded within each machine cycle. Besides, a feedback device installed for each individual pin position feedback is impractical and very 6 difficult to integrate in the molds and hot runner assemblies. Even if, and when used, the closed-7 loop servo motor powered valve pin, must maintain the valve pin in a closed position when the 8 operator's gate is open to prevent hot plastic melt spray and injury to operators entering the mold 9 area. The servomotor must maintain positioning accuracy and stiffness throughout the injection cycle, and that means high current is required to just maintain the position.
Motors must be rated 11 for 100% duty cycle. It is easy to see how overheating of the electric motor can occur, and then 12 additional complexities must be built into a servo system to overcome that. Molders today just 13 are not ready to put up with maintenance and servicing requirements of hundreds of the 14 individually controlled servo motor systems, despite the valve pin positioning accuracy and associated benefits of the accurate individual valve pin positioning.
16 [00048] U.S. Patent No. 5,556,582 describes the system wherein an adjustable valve pin is 17 operated by the servo controlled motor. The valve pin can be dynamically adjusted by a 18 computer according to pressure data read at or near the injection gate.
If multiple valves are used, 19 each is independently controlled. A hot runner nozzle is not provided.
Also, as the system is used, the repetitive actions of the valve pin cause significant wear on the tip of the valve pin.
21 This wear, is a result of the repeated impact with the mold cavity.
Basically, an adjustable valve 22 is provided that is adjusted by the close loop servo system, while the plastic melt material is 23 flowing through the gate into the mold cavity. The computer controls the servo motor, based on a 24 sensor in the cavity, preferably stated as being cavity pressure closed loop servo system. This control is complex and not easy to implement in large cavitation molds.
26 [00049] U.S. Patent No. 6,294,122 B1 describes the system of driving the pin axially 27 along the nozzle melt channel in a closed-loop control by operatively connecting the pin with 28 linearly moving mechanical transmission assembly, which converts the rotational movement of 29 the motor assembly into linear motion. The conversion assembly is a gearbox or screw and a nut threadingly engaged with each other or alternatively driven by the rack and pinion gear 31 assembly. Positioning is based on the proportional integral and derivative (PID) controls getting 22928473.1 CA Patent Application Agent Ref: 76798/00016 1 position feedback from an encoder. This approach, while sophisticated and allowing for very 2 precise valve pin positioning, is complex and overly sophisticated for the applications and the 3 current state of the art in the plastic industry today. Besides, having transmission elements 4 between an electrical rotor and vale pin introduces unacceptable response delay. The motor gear assembly use is therefore limited to large molds often used in automotive applications where fast 6 pin closing moves are not required. Besides, having bulky motor and a gearbox between the 7 molding platens of the injection machine limits the opening stroke and type of the parts that can 8 be molded with this arrangement. Again this is generally not practical for high cavitation counts 9 and high production rates.
[00050] In a similar attempt to operate a valve pin with a clean electrical motor and 11 accommodate a large number of drops, U.S. Patent Publication No.
2011/0293761 Al describes 12 a system where a plurality of pins is attached to an electro-magnetically driven plate so the valve 13 pins are movable responsive to movement of the actuation plate. No proposed driving logic is 14 offered as to how to control the largely uncontrollable force at the end of the stroke. When two magnetic assemblies of the type shown in this published application get very close together, the 16 impact and noise generated by the plate contact is likely to damage connecting elements of the 17 valve pin if not mitigated with additional complexities. It would also likely result in a very slow 18 movement of the valve pin assembly because it would take substantial time to establish a 19 magnetic field in a large magnetic storage like electromagnets, and to subsequently reverse that field. To collapse the electromagnetic field in assemblies of this size is a lengthy and involved 21 process, even when sophisticated electronic devices are used. The magnetic structures of this size 22 and mass do not allow for fast current switching, because the collapsing magnetic field and 23 changing polarity will generate back electromotive force of significant proportions. Simply, a 24 large mass does not lend itself for fast opening and closing valve pins.
[00051] At the opposite spectrum of valve pin actuations, small electromagnetic actuators 26 have been proposed and tested. The most promising method of direct pin activation is the method 27 of controlling pin closed and pin open position with two solenoids but aided by a spring: one to 28 hold the valve open and one to hold the valve closed. Since the electromagnetic solenoid 29 actuators are inherently unidirectional and a large force is required at the end of the stroke, it is very difficult to electronically control the movement of the pin.
31 [00052] Also the force exerted by these solenoid actuators is proportional to the square of 22928473.1 CA Patent Application Agent Ref: 76798/00016 1 the current input, and decreases as the function of the air gap between the actuator and the 2 armature. Therefore, as good as these actuators are, their control is difficult for consistent 3 operation. Having a mechanical spring is also an undesirable feature.
4 [00053] It is critical for the valve pin to arrive exactly at the end of the stroke with exactly near zero velocity. This is often defined in state of the art as perfect "soft landing". The receiving 6 end actuator must do exactly as much work as was done against friction and adhesive force of 7 the melt along the entire transition from open-to-close or vice versa. If the actuator does not do 8 this much work, the valve pin will stop before the end of the stroke. If the actuator does any 9 more than the exact correct work, the valve pin arrives at the end of the stroke with non-zero velocity where it can impact valve seat if contacts it, or imparts the shock and vibration on the 11 valve pin assembly by impacting against a hard stop. The non-uniform force, and other effects 12 cause disturbances of the valve pin assembly and make this system very difficult to control.
13 [00054] None of the foregoing valve pin activation and control techniques offer 14 individually controlled valve pin structure in a small and compact size that will move the pin axially along the melt flow channel, without any interconnecting, converting mechanical 16 transmission elements to reduce speed, or convert power or convert torque. These and other 17 systems require installation of the position or process feedback devices in areas that has limited 18 space. The mechanical structures have very small structural safety margins, and any additional 19 requirements for installation of any feedback devices make the systems very complex. This is particularly difficult in applications with an increased number of cavity drops and reduced drop 21 to drop (pitch) spacing.
22 [000551 The above discussion exemplifies the challenges in controlling a manufacturing 23 process to produce components at a high production rate with a consistently high quality.
24 Attempts to maintain machine operations with fixed input parameters inevitably leads to a variation of quality as ambient and input conditions change. For example a small change in the 26 characteristics of the feedstock, or a change in the ambient temperature in which the process is 27 performed, may produce a deleterious effect on the component being produced, which will lead 28 to rejected components if an adjustment is not made.
29 [00056] Attempts to compensate for such variation utilize feedback controls, either open loop or closed loop systems. Feedback systems have part of their output signal "fed back" to the 31 input for comparison with the desired set point condition. The type of feedback signal can result 22928473.1 CA Patent Application Agent Ref: 76798/00016 1 either in positive feedback or negative feedback.
2 [00057] In a closed-loop system, a controller is used to compare the output of a system 3 with the required condition and convert the error into a control action designed to reduce the 4 error and bring the output of the system back to the desired response.
Then closed-loop control systems use feedback to determine the actual input to the system and can have more than one 6 feedback loop.
7 [00058] Closed-loop control systems have many advantages over open-loop systems. One 8 advantage is the fact that the use of feedback makes the system response relatively insensitive to 9 external disturbances and internal variations in system parameters such as temperature. However it has been found through experiments and customer feedback that close controls executed in a 11 classical way do not produce the adjustments necessary to maintain good quality articles. For 12 example, in the case of temperature, the sensor is the thermocouple.
(TC). It measures 13 the temperature at the single point and controller is attempting to maintain the same temperature 14 at that point. Even if this is possible, quality of the final product may not be good, because of the very many variables that can affect the part physical dimensions, and perhaps flow characteristic 16 of the melt.
17 [00059] It is therefore an object to the present invention to obviate or mitigate the above 18 disadvantages.

SUMMARY
21 [00060] In an aspect of this invention, there is provided a method of controlling a 22 manufacturing process having a machine to form a material in to a component. The method 23 comprises the steps of establishing an initial set of operating parameters for the machine, 24 producing an initial component from the machine, inspecting the component to determine its acceptability relative to a desired component, determining a variation in the operating parameters 26 to improve the acceptability of the component, effecting changes in the operating parameters and 27 inspecting subsequent components to determine their acceptability.
28 [00061] According to a further aspect of the present invention there is provided a machine 29 to form a component. The machine includes a production system to form the component from a material, and an inspection system to determine acceptability of the component relative to a 22928473.1 CA Patent Application Agent Ref: 76798/00016 1 reference. A data store retains information obtained from the inspection system, and a knowledge 2 management system for analyzing and predicting data accesses the data store and determines 3 variations in operating parameters of the production system.

BRIEF DESCRIPTION OF THE DRAWINGS
6 [00062] FIG. 1 is a sectional view of a multi-cavity valve gated hot runner injection 7 molding system or apparatus to mold plastic articles like PET preforms;
8 [00063] FIG. 2 is a section on the line II-II of FIG. 1;
9 [00064] FIG. 3 is an enlarged section of an upper portion of FIG. 2;
[00065] FIG. 4 is an enlarged section of a lower portion of Figure 2;
11 [00066] FIG. 5 is an enlarged section of a portion of FIG. 4 showing a valve pin closed 12 portion;
13 [00067] FIG. 6 is a section on the line VI-VI of FIG. 3;
14 [00068] FIG. 7 is an enlarged section of a portion of FIG. 3 showing a valve pin locking assembly;
16 [00069] FIG. 8 is a view on the line VIII of FIG. 6;
17 [00070] FIG. 9 is a view on the line IX ¨IX of FIG. 7;
18 [00071] FIG. 10 is a simplified block diagram of an electronic valve gate drive controller 19 used with the apparatus of FIG. 1;
[00072] FIG. 11 is a plot showing the relationship between force and current over time 21 provided by the actuator of FIG. 2;
22 [00073] FIG. 12 is an enlarged section similar to Figure 3 of an alternative embodiment of 23 a Lorentz force actuator assembly;
24 [00074] FIG. 13 is a section similar to FIG. IA showing a further embodiment of a force actuator assembly;
26 [00075] FIG. 14 is a section of the multi-pin array actuator operated by a single Lorentz 27 force actuator assembly of FIG. 1;
28 [00076] FIG. 15 is a flow diagram of a single iteration of waveform control signals;
29 [00077] Fig. 16 is a block diagram showing an open loop positive feedback control production system;

22928473.1 CA Patent Application Agent Ref: 76798/00016 1 [00078] Fig. 17 is a block diagram of a negative feedback control system;
2 [00079] Fig. 18 is a representation of a data mining model suitably used with the system 3 and apparatus of Fig. 16;
4 [00080] Fig. 19 is a schematic diagram of MAXIcastTM casting machine working in an open loop (operator modifies process parameters based on bad wheel x-ray images) with 6 WHEELinspectorTM (ADR); and 7 [00081] Fig. 20 is a schematic diagram of MAXIcastTM casting machine working in an 8 open loop positive feedback with WHEELinspectorTM (ADR-Automatic Defect Recognition).
9 [00082] Corresponding reference characters indicate corresponding components throughout the several figures of the drawings. Elements in the several figures are illustrated for 11 simplicity and clarity and have not necessarily been drawn to scale. For example, the dimensions 12 of some of the elements in the figures may be emphasized relative to other elements for 13 facilitating understanding of the various presently disclosed embodiments. In addition, common, 14 but well-understood, elements that are useful or necessary in commercially feasible embodiments are often not depicted in order to facilitate a less obstructed view of the various embodiments of 16 the present disclosure.

19 [00083] Referring initially to Figure 1, a manufacturing process is exemplified by a hot runner system (100) that receives plastic in a molten state from an injection nozzle (10).
21 [00084] The nozzle (10) is part of an injection machine that includes a hopper (12), heater 22 (not shown), and feed screw (14) as is well known in the art. A control (180) controls operation 23 of the machine to perform the required sequence of operations to produce molded product. The 24 feed screw (14) delivers the plastic material under pressure to the nozzle (10) from where it is delivered through interconnected melt passages (16) to respective mold cavities (141).
26 [00085] The cavities (141) are formed within mold plate assemblies (140) that meet along 27 a common face. The mold plate assembly (140) includes a movable part (140a), and fixed part 28 (140b) may be separated to allow access to the cavity (141) for ejection of a molded article, and 29 are held closed during molding to contain the molten plastic.
[00086] The fixed part (140b) is connected to a cavity plate assembly (119) that includes a CA Patent Application Agent Ref: 76798/00016 1 manifold plate 120 to define the melt passages 16. A backup plate (121) supports the cavity plate 2 assembly (119) to permit the cavity plate (119, 120) to be changed readily without dismantling 3 the entire hot runner system.
4 [00087] Flow through the melt passages (16) in to the cavities (141) is controlled by a gate valve assembly (18) that is located in the backup plate (121) and extends through the cavity plate 6 assembly (119) to the cavities (141). Alternatively, the gate valve assembly may be incorporated 7 in the moveable part (140a) of the mold plate assembly (140) where the configuration of the 8 molded article permits.
9 [00088] As shown in FIG. 1, the cavity plate assembly (119) is configured for a multi cavity gated arrangement with a gate valve assembly (18) associated with each cavity (141).
11 [00089] The gate valve assembly (18) is shown in greater detail in Figure 2 and includes a 12 melt flow modulating assembly (102) and an actuator assembly (101).
13 [00090] The modulating assembly (102) includes a manifold bushing (132) that connects 14 through a manifold (131) located in a cavity of manifold plate (120) to the melt passages (16).
The bushing (132) is connected to an injection nozzle (113) that includes a melt flow channel 16 (112) to convey melt to a nozzle tip (114).
17 [00091] A backup pad (130) supports the bushing (132) against axial displacement.
18 [00092] A valve pin (110) extends from the actuator assembly (101) through the 19 modulating assembly (102) to control the melt flow from the nozzle tip (114).
[00093] The manifold plate (120) is used to house the manifold (131) and to distribute 21 molten plastic to each drop, as represented by an injection nozzle (113). The injection nozzle 22 (113) is sealably attached to manifold bushing (132) via a seal off (115) and detachably connects 23 the manifold bushing (132) with the injection nozzle (113).
24 [00094] Each the injection nozzles (113) is heated and the melt flow channel (112) extends therethrough from the rear end to the front end, and flowing into a mold gate (160) 26 located at the interface of the mold plate (140) and the cavity plate assembly (119). The mold 27 gate (160) is defined by a recess at the intersection of cavity plate assembly (119) and mold plate 28 assembly (140) that may have a conical shape. The frontal end of the drop is the nozzle tip (114) 29 and is a commonly replaceable part of the injection nozzle.
[00095] In the embodiments shown, the actuator (101) is illustrated as a Lorentz force 31 actuator assembly (101) hereafter referred to as the LFAA assembly (101). A Lorentz force 22928473.1 CA Patent Application Agent Ref: 76798/00016 1 actuator provides a linear force output proportional to a drive current and thereby allows the 2 force generated by the actuator to be modulated by modulating the current supplied.
3 [00096] The LFAA assembly (101) is generally placed in the metal pocket machined in 4 the backup plate (121). The backup plate (121) is water-cooled, and the LFAA assembly (101) is in at least partial thermal communication with the backup plate (121).
Preferably, the LFAA
6 assembly (101) is thermally communicating with the backup plate (121) via a partially threaded 7 connection or other type of connection means to permit thermal transfer.
8 [00097] By way of example, when cylindrically shaped, the LFAA
assembly (101) can be 9 placed in the pocket by partially or fully threaded connection that acts as a thermal bridge and improves cooling of the LFAA assembly (101). Alternatively, as shown in FIG.
6, a square-11 shaped configuration of the LFAA assembly (101) can be installed with an interference fit 12 generated by the operating temperature of the Lorentz force actuator assembly (101) and the 13 backup plate (121), or by partially connecting the structure of the LFAA
assembly (101) with the 14 backup plate (121) by cover plates (252) extending across the pocket.
[00098] A major advantage of this type of the installation is that the LFAA
assembly (101) 16 is accessible from the back of the backup plate (121) but at the same time, the installed LFAA
17 assemblies (101), being solid steel structure, strengthen the backup plate (121) at the point where 18 the manifold backup pad (130) (FIG. 2 and 4) transfers the seal-off forces generated by the melt 19 through the manifold bushing (132) from an injection nozzle seal off interface (115).
[00099] With the recent advent of high energy density rare-earth magnets, such as 21 Neodymium, Iron and Boron (Nd-Fe-B), and by modifying the electrical coil (104) accordingly, 22 it is now possible to construct a quite compact, yet powerful, valve gate actuator, such as the 23 LFAA assembly (101), that can under short duty cycle generate substantial axial linear force. As 24 will be described below, this short duty cycle (pulsed) force generation is used to position the valve pin assembly (110) in a desired position along a melt flow path (112).
26 [000100] The LFAA assembly (101) has at least two distinct assemblies: the magnetic 27 closed circuit assembly, and the electrical closed circuit assembly.
28 [000101] As shown in Figure 2, 3 and 6, the LFAA assembly (101) has generally two 29 distinct mechanical assemblies acting as the magnetic closed circuit assembly as well as structural parts. A yoke magnetic assembly (105) is used to conduct magnetic force, but also 31 provides a peripheral wall that bounds all parts of the LFAA assembly (101). A core magnetic 22928473,1 CA Patent Application Agent Ref: 76798/00016 1 assembly (106) is located within the yoke magnetic assembly 105. An annular air gap (116) is 2 located between the core magnetic assembly (106) and the yoke magnetic assembly (105). An 3 electrical coil (104) is located in the air gap (116). The coil (104) is wound on a bobbin (108) 4 that passes between the assemblies (105, 106) and has a base (145) connected to one end of pin (110), as described below. A yoke permanent magnet assembly (107) is placed between the inner 6 surface of the yoke magnetic assembly (105) and the core magnetic assembly (106), and thus 7 creates the magnetic flux.
8 [000102] The core magnetic assembly (106) and the yoke magnetic assembly (105) are 9 preferably made from the electronic and magnetic alloys with high magnetic permeability; the higher the permeability, the better the magnetic performance of the magnetic material. The core 11 magnetic assembly (106) presents a low magnetic resistance return path for the magnetic 12 induction generated by the strong permanent magnets. These two parts are operatively connected 13 to make up mainly uninterrupted closed magnetic circuit for the magnetic induction from the 14 strong permanent magnets to pass perpendicularly through.
[000103] High saturation properties of the yoke magnetic assembly (105) allow for higher 16 peak current in the coil, and therefore higher induction values before saturation is reached. This 17 allows for the designs of the LFAA assembly (101) that will function with greater force and 18 efficiency, but maintain a linear relationship between current and generated force, according to 19 the Lorentz Force Law.
[000104] Some of the exemplary magnetic alloys suitable for high force applications are:
21 Iron-Cobalt or Nickel¨Iron alloys with high magnetic permeability and high flux density. Uses of 22 the 430FR type of the ferritic chromium steel alloys have demonstrated good usability of the 23 application in the preferred embodiments.
24 [000105] It is highly desirable that the magnetic flux path of the core magnetic assembly (106) and the yoke magnetic assembly (105) is arranged so that the magnetic flux generated by 26 the core magnetic assembly (106), or the yoke magnetic assembly (105), or both the yoke 27 magnetic assembly (105) and the core magnetic assembly (106), which may be permanent 28 magnets, are perpendicular to the electrical coil (104) within the air gap (116), so that when an 29 externally applied current conducts through the electrical coil (104), the electrical coil (104) will be displaced axially along the axial magnetic air gap (116), and the amount of the displacement 31 is linearly proportional to the applied current.

22928473.1 CA Patent Application Agent Ref: 76798/00016 1 [0001061 The yoke permanent magnet assembly (107) can be made from any high quality 2 permanent magnets in the form of the magnet bars or elongated arcuate segments, magnetized 3 through the thickness of the bars or segments, and suitably arranged to cover the inner surface of 4 the yoke magnetic assembly (105) in a way to create a uniform unipolar field in the axial magnetic air gap (116). Preferably, a neodymium magnet (also known as NdFeB, NIB, or Neo 6 magnet), is used in the preferred embodiments, and it is a strongest type of rare-earth permanent 7 magnet. The neodymium magnet is a permanent magnet made from an alloy of neodymium, 8 iron, and boron to form the Nd2Fe14B tetragonal crystalline structure resistant to 9 demagnetization. General Motors and Sumitomo Special Metals developed these neodymium permanent magnets in 1982, but only recently are these magnets being made readily available.
11 The neodymium magnet has replaced other types of magnets in the many applications in modern 12 products that require strong permanent magnets. The most preferred type is in the class N52, 13 specifically designed for demanding mechatronic applications and is readily available. This type 14 of the permanent magnet is not susceptible to demagnetization due to high current flow in the electrical coil (104).
16 [000107] The permanent magnets (107) could also be placed on the core magnetic assembly 17 (106). The permanent magnet assemblies could also be distributed between the core magnetic 18 assembly (106) and the yoke permanent magnet assembly (107). These assemblies can be 19 paralleled by using sets of the electrical coils in parallel, where multiple coils would be acting on the valve pin assembly (110) and thus increasing the axial force.
21 [000108] A high density magnetic field in the axial magnetic air gap (116) is achieved by 22 the LFAA assembly (101). Other types of the magnetic structures like a ring magnet with radial 23 magnetization can be used as long as a uniform, high density, unipolar magnetic field is 24 produced within the axial magnetic air gap (116). In the preferred structure of this embodiment, as seen in the section of Figure 6 square bar magnets (117) are used to create a uniform magnetic 26 field in the axial magnetic air gap (116) so that force due to a current in the electrical coil (104) 27 is acting in the axial direction and is centered to drive the valve pin assembly (110) axially along 28 the melt flow channel (112), and not imparting any side forces on the valve pin assembly (110).
29 Bar magnets are used because no strong permanent magnet is available today in the form of the radially magnetized ring due to difficulty in manufacturing, but these may be available for 31 consideration in a preferred embodiment in the future.

22928473.1 CA Patent Application Agent Ref: 76798/00016 1 [000109] Referring back to FIG. 2 and 3, an electrical coil (104) is wound onto a coil 2 bobbin (108) that is positioned between the core magnetic assembly (106) and the yoke magnetic 3 assembly (105), and is free to move along the axial magnetic air gap (116). The electrical coil 4 (104) is preferably made from a highly conductive material like silver and/or copper but an aluminum coil may be used in some applications. The electrical coil (104) may be wound on the 6 coil bobbin (108), which is made from plastic composites. A fine balance is made between the 7 number of turns and the required axial force. Preference is given to structures without a separate 8 coil bobbin (108) where the electrical coil (104) has a winding and coil binder, or hardening 9 resins act together as self-supporting structures with solid integrity.
The electrical coil (104), made from the rectangular profile wire, intertwines with the Kapton ribbons or fibers and baked 11 in a high temperature resin, has the strength to support impulse forces expected in the preferred 12 embodiment of this invention. An electrical coil can be made from segmented individual turns, 13 preferably flat stamped, and only when assembled together are all the turns (of the coil) 14 electrically interconnected. As well, the electrical coil (104), in form of a "slinky", and/or made from silver or pure copper material, can provide a distinct advantage in the preferred 16 embodiments because the electrical coil (104) may provide a structure without the coil bobbin 17 (108).
18 [000110] Among the materials suitable for application to improve the structural integrity of 19 the electrical coil (104) is one of the DuPont (TRADEMARK) Kapton (TRADEMARK) MT
polyimide film, a homogeneous film possessing three times the thermal conductivity and cut-21 through strength of the standard Kapton (TRADEMARK) HN film. This polyimide film has 22 thermal conductivity properties that make it ideal for use in dissipating and managing heat in 23 electronic assemblies, such as printed circuit boards and electrical coils with high integrity 24 windings. It is anticipated that other materials and forms can be used for making reliable coils and, therefore, expand the applicability of this invention. High coil integrity is required due to 26 high acceleration and deceleration rates of the electrical coil (104).
27 [000111] The material used in the electrical coil (104), in some embodiments, could be 28 made from highly conductive soft magnetic alloys, to reduce the effective air gap and to increase 29 the valve-pin closing force. However, the opening force may be nonlinear and may be reduced with the use of such a coil.
31 [000112] Also, highly conductive graphite used in the electrical coil (104) when combined 22928473.1 CA Patent Application Agent Ref: 76798/00016 1 with oriented thermally-and-electrically-conductive nano-material structures with high axial 2 integrity may be used to support the axial force, and may be an example for some embodiments.
3 To modify coil performance, the air gap may be filled in with nano-magnetic fluids.
4 [000113] It is likely that some applications may require an electrical coil made over a bobbin by techniques well known in printing with thin and/or thick film or even by deposition of 6 the conductive coil material over layers of dielectric by spray techniques also known in the 7 industry. Other embodiments may select to photo etch the coil patterns, or even plate the coil 8 patterns but all of these and other techniques are anticipated by this invention.
9 [000114] Flexible power leads connect the electrical coil (104) to the control (180) that provides a current pulse electrical signal. Suitable flat litz wire or flat flexible ribbon can be used 11 for this application.
12 [000115] Referring again to FIG. 2, a mechanical pin locking assembly (126) is installed in 13 the bottom portion of the yoke magnetic assembly (105). As illustrated more fully in FIGS. 7 ¨ 9, 14 a pin-locking slide (122) is made to move laterally along lock slide guides (125). As can be seen in Figures 7 - 9, the assembly (126) includes a pair of jaws 502 that are slidably mounted on the 16 guides (125). Each of the jaws (502) has a recess (503) configured to conform to the outer 17 surface of the sliding pin (110). The opposed faces of the jaws (502) have magnetic holding 18 springs (501) embedded therein to provide an attractive force in the direction of closing around 19 the valve pin assembly (110). As can be seen in Figure 2, the pin (110) includes surface formations such as a lock thread 330, intended to be engaged by the jaws (502) and prevent the 21 valve pin assembly (110) from moving axially when operating power from the LFAA assembly 22 (101) is removed within the cycle of the molding of the plastics parts.
The axial length of the 23 surface formations is equal to an approximate length of the total axial stroke of the valve pin 24 assembly (110).
[000116] A pin-locking coil (123) is placed inside the yoke magnetic assembly (105) just 26 below the pin-locking slide (122). A set of pin locking permanent magnets (124) is carried by 27 each of the jaws (502) so as to be horizontally disposed above the coil (123). Energization of the 28 coil (123) causes the magnets (124) to apply a force to separate the jaws (502) and thereby 29 release the valve pin assembly (110) to allow axial movement when the LFAA assembly (101) is energized. In the arrangement shown, it is advantageous to install the pin-locking coil (123) 31 below the pin-locking slide (122) to facilitate assembly and improve cooling of the pin-locking 22928473.1 CA Patent Application Agent Ref: 76798/00016 1 coil (123).
2 [000117] As noted above, the valve pin assembly (110) has a surface formation indicated at 3 330 that provides a locking feature. The pin-locking feature (330) is formed as a thread and is 4 positioned some distance from the distal end of the valve pin assembly (110), and is located in the axial position of the mechanical pin locking assembly (126). In the preferred embodiment, 6 the pin-locking feature (330) is operatively engaged by the jaws of the lock slide guides (125) to 7 prevent any axial movement of the valve pin assembly (110) when the LFAA
assembly (101) is 8 de-energized. In this way, the pin-locking feature (144) operatively arrests motion of the valve 9 pin assembly (110) during a gate open condition or a gate closed condition within a molding operation.
11 [000118] Because of the short duration of the axial valve pin movement (less than 35 ms or 12 milliseconds), a relatively high current pulse can be used and not overheat the coil windings.
13 [000119] An active short duty cycle of the valve pin assembly (110) allows for long power 14 off time with a separate instance of a mechanical pin locking assembly (126) as will be discussed below.
16 [000120] The pin-locking coil (123) and the electrical coil (104) can be energized in a 17 required sequence determined by the mold sequence controller (180), or can be energized at the 18 same time to open the pin-locking slide (122) and move the valve pin assembly (110) axially.
19 [000121] The control of the coils (104,123) is provided by the electronic valve gate drive controller (400) that is part of the controller (180) and is shown as simplified block diagram in 21 Figure 10.
22 [000122] An energy storage capacitor (405) is provided and is capable of storing and 23 discharging, on demand, a certain calibrated amount of electrical energy into the electrical coil 24 (104) of FIG. 1. The energy storage capacitor (405) is operatively connected to conduct an electrical charge via the electrical wire supply (406) and the electrical wire return line (407) to 26 the electronic switch Q1 (410) and the electronic switch Q3 (430). The electronic switch Q1 27 (410) and the electronic switch Q3 (430) operatively control the directional flow of the electrical 28 current through the actuator coil (403) of the LFAA assembly (401). A
second set of electronic 29 switches Q2 (420) and Q4 (440) facilitates the flow of the electrical current through the actuator coil (403) by being operatively switched on and off in a precisely determined order and with 31 particular timing based on the input from the control computer.

22928473,1 CA Patent Application Agent Ref: 76798/00016 1 [000123] The electronic switch Q1 (410) and the electronic switch Q2 (420) cannot be in an 2 operatively ON state at the same time; this will cause a short circuit to the energy storage 3 capacitor (405). Also, the electronic switch Q3 (430) and the electronic switch Q4 (440) cannot 4 be closed in the ON state at the same time.
[000124] A duty cycle controller Qo (402) is provided to optimize and control the level of 6 charge in the energy storage capacitor (405). The duty cycle controller Qo (402) charges the 7 energy storage capacitor (405) via the electrical conductors that are suitably connected from the 8 energy storage capacitor (405) to the duty cycle controller Qo (402). The duty cycle controller 9 Qo (402) operatively charges the energy storage capacitor (405) in a predetermined and controlled sequence, operatively based and referenced to the operational cycle of the molding 11 apparatus. This is done in a way that the LFAA assembly (101) will be energized only when 12 axial motion of the valve pin assembly (110) is requested, with a particularly controlled duty 13 cycle, and this arrangement prevents damage to the LFAA coil assembly (104) due to 14 overheating temperature of the actuator coil (403) due to the high current. Recharging the energy storage capacitor (405) is required after each single axial movement of the valve pin assembly 16 (110) to ensure accurate capacitor charge and improve accuracy in the positioning of the valve 17 pin assembly (110). The LFAA assembly (101) is intended to operate only with a limited duty 18 cycle. In the preferred embodiment, the duty cycle should not exceed 25%. In another 19 embodiment, process demand for a short cooling time duty cycle may be less than 10%. The axial move time is preferably less than 10 ms (milliseconds).
21 [000125] The duty cycle indicates both how often the LFAA
assembly (101) will operate 22 and how much time there is between operations. Because the power lost to inefficiency 23 dissipates as heat, the actuator component with the lowest allowable temperature, usually the 24 actuator coil (104), establishes the duty-cycle limit for the complete instance of the LFAA
assembly (101).
26 [000126] The duty cycle is relatively easy to determine if the LFAA assembly (101) is used 27 on a molding machine, since the repeatable cycle of the molding machine has intervals when the 28 LFAA assembly (101) is demanded to be energized (during valve closing or opening only), and 29 de-energized (during mold cooling and part handling time). The provision of the pin locking assembly 126 enables a very short electrical actuator power ON time as there is no longer a need 31 to maintain power to the actuator coil (104) once the valve gate (110) is closed or the valve gate 22928473.1 CA Patent Application Agent Ref: 76798/00016 1 (160) is opened. The actuator coil (104), is ON only during the axial translation of the valve pin 2 assembly (110) from the first preferred position (usually open), to the second preferred position 3 (usually closed). During all other process times, the LFAA assembly (101) is locked into 4 preferred positions with no demand for power. The pin-locking stroke of the jaws (122) is very short, usually only as much as is required to maintain the arrest position of the valve pin 6 assembly (110). It is anticipated that the opening time of the pin locking assembly (126) is 7 scheduled before the valve pin assembly (110) is directed to move, although some overlapping in 8 sequence may be conceivable.
9 [000127] Operating on the edge of the molding's power curves, i.e. shortest possible mold cycle time, might incur the risk of the LFAA assembly (101) running hot.
However, the generous 11 cooling time available for solidification of the plastic in the mold enables heat in the coil (104) to 12 be dissipated. In most applications, molding PET or other food and medical moldings, where the 13 duty cycle is 5% or less, the LFAA assembly (101) can run to the limit of its power curves, once 14 the backup plate cooling is effective. The duty cycle of the electrical locking coil (123) has no limit on duty cycle because the coil impedance limits the excessively high current flow to cause 16 any overheating. The longer part of the operational cycle of the valve pin assembly (110) is 17 normally maintained by the permanent magnets, and all coils are without power and are self-18 cooled and are getting ready for the next movement cycle of short duration.
19 [000128] Referring back to FIG. 10, the actuator axial force is generated upon a request from the controller (180) for the energy storage capacitor (405) to discharge.
The controller 21 (180) implements computer software instructions to effect control (as is well known to those 22 skilled in the art of computers) of the gate drive controller (400). The controller (180) may 23 request closure of the electronic switch Q1 (410) and the electronic switch Q4 (440). The energy 24 storage capacitor (405) is operatively short-circuited by the electronic switch Q1 (410) and the electronic switch Q4 (440) and connected with the electrical wire supply (406) and the electrical 26 wire return line (407), will discharge a certain calibrated amount of electrical charge to move the 27 valve pin assembly (110) of FIG. 1 to the preferred position, be it in the open direction or the 28 closed direction. The amount of charge in the energy storage capacitor is selectable by the 29 operator, via the valve gate controller logic and the operator interface screen (182). At the same time, the coil (123) is energized to release the jaws (502).
31 [000129] Once a high current flows through the actuator coil (403), an axial force in the 22928473.1 CA Patent Application Agent Ref: 76798/00016 1 direction of the air gap (116) will push the valve pin assembly (110) axially to the desired 2 position. Once the position is reached with a slow speed (i.e., below 5 to 25 mm/s (millimeters 3 per second)). As shown in FIG. 11, the current pulse, shown by the dark line, progressively 4 increases and then decreases over the duration of the pulse. The corresponding force generated is shown in the lighter line and follows the profile of the pulse. The force applied to the pin (110) 6 initially accelerates the pin (110), causing it to move toward the closed position. The acceleration 7 is opposed by the inertia of the pin (110) and by the resistance of the motion of the pin (110) 8 through the plastic in the melt flow channel (112). As the current is reduced, the force generated 9 by the actuator (103) is correspondingly reduced and the velocity of the pin (110) progressively reduces, due to the resistance of the plastic melt, so that, as the pin attains the closed position, its 11 velocity approaches zero.
12 [000130] The shape of the pulse is selected to provide an optimum velocity profile in both 13 the closing and opening direction. It will be appreciated that a more aggressive declaration can 14 be obtained by reversing the direction of current flow in the coil 104 during movement, and that the pulse shape may be different between opening and closing directions.
16 [000131] Upon attainment of the open and closed position, the pin locking assembly (126) 17 will de-energize and lock the valve pin assembly (110) in the targeted preferred position. No 18 power is applied, nor is required, for the locking coil to hold the valve pin assembly (110) in the 19 arrested position. In a preferred embodiment of this invention, the force of permanent magnets (501) locks the valve pin assembly (110). When the valve pin assembly (110) is locked, the 21 electronic switch Q1 (410) and the electronic switch Q4 (440) are open (OFF).
22 [000132] Next, the duty cycle controller Qo (402) requests re-charging of the energy 23 storage capacitor (405) from the suitable bus voltage power supply (480) according to demanded 24 charge levels.
[000133] Once charged back to a demanded energy level, the request for movement of the 26 pin (110) from the closed position to the open position may be initiated by closing the electronic 27 switch Q3 (430) and the electronic switch Q2 (420). The locking coil is energized to release the 28 latch and permit movement of the pin (110). The controller (402) provides a current pulse to 29 move the pin (110) to the closed position and decelerate it at the closed position. The latch is released to hold the pin (110).
31 [000134] Since the transition time of the coil and energizing coil (123) in the modem power 22928473.1 CA Patent Application Agent Ref: 76798/00016 1 switching device is a fraction of a microsecond, modulation and intervention in the shape of the 2 energy pulse is possible to ensure formation of an accurate and most desirable pulse shape.
3 [000135] Total stroke time for the axial distance of about 7 to about 9 mm (millimeters) is 4 demonstrated to be about 5 to about 10 milliseconds, and is largely dependent on the size of the coil assembly of the LFAA assembly (101).
6 [000136] The control pin movement (110) by coil (104) and the high forces available make 7 it possible, in the preferred embodiment, to profile the end of the stroke to best meet the 8 demanding quality of the gate vestige without using complex servo controlled positioning based 9 on the position feedback device.
[000137] The nature and the application of the preferred embodiment for an injection mold 11 of the hot runner application allows for good vestige of the molded parts to be examined by the 12 operator for each cycle of the machine during setup and pre-run verification, and suitable 13 correction to the inputs can be made during the setup process to modify the pulse shape and the 14 closed pin position. As shown in Figure 11, adjusting the pulse shape at each iteration allows incremental changes in the position of the pin (110) until the optimum position is attained. Once 16 attained, the shape may be saved in the controller (402) for repetitive molding operations. If 17 inspection shows deterioration of the vestige, an adjustment can be made by the operator through 18 the interface (182).
19 [000138] The primary parameter for controlling movement of the pin (110) is the current supplied from the capacitor Q5. The rate of current is 50 to 100 A/millisecond (amperes per 21 millisecond), and the limit for the peak current is set by comparing the current feedback from a 22 current feedback device (455) at the electronic valve gate drive controller (400) and the set peak 23 current. The set peak current is an operator-controlled set point from the operator machine 24 interface computer (182) and is based on the preferred axial position for the valve pin assembly (110). This input, the capacitor charge, may be generated by the operator based on the part 26 quality observed or it may be automatically selected from a molding parameter set data matrix.
27 The log matrix in the form of the lookup table can be implemented to compare the vestige 28 quality with the valve pin position, as attained using a selected current pulse shape, and use this 29 information as a teaching tool for an optimum new position set point of the valve pin assembly (110). Controlling the current through the electrical coil based on a certain set value may be 31 accomplished by implementing the hysteretic control model where the hysteretic control circuit 22928473.1 CA Patent Application Agent Ref: 76798/00016 1 in the valve gate controller turns the electronic switch Q1 (410) or the electronic switch Q3 (430) 2 OFF when the current amplitude reaches the upper set point value, and then turns the electronic 3 switches back ON when it reaches the preset lower values amplitude point.
This control scheme 4 may be used in a standalone manner to improve valve pin positioning and therefore improve the part vestige, or in combination with other methods like iterative learning control (ILC) as 6 discussed more fully below. It will be appreciated that the control 400 will incorporate memory 7 buffers, set point comparators, timers and devices like digital microprocessors in the valve gate 8 driver circuit of FIG. 10, in various known ways, and packaged in the electronic valve gate drive 9 controller (400) suitably built to operatively interface with the operator input device and a molding machine computer to logically control the vestige quality by accurately positioning each 11 instance of the valve pin assembly (110). Control monitoring of the electrical current pulse by 12 current-sensing power MOSFETs provide a highly effective way of measuring load current 13 through the electrical coil (104) in FIG. 1 of the LFAA assembly (101).
14 [000139] The current plot of the current is compared to the preferred plot to reduce positioning error for the valve pin position. Valve pin positioning accuracy of plus or minus five 16 micrometers can be achieved by implementation of-Iterative Learning Control (ILC) in valve pin 17 positioning in the hot runner systems or injection molding, and can improve valve pin 18 positioning.
19 [000140] The use of ILC is shown schematically in Figure 15 and is a method of improving control parameters for systems that function in a repetitive manner. The repetition involved in 21 valve positioning provides typical conditions to use iterative learning techniques. There is room 22 for significant positional accuracy improvement and, therefore, improvement in the vestige 23 quality of the moldings when using the Lorentz force actuator assembly (101) in FIG. 2. ILC can 24 be implemented in any system that is required to perform the same action for millions of cycles with high precision. Each repetition or cycle allows the system to improve tracking accuracy, 26 gradually learning the required input needed to track the reference to a small margin of error.
27 The learning process uses information from previous repetitions to improve the control signal, 28 ultimately enabling a suitable control action to be found.
29 [000141] Through iterative learning perfect tracking of the valve pin position can be achieved. Perfect tracking is represented by the monotonic convergence of the mathematical 31 model. Iteration allows for monotonic convergence to achieve more accurate positional accuracy 22928473.1 CA Patent Application Agent Ref: 76798/00016 1 of the valve pin assembly (110) in a molding application of the hot runners. Experiments 2 demonstrate convergence within 5 to 10 pin cycles. After achieving convergence the valve pin 3 assembly (110) will be able to operate in a stable state.
4 [000142] To improve the speed of convergence fuzzy logic can be implemented as part of the ILC. Improved parameters for the ILC algorithms can be attained through the use of external 6 sensor feedback. These sensors could include; x-ray sensors, electromagnetic sensors, or other 7 appropriate sensors that would provide meaningful information. The ILC
algorithms and fuzzy 8 logic parameters can be updated in real-time or through analysis of previously collected and 9 stored data, as will be described more fully below with respect to an alternative embodiment of molding apparatus. In the present embodiment the result of the combination of real-time dynamic 11 parameter modification is a self-tuning system that will have automated tracking accuracy of the 12 open loop valve pin positioning. The fuzzy system is used to precisely position the valve pin tip 13 of FIG. 3 without installing any physical motion or position feedback device or structure within 14 the valve pin stroke. Turning back to FIG. 10, the inherent characteristic of the actuator coil (403) of the LFAA assembly (104) to slow down in the magnetic field when coil terminals are 16 short-circuited can be operatively used to slow down the axial movement of the coil of the LFAA
17 assembly (104). This can be accomplished by opening the electronic switch Q2 (420) and the 18 electronic switch Q4 (440), and closing the electronic switch Ql (410) and the electronic switch 19 Q3 (430). The flyback diodes (451, 452, 453, 454) functionally support the switching operation of the electronic valve gate drive controller (400).
21 [000143] The operation of the gate valve assembly will now be described, assuming 22 initially that the pin (110) is held in an open, i.e. retracted position by the jaws 502 engaging a 23 lower portion of the screw thread (144). The coils (104, 123) are de-energized and the jaws held 24 closed by action of the magnets (501). The first step in the operation of the LFAA assembly (101) involves application of the suitable electrical current pulse through the electrical coil (104) 26 (FIG. 2), and at the same time the application of the voltage to the pin-locking coil (123). The 27 pin-locking coil (123) reacts with the pin locking permanent magnets (124) and separates the 28 jaws to unlock the valve pin assembly (110). The current applied to the electrical coil (104) 29 operatively reacts to a magnetic field from the permanent magnets, generates a force perpendicular to the current flow through the electrical coil (104), i.e. in the axial direction of the 31 pin 110. Therefore accelerates the valve pin assembly (110) in the direction of the actuator force 22928473.1 CA Patent Application Agent Ref: 76798/00016 1 pushing downward to close the mold gate as per FIG. 1.
2 [000144] The next step in the operation of the LFAA assembly (101) is to decelerate the 3 valve pin assembly (110) to the gate closing point, but not to impact the mold gate (160) and to 4 cause any damage by hard stops. The electronic valve gate drive controller (400) modulates the current pulse to follow the shape of the current profile in the controller memory predetermined 6 by experiments for the type of the product that is being molded. The electronic valve gate drive 7 controller (400) follows a shape of the current signal already stored in the controller memory 8 within the proportional hysteresis bandwidth and based on the current feedback from the 9 electronic drive controller, and determines optimal deceleration slope of the current pulse. The electronic valve gate drive controller (400) has the ability to brake by shorting the electrical coil 11 (104) by switching the appropriate electronic switches in FIG. 10 (specifically, the electronic 12 switch Q1 (410) and the electronic switch Q3 (430) are set OFF, and the electronic switch Q2 13 (420) and the electronic switch Q4 (440) are set ON) to accurately stop the valve pin assembly 14 (110) and "soft land" the valve pin assembly (110) into a gate closed position or any preferred position within the axial stroke of the LFAA assembly (101).
16 [000145] When it is anticipated that the valve pin assembly (110) has arrived at the 17 preferred position, the pin-locking coil (123) de-energizes, and the jaws (502) moved under the 18 influence of the magnets (501) to operatively engage the threaded portion (144) (a high friction 19 area) of the valve pin assembly (110) by the attractive force of the magnetic holding springs (501) in FIG. 5. Any residual motion of the valve pin assembly (110) is arrested. In this 21 condition, no current is required to hold the valve pin assembly (110) in a gate-closed position, 22 as shown in FIG. 5.
23 [000146] In this position, the valve pin assembly (110) extends through the mold gate (160) 24 and blocks the flow of the molding material through the mold gate (160) (or the mold gate channel).
26 [000147] The next step involves cooling of the moldings in the mold cavity (141), ejecting 27 the molded part from the mold cavity (141) by opening mold (140), and closing the injection 28 mold. The cooling process of plastic parts takes time. Plastic solidification and part removal from 29 the mold cavity (141) is at best five to ten times longer than the time to inject molten material into the mold. Thus, there is a substantial amount of time where the valve pin assembly (110) is 31 resting in a closed position and is de-energized. The mold core portion (140a) and the mold 22928473.1 CA Patent Application Agent Ref: 76798/00016 1 cavity portion (140b) are movable relative to each other, and when the part has solidified, the 2 mold is opened and the part ejected. After ejection, the mold core portion (140a) and the mold 3 cavity portion (140b) are positioned to abut each other so that the mold cavity (141) is formed, 4 and the resin or the molding material may again be injected into the mold cavity (141).
[000148] The controller (400) thus energizes coil (104) to retract the pin (110) and the coil 6 (123) to release the jaws (502). The pin (110) is retracted and braked by the current pulse from 7 the controller (400) and the jaws again engaged to hold the pin (110) in the open position.
8 [000149] Thus the mold gate (160) is opened by moving the valve pin assembly (110) 9 upward in the preferred position within the axial stroke of the valve pin assembly (110). The valve pin assembly (110) is designed to open and/or increase the cross-sectional area of the mold 11 gate (160) with the coil bobbin (108) and the pin-locking coil (123) energized, to allow the flow 12 of molten resin into the mold cavity (141).
13 [000150] In some embodiments, the axial stroke can be 8 to10 mm (millimeters) which is 14 deemed sufficient to avoid adverse effect of the annular flow for most medical moldings and the molding PET preforms.
16 [000151] The LFAA assembly (101) exploits the inherent characteristic of the injection 17 molding process and the hot-runner system (100), where the plastic cooling takes a much longer 18 time in the process than the injection of the polymer the molding material) into the mold cavity 19 (141). Therefore, it is possible to operate the LFAA assembly (101) in a condition of significant current pulse overdrive, limited only by the thermal limitations of the LFAA
assembly (101).
21 The method for modulating melt flow within the hot-runner system is obtained by generating a 22 significant axial electrical force for a very short time lasting 5 to 20 ms (milliseconds). In a 23 preferred embodiment, the valve pin assembly (110) moves along the melt flow channel (112) by 24 the LFAA assembly (101) when powered from the current pulse power supply (480). Thus, positioning is accomplished by controlling the pulse current amplitude as a function of time. This 26 allows the actuator operation only during axial movement of the electrical coil (104), leading to 27 reduced operational time within the thermal limitation of the electrical coil (104). During the 28 cooling part of the cycle, the electrical coil (104) is de-energized but locked by the magnetic 29 holding springs (501) of the permanent magnet. The magnetic holding springs (501) rely on magnetic attraction or repulsion to control the force of the locking mechanism. The magnetic 31 holding springs (501) have a significant life and are a very consistent and reliable means of 22928473.1 CA Patent Application Agent Ref: 76798/00016 1 creating a spring force.
2 [000152] It will be noted that no mechanical spring is utilized in the preferred embodiment 3 shown in Figures 2 through 10.
4 [000153] Experimentation with and measurements of the embodiments shown demonstrated operational efficiency with the duty cycle of the LFAA assembly (101) up to 25%
6 for an application involving the packaging molding processes, like PET
preforms, closures and 7 coffee cap multi-material moldings.
8 [000154] An alternative embodiment of a Lorentz force actuator assembly is shown in 9 Figure 12, in which like components are identified by like reference numerals with a suffix "a"
for clarity. In the embodiment of FIG. 12, provision is made for enhanced guidance of the pin 11 assembly (110a). Referring therefore to FIG. 12, a Lorentz force actuator assembly (101a) is 12 operatively connected to a valve pin assembly (110a) having a yoke magnetic conductor (105a) 13 and a core magnetic conductor (106a). The core magnetic conductor (106a) and the yoke 14 magnetic conductor (105a) operatively support permanent magnets facing an axial air gap (116a). A yoke permanent magnet assembly and a core permanent magnet assembly (209) are 16 each facing an electrical coil wound as a self-supporting structure or being structurally supported 17 by the coil bobbin (108a). To close the magnetic circuit, a base plate magnetic conductor (206) is 18 provided. The electrical coil (104a), designed to conduct a high current pulse, is placed on the 19 coil bobbin (108a) that is designed to operatively move along the air gap (116a) in the axial direction under an electrical force acting in concert with a permanent magnet field in the air gap 21 (116a), thus creating suitable a condition to generate an electrical force that is perpendicular to 22 the coil current flow and, by definition, in a direction according to polarity of the current pulse.
23 This force is directly and linearly proportional to the current amplitude and increases in the valve 24 pin closing direction.
[000155] A pin locking assembly (126a) is positioned at the distal end of a valve pin 26 retainer (240), and the pin locking assembly (126a) operatively arrests any movement of the 27 valve pin assembly (110a) when the electrical locking coil (123a) is de-energized. There is a 28 locking slides air gap (220) between the electrical locking coil (123) and the locking magnet 29 (124a). The pin locking jaws (502a) are guided by the locking slide bearings (125a). When the electrical locking coil (123a) de-energizes, the permanent magnet assembly attracts the two jaws 31 toward each other to close and engage the pin locking rib (230) formed on the pin 110a and to 22928473.1 CA Patent Application Agent Ref: 76798/00016 1 arrest any motion of the valve pin assembly (110a). The pin locking assembly (126a) can be 2 placed along a length of the valve pin assembly (110a), as well as attached to any axially moving 3 part. Magnetic holding springs as shown in Figure 5 are used for attracting the pin locking jaws 4 (502a) around the valve pin assembly (110a).
[000156] Referring now to FIG. 12, once placed in the backup plate (121a), an actuator 6 cover plate (252) is installed. The actuator cover plate (252) is also preferably manufactured 7 from a soft magnetic material. The actuator cover plate (252) has an opening 253 where a valve 8 pin retainer (254) can be guided during the axial movements of the valve pin assembly (110a).
9 This opening can also be used as an access for accurate pin height adjustments. Additionally, the electrical coil bobbin guides (250) are provided to prevent potential side loading on the valve pin 11 assembly (110a) and improve axial motion of the electrical coil (104a).
Pin upper position 12 holding magnets (251) are provided for some embodiments to hold the pin (110a) in an open 13 position.
14 [000157] To ensure accurate and precise alignment for the valve pin assembly (110a) and the electrical coil (104a) generating the axial motion, additional guides (250) are incorporated in 16 the bobbin (108a).
17 [000158] Rib (230) is formed as a conical portion of progressively enlarged diameter which 18 thereby provides a radial abutment surface facing the actuator (101a). A
number of such ribs 19 may be provided a discreet location on the pin (110a) to provide multiple stable positions.
[000159] In use, the jaws (502a) operatively engage the valve pin assembly (110a) through 21 the pin locking rib (230) and are separated as the pin (110) moves toward the closed position by 22 the electrical coil (104a). The coil (123a) I used to separate the jaws (502a) to release the pin 23 (110a) to move to the open position. The electrical locking coil (123a) is placed below a locking 24 magnet to ensure better coil cooling by the backup plate (121 from FIG.
1). The locking magnet may be a permanent magnet.
26 [000160] In the embodiment of Figure 12 the pin locking assembly (126a) operatively 27 engages the valve pin assembly (110a) against the pin locking rib (230) in a pin closed position.
28 When the valve pin assembly (110a) is in a fully opened position, the pin upper position holding 29 magnets (251) are used as an alternative to bi-directional locking.
[000161] A further embodiment is shown in FIG. 13 where like elements are identified with 31 like reference numerals with a suffix "b" for clarity. Referring now to the example of FIG. 13, 22928473.1 CA Patent Application Agent Ref: 76798/00016 1 there is a valve pin retainer (254b) shown at the distal end of the valve pin assembly (110b) with 2 a suitably arranged structure for precise adjustments of the valve pin protrusion, in a form of a 3 valve pin height adjustment assembly (651). A valve pin height adjustment assembly (651) 4 comprises a valve pin guide bushing (604) and a fine-tuning pin indicator providing auditory and visual measure of adjustments in certain "clicks". Each incremental position represents 6 selectively 5 to 25 micrometers of linear movement of the valve pin assembly (110b). The 7 audible "clicks" are generated by the valve pin sound feedback lock (601) in the valve pin 8 adjusting assembly (600) made with the permanent magnet spring (602) shown in FIG. 6. This is 9 a technical structure that will benefit fast service and maintenance of the valve pin assembly (603). A slot (606) is a slot for inserting an adjusting tool for tensioning the magnetic holding 11 springs (501).
12 [000162] As an alternative a method for arresting or locking the axial movement of the 13 valve pin assembly (110), a rotary arrangement using rotary locking slides may be used. This can 14 include a rotating lock using a collet assembly that can effectively maintain the valve pin assembly (110) at rest when the power to the LFAA assembly (101) is turned off.
16 [000163] As a further alternative, a method for arresting or locking the axial movement of 17 the valve pin assembly (110) can be done by utilizing principles of a smart material that changes 18 the volume or the linear dimension by the application of the electrical signal. Some materials of 19 this nature are crystals like quartz, often used to generate and receive a signal. Other well-known materials sensitive to a voltage charge are lead zirconate-titanate or well known as PZT. The 21 science and art of smart materials are replete with materials that can change size by the 22 application of an electrical charge. Material properties like electrostriction, or crystal material 23 matrix reconfiguration, or molecular or particular realignment, are some of the properties of 24 materials like Nitinol, electro rheological fluids or nano-ferofluides, etc.
[000164] Since various aspects and structures are possible to use to arrest the movement of 26 the valve pin assembly (110), one example of the preferred arrangement is mentioned, where the 27 valve pin assembly (110) operatively incorporates an electrically-charge sensitive smart material 28 that contracts radially when the electrical signal is applied to opposite axial sides of the smart 29 material (in the form of an insert) to maintain easier axial motion of the valve pin assembly (110). When the signal is not present, the smart material recovers in the normal condition of zero 31 energy state where the majority diameter is larger than the average diameter of the valve pin 22928473.1 CA Patent Application Agent Ref: 76798/00016 1 assembly (110). This may create a friction force against a tightly fitted valve pin assembly (110), 2 and the surrounding coaxial structure may arrest any motion of the valve pin assembly (110).
3 This is a simple and effective way to arrest the axial motion of the valve pin assembly (110) in 4 situations where the mechanical perpendicular obstruction is not desirable for whatever reason.
Opposite arrangements, where the contracting ring inner diameter may arrest the movement of 6 the valve pin assembly (110) in position is envisioned.
7 [000165] The above description of the arrangement and operation of a molding system and 8 apparatus has demonstrated the collection of data and the use of that data to the control of the 9 injection apparatus to control the quality of the molded component. While this approach is possible, it requires a very skilled level of people involved with setup of the machine and 11 process. Firstly, the person interpreting images of the gate vestige or complete part(s) must be an 12 expert in analyses of the x-ray imagery or IR imagery or other very specialized qualitative 13 inspection method, and then the same person or different must be expert in process adjustments 14 and interpretation of the inspection findings. The second person must take the quality inspected part and convert the knowledge about the actual part into a set of relevant parameters that can be 16 applied to input of the machine to hopefully produce new part with better and more exact quality 17 characteristics.
18 [000166] To mitigate these challenges, a further embodiment of a molding system and its 19 control system is shown in Figures 16 to 20. In that embodiment, the manufacturing process is exemplified as a metal casting machine, such as is used in the production of "alloy" wheels, 21 although the general applicability to other processes, particularly the molding of other parts, 22 including plastic parts will be appreciated.
23 [000167] In general terms, the molding machine will produce a part and transfer it into an 24 inspection system (IS) with at least one part inspection sensor from those well known in the state of the art, like X-ray, UV, IR or visible light. The inspection system IS will look into, for 26 example, gate vestige of the molded article, which is related to the pin position, and create an 27 image or set of digital data that characterizes this specific part vestige quality. Assuming that, 28 when compared to drawing digital data from the design, the part is good and acceptable, the 29 process will automatically transfer serialized set of digital data created by the inspection system and store it in the Good Part Data Set block (GPDS) shown in Fig. 16. At the next step this 31 GPDS data will also be shared with the KSDAP (Knowledge System for Data Analyzing and 22928473.1 CA Patent Application Agent Ref: 76798/00016 1 Prediction) which function block contains all mathematical algorithms for retrieving necessary 2 information to comparing data and running algorithms to determine if a change to the process is 3 required. Since in this step no changes are warranted because the process has generated good part 4 vestige, information about part will be transferred into Parameter Generator (PG) to create new relevant set of inputs that do not change the process at this moment. The machine continues to 6 produce parts with the valve stem positioning as before where the iterative learning controls 7 maintains status. The process continues, and inspection continues, until the inspection system 8 detects variations in the gate vestige. The gate vestige may be evaluated for example by using a 9 visible image of good vestige as an image template and correlating this image with proper valve stem position. Subsequent images are then compared to the image template and the information 11 transferred to the Parameter Generator (template matching).
12 [000168] If inspection shows the vestige is unacceptable and part is no longer good quality, 13 information about this part is transferred in to Bad Part Data Set base (BPDS). The KSDAP
14 takes BPDS and compares it to information from GPDS and creates set of parameters indicating need to adjust the valve stem to improve vestige. This information is now transferred to PG to 16 generate new set of inputs for the machine (of any kind, molding, casting, CNC, additive mfg. or 17 valve positioning) to iteratively adjust, in particular example, stem position and injection 18 pressure. After Production system (PS) makes new part, and IS inspect the new part, it was found 19 that next part is good and cycle continues.
[000169] This arrangement may of course be applied to inspection of any produced part 21 and use of the data from the inspection, compare it to data from data bases of good and/or bad 22 parts and compare this to original data set for the design model data set (DMDS) and determine 23 in the KSDAP how to increment the process parameters to create new set of inputs to the 24 machine that will produce next good part and continually do that for the production run, until interrupted by the operator locally at the machine or remotely stopped by the intelligent system 26 operator.
27 [000170] A more detailed example of the application of the invention is its use in the die-28 casting of an alloy wheel rim shown in the Fig 19. A production system PS represents the 29 typical well known die-casting machine. This machine may be any machine according to standard state of the art, high pressure, low pressure, cold chamber, hot chamber, or 31 thixomolding injection molding machine as well as MAXIcastTM semi-solid casting machine as 22928473.1 1 described in US patent application US 2015/0266086.
2 In the embodiment shown in Fig. 19, the material is supplied to the mold 3 cavity from a solid feedstock which is melted to slurry and processed to the runner system 4 associated with the mold used to produce the product. The production system PS has the controls necessary to adjust the operating parameters of the die casting machine as described above, 6 including an iterative learning control to adjust operation of the valve.
Once cast, the die cast 7 wheels are ejected from the production system PS and transferred to an inspection system IS.
8 [000171] The Inspection system(s) accepts the cast parts and proceeds with automatic 9 inspection according to one of the known or customized testing and inspection protocols. The items are inspected for critical dimension or other critical features of the part. If for example, 11 linear dimension of the item is critical for quality, the linear dimension of the parts is measured 12 and digitized preferably by electromagnetic measurements like laser. If however surface 13 appearance is critical to quality feature measurement for surface irregularities with one of the 14 well-known methods is utilized. Usually, the surface profile and roughness of a machined work piece are two of the most important product quality characteristics and in most cases a technical 16 requirement for mechanical products. Achieving the desired surface quality of the finished 17 castings is of great importance for the functional behavior of a part.
The process-dependent 18 nature of the surface quality mechanism along with the numerous uncontrollable factors that 19 influence pertinent phenomena, make it important to implement an accurate prediction model.
Final improvements in prediction of surface profile using modern neural network may be used to 21 ensure desired high quality surface of the part.
22 [000172] It is also necessary to inspect parts for internal structural abnormalities like 23 discontinuation or large gas inclusions that may under operating conditions create stress 24 concentration or will prevent heat treatment of the casted part in a subsequent step. Ultimately, it may be necessary to determine the micro structure of the part volume by digitizing caste(' parts 26 and automatically determining if part is suitable for use for the intended purpose. The parameters 27 measured by the inspection stations IS are communicated to a computer system CS which also 28 communicates with the operator GUI and the TT system of the plant. The control of the 29 production system is available through the GUI, but as will be explained below, can be enhanced through the integration with an integrated knowledge management system, IK.MS, as shown in 31 Figure 16.

Date Recue/Date Received 2022-11-08 CA Patent Application Agent Ref: 76798/00016 1 [000173] As shown schematically in Fig. 16, the integrated knowledge management 2 system IICMS is organized as a network of interconnected functional modules that co-operate to 3 produce molded components, assess their quality according to a variety of criteria, collect and 4 store data on the components and adjust the production process as necessary to attain the desired quality. The system includes the production cell PS that produces the molded component and 6 delivers it to an inspection system IS. The inspection system collects data indicative of the 7 quality of the component and associates the information with a unique identifier indicative of the 8 component that is stored in the part identifier PI. The part identifier PI delivers the data to a 9 knowledge system KSDAP which communicates with a summing junction where the manual input at the process start from the user interface is provided. The output of the summing junction 11 is used to generate operating parameters PG used as inputs to the production cell PS.
12 [000174] The part identification PI and knowledge system KSDAP
exchange information 13 with supplementary data sets that include an original production data set OPDS that represents 14 the original input parameters determined by the operator at the process start, and a good part data set GPDS which represents the corresponding parameters for those parts that pass the quality 16 criteria at the inspection system IS. Similarly, a bad part data set is compiled from the parameters 17 of components that fail the criteria set by the inspection system IS. A
DMDS data set containing 18 a 3D representation of the component is also available to the knowledge system and the part 19 identification.
[000175] The inspection system may also communicate via communication module WiC to 21 remote locations to share data on similar production facilities.
22 [000176] With such a system, there is a need to see/check/measure micro structure density 23 of the melt or semi-solid slurry before injecting into a mold. This may be achieved using a real-24 time 2D x-ray grayscale image of melt/semi-solid slurry can be correlated to pre-stored optimal grayscale image (template matching). Or 1D density measurement using x-ray, gamma or 26 radioactive sources with proper signal evaluation detector can be also used. The data signals 27 obtained are used to control temperature of the main reactor at MAXIcastTM semi-solid casting 28 machine in a close loop manner fully automatically to get optimum content of solids in liquids.
29 These data signals are fed back to Parameters Generator (PG) to generate improved input(s) at the casting machine.
31 [000177] The Production System (PS) may not need to be modified or any different than 22928473.1 CA Patent Application Agent Ref: 76798/00016 1 the ones readily available on the market today which typically have suitable interface 2 communication channels, to ensure fast data exchange. Data exchange over the communication 3 channel is used to accept input parameter, transfer output parameters to inspection system and 4 communication with a part identification module (PI), Knowledge System as well as various data set data basis as shown in the Fig 16. It is understood that Production System (PS) may include 6 all required auxiliary equipment commonly associated with Production System inputs and 7 outputs.
8 [000178] An event log of the process parameters for each cast part is made available to 9 Original Production Data Set (OPDS) for warehousing and future processing. Each part cast in the PS will get assigned unique identification code for tracking and further processing. Once 11 cast, the part is serialized and ejected from the machine, it is then sent to at least one inspection 12 system(s) for inspection and analysis. It is understood that inspection system (IS) can be remote 13 or distributed as well as fully integrated in each production system (PS).
14 [000179] Even if inspection of the part indicates it is not of satisfactory quality, this data set of the bad part will be stored in the Bad Part Data set (BPDS) and some valuable information 16 from the bad part will be used to accelerate convergence by using in the new parameters 17 generation for the subsequent casting.
18 [000180] Output from the inspection system module is in digital or analog format suitable 19 to be transferred to part identification module to associate information about part received from inspection system (IS) and store it with production System (PS) original data set (OPDS). It is 21 also envisioned that some of the inspection and OPDS data sets will also be transferred to 22 centralized or distributed data centers via wireless communication networks or Internet for 23 further comparisons or processing. Some other location may be molding similar or same part and 24 inspection system from this location may be beneficially used to improve quality of the remote castings.
26 [000181] The parts identification (PI) module associates production data set for each part 27 and inspection system data sets and pass this to other modules as required. The PI data set will 28 contain critical to quality parameters and defect recognition data based criteria's. The part 29 machine parameters and inspection system parameters are available to other parameters data base structures for further processing and assignments via regular communication channels indicated 31 with thin line. Critical communication protocols are used for high volume data exchange 22928473.1 CA Patent Application Agent Ref: 76798/00016 1 between data buses and production and inspection system and are shown in Fig. 16 with a thick 2 line.
3 [000182] Any information useful for the process is shared by remote users via wireless or 4 cable internet (intranet) communications protocols and stored in the Knowledge System (KSDAP) data block. Knowledge System for Data Analyzing and Prediction (KSDAP) is a tool 6 box with access to a library of algorithms that can be called to analyze and respond to the data 7 presented. Suitable algorithms include data mining, artificial intelligence, neural networks, K-8 Nearest Neighbor algorithm (KNN), Restricted Coulomb Energy algorithm (RCE), SPC, QLF by 9 Taguchi, fuzzy logic and 1LC. These algorithms may be used alone or in any other combination to indicate to the parameter generator PG, the adjustments to the inputs to the production system 11 PS.
12 [000183] As indicated in Figure 18, the Knowledge System for Data Analyzing and 13 Prediction (KSDAP) manipulates the data received to extract information applicable to the 14 control of the production system PS. The KSDAP compares the digital information data set for each part from the inspection systems with prior similar parts information data sets and 16 determines best strategy for part quality improvements. The KSDAP
ensures that only relevant 17 data sets are modified and prepared to be iteratively varied so that changes to some parameters 18 does not lead to worsening of others and convergence takes much longer to be attained. Various 19 algorithms are used to prepare data set for Parameter Generator (PG).
[000184] The Parameter Generator (PG) utilizes vector Iterative learning control (ILC) to 21 modify relevant parameters determined in KSDAP data base to ensure that next iterative process 22 leads to improved part quality. An example is provided above where the single axis positioning 23 of the valve pin operated by the electromagnetic actuator in molding apparatus implements 1LC
24 to slowly approach desired pin position or control melt flow through the gate. In a similar manner, the Parameter Generator PG may influence other multiple machine axes and processes 26 like heat and pressure to attain the required quality of the component.
27 [000185] Finalized input parameters are generated as machine inputs at the parameter 28 generator PG and directly communicated to Production System to execute as new iteratively 29 improved set of inputs.
[000186] During iterative learning process at least one of the set of data for following data 31 bases is used: Good Part Data Set (GPDS), Bad Part Data Set (BPDS), Designed 3D model Data 22928473.1 CA Patent Application Agent Ref: 76798/00016 1 Set (DMDS) as well as Original Production Data Set (OPDS) for generating and iteratively 2 improving at least some machine inputs to ensure higher quality part is casted in subsequent 3 cycle.
4 [000187] The GPDS contain all digital information related to an acceptable part. All inspected parts that are acceptable to users are stored in this database. The BPDS contain all 6 digital information related to defective parts. Defect may be dimensional, surface or volume and 7 may contain information on the part micro structure and structural inner inclusions or voids. The 8 DMDS contains data information about part original design and critical to part inspection items.
9 The OPDS data set contains all production parameters related to each produced part be it good or bad.
11 [000188] During iterative process these and other information will be used to improve 12 production system input parameters. These production parameters related to high quality part 13 will be available as a record to be printed when part enters the stream of commerce and requires 14 traceability. The design 3D model data set (DMDS) is used as a reference for desired outcome.
All critical dimensions and digitized vector information about parts will be stored in this data set.
16 [000189] The IKMS is incorporated in the production system PS
as indicated in Fig. 20.
17 From which it can be seen that the control input to the production system PS is provided from the 18 parameter generator PG and the GUI communicates through the KSDAP, where changes made 19 by the operator can be evaluated and adjustments implemented accordingly.
[000190] On startup the operator will turn power to the production system, follow initial 21 checklist related to supply of the material and initial parameters and safety verification and press 22 the start button. The production system PS will start casting the first part and upon completion 23 the component is output from the production machine and transported to the inspection system 24 IS. The inspection system IS determines the quality of the part. If part is bad, information is stored in bad part data set data base BPDS. If part is good all information about this good part 26 will be stored in good part data set data base GPDS. Assuming that the part is bad, the 27 knowledge system KSDAP will determine which set of parameters needs to be adjusted 28 iteratively to improve part deficient characteristics and a new set of inputs is generated in 29 parameter generator PG. By way of a hypothetical example, the inspection may indicate that the wheel is not completely formed, indicating insufficient feedstock injected in to the mold. The 31 inspection system produces a set of image(s) to construct 3D model or x-ray 3D-Computed 22928473.1 CA Patent Application Agent Ref: 76798/00016 1 Tomography model that when compared to the DMDS (3D design model) by the KSDAP there 2 is an indication of a missing portion. The KSDAP interrogates the data base for similar deficient 3 castings and determines the corrective action is to delay the closing of the pin, KSDAP
4 communicates to the PG which iteratively instructs the parameter generator to delay the closing of the valve pin and increments the open period of the pin to provide a longer injection period.
6 [000191] The next machine cycle is executed, serialized and passed to the inspection 7 system. More data from the inspection system is generated and more information allows for 8 faster convergence to a set of parameters that yields a good quality part. Let us suppose that this 9 time part linear and surface dimensions are acceptable but an x-ray interrogation uncovered air entrapment in the body of the part with diameter 300 micrometer. The knowledge system 11 determines an appropriate step in the iterative process will be to increase vacuum in the mold and 12 increase pressure of the injection ram.
13 [000192] Again, the parameter generator PG will communicate through the iterative 14 learning control to modify suitable relevant parameters and produce 3rd part. When this part is again x-ray inspected it is determined that inclusion is no longer there. The component is 16 acceptable and information about good part is stored in the good part data base GPDS. The 17 production system PS will now use those input parameters that produced a good part as a new 18 reference set of good inputs and maintain and inspect all new parts until deviation from good 19 quality is observed. At that point, the knowledge system KSDAP may access the data collected and determine the appropriate parameters to be adjusted through the iterative learning controls.
21 All information stored will be retrievable and subject to comparisons with design requirements.
22 Information about good set of parameters that produced good parts at similar conditions may be 23 then made available to any system worldwide casting same wheel type.
24 [000193] Significant savings can be had with this approach for very large production series of the identical parts.
26 [000194] Having described preferred embodiment of process of producing alloy wheel by 27 closing the loop with data from an inspection system and accelerating convergence with data sets 28 from good parts and bad parts as well as 3D model of the original design, it will be appreciated 29 that this approach allows processes to produce high integrity molded components with minimal rejects. It is believed that other modifications, variations and changes will be suggested to those 31 skilled in the art in view of the teachings set forth herein. It is therefore to be understood that all 22928473.1 CA Patent Application Agent Ref: 76798/00016 1 such a variations, modification and changes are believed to fall within the scope of the present 2 invention as defined by the appended claims. Although specific terms are employed herein, they 3 are used in their ordinary and customary manner only, unless expressly defined differently 4 herein, and not for purposes of limitations.

22928473.1

Claims

1. A method of controlling a manufacturing process having a production system to form a material into a component, said method comprising:
a) establishing an initial set of operating parameters for said production system, the initial set of operating parameters based on an original production data set and a design model data set stored in a data store, and forwarding the initial set of operating parameters from the data store to a parameter generator;
b) the parameter generator generating machine inputs from the forwarded set of operating parameters, wherein the parameter generator directly communicates the machine inputs to the production system;
c) the production system producing the component from the material;
d) an inspection system inspecting the component, wherein the inspection system uses at least a sensor to collect data specific to the component and accesses the data store to compare the collected data with at least the design model data set to determine whether the component is acceptable or unacceptable;
e) storing the collected data in the data store in either a good part data set if the component is acceptable or in a bad part data set if the component is unacceptable;
f) a knowledge management system accessing the data store, wherein the knowledge management system automatically utilizes one or more algorithms to correlate the collected data with one or more data sets in the data store, including the original production data set, the design model data set, the good part data set, and the bad part data set, to create a new set of operating parameters for the production system without human intervention;
g) transferring the new set of operating parameters from the knowledge management system to the parameter generator; and h) iteratively repeating steps b) to g) one or more times to refine quality of further components.

2. The method of claim 1 wherein each component produced by the production system has a unique identifier associated with the component and its collected data and said unique identifier allows each component to be individually identified and reproduced based upon the collected data corresponding to its unique identifier.

3. The method of claim 1 wherein said knowledge management system accesses a digital representation of said component to help create the new set of operating parameters.

4. The method of claim 1 wherein said knowledge management system accesses said initial set of operating parameters to help create the new set of operating parameters.

5. The method of claim 1 wherein an operator interface is provided to optionally effect adjustment of said operating parameters, said operating interface providing an input to said knowledge management system and said knowledge management system deteimining changes of said operating parameters as a result of inputs from said operator interface.

6. The method of claim 1 wherein the collected data is communicated to other systems producing similar components.

7. The method of claim 1, wherein the collected data includes dimensional, surface, volume, and micro structure data and data on structural inner inclusions or voids.

8. The method of claim 1, wherein the one or more algorithms include data mining, artificial intelligence (AI), neural networks, K-Nearest Neighbor (KNN), Restricted Coulomb Energy (RCE), or iterative learning control (ILC).

9. A machine to form a component, said machine comprising a production system to form said component from a material based upon an initial set of operating parameters, an inspection system, wherein at least one sensor of the inspection system inspects the component to collect data to allow the inspection system to determine acceptability of said component relative to reference data, a data store in electonic communication with the inspection system, the data store containing reference data and storing the collected data obtained from said inspection system, wherein the reference data includes data of operating parameters associated with an acceptable component stored in one data set and data of operating parameters associated with unacceptable component stored in another data set, and wherein the collected data is stored in either the one data set or the another data set according to whether the component is acceptable or unacceptable, and a knowledge management system in electronic communication with the data store, wherein the knowledge management system automatically accesses said data store after collection and storage of the collected data to correlate the collected data and the reference data in order to automatically create a new set of operating parameters used to create a subsequent component, wherein the knowledge management system iteratively accesses the data store, correlates collected data from a most recently produced component and the reference data, and automatically creates a new set of operating parameters between each successive component production cycle.

10. The machine of claim 9 wherein said collected data includes a digital representation of said component.

11. The machine of claim 9 including an operator interface communicating with said knowledge management system.

12. The machine of claim 9 including a communication module to transfer data between the machine and one or more other machines.

13. The machine of claim 9 wherein the knowledge management system selectively operates on the reference data and collected data using at least one of a plurality of data manipulation algorithms to determine appropriate variation in said operating parameters.