BACKGROUND OF THE INVENTION

The present invention relates generally to the field of forming three-dimensional

objects. Specifically it relates to volumetric feed control of a build material element in three-

dimensional modeling.

Three-dimensional modeling machines in the current art often use build elements made

from a build material such as a filament, often wound on a roll, that are fed to the machine by the use of feed rollers or other devices. In the modeling machine, the material is melted

before being dispensed at an application tip. The filament is preferably made of a material that

will rapidly solidify upon a small drop in its temperature. Materials often used for modeling

include waxes, thermoplastic resins, and various metals.

Build material elements have varying configurations, but typical elements used in the

field of three-dimensional modeling must currently be manufactured to a high degree of

accuracy due to the tolerance limits imposed by current modeling machine technology. The

tolerance limits imposed on build material elements by the current technology are in the area

of 4 per cent of volume. Typically, the build material element is in the form of a cylindrical

filament approximately 0.070 inches in diameter, with a required tolerance of +/- 0.0015 inches. In current modeling systems using this filament, if the filament exceeds tolerance by

either being too large or too small, the machine may plug or the part become corrupted. Also in current systems, the accuracy of the model created is dependent upon the accuracy of the

filament supply. When the filament varies in cross section, the volume of material delivered

to the application tip of the machine does not remain constant, and therefore the flow rate of

material fluctuates. When the filament is too large, there may be a build-up of excess filament at the application tip, resulting in a model layer that is too thick. Similarly, when the filament

is too small, there may be a shortfall or lag of filament at the application tip, resulting in a

deposited layer that is too thin, or that leaves gaps in the model. Both situations may easily

lead to model failure. An uncontrolled flow rate, even when the fluctuations are small, can contribute to large variations in the quality of the final model, including gaps and excess

thickness.

The three-dimensional modeling machines of the current art use flexible filament, often stored on a roll, that is fed to an application tip by the use of feed rollers. Although attempts

have been made at constant volumetric control of melted filament at the application tip,

proposed solutions have not controlled the problem enough to allow a tolerance level change in the filament used. An excess of material leads to overthick layers, and a shortfall leads to

unacceptable gap or thin wall errors in modeling.

Various prior art designs for automatic feed control and variable speed feeding exist. For example, U.S. Patent No. 3,017,060 to Miller discloses an automatic feed control mechanism for wire filaments. Miller teaches an intermittent feed of the filament, the filament

having segments of varying cross-sectional area. No continuous measuring of the wire

diameter, cross sectional area, or effective cross section is taught by Miller. No continuous adjustment of filament feed speed in response to changes in wire diameter is taught by Miller.

U.S. Patent No. 4,932,581 to Ohle et al. discloses the controlled speed feeding of a predetermined length of solder wire to a soldering location. Ohle et al. uses force sensors to

determine feed force or speed by converting force to a proportional electric signal controlling an electric motor. Ohle et al. does not teach any type of sensing device for measuring the

diameter, cross sectional area, or effective cross section of the solder wire. Adjustment of the feed controller for solder wire thickness is done manually, and only for a nominal diameter.

No adjustments are made due to any variance from nominal diameter.

Several patents disclose various devices to control feeding of fibers or yarn. For example, U.S. Patent No. 4,019,691 to Wroclawski et al. discloses winding of yarn onto a

bobbin at a variable and controlled speed. The speed is dependent upon the volume of yarn

held in a receptacle. Sensors in the receptacle are of variable type, including photoelectric

sensors. The sensors determine the volume of yarn in the receptacle. The winding shaft in

Wroclawski et al. rotates at a constant speed. The variable yarn winding speed is

accomplished by use of a cone-shaped bobbin. A different contact point on the bobbin results

in a different yarn winding speed. Wroclawski et al. discloses no monitoring or sensing of the diameter of the yarn to control feed speed. The feed speed of the yarn is constant.

U.S. Patent No. 4,820,281 to Lawler, Jr. discloses measurement of the volume of a

drop of liquid and control means for regulating flow of a liquid. The volume of drops of

liquid in an IV tube is calculated by timing circuits activated by photoelectric sensors to

measure the time it takes for the drop to fall a known distance. Volumetric flow rate is

determined by multiplying drop volume and drop rate, and is adjusted according to a

predetermined flow rate. No continuous feeding of a solid element is taught by Lawler, Jr.

No continuous monitoring of diameter, cross sectional area, or effective cross section of a

filament is taught by Lawler, Jr. The teaching of Lawler, Jr. is unworkable when dealing with

solids.

It would be desirable to provide an apparatus for continuous volumetric control of build

material fed to the extrusion head of a three-dimensional modeling machine. Such an apparatus should be adaptable to be added to an existing machine, as well as added at the manufacturing stage.

BRTEF SUMMARY OF THE INVENTION

It is an object of the present invention to provide a volumetric feed control system that

will allow for an increased tolerance range in build element effective cross section, in other

words, a loosening of the strictness of the tolerance requirements.

It is another object of the present invention to improve the constancy of the flow rate

of fluid delivered through the application tip of a three-dimensional modeling machine.

It is yet anotner object of the present invention to provide a feed control system suitable

for installation on an existing three-dimensional modeler including the type disclosed in U.S.

Patent No. 5, 121 ,329 to Crump, which is hereby incoφorated by reference, and which will

require little modification of existing equipment.

The present invention achieves these objects by providing a system for continuously measuring, computing, and monitoring the effective cross section of a build material element

such as a filament as the element is fed to the melting unit or application tip of the modeling

machine, and adjusting the speed at which the element is fed to the melting unit or the

application tip to ensure a more constant flow rate of fluid at the application tip.

Build material element is fed to a modeling machine by an advancement mechanism including a motor. The motor can be a stepper motor or a DC servo motor. In the present invention, the build element is advanced to the dispensing head by the advancement

mechanism. The effective cross section of the element is determined through the use of

sensing means and a central processing unit. Various configurations for determination of effective cross section may be used, for example infrared emitter and detector pairs, tungsten filament and infrared detector arrays, pinch rollers and linear variable differential transformers

(LVDT), and capacitive measurement.

There is usually a spatial gap between the sensing means and the application tip of the three-dimensional modeler. Therefore, a lag response exists in the system. A variance in

filament cross section will not immediately affect the volume of filament material present at the application tip. To combat this delay, the distances between the sensing means and the

application tip as well as between the sensing means and the center of the modeler

advancement mechanism are provided to the central processing unit. The sensing means

provide continuous signals to the central processing unit which can then compute the effective

cross section of the filament. The motor of the advancement mechanism is connected to the

central processing unit to allow the central processing unit to know the speed of the motor at

all times. Since the central processing unit knows the speed of the motor and receives measurements from the sensing means, the effective cross section of the element may be

continuously computed. Necessary changes to motor speed and hence feed roller speed and

the proper times to effect the changes in speed may therefore be controlled by the central processing unit to ensure that the lag response is continuously and properly compensated.

The volumetric feed control apparatus of the present invention will allow the tolerance

of flexible filament to be increased. In other words, the filament diameter accuracy level required by the present invention is less than that previously required by the prior art. The reason for this is the use of more accurate methods of providing a constant flow rate of

filament to the modeling machine application tip. This relaxed tolerance requirement will considerably cut the manufacturing costs of filament, and therefore of models created by

modeling machines.

Further, any existing system using a roller feed method of feeding a flexible filament

to a modeler may easily use the present invention. Insertion of the sensing means into an existing system may be easily accomplished. The central processing unit may be located

anywhere, provided that proper connections are made to the sensors and to the motor controlling the advancement mechanism. No changes other than the drive control of the

motor, and the possible addition of an encoder f edback to the motor are necessary.

Typically, the build material element is configured in the shape of a cylindrical

filament, normally wound on a spool or stored in a roll. However, the element may be

supplied in a wide variety of configurations, including ribbons, tubes, extrusions of triangular,

trapezoidal, or pentagonal shape, and the like. The build material element may also be supplied in discrete quantities, and need not be wound on a spool.

These and other objects and benefits of the present invention will become apparent from

the following detailed description thereof taken in conjunction with the accompanying drawings, wherein like reference numerals designate like elements throughout the several

views.

DESCRIPTION OF THE DRAWINGS

Fig. 1 is a perspective view of an embodiment of the volumetric feed control in place on a

machine;

Fig. 2 is a view of an embodiment of the sensing system of Fig. 1 , taken along line 2-2

thereof;

Fig. 3 is a view of an alternative embodiment of the sensing system of Fig. 2; Fig. 4 is a view of another alternative embodiment of the sensing system of Fig. 2; Fig. 5 is a view of yet another alternative embodiment of the sensing system of Fig. 2; and

Fig. 6 is a partial block diagram of a control system for the volumetric feed control.

DESCRIPTION OF THE PREFERRED EMBODIMENT

Referring now to the drawings, Fig. 1 shows the volumetric feed control 10 in place on a three-dimensional modeling machine 12, the machine 12 having an element advancement

mechanism such as feed rollers 14, a dispensing head 16 with an application tip 17, and a frame 18. A build material element such as filament 20 is fed to machine 12 from a build

element source, not shown, which may be a spool or roll of material, or another such storage

device. Build material is often formed as a filament such as filament 20, but may take other

configurations. Such other configurations include those of different cross sections. Further,

the build material element may be supplied in discrete quantities, and need not be wound in a roll or on a spool.

Volumetric feed control 10 may include dimension control ring 22 and DC servo motor 24 with encoder feedback. Dimension control ring 22 contains sensing means 23 to determine

the effective cross section of a build material element such as filament 20 being fed through

the dimension control ring 22.

Cross sectional area is often used to estimate volume. Also used are diameter

measurements. These methods of determining volume only work when no anomalies are

present. Such methods of measurement, especially using conventional instruments and techniques, fail to take into account that if only a diameter measurement is made, a volume computation is meaningless if the element being measured is hollow. The effective cross

section is a measurement that takes into account such factors as the cross sectional area of the

element, known quantities such as length, width, and diameter, and other information, such

as whether the element is hollow, oblong, or the like. The use of effective cross section

reduces the possibility for miscalculation due to various physical factors of the element. A

more accurate volume calculation may then be made. The sensing means 23 is operatively

connected to a central processing unit 74, and continuously gathers data and feeds information

signals measuring the build material element being fed therethrough to the central processing

unit 74. The central processing unit is also operatively connected to a motor 24, which may be a DC servo motor or a stepper motor. Motor 24 is in turn operatively connected to feed rollers 14 or an other element advancement mechanism which may be used on the modeling

machine. The central processing unit 74 controls the speed at which the motor 24 turns, and

therefore the speed at which the advancement mechanism rotates to pull the build material element through the dimension control ring 22. Central processing unit 74 adjusts the speed

of motor 24 in order to provide a constant flow rate of build material element to dispensing

head 16 and application tip 17.

Sensing means 23 need not be housed in a dimension control ring 22. It may be

mounted at any place in which it will be positioned to make measurements on element 20.

Sensing means 23 is preferably positioned between any driving contact means such as feed rollers 14 and the application tip 17. This is because the driving contact means may affect the effective cross section of the element due to forces imparted by the feed rollers 14 or other

mechanical driving means. Sensing means 23 may even be incoφorated into the element

advancement mechanism, such as by mounting the feed rollers to communicate with a linear variable differential transformer as described below.

Referring now also to Figs. 2-4, various embodiments of the sensing means 23 may be seen. In Fig. 2, filament 20 is shown substantially centered in dimension control ring 22. As has been mentioned, dimension control ring 22 is not required. The sensing means 23 may

be mounted elsewhere. Although in this configuration it is preferred that filament 20 be centered in dimension control ring 22, it is not necessary. In Fig. 2, a tungsten halogen lamp

26 is oriented with its lamp filament 28 parallel to and a fixed predetermined distance 30 from

the axis 32 of filament 20. The emitted light 34 from the lamp 26 creates a shadow of the filament 20 on a linear CCD array 36. The shadow width is designated as letter A (Fig. 2).

Shadows of different widths A will cause a variation in the number of array pixels 38 that are

illuminated by the light 34 from lamp 26. The width 21 of filament 20 may then be

determined using standard clocking and preamplification techniques known in the art and not further described herein. These measurements are used to generate an effective cross section

of filament 20.

Referring to Fig. 3, a sensing means 23 using a three-coil linear variable differential transformer (LVDT) 40 and a pair of dimension rollers 42 is shown. Dimension rollers 42 are

mounted to shafts 44. Each shaft 44 is flexibly attached to rigid mount 46 by a suitable

mounting apparatus, such as pins 48. Rigid mount 46 is attached to the interior surface 72 of

the dimension control ring 22, but may also be attached at any fixed point on the frame of the

modeling machine. The dimension rollers 42 are biased toward each other by spring 50

attached between shafts 44. Each shaft 44 also carries an attachment for LVDT 40 which is

attached therebetween. The terminals of LVDT 40 are operatively connected to rollers 42, one terminal to each roller. As has been described above, tf oilers 42 may be used with multiple functions, such as to also serve as feed rollers for eicment advancement, or encoding the

element velocity.

Filament 20 passes between dimension rollers 42. As the filament diameter changes, the dimension rollers 42 respond by moving, causing motion of the shafts 44. The movements of the dimension rollers 42 cause the core 52 of LVDT 40 to change position within LVDT

40. Sense circuits known in the art may be employed to translate the relative changes in core 52 position and inductance in the LVDT coils to an analog voltage which may be used to compute the effective cross section of the filament 20.

Fig. 4 shows a third embodiment of the sensing means 23 used in the dimension control ring 22 or alternatively mountable directly to the modeling machine. In this embodiment, the filament effective cross section is calculated using information from a capacitive sensor 54.

The sensor 54 is comprised of a pair of conducting plates 56 sandwiching a thin insulator 58

therebetween. A typical insulator 58 is mylar, approximately 0.002 inches in thickness. Insulator 58 lies in a plane 60 through which a hole 62 is drilled, the hole 62 being coaxial

with the axis 32 of build filament 20 and coplanar with plane 60, and exposing the conductive

plates 56. The diameter 64 of hole 62 is slightly larger than the diameter of the largest

filament 20 expected to be used with the modeling machine. A capacitance meter 66 is attached to the conductive plates 56. Changing diameter of the filament 20 will cause a change in capacitance that may be translated in known fashion to the cross sectional area of the

filament. This method of determining the effective cross section of the filament 20 is independent of its cross sectional shape, since the capacitive sensor 54 essentially measures the

percentage of the hole 62 that is filled with filament 20 as opposed to air. The capacitance

measurement scheme may also be used to detect the presence of absorbed water in the filament

20, since the filament 20 material and water have different capacitive properties.

Yet another alternative embodiment (Fig. 5) of the sensing means 23 comprises at least

one set of emitter/ detector pairs 68, 70 spaced 180 degrees apart around the axis of the filament 20. The emitter/detector pair 68, 70 is situated so that the filament 20 will pass

through the beam from emitter 68 as the filament 20 is advanced by feed rollers 14. The

emitter 68 sends out a light signal or beam that may be partially occluded by filament 20 and

which is detected by the opposite detector 70. If the filament 20 changes in size, the intensity of the beam detected by the detector 70 will change, allowing a measurement which may be

used to calculate the filament effective cross section. Preferably, a plurality of emitter/detector pairs are spaced at equal intervals around the axis of filament 20 in order to more accurately

detect and measure the effective cross section of the filament 20 as it is advanced by the

advancement mechanism.

Other types of sensors are also compatible with the volumetric feed control 10, and

could be used in place of the emitter/detector pairs. Such other sensors include but are not limited to opto-electric sensors, miniature CCD cameras, fiber optic sensors, mechanical

bladder ring sensors, or pinch roller sensors. All could send information that may be used to

determine the effective cross section of the flexible filament to the central processing unit 74.

The various sensing means 23 described above each send a signal corresponding to the

filament effective cross section to central processing unit 74. The central processing unit 74

computes the effective cross section and the required feed speed of the filament 20 to provide a constant flow rate of build material to the application tip 17 of the dispensing head 16, and

adjusts the speed of motor 24 which controls the speed of filament feed by controlling the

speed of the advancement mechanism such as feed rollers 14. This adjustment ensures that any changes in the effective cross section of the filament 20 will result in a corresponding change

in the speed at which the filament 20 is fed.

If further accuracy of effective cross section measurement and testing of properties of

the filament 20 is desired, two or more sensing means may be combined or used in sequence to determine multiple properties of the filament 20 and provide a reference for averaging

computed values. For example, an LVDT sensing means as described above may be used along with a capacitive sensor as also described above, in order to determine both the outer

dimension and the water content of the filament 20. Consequently, the tolerance of a filament

with nominal diameter of 0.070 inches can be increased substantially from +/- 0.0015 inches.

Another way to increase the accuracy of the effective cross section is to more closely monitor the effective velocity of the element. Although the rotational speed imparted by the

element advancement mechanism provides a good approximation of the element velocity, the

approximation may be in error due to a number of factors. The element advancement mechanism itself often requires a groove such as that shown in Fig. 3 on rollers 42 in order to seat the build material element properly. The angular velocity of the rollers and accordingly

the linear velocity imparted by the rollers differs at the outer periphery and at the central most portion of the groove. To combat such inaccuracies, the element advancement mechanism may

be encoded for element velocity. When such an account is made of effective element velocity,

this measure may be combined with effective cross section to more accurately compute the volumetric flow rate.

Build material elements may be fabricated in a variety of cross sectional shapes and

geometries, including ribbons, tubes, extrusions of triangles, trapezoids, pentagons, and other

polyhedra. The measurement techniques described above may be adapted to these alternative geometries.

In operation, the volumetric feed control 10 works as follows. The sensing means 23 sends measurement signals to central processing unit 74. Central processing unit 74 uses the

signals to continuously compute the effective cross section of the build element. In response to this known quantity, central processing unit 74 is able to adjust the feed speed of the build

element by adjusting the speed of motor 24 controlling the feed rollers 14 that advance the build element toward the dispensing head 16 and application tip 17. The adjustment is made

so that a constant flow rate of build material is fed to the dispensing head 16.

A time lag exists between the computation of build element effective cross section and the arrival of the particular measured portion of the element at the application tip 17.

Consequently, a change in effective cross section of the build element at the sensing means 23

does not immediately translate to a change in volume of material at the application tip 17. The

change in build element effective cross section will affect the volume of material at the

application tip at some point in time after it passes the sensing means 23.

To account for this lag, the central processing unit 74 must know the distance 76

between the application tip 17 and the sensing means 23, as well as the distance 78 between the sensing means 23 and the center of feed rollers 14. The connection of motor 24 to the

central processing unit allows it to know or compute at all times both the effective cross section of the build element and the speed of the feed rollers 14. The central processing unit

74 can therefore correct for the lag response in the system and provide a constant volume of material at the application tip 17 regardless of variances in the effective cross section of the build element.

The detailed description outlined above is considered to be illustrative only of the principles of the invention. Numerous changes and modifications will occur to those skilled in the art, and there is no intention to restrict the scope of the invention to the detailed description. The preferred embodiment of the invention having been described in detail, the

scope of the invention should be defined by the following claims.