TECHNICAL FIELD The present invention relates, in general, to a novel and useful apparatus and method for controlling the outflow of molten metal from a holding vessel or furnace, and more particularly, to an apparatus and method for controlling pressure-based pouring of molten metal from a holding vessel.

BACKGROUND ART In the past, molten metals have been delivered to molds and casting machines from furnaces by two general methods. One method employs a mechanical apparatus in which pouring from the furnace or molten metal holding vessel is accomplished by tilting the furnace or vessel or dislodging a plug in the bottom of the furnace or vessel. Such apparatus, although straightforward, are inaccurate and not easily controlled. They often result in spillage, the production of unusable parts, and injury to personnel. In addition, the molds or other containers which are employed to accept the molten metal are typically under filled or overfilled. Moreover, mechanisms employed to effect such mechanical emptying of metal holding vessels are expensive to manufacture and operate. Because of such problems, the mechanical tilting mechanisms for emptying molten metal furnaces or holding vessels have largely fallen into disuse, while stopper plug devices remain expensive to operate and maintain. Another method for pouring molten metals from a furnace or holding vessel uses pressurized gases. This method offers several advantages over the older mechanical pouring mechanisms. Namely, pressure pouring apparatus are safer to use since a loss of pressure during such process results in the molten metal remaining in the furnace and thereby does not tend to endanger personnel. Although useful in a batch delivery mode, pressure pouring furnaces or vessels still encounter problems with respect to pouring accuracy. Moreover, it is important when successive pours of molten metal take place that each pour include a specific quantity of metallic material poured in the same specific interval of time. In the past, pressure pouring systems have suffered in this regard due to the fact that, as the molten metal level in the vessel decreases, there is a corresponding lowering of the metal level in the pouring spout or tube from which the metallic material egresses. Without compensation for this effect, serial pours of declining quantity over increasing time intervals would result. In the past, attempts to solve this problem have suffered or failed due to the fact that detectors, compensation devices and controls are adversely influenced and even destroyed by the molten metals, and are otherwise unable to accommodate the changing internal geometry and dynamics of the process.

Accurately increasing the gas pressure in the molten metal holding vessel during pouring operations has been recognized as necessary to maintain precise pouring through the spout or pour outlet. In the past, such pressurizing has been accompanied by the use of submerged mechanical valves or the placement of an open orifice in the vessel of smaller size than the pouring spout. Such devices, although theoretically functional, often cease to operate due to the malfunction of mechanical parts and to erosion or clogging due to slag, sludge, dross, or other debris commonly found in molten metals.

Serial pouring of molten metal into molds has become extremely important in industry as a result of the development of high speed sand casting molding machines, fast indexing permanent mold turntables, and rapid cycling high pressure die casting machines. For example, high production molding lines are now capable of producing over 500 uncored sand molds per hour. Filling these molds at a commensurate rate requires serial pours from a furnace that are rapid, very accurate, and highly repeatable. Current pressure-based molten metal pouring apparatus have not been able to meet these demanding production goals, and pouring from the holding vessel has become a limiting factor in maintaining the high •production rates necessary to achieve the desired economic benefits in such molding lines.

The patent literature includes numerous ^teapot11 pressure pouring metal holding assemblies, often configured in the characteristic v""teapot'' design For example, United States Patent 3,058,180 discloses a teapot holding vessel system in which a float is employed at the outlet spout of the holding vessel. The float detects the level of the molten metal in the spout of the vessel and generates a signal which increases or decreases the pressure of the gas in the vessel to maintain a certain level of metal in the spout. The float, however, is easily fowled by slag, sludge, dross, and other common metal impurities, as well as being eroded or destroyed by the heat of the molten metal.

United States Patent 3,998,365 discloses a molten metal dispensing system for serial pouring of metal in which the gas pressure in the molten metal container is successively increased by a predetermined amount with each pour of metal. A contact probe is lowered into the pour spout to just below a level at which pouring will occur and the pressure in the sealed vessel raised until the molten metal reaches the probe. The probe is then retracted and an incremental pressure is applied to the vessel for a timed interval to produce pouring of a desired quantity of metal. This type of pour spout level sensor is also known as an electrical ^"contact and withdraw11 level sensor. At the end of the pouring time pressure is vented to lower the metal in the pouring spout, and thereafter the cycle is repeated. While this has some appeal as a solution, controlling the pressure of a compressible fluid is generally inaccurate even when attempts are made to compensate for lowering levels of metal in the pressure vessel. The gas is a compressible spring whose properties change, and the much greater mass of the molten metal can cause surging that greatly complicates achieving accurate pours by trying to use pressure-based control systems.

In United States Patent 4,220,319 a molten metal holding vessel is provided in which an electrical contact and withdraw sensor is used at the discharge spout to determine when molten metal has reached the pour threshold. The pressure at which the pour threshold is reached is then maintained in a differential pressure sensor assembly and a pressure increment applied to the vessel and sensed. Such differential sensing allows a pressure adjustment in the vessel for lowering metal levels while maintaining the level of the molten metal at the discharge spout. This also is an example of a pressure pouring system which employs an open loop pressure control in which the pressurized gas is exhausted from the vessel chamber to atmosphere between successive pours, thus wasting compressed gas and increasing the time required to complete each pouring cycle. Such systems are more costly and less efficient and productive to operate. They are also unable to automatically compensate or self-correct for normal system variables, such as gas leakage from the vessel chamber. Additionally, they attempt to control based upon gas pressure sensing, with the inherent loss of accuracy that compressibility produces.

The inability to control pours using the apparatus of United States Patent No. 4,220,319 was commercially addressed by adding a load cell to the assembly. The load cell measured the weight of the containment vessel and molten metal, and weight changes as a result of each pour were used to determine changes in the level of molten metal in the vessel . This level determination was used to attempt to improve the setting of the differential pressure used to make the pour. Nevertheless, pour accuracy was not as high as desired, particularly when the vessel was being refilled, and weight-based level sensing has its disadvantages, particularly with low density metals, as described below.

In German Patent No. DE 40 29 386 C2, a pressure pour furnace is disclosed which includes a chamber pressure sensor and a molten metal level sensor in the outlet or pour spout. An integration module also is provided that integrates the chamber pressure over time, with the integration starting when the metal level reaches the level sensor in the pour spout . The pressure integration is used to determine pour volume, but again the accuracy of the pours was not as high as would be desirable because the system was controlling based upon pressure sensing.

My United States Patent 3,499,580 employs a submerged ceramic bubbler tube to sense the pressure at a fixed level below the surface of the molten metal in a vessel . Continuous emission of gas bubbles enables a determination of the height of the metal in the holding vessel pouring tube. Such measurement is then employed to control the pressure of the gas in the vessel chamber, forcing the metal from the vessel in order to maintain a constant pour rate. Although effective, the bubbler tube is susceptible to erosion, clogging, and blockage if gas is not supplied to the bubbler on a continuous basis.

United States Patents 3,412,899, 4,445,670, and 4,730,755 recognize the need to increase the gas pressure as pouring continues from a vessel on a batch basis. Control systems in these references are based on load cell weighing systems which measure the quantity of molten metal in the vessel for each pour.

In U.S. Patent No. 3,412,899 the sensed weight using the load cell is combined with the sensed chamber pressure, but the load cell weight sensing is only an indirect and relatively insensitive method of attempting to determine the level of the molten metal charge. Pressure pouring furnaces which employ a load cell to determine the metal level in the holding chamber are inherently less sensitive for dispensing low density" metals, such as aluminum. This is because load cell accuracy is always a percentage of the total amount weighed, and the load cell must sense the combined weight of the holding vessel with all attachments including any heating apparatus as well as the molten metal charge. Thus with low density metals, the weight of the molten metal charge, which is a smaller portion of the total weight sensed by the load cell, cannot be determined with sufficient accuracy to precisely infer the level of molten metal in the vessel. As a result, the accuracy of individual pours of low density metals is usually not commercially acceptable. Additionally, if the volume of the holding chamber of the vessel is not uniform, or if the heating apparatus affects the volume over the height of the chamber, the load cell output will be non-linear and must be compensated for if the level of the charge in the holding chamber is to be inferred from the load cell output.

In U.S. Patent No. 3,412,899, therefore, the relatively insensitive attempt to determine the location of the top surface of the metal in the chamber is combined with the pressure measurement to attempt to infer when the top surface of the metal in the pour spout reaches level N, the pour threshold. The system of the 3,412,899 patent is "controlled" by sensing the metal level in the mold and shutting down the pour, rather than any attempt to control the pour volume using the furnace controller.

In addition, prior art control systems including that of U.S. Patent No. 3,412,899 have not corrected for the increase in time required to fill the gas volume in the sealed holding vessel above the molten metal as the level of such molten decreases due to pouring. Thus, such prior art systems are limited by the maximum capacity of the gas inlet valve for a given fixed source pressure, and they continuously lose cycling speed and their pouring rates steadily diminish as the number of pours increases in serial pour applications, which makes these systems commercially unacceptable for today's high production requirements.

Yet another practice has been attempted in this field, namely, to ignore the changing level of molten metal in a vessel due to pouring, and, instead, to have the system operator manually attempt to maintain a constant level of molten metal in the holding vessel by frequently adding fresh molten metal to the vessel or furnace during the pouring operation. However, such a refilling process still results in inaccurate metal pours, and it is also inconvenient and labor intensive.

Accordingly, it is an object of the present invention to provide a pressurized gas based, molten metal pouring apparatus, and method for controlling the operation of the same, which meters the outflow of molten metal from the holding vessel rapidly, accurately, and in a repeatable amount of time, regardless of the level of molten metal in the holding vessel.

Another object of the present invention is to provide a molten metal pouring apparatus and method which is well suited for use in high production, serial pour, metal molding lines. Another object of the present invention is to provide an apparatus and method for controlling the outflow of molten metal from a holding vessel or furnace which reduces costly spillage and reduces the under- filling and over-filling of molds, casting machines, and other containers receiving the outflow from the vessel.

A further object of the present invention is to provide an apparatus and method for controlling the outflow of molten metal from a holding vessel or furnace which is not susceptible to clogging due to impurities found in the molten metal.

Another object of the present invention is to provide an apparatus and method for controlling the outflow of molten metal from a holding vessel or furnace which is safer to operate.

Another object of the present invention is to provide a molten metal pouring apparatus and method which is usable with many different kinds of molten metals including relatively hot and dense molten metals, such as copper based alloys, iron and steel, as well as relatively lower density metals, such as aluminum, zinc and magnesium.

Another object of the present invention is to provide an apparatus and method for accurately controlling the outflow of molten metal from a holding vessel which minimizes dynamic control system errors including metal surging and dynamic gas pressure oscillations.

A further object of the present invention is to provide an apparatus and method for accurately controlling the outflow of molten metal from a holding vessel which utilizes no moving or active parts or sensors in direct contact with the molten metal.

Still another object of the present invention is to provide a pressure-based molten metal pouring apparatus and method which is relatively inexpensive to construct and operate, overcomes gas leakage, requires minimal maintenance and repair, and is adaptable to a variety of applications.

The molten metal pouring apparatus and control method of the present invention have other objects, features and advantages which will become apparent from, or are set forth in more detail in the following Best Mode of Carrying Out the Invention and the accompanying drawing.

DESCRIPTION OF THE INVENTION The molten metal pouring apparatus of the present invention comprises, briefly, a molten metal holding vessel having a sealable chamber with a pour passageway extending from the chamber to a pour spout outlet; a gas supply assembly formed and coupled to supply gas to the chamber to pressurize the chamber over the molten metal in order to control the level of molten metal in the pour passageway; and a pour control assembly including a pressure sensor positioned to sense gas pressure in the chamber, a chamber distance sensor formed and positioned to sense the distance to a top surface of the molten metal in the chamber, a pour distance sensor formed and positioned to sense the distance to a top surface of the molten metal in the pour passageway, and a controller coupled to the pressure sensor, the chamber distance sensor and the pour distance sensor to receive signals therefrom, and the controller being formed to be responsive to the signals from the sensors to control the supply of gas from the gas supply assembly to the chamber.

The pressure sensor ge.nerates a signal representative of the gas pressure detected in the chamber. The controller advantageously includes a summing junction which adds the values of the pressure sensor signal and the chamber distance sensor molten metal level signal to produce a process variable signal, which is representative of the level of molten metal in the passageway. A set point module generates a signal indicative of a desired level of molten metal in the pour passageway. Such desired level may be to the point of pouring or at a point ready to pour. The controller compares the process variable signal from the summing junction with the set point signal from the set point module and generates a corresponding output signal.

The gas supply assembly includes a valve assembly and pneumatic circuit that regulates the flow of pressurized gas from the source of pressurized gas to the vessel chamber, as well as from the vessel chamber to atmosphere, preferably through a single conduit. The valve assembly is selected to include a relatively low flow rate capacity inlet valve so that controlling the inflow of gas to the chamber can be accomplished with minimal pressure oscillation and the valve assembly is responsive to the output signal from the controller to effect regulation of the flow of pressurized gas to and from the vessel. In the present invention, compensation for the low flow rate capacity of the inlet valve is provided by a gas pressure booster pneumatic circuit and control mechanism. The booster circuit increases the rate of flow of gas from a source of pressurized gas through the low flow rate capacity inlet valve to the vessel chamber as the level of molten metal in the vessel chamber diminishes. Such booster apparatus overcomes the maximum flow rate limitation imposed by the turndown ratio of the low flow capacity inlet valve as higher flow rates are required to maintain the production pace. A computation control module, which receives the molten metal level signal from the chamber distance sensor, is preset to generate an inverse output signal which passes to the pneumatic booster circuit and increases gas flow through the inlet valve as the chamber metal level diminishes.. A converter may be used to transduce an electrical computation module signal to a pneumatic signal, if desired.

In a further aspect of the present invention, a pour spout distance sensor is provided which is formed and positioned to sense the level of the top surface of the molten metal in the pour passageway. The distance level sensor preferably senses the top molten metal surface without contacting the same, and most preferable is a wave-operated sensor, such as a radar wave-based sensor which is less sensitive to smoke, dust, steam, variable air temperatures and the like over the surface of the metal. The pour passageway distance sensor also may be a contact and withdraw distance sensor, particularly if lower temperature molten metals are being poured. If a pour passageway distance sensor is employed, a load cell based inferential determination of the level of the charge in the holding chamber can be employed, particularly if the holding chamber has a relatively uniform geometry over its height. Pour passageway distance sensing can be used to overcome small load cell level determination inaccuracies.

The method of controlling the pouring of molten metal from a holding vessel of the present invention is comprised, briefly, of the steps of sensing the distance to the top surface of a charge of molten metal in a sealed chamber of the holding vessel, sensing the pressure in the chamber over the molten metal, and pressurizing the chamber over the molten metal in response to the combination of the sensed distance and the sensed pressure. Additionally, it is preferable that the present control method includes the step of sensing the distance to the top surface of the molten metal in the pour passageway from the chamber. Both distance sensing steps are preferably accomplished using a radar distance sensing device that can sense the level of the metal without contacting the same.

In another aspect of the method of the present invention a control method is provided in which the flow rate of compressed gas to the sealed chamber is controlled by a low flow rate capacity inlet valve which can have its flow rate capacity increased or boosted as the metal level in the chamber diminishes.

In a final aspect of the present invention, sensing of molten metal levels in one or both of the holding vessel chamber and the pour passageway is accomplished using a distance sensor not requiring contact with the molten metal, and most preferably level sensing is accomplished using a radar based distance sensor.

DESCRIPTION OF THE DRAWING FIG. 1 is a schematic, side elevation view, in cross section, of a molten metal pouring apparatus constructed in accordance with the present invention.

FIG. 2 is an enlarged, fragmentary view, in cross section, of the pour spout portion of the holding vessel apparatus depicted in FIG. 1.

FIG. 3 is a graphical representation of levels in the pour spout of the holding vessel at specific times during a pour cycle.

FIG. 4 is a block diagram corresponding to FIG. 1 depicting schematically the function of the control system of the present invention.

For a better understanding of the invention reference is made to the following detailed description of the preferred embodiments thereof which should be referenced to the prior described drawings.

BEST MODE OF CARRYING OUT THE INVENTION Turning now to FIG. 1 a molten metal holding and pouring apparatus, generally designated 10, is shown in which a molten metal holding vessel 12 is depicted as a conventional pressure tight holding vessel. Vessel 12 is properly insulated to hold molten metal charge 14 therewithin, and includes a pressure tight top 16, bottom 18 and side 20 defining a molten metal holding chamber 22. Vessel 12 can be a furnace and most typically will have a heating assembly, not shown for simplicity of illustration. Vessel 12 also includes a charging inlet 24 which accepts a metal charge 14. Pour passageway or spout 26 extends from chamber 22 to a pour spout outlet 28 to provide a path for pouring of metal charge 14 from chamber 22 of vessel 12. Mouth 28 of pour passageway 26 lies a certain distance above a floor 30 of bottom 18 of the holding vessel. Mouth 28 preferably takes the form of a transversely extending weir or orifice through which molten metal can flow at a volumetric rate that can be precisely determined.

Vessel or furnace 12 is of "■"'teapot'' configuration which is well known in the art. In most embodiments of the invention, the vessel 12 will sit upon a platform 32 which is supported by a fixed or moveable structure (not shown) attached to the ground surface 34, which is designed to accommodate the attendant molding machine or casting process. In certain other embodiments of the invention, which will be discussed hereinafter, vessel 12 may sit upon a platform 32 which is itself supported above ground surface 34 by fulcrum support member 36 and a load cell 38 (depicted in broken lines) . Although a single load cell 38 is depicted in FIG. 1, multiple load cells may be employed with apparatus 10 of the present invention. Moreover, as also will be explained, load cell 38 is not a required nor preferred element of the broadest aspect of the present invention.

The molten metal holding vessel of the present invention employs pressurized gas to effect pouring of metal from pour spout 26. It will be seen, therefore, that a gas supply assembly is provided including a pneumatic circuit or conduit array. Conduit 40 of that array can be seen to extend from a pressurized source 42, here illustrated as a pressurized gas container. Source 42 may be provided by dry compressed air, nitrogen, argon, or any gas which is compatible with molten metal charge 14 within chamber 22 of vessel 12, that is, a gas that will not oxidize or otherwise react with and degrade the charge. The gas supply assembly also includes a valve assembly, generally designated 44, which regulates the flow of compressed gas from source 42 through conduit 40 and into chamber 22 of holding vessel 12. The top surface 39 of molten metal 14 in the vessel chamber 22 lowers as gas pressure in chamber 22 forces molten metal 14 out of the chamber 22 through pour passageway 26 and out the pour spout outlet 43.

Valve assembly 44 of the present invention is preferably selected to include an inlet valve 45 which has a relatively low flow rate capacity as compared to the volume of chamber 22. This low flow rate capacity provides good control at the low end of flow rate demand, and a booster assembly 88 is also provided to enable increasing or boosting of the flow rate through valve 45 as flow rate demand increases, as will be set forth in detail below.

Thus, valve assembly 44 may include an inlet valve 45 having a valve actuator 47, such as a pneumatic or electrical actuator. A pneumatic actuator is shown. In addition, valve assembly 44 also preferably includes an exhaust valve 51 having a pneumatic actuator 53. Valves 45 and 51 preferably operate in a ^"split range11 configuration. That is to say, inlet valve 45 and exhaust valve 51 are both closed at a certain intermediate pneumatic control pressure value (e.g., 9 PSI). Inlet valve 45 progressively opens when its actuator 47 receives an increasingly higher pneumatic control signal (e.g., ranging from about 9 to about 15 PSI) , while exhaust valve 51 progressively opens when its actuator 53 receives an increasingly lower pneumatic control signal (e.g., ranging from about 9 to about 3 PSI) . Between zero and 3 PSI exhaust valve 51 is fully open and inlet valve 45 is fully closed. It is a feature of the present invention, therefore, that the split range configuration of valves 45 and 51 results in a fail¬ safe condition in which chamber 22 is not pressurized in the absence of a pneumatic control signal.

Many prior art molten metal holding and pouring apparatus have attempted to control pouring from the holding vessel by primarily sensing and using the gas pressure in the metal containment chamber. The gases used to pressurize the chamber, of course, have little mass and are compressible, which makes the gas behave like a spring, making reliance on the chamber pressure alone unsatisfactory. The problem is exacerbated by the fact that the molten metal has great mass and inertia, and is incompressible, so that the use of """springy11 compressed gas to bring about pouring can easily cause undesirable surging and oscillating, making attempts at automatic controls more difficult. When the gases in the chamber are exhausted to atmosphere and a new pour cycle is quickly attempted, gas compression can cause metal surging making the pour controller lose its ability to achieve accurate pours.

Prior art attempts to supplement the pour controller with a sensor which senses the level of the metal in the pour spout, usually by means of an electrical contact and withdraw sensor or a float, has only marginally improved the accuracy of the pours. Such prior art pour spout level sensors, which are only used with low temperature molten metals like aluminum and zinc, cannot be used with higher temperature ferrous and copper base molten metals. Finally, as previously described, the use of indirect level sensing of molten metal in the pour spout has proven to be insensitive, ineffective and inaccurate.

In one aspect of the molten metal pouring apparatus of the present invention, therefore, a controller assembly is employed in which direct distance sensing of the level of the molten metal in the chamber is combined with pressure measurements to effect more accurate control of metal pours. In another aspect, still further improvement results with level sensing of the metal in the pour passageway. In still a further aspect, a pressure control assembly is employed in which a low flow rate capacity inlet valve is combined with a flow rate booster assembly to achieve improved pressure control, resulting in reduced surging and oscillation at both the low and high ends of flow rate demand.

The control assembly of the present invention, therefore, includes a sensitive, fast-acting pressure sensor or transducer 54 formed and positioned to sense and monitor the gas pressure in conduit 40 and thus in chamber 22 above molten metal charge 14. For example, a pressure-to-electric transducer manufactured by Transicoil of Norristown, PA would suffice in this regard. Transducer 54 could also be located in chamber 22, and it preferably is formed to produce an electrical signal which is communicated via line or conductor 56 to a summing junction 58.

One key to accurate performance of the control assembly of the present invention is that direct sensing of the level of molten metal is employed, rather than indirect sensing through the use of a load cell or bubbler tube. Thus, the control assembly of the present invention includes a vessel chamber distance sensor or detector 50 which directly senses the distance from the sensor to the top surface 39 of the molten metal in chamber 22. Sensor or transducer 50 thus effects a distance sensing measurement without contacting the corrosive and destructive molten metal. Sensor 50 can include a pressure-tight observation tube 55 which is mounted to vessel cover 16 and extends into chamber 22 of vessel 12. Chamber distance sensor 50 may advantageously be a wave-operated distance sensing device such as a radar based sensor, which will be discussed in more detail below. Sensor 50 determines the level of the top surface 39 of molten metal charge 14 in vessel chamber 22 throughout the pouring cycle.

In addition, holding vessel 12 will normally include a metal receiving inlet 41 with filling passageway 24 positioned higher than pour spout outlet 43 of pour passageway 26 to allow recharging of chamber 22 while it is pressurized. Chamber distance sensor 50 is formed to produce a signal which is communicated through conductor 57 to a summing junction 58 and to a computational module 90. The chamber distance sensor signal is adjusted to decline in proportion to the dropping of the level of metal surface 39. Thus, accurate metal level sensing in chamber 22 determined by non-contact distance measurements can be combined with the chamber pressure measurements to produce enhanced control of the pour cycle. The precision of successive pours can be further enhanced by employing a molten metal level sensor at the top of pour passageway 26. The control assembly of the present invention, therefore, also preferably includes a pour spout distance sensor' 46 formed and positioned to sense the distance to the top surface of the metal in the pour spout. Pour spout distance sensor 46 preferably includes an observation tube 48 which is mounted to vessel 12 and extends into mouth 28 of pour passageway 26. Sensor 46 determines the level of molten metal in pour passageway 26 and generates a signal representing the same. The pour spout sensor signal is communicated down line or conductor 49 to a timer 76 or, alternately, to integration module 77 (shown in broken lines) . In particular, sensor 46 senses the level of molten metal in pour passageway 26 ranging between a ""ready- to-pourM level and a maximum ""pour11 level, which will be discussed hereinafter. It will be appreciated that a ""contact and withdraw1' level sensor 46, as for example is shown in U.S. Patent Nos. 3,998,365 and 4,220,319, also can be used as a pour spout distance sensor, particularly when lower temperature molten metals, such as aluminum and zinc, are being poured.

Sensors 46 and 50 are most preferably distance measuring sensors which can sense the level of the top surface of the molten metal without having to contact the same so as to be less susceptible to damage by the high temperature molten metal charge 14. Advantageously, sensors 46 and 50 are wave- operated sensors which may employ radar, a laser beam or other optics, or sonic or ultrasonic waves, for example, in making level determinations. A capacitance based sensor also could be employed. Radar-type sensors available under the trademark Sitrans LR, manufactured by Siemens, are particularly well suited for both sensors 46 and 50. Radar sensors are less affected by smoke, dust, steam, slag, and high molten metal temperatures than many other remote sensing transducers.

While a radar sensor is preferred for use at pour spout 26, in the broadest aspect of the present invention electrical """contact and withdraw" type sensors, as shown in the prior art, also can be employed to sense any of the discrete ^ready-to- pour, '' "^threshold of pouring,1' or "vpourM levels that are used in combination with direct level sensing and pressure sensing in chamber 22 in controlling pouring from vessel 12. Chamber distance sensor 50, however, preferably should not employ electrical contact distance sensing because of the unsuitable environment for such sensors, and since there are an infinite number of levels in chamber 22 which need to be sensed for maximum accuracy.

In order to achieve a process variable control signal that is equivalent to the level of molten metal in pour passageway 26, the control assembly of the present invention includes an electrical summing junction 58 for combining the chamber distance level sensor signal with the chamber pressure sensor signal. Accordingly, electrical summing junction 58 receives an electrical signal via conductor 57 from vessel chamber distance sensor 50, as well as the signal from pressure transducer 54, representative of the gas pressure in conduit 40 and thus in chamber 22, through conductor 56. The output of summing junction 58 is the process variable signal that is communicated through conductor 60 to controller 80. This process variable signal is the electrical equivalent of the height of molten metal in pour passageway 26. That is to say, when the pressure transducer signal equals zero, the process variable signal solely equals the level 39 of molten metal 14 within chamber 22, as sensed by the chamber distance sensor 50, which will be the same level as the molten metal in pour spout 26. However, as the pressure of gas within chamber 22 of vessel 12, as monitored by pressure transducer 54, is increased, the height of molten metal 14 in pour passageway 26 increases directly in proportion to the sensed gas pressure, while the height of molten metal in the chamber decreases. Such changing in level of molten metal 14 in pour passageway 26 and reduction in the height in chamber 22 will be reflected in the process variable signal from summing junction 58.

A set point module 62 may be used to produce an adjustable ""set point" output signal in conductor 78, which set point signal is electrically proportional to the desired height of molten metal in pour passageway 26. Set point module 62 includes set point 64 representing a ""ready-to-pour" level 66 as shown in broken lines in FIG. 2. In addition, a ""pour level11 set point 68 is provided in set point module 62 and corresponds to a "spour level11 70 of metal 14, illustrated in broken lines in FIG. 2. Either set point 64 or set point 68 is electrically selected by relay means 72 within set point module 62. The quantity of metal egressing from mouth 28 of pour passageway 26, per unit time, may be adjusted by varying set point 68, which changes pour level 70 of molten metal in pour passageway 26. In addition, set point 68 also may be varied to change the rate of pouring as molten metal egresses from mouth 28 of the pour passageway 26.

In order to enhance pouring control accuracy, the signal from pour spout distance sensor 46 may be communicated through conductor 49 to a timer 76. When an external signal to begin pouring is received by the controls, relay 72 in set point module 62 switches from the ""ready to pour" set point 64 to the ""pour level11 set point 68. As pouring commences, the pour spout level signal 49 from transducer 46 provides electrical confirmation that molten metal in pour passageway 26 has reached the ""threshold of pouring11 level, namely level 74 shown in broken lines in FIG. 2, also known as the ""drip point.'1 Timer 76 is used to determine the quantity of molten metal poured. In this regard, timer 76 receives the level signal from pour spout distance sensor 46, indicating that molten metal has reached the ""drip point'1 or ""threshold of pouring11 level 74, which initiates the pouring time interval of timer 76. Upon the expiration of its preset time interval, timer 76 signals internal relay 72 of set point module 62 to return to ""ready-to-pour" level 66 as determined by set point 64.

Alternatively, pour spout distance sensor 46 may send a level signal to an integration module 77, shown in broken lines in FIG. 1, instead of to timer 76. The continuous output signal from level sensor 46 would then be used by integration module 77 to produce an accurate quantitative measurement of the amount of molten metal poured, beginning the moment that molten metal in pour passageway 26 exceeds the ""threshold11 elevation 74. FIG. 1 depicts such alternate arrangement in broken lines wherein integration module 77 connects to set point module 62. When integration module 77 reaches its preset limit, thereby indicating that the desired amount of molten metal has been poured from pour passageway 26, integration module 77 signals relay 72 to switch from ""pour11 level set point 68 back to ""ready-to-pour11 level set point 64.

The output signal from set point module 62 is communicated to controller 80 through conductor 78, and the process variable signal from summing junction 58 is also input to the controller through conductor 60. Controller 80 preferably is a suitable 3-mode (rate, reset and proportional band) controller which is capable of providing an electrical output that varies with the rate of change, integral over time, polarity and magnitude of the difference of the two input signals received. Of course, controller 80 may produce other types of output signals, e.g. , pneumatic signals, etc. In other words, controller 80 compares the process variable signal from summing junction 58 to the selected set point signal from set point module 62. Any difference between the two signals is the ""error11 or differential signal, which causes controller 80 to produce an output signal in conductor 82 that travels to a fast-acting electric to pneumatic converter 84 via conductor 82.

Converter 84 directly controls valve assembly 44, specifically valves 45 and 51 through actuators 47 and 53, respectively, as heretofore described. As shown in the preferred embodiment of FIG. 1, converter 84 is powered by compressed gas from source 42 via regulator 86. A separate source of compressed gas also may be employed to operate regulator 86. Also, a motor operated pressure regulator with a position indicating slide wire may be employed as converter 84 and valve assembly 44. Alternatively, valves 45 and 51 may be solenoid operated by an electrical signal directly from controller 80. An electro-pneumatic converter Model 870022, manufactured by DeZurik, can be employed as the fast- acting electric to pneumatic converter 84.

A reduction in pressure oscillations and surging in chamber 22 also is achieved in the present invention by selecting an inlet valve 45 that has relatively low flow rate capacity, and then boosting the flow rate of the valve, as metal level 39 diminishes, with a separate booster assembly. This combination widens the valve's operating range to include the entire volume of vessel chamber 22 when it is nearly empty, effectively circumventing the valve's finite turndown ratio. The relatively low flow rate capacity inlet valve permits controller 80 to make minute corrections without directing too much or too little gas into chamber 22, thus reducing or eliminating surging and oscillations that would otherwise, as in the prior art, be caused by applying these same corrections to a necessarily larger flow capacity valve not equipped with a separate booster assembly. During the time intervals between successive pours, controller 80 directs converter 84 to maintain a typical pressure control signal to valves 45 and 51 between 9.5 and 10.5 PSI. Because of the split range configuration of valves 45 and 51, this means that exhaust valve 51 is completely closed and inlet valve 45 is slightly open, which maintains adequate pressure in chamber 22 to steadily keep the level of metal in pour passageway 26 at the v"~ready-to-pour' ' level 66. When there are leaks from chamber 22, controller 80 will automatically increase the pressure control signal by whatever amount is needed to simultaneously maintain the ^ready-to-pour' ■ level 66 and overcome the leaks.

Such stable control can best be achieved by using an inlet valve 45 which has a low flow rate capacity relative to the volume of chamber 22. For example, in one embodiment a chamber 22 having a volume of 36 cubic feet was controlled by an inlet valve 45 having a flow coefficient, or Cv, of 3.7. Such a low flow coefficient, however, becomes a limiting and disadvantageous feature as the demand for flow increases, as is described below. However, if inlet valve 45 is made larger in order to satisfy the high flow rate required when metal level 39 in vessel chamber 22 is low, making the gas volume above the metal correspondingly large, pressure oscillations due to control feedback are experienced at the low end of flow rate demand, where the system needs to operate most of the time, because of the larger valve's inability to precisely modulate lower levels of gas flow when directed by controller 80. It has been found that the flow coefficient or Cv of inlet valve 45 will substantially scale relative to the chamber 22 volume according to Boyle's Law, namely, (Cv1 X T1)ZV1 = (CV2 X T2) /V2 Thus for a constant temperature, the above example can be scaled up or down with changes in the volume of chamber 22 in order to select a ""relatively small'1 or "low flow rate capacity" inlet valve 45. The above example, however, is not necessarily regarded as being absolutely optimal, and one skilled in the art will recognize that both larger and smaller ratios of Cv to V can be employed. But as the ratio of Cv to V increases, at some point the Cv will be large enough to begin to introduce pressure oscillations in chamber 22 that will interfere with, or degrade, the pressure control. As used herein, therefore, the expression vvrelatively low flow rate capacity11 shall mean an inlet valve having a flow rate capacity which is sufficient to minimize or substantially eliminate pressure oscillations in chamber 22 due to control feedback in operation of valve 45.

While use of a relatively low flow rate capacity inlet valve allows the low end of flow control to be enhanced, such an inlet valve by itself cannot adequately respond to system high end flow demand. One problem which has existed in prior molten metal pouring apparatus is that, as the level of molten metal drops in the vessel chamber, it requires more time for a larger quantity of gas to flow from the source of pressurized gas, in order to reach the pressure in the chamber necessary to produce the desired pour level in the pour spout. Low flow rate inlet valve 45, like all such valves, has a limited range of operation, commonly known as its v"turndown ratio. M Without a booster mechanism to effectively extend its range, this ^turndown ratio11 will restrict the ability of valve 45 to admit increasingly larger quantities of compressed gas to chamber 22, in the same amount of time during each pour, as the level of metal 39 diminishes due to pouring. The pressure source might, for example, provide compressed gas at 75 to 150 PSI, but a fixed regulator or regulators would normally be employed in the prior art to reduce the pressure to say 8 to 10 PSI to avoid molten metal surging and pour overshoots. Then, if the level of molten metal in chamber 22 were to drop sufficiently low, controller 80 would open valve 45 until it was wide open, delivering its maximum flow rate to the chamber. However, because valve 45 is selected to be a low flow rate valve for stable low-end flow rate control, this flow rate is highly likely to be insufficient to provide adequate gas flow to chamber 22 for low metal levels because of the limitation imposed by the turndown ratio of valve 45. For this reason the apparatus of the present invention preferably includes a pressure booster or variable regulator assembly to overcome this problem.

Booster mechanism 88 may include a computation module 90 which receives the chamber level signal from sensor 50 via conductor 57. Computation module 90 produces an electrical first order linear output signal which is communicated through conductor 92 to an electric-to-pneumatic converter 94. The linear output signal is inversely proportional to the input from chamber distance sensor 50 according to the equation: y = mx + b where ^y'1 is the output of computation module 90, ^x'1 is the input from chamber level sensor 50, ^m11 is the adjustable slope of a curve (which will be negative), and "*vb!I is the adjustable Y-intercept which determines the module 90 output when the input from sensor 50 is zero, corresponding to the lowest acceptable level of metal in chamber 22 of vessel 12.

The electrical output signal from computation module 90 is sent to electric-to-pneumatic converter 94, similar in structure to converter 84. Converter 94 may share pneumatic power with converter 84 via regulator 86. A gas volume booster relay 96 produces a high volume flow rate at a pressure equal to a reference pressure input signal communicated from converter 94 by conduit 98. When signaled by electric-to-pneumatic converter 94, volume booster relay 96 increases the pressure at the input side of inlet valve 45, which permits compressed gas from source 42 to flow through control valve 45 at a very high rate, for example, well above the rate limited by the turndown ratio of valve 45 at the unboosted pressure which would otherwise be supplied to the valve. A volume booster relay Model 4500, manufactured by Fairchild can be used as volume booster relay 96. Pneumatic converter 94 would take the same commercial form as converter 84.

Booster mechanism 88, therefore, greatly extends the useful range of operation of valve 45 and reduces the time required for pour cycles when the metal level in chamber 22 drops to lower levels by increasing the gas pressure available to valve 45. The increase from booster 88, however, is gradual and does not increase the pressure available to valve 45 to a level causing pressure oscillations at the low end of flow rate demand. The booster, however does permit valve 45 to admit the required volume of compressed gas into chamber 22 in the same amount of time during each pour at both the high and low ends of flow rate demand, with a corresponding improvement in overall system pouring accuracy and efficiency. As previously noted, the function and operating range of booster mechanism 88 is determined by the settings ""m11 and "^b" in computational module 90. In operation, holding vessel 12 receives a charge 14 of molten metal of any desirable genre via charging inlet 24. To minimize surging of molten metal during successive pours, and the disadvantages accruing thereto, it is desirable to initially hold the molten metal in vessel 12 under a pressure in chamber 22 causing the metal in pour passageway 26 to be raised to ""'ready-to-pour" level 66 in FIG. 2 prior to pouring. Accordingly, when pouring apparatus 10 is initiated, an "error11 signal is immediately detected at the input of electronic controller 80, based on the fact that the process variable signal is less than v"ready-to-pour'' set point 64. A rising output signal from controller 80 is received by electric-to- pneumatic converter 84. This produces a rising pneumatic signal from converter 84 which, in turn, causes exhaust valve 51 to close and gas inlet valve 45 to open. This results in an increase in gas pressure in chamber 22 of vessel 12, and a corresponding ris.e in the level of molten metal 14 in pour passageway 26. The gas pressure within chamber 22 will continue to increase until the process variable signal 60 entering controller 80 exactly equals ^vready-to-pour'' set point signal 78 from set point module 62, and molten metal 14 reaches "-ready- to-pour11 level 66 in passageway 26. When this occurs the output of controller 80 will stop the increase of pressure within chamber 22 of vessel 12 and molten metal 14 will stay at ""ready-to-pour" level 66 in pouring tube 26.

The command to actually pour molten metal 14 from pour passageway 26, directional arrow 100 in FIG. 1, would commence upon the receipt of an external signal from an adjacent molding or casting machine or other casting equipment (not shown) . Such command causes the set point relay 72 within set point module 62 to switch from "ready-to-pour" set point 64 to ""pour11 set point 68. The resulting new ""error'1 signal is detected by controller 80 which generates, via converter 84 and valve assembly 44, an increase in gas pressure through conduit 40 to chamber 22 of vessel 12. When the molten metal in the pour passageway reaches the threshold level 74, the presence of molten metal at this level is sensed by pour passageway level sensor 46, which then starts electronic timer 76, and pouring actually begins. Pouring continues for a user-preset interval of time and at a user-preset pouring rate determined by the ""pour" set point 68 which controls pour level 70. FIG. 3 schematically depicts the control process above described.

It should be noted that pour passageway distance level sensor 46 eliminates a major component of dynamic pouring error by starting the pour interval timer 76 when the molten metal in passageway 26 actually reaches ""threshold of pouring11 level 74 instead of beforehand, i.e., instead of when a signal to begin pouring is first received, as in the prior art. This feature greatly reduces dynamic system errors derived from variations in the time required for molten metal to move from ""ready-to-pour11 level 66 to ""threshold of pouring" level 74.

FIG. 3 illustrates the advantage of holding the metal at the ""ready-to-pour" level and not starting the timer or integration until threshold of pouring level 74 is actually reached. Abscissa 102 in FIG. 3 represents time, generally measured in seconds, and ordinate 104 represents the height of molten metal 14 in pour passageway 26 as detected by pour passageway distance sensor 46. At origin 106, t0 is defined as the instant that action to initiate the pouring cycle begins, and h0 is the v"ready-to-pour' ' level 66 in pour passageway 26. The height hx is the top of weir 108 of pour passageway 26 when the metal is at threshold of pour level 74 as shown in FIG 2, and h2 represents the user input or selected """pour11 level 70. It should be realized that the flow of any liquid including molten metal over weir 108 or, alternatively, through an orifice (not shown) is directly proportional to the height of molten metal above the weir 108 or orifice, as a function of time.

The dynamic time intervals during which the metal level in pour passageway 26 moves from level 66 to level 70 and back again are defined by the time intervals t0 to t2 and t3 to t5 respectively. These dynamic time intervals are more difficult to control, on a repeatable basis, than the steady state time intervals such as those between t2 and t3 , after t5, and prior to t0. This means that the amount of metal 14 leaving mouth 28 of pour passageway 26 during the dynamic control phases will contain a larger component of error per unit time than the amount of metal 14 poured during steady state control phases . However, pour passageway distance sensor 46 is able to detect level 74, the threshold level, enabling timing device 76 to start at time tx rather than at t0, as found in the prior art. In this way, all dynamic control errors between h0 and hx are eliminated. Since this is the largest error component of the dynamic control phase, overall pouring errors are significantly reduced. As above noted, timer 76 may be replaced by- electronic integration module 77 that operates in conjunction with a continuous output signal originating from pour passageway level sensor 46. Again, an adjacent molding machine or casting machinery (not shown) would initiate pouring by causing relay 72 to switch from ""ready-to-pour'• set point 64 to the ""pour11 set point 68. When the level of metal in pour passageway 26 increases above the ""drip point11 or ""threshold of pouring11 level 74, (hi of FIG. 3) , integration module 77 begins calculating the amount of molten metal actually dispensed or poured ""on the fly1' by integrating the molten metal level h (FIG. 3) over time as continuously reported by output signals from level sensor 46. When integration module 77 determines that the quantity of molten metal, preset by the user, has been dispensed, it signals the electric relay 72 to switch back to the ""ready-to-pour" set point 64. Thus, the smaller steady state control system errors that affect the actual pour level h2 when only a timer is employed are also reduced by use of integration, further increasing overall system accuracy. When a desired quantity of metal to be dispensed is preset in integration module 77, an allowance must be made for the small but repeatable quantity of metal represented by area 110 on FIG. 3, which is poured when the metal in pour passageway 26 drops from h2 back to Ji1. The integration module precisely terminates the pour short of the desired volume by this amount, as represented by area 110.

Controlling the quantity of metal poured through this integration technique is particularly useful when pour level h2 is purposely changed during the pour itself to accommodate individual mold pouring requirements.

Because of the small height differential between ssready-to-pour" level 66 and "'"pour11 level 70, FIG. 2, the partial pressurization of chamber 22 between successive pours, and the behavior of closed loop, servo controller 80 when properly adjusted, the transition between these two levels is rapid and very smooth. Thus, the quantity of compressed gas needed to complete each pour together with the pouring cycle time are considerably reduced, and pressure oscillations and molten metal surging are also minimized.

It should be noted that when the system returns to ^"ready-to-pour" set point 64 in set point module 62, it begins to operate in reverse since an """error11 signal of opposite polarity is received by controller 80. Controller 80 then decreases its output signal 82 to converter 84, which in turn causes exhaust valve 51 to dump only a small portion of the compressed gas contained in chamber 22 of vessel 12 via conduit 40. This action takes place until metal 14 in pouring tube 26 returns to """ready-to-pour' ' level 66.

When molten metal is charged into chamber 22 of holding vessel 12 via inlet passageway 24, the higher level of molten metal 39 in chamber 22 is detected by vessel chamber distance sensor 50. The process variable signal then increases resulting in a decrease in the gas pressure supplied to chamber 22 of vessel 12. In other words, the process variable signal 60 remains equal to the desired height of molten metal in pouring tube 26 and, via controller 80, equal to the set point selected in set point module 62, even when the vessel is being refilled.

It will be apparent, therefore, .that the pressure control system described herein is a true closed loop, servo-type control system which will act to self-correct for a number of system variables, including naturally occurring gas leaks from the vessel chamber, which significantly contributes to its reliability and accuracy.

The present invention makes it possible to pour very accurate quantities of molten metal in a wide range of flow rates because the flow rate of metal from the pour passageway 26 of vessel 12, as well as the time required to induce that flow rate, remains constant, regardless of the level of molten metal in the vessel.

While the preferred embodiment of the present invention is based upon a chamber distance sensor 50 which directly senses the level of molten metal charge in chamber 22, it is also possible to attain many of the advantages of the present invention by employing indirect sensing of the level of molten metal in chamber 22, particularly when the chamber configuration is uniform so that a substantially linear relationship of weight-to-charge level exists. Thus, instead of vessel chamber distance sensor 50, the apparatus of the present invention can employ a load cell 38 which automatically determines the weight of charge 14 by subtracting out the tare weight of vessel 12 and its components from the gross weight reading obtained. Load cell 38 then produces an electrical signal analogous to the level of molten metal in the vessel chamber 22 which is communicated via conductor 52 to summing junction 58 and computational module 90. It should be noted that load cell 38 and pressure transducer 54 require accurate calibration and adjustment upon installation in system 10 of the present invention with each particular application. After initial calibration, very little adjustment is required to maintain the balance of apparatus 10 of the present invention. Use of indirect or load cell sensing of the level of molten metal in chamber 22, however, is not preferred and can be problematic for low density metals, as above noted. However, error can be minimized by combining pour passageway level sensing with load cell based chamber level sensing.

Turning now to FIG. 4, a block diagram of the preferred embodiment of the above-described molten metal holding and pouring apparatus 10 is shown. Vessel 12 has a chamber 22 for charge 14, and direct distance sensors 46 and 50 are provided to sense the molten metal levels in the pour passageway and the chamber, respectively. The sensed level in the chamber is combined with the pressure in the chamber sensed by transducer 54 at summing junction 58 to produce a process variable signal. Controller 80 compares the process variable signal to set point signals from set point module 62. The controller controls operation of gas valve mechanism or assembly 44 to admit gas from source 42 into chamber 22. Level signals from sensor 46 determine the level of molten metal in the pour passageway and start a timing or integration process (76/77) which, at its conclusion, signals the set point module 62 to cause controller 80 to exhaust gas from vessel chamber 22 at the end of the pour cycle. Finally, computational module 90 causes booster 96 to increase the flow of compressed gas through valve assembly 44 to vessel chamber 22 as the level of molten metal charge 14 diminishes, as determined by vessel chamber level sensor 50.

From the foregoing description of the present apparatus and its operation, it will be seen that the method of controlling pouring of molten metal from metal holding vessel 12 includes the steps of sensing the distance to top surface 39 of metal charge 14 in chamber 22, sensing the gas pressure in the chamber, and pressurizing the chamber over the molten metal in response to a combination of the sensed distance and pressure by an amount producing a controlled outflow or pour of molten metal from chamber 22 through pour passageway 26. The distance sensing step is preferably accomplished without contacting the metal charge in chamber 22 by employing a wave-operated distance sensor such as a radar sensor.

The present method also preferably includes a sensing step in which level of the molten metal in the pour spout is also sensed and used to precisely detect the beginning and control the duration of the pour. Again this is most preferably accomplished using a wave-operated distance sensor 46 which does not contact the molten metal.

In another aspect of the present method, a step of selecting a relatively low flow rate capacity inlet valve is accomplished for low end flow rate control and the step of boosting the rate of the admission of gas into chamber 22 is undertaken as the level of molten metal 14 in chamber 22 diminishes for high end flow rate control. This pressure boosting step allows constant pour cycle times to be maintained as the charge in the holding vessel decreases.

It should be noted that, in the interest of simplicity and clarity, the above system has been described as being comprised of individual, discrete components and modules. However, this description is not intended to be limiting in this regard since many of these components and modules can be readily obtained as part of commonly used industrial programmable logic controllers (PLCs) or industrial grade computers, which are highly recommended for this and similar applications.

It will thus be seen that the objects set forth above, among those made apparent from the preceding description, are efficiently attained and, since numerous changes may be made in the above construction without departing from the spirit, scope and principles of the invention, it is intended that all matter contained in the above description and shown in the accompanying drawings shall be interpreted as illustrative and not in a limiting sense.