US3605528A - Incremental construction of three-dimensional objects having premachined rod elements and method for forming the same - Google Patents

Abstract

A THREE-DIMENSIONAL ARTICLE AND A METHOD OF MANUFACTURING IT, SAID ARTICLE COMPRISING INDIVIDUAL INCREMENTS IN THE FORM OF RODS THAT ARE PREMACHINED OR PREFORMED AND ASSEMBLED IN REGISTRY, SAID RODS BEING PRECUT TO DISCRETE LENGTHS, ONE END OF EACH ROD FORMING AN INCREMENT OF A PRECALIBRATED SURFACE CONTOUR, EACH INCREMENTAL SURFACE BEING LOCATED IN AN OPTIMUM PLANE TANGENT TO THE PRECALIBRATED SURFACE, THE ROD LENGTHS AND THE CUTTING ANGLES FOR THE INDIVIDUAL RODS BEING DETERMINED NUMERICAL CONTROL TECHNIQUES.

Description

sept. 2o, 1971 F. E. wHrrAcRE ETAL 3,605,528

INCREMENTAL CONSTRUCTION OF THREEl-DIIVIENSIONAI.;` OBJECTS HAVING PREMACHINED ROD ELEMENTS AND METHOD FOR FORMING THE SAME 8 Sheets-Sheet l Filed Aug. 2, 1968 .5: aff/c Sept. 20, 1971 EE. wHlTAcRE ETAL 3,605,528

INCREMENTAL CONSTRUCTION 0F THREEDIMENSIONAL OBJECTS HAVING PREMACHINED ROD ELEMENTS AND METHOD FOR FORMING THE SAME Filed Aug. 2, 1968 8 Sheets-Sheet 2 45T" l 3?' l @j fg /i N l I :L5 L/f ffl Sept. 20, 1971 F. E. wl-n'rAcRE ETAL 3,505,523

INCREMENTAL CONSTRUCTION 0F THREEDIMENSIONAL OBJECTS HAVING PREMACHINED ROD ELEMENTS AND METHOD FOR FORMING THE SAME Filed Aug. 2. 1968 v8 Sheets-Sheet 5 E1 NM NN J *N I wmf @MW@ l 1 3,605,528 s HAVING sept. 2o, 1971 I EMENTAL CONST EMACHINED ROD ELEMENTS AND METHOD FOR F Flled Aug 2 1968 F. E. WHITACRE E:TAI- RUGTION OF THREE-DIMENSIONAL OBJECT ORMING TH E SAME 8 Sheets-Sheet 4.

Sept. 20, 1971 F, E, wHrrACRE ETAL 3,605,528

INCREMENTAL CONSTRUCTION 0F THREE-DIMENSIONAL OBJECTS HAVING PEEMACHINED Roo ELEMENTS AND METHOD FOR FORMING THE SAME Filed Aug. 2, 1968 8 Sheets-Sheet 5 Sept. 20, 1971 F E wHrrACRE ETAL 3,605,528

INGREMENTAL CONSTRUCTION OF THREE-DIMENSIONAL OBJECTS HAVING PREMACHINED ROD ELEMENTS AND METHOD FOR FORMING THE SAME Filed Aug. 2, 1968 8 Sheets-Sheet 6 afi;

@ irl/iig;

E] M 59 l Sept. 20, 1971 F. E. wHlTAcRE ErAL 3,605,528

INCREMENTAL CONSTRUCTION 0F THREE-DIMENSIONAL OBJECTS HAVING PREMACHINED ROD ELEMENTS AND METHOD FOR FORMING THE SAME Filed Aug. 2. 1968 l 8 Sheets-Sheet 7 INVENTO S.'

y rar/VIV;

Sept. 20, 1971 F, E, wHn-ACRE ETAL 3,605,528

INCREMENTAL CONSTRUCTION OF THREE-DIMENSIONAL OBJECTS HAVING PREMACHINED ROD ELEMENTS AND METHOD FOR FORMING THE SAME Filed Aug. 2, 1968 8 Sheets-Sheet 8 Jzzz Jazz 2C z: 2z m m M'efz 5 xy irri/2W figa@ ,7g4/vf! ,7j/MW( f' ff: rf E f. I If/754' 325.257 Zi/f agg/M1 g2g/fig 'rur l l l v| l l I l I I l l I I 1 l I I l I I i l( 7 I l i l l l l I I l I I I I y I l i (E) 1E) E I I azz al I um I ig I jid/ff U//zr II IW?! i g I EWE@ kfw fl/ D G) (D United States Paten INCREMENTAL CONSTRUCTION F THREE- DIMENSlONAL OBJECTS HAVING PRE- MACHINED ROD ELEMENTS AND METHOD FOR FORMIN-G THE SAME Foster E. Whitaere, Farmington, Archie A. Pearson, Dearborn, Harold N. Bogart, Farmington, and Norman W. Hopwood, Jr., Dearborn, Mich., assignors to Ford Motor Company, Dearborn, Mich.

Filed Aug. 2, 1968, Ser. No. 749,685 Int. Cl. B21k 5/20 U.S. Cl. 76-107 12 Claims ABSTRACT OF THE DISCLOSURE A three-dimensional article and a method of manufacturing it, said article comprising individual increments in the form of rods that are premachined or preformed and assembled in registry, said rods being precut to discrete lengths, one end of each rod forming an increment of a precalibrated surface contour, each incremental surface being located in an optimum plane tangent to the precalibrated surface, the rod lengths and the cutting angles for the individual rods being determined by numerical control techniques.

GENERAL DESCRIPTION OF THE INVENTION Our invention is related generally to the manufacture of three-dimensional objects, such as metal forming dies, although the teachings of our invention can be applied also to construction of other objects such as die casting dies, plastic molds and other objects of various geometry.

It is possible also to employ the teachings of our invention in forming objects that are related indirectly to the metal forming art. Examples of such objects might be tooling aid devices, fixtures and objects which might otherwise be formed by casting techniques.

'We prefer to use metal rods of hexagonal shape in building a die or other article. The rods are precut to length using a numerically controlled cut-off machine, the ends of each of the rods forming surface increments that are optimally tangent to a curved surface contour of a precalibrated surface. The point of tangency between the incremental tangent surface and the curved surface is located at the rod center.

An automated, numerically-controlled identification and assembly procedure makes it possible to identify individual rods and to position them in proper registry with the rod ends forming the equivalent of a rough-machined surface for a casting.

The individual rods may be bonded together by brazing, diffusion brazing, adhesive bonding, chemical bonding or other bonding procedures to form a solid section. The section then can be nish machined to a required surface contour.

In the case of a metal forming die, the upper die section and the lower die section can be formed simultaneously. The basic procedure used during the numerically controlled rod cutting step produces registering elements of the upper and lower die sections, the matching ends of the twoy rods forming increments of the rough machined die surface equivalent for their respective die sections. The die surface angle formed during each angular cut for the rod for one die section is necessarily the proper die surface angle for the corresponding rod for the remaining complementary die section.

rIhe numerical definition of the finished die surface is determined with computer assisted data processing steps. The surface itself is normally defined initially in two dimensions on a surface layout, although other techniques have been used. A series of points on character lines and section lines on the two-dimensional surface draft is translated by interpolation into a mathematically continuous surface. With aid of a computer, a network of mathematically computed points in space can be obtained with sufiicient density so that the points define the finished surface in three dimensions.

Tangent vectors for each of the points thus computed can be determined by partial differentials with reference to each of two coordinate planes. Having determined these vectors, a surface normal can be determined mathematically by using, for example, a cross-product method. It is this normal that determines the cutting angle for the die surface end of the rod and the point at which each normal is determined is made to correspond to a point lying on the axis of the individual rod. Where compound curvature precludes a unique common plane, the center is translated to assure sufficient machining stock. Similarly, the location of extreme angles, determined to be still in saw range, are adjusted for variations in saw kerf.

The computed data is prepared with known data processing procedures and used to generate a control tape which is adapted for controlling a numerically-controlled cut-off machine. Intelligence stored in the tape includes data necessary to obtain the proper rod length for each of the individual rods, the die surface angle for each of the rod ends and the identification number for each rod. It is desirable to side-stripe the rods to assist in orientation during final assembly.

It is possible to include rods of different hardness in the completed die. Rods of varying chemical composition may be used to provide variable wear resisting qualities in selected areas of the die surface if this is desired. Inl this instance the preprocessing steps in the data processing procedure for preparing the control tape can be modified to include necessary instructions to provide material selection at the outset before the angular cuts are made.

It is possible also under some circumstances to use rods other than hexagonal rods. For example, triangular rods, rods with trapezoidal cross sections or rods with cross sections in the shape of modified squares or parallelograms also may be used. The rods should have, however, a selforienting feature in order to permit a registry that establishes structural rigidity.

Although the instant disclosure is directed to a construction for a metal forming die and to a method for making them, corresponding procedural steps can be used also to form objects other than metal forming dies, such as die casting dies, casting equivalents of objects such as fixtures, certain tooling aids, machine bases, et cetera.

In making a numerical definition of the die surface, a process drawing is prepared concurrently with the numerical processing of the panel surface. This includes all the lines necessary to describe the surface. Each line is given an identifying number on the drawing. The numbered lines on the drawing then are identified on a body draft, showing the surface in orthogonal representation. The lines necessary to describe the surface are superimposed on the body draft. All lines are digitized and used as raw input data for the computer assisted processing steps. The mathematical equations of the characteristic lines then are determined, which information is used to develop a matrix of points in space.

Other factors necessary for the building of the incremental die, including holes, edges or other special conditions such as the size and shape of the bars, and the symmetry or lack of symmetry of the die then are determined.

Special regions of the die may be defined if this is necessary. For example, it may be necessary to use a stop-off material to prevent brazing or adhering of certain incremental rods at one location, but not at another. Also the rods may be required to be shortened at one location, or removed, or elongated. Different rod material in certain regions might be required. The line enclosing the region in question must be defined numerically, and the relationship of that line to the other character lines on the surface must 'be determined,

Information regarding identification of the rods in certain regions must also be determined numerically. Such information must include instructions for slabbing or grooving of the corners of the rod or permanent numbering of the rods.

A two-dimension grid of points representing the corners and centers of all the rods required to make up the incremental die is determined. The points on this grid then are projected onto the numerical denition of the die surface which already is determined, as explained above, thereby providing three-dimensional data and surface normal data.

After the rods are cut and oriented, a reading head scans the identification of the individual rods and identifies the pieces of the rods. Odd numbered pieces are channeled in one direction, and even numbered pieces are channeled in another direction. This in effect separates the hex rod elements into complementary groups, which are the principal components of the punch half of the die sections and the principal components of the die half of the die sections.

Following this stage, both the odd numbered pieces and the even numbered pieces are arranged in numerical order. This operation is automated under numerical control. If the numerical arrangement is interrupted by reason of a missing piece, a gap in the row .will appear and a signalling device will alert the operator so that the missing piece may be located and inserted in the gap, thus allowing a continuous iiow of rods. If necessary, a dummy piece may be inserted into the gap thereby allowing operation to continue. The dummy then can be replaced in an olf-line operation with an actual rod.

After the rods are arranged in numerical order, they are assembled into a xture of special height, width, and depth to permit assembly of all the rods for the particular die half involved. All of the faces of the rods that are identified by slabbing are oriented in a common direction.

The automated portion of the system may be a transfer line type of machine into which the raw material in the form of the hex bar stock can be introduced and fed through the several stations. Changes occur in this raw material in progressive stages, as explained above, as it is transferred from one station to the next. IIt finally emerges from the terminal end of the system as hex rod elements of varying dimensional characteristics. The processing that occurs at the servo stations is initated and controlled primarily through the use of numerical data, although this is supported by automatic cycling devices.

The numerical data is the result of the merging of the process data required for the incremental die and the numerical representation of the die surface, as indicated in the schematic diagram of FIG. 1.

BRIEF DECRIPTION OF THE FIGURES OF THE DRAWINGS FIG. 1 is a schematic diagram of the method steps used in making incremental metal forming dies.

FIG. 2 is a plan view of an assembled incremental die.

FIG. 3 is a cross section view taken along the plane of section line 3 3, FIG. 2.

FIG. 4 is a cross sectional view taken along the plane of section line 4--4 of FIG. 2.

FIG. 5 is an isometric, sectional view taken along the plane of section line y5---5 of FIG. 2.

FIG. 6 is a cross sectional, geometric representation of a die surface formed by the ends of the rods. the ends being situated in planes that are tangent to the finished surface contour at the location of the rod center d lines. It is taken along the plane of section 6-6 of FIG. 8.

FIG. 7 is a cross sectional fview of cooperating upper and lower die sections of a metal forming die, the plane of the section being parallel to the center lines of the incremental rods.

FIG. 8 is a partial plan View of the incremental rods of FIG. 7 as seen from the plane of section line 8 8 of FIG. 7.

FIGS. 9A, 9B and 9C is a process flow chart showing the method for constructing a numerically-controlled, incremental, rough-machined casting equivalent.

FIG. 10 shows a typical machine base made in a form of a hexagonal rod construction.

FIG. l1 is an automotive fender stamped from a die of the type illustrated in FIG. 7.

PARTICULAR DESCRIPTION OF THE INVENTION In FIGS. 2, 3, 4 and 5 we have illustrated one of two die sections for forming sheet metal. In this instance, the die section has a cooperating die surface in the form of a concavity. The cooperating die section would have a convex die surface that would register with the concavity of the die section of FIG. 2.

Numeral 10 designates in FIG. 2 a die housing, which may be in the form of a steel box. The housing may be of any shape desired although in the FIG. 2 embodiment it is square.

Situated in the housing in close registry are hexagonal rods 12. Each rod is cut to form an upper surface, such as that shown at 14, which generally cooperates with an adjacent end surface of the adjacent hexagonal rod. The surfaces directly adjacent surface 17 are designated by reference characters 16, 18, 20 and 22 in FIG. 5. The height of each rod as well as the angularity of the end surfaces are chosen so that each individual surface forms an increment of a larger surface having a contour that approximates the contour of the desired iinished die surface.

In the embodiment shown in FIGS. 2, 3, 4 and 5, the row of hexagonal rods directly adjacent the vertical walls of the housing 10 defines a pilot ridge 24 which extends around the periphery of the die section. Each individual rod is formed with a segment of the ridge 24 so that when the individual segments are joined side-by-side they dene the peripheral ridge 24. This ridge registers with a peripheral groove in the registering die section so that the die sections are piloted, one with respect to the other, into perfect registry.

When the die sections are brought together into registry, eac'h increment end surface of each rod 12 registers with a cooperating end surface of a companion rod for the other die section. The angle of the surface for any given rod 12 in the die section of FIG. 5, for example, would have an identical angle on the cooperating end surface for the companion rod of the other die section. A clearance space between these two end surfaces accommodates the metal thickness of the sheet metal that is Worked by the dies.

In FIG. 7 We have shown in cross section form a pair of die sections, each having a die surface that registers with the die surface of its companion section. The space betweenithe companion end surfaces of the upper and lower rod segments of the upper die section 27 and lower die section 29 is occupied by the sheet metal 25.

The die sections shown in FIG. 7 have been finish machined to provide continuous, smooth, die surfaces. This machining process employs numerically-controlled, multiple-axis, milling machines in a manner described in copending application Ser. No. 577,997, tiled Sept. 8, 1966, which is assigned to the assignee of this invention. Reference may -be had to application Ser. No. 577,997 for the purpose of supplementing this disclosure.

Prior to the finishing machining operation, each end of the hexagonal rods forms an end surface that is tangent to the die surface contour at a point that falls on the center line of the rod itself. This is illustrated in FIG. 6 where the individual rods 12 include surfaces 26, 2S and 3f) which form tangent planes for the die surface contour shown at 32. The points of tangency for the surfaces 26, 28 and 30 coincide with the points of intersection of the center lines 34, 36 and 38 for the rods 12 with the nished die surface 32. The region between the surface 32 shown in FIG. 6 and the individual tangent planes provided by the surfaces 26, 28 and 3()` represents excess metal that is removed during the finish machining operation. The tangent planes themselves approximate the contour 32 after the rods are assembled in registry. This is the equivalent of a rough-machined casting of the die section.

It is not necessary, as in the case of finish machining of cast die sections, to remove large quantities of metal during the machining operation. No rough machining is required. Only a single finish machining operation is needed to transform the approximate surface contour provided by the tangent planes of the rod end surfaces into the finished contour represented by reference character 32.

The surface designated by reference character 32 in FIG. 6 should correspond to the mathematically determined surface defined by the computed points in space described in the preamble portion of this specification. Suitable dowlines, or proportionately defined lines on the surface 32 can be computed by mathematically interpolating appropriate points on the specified surface.

The tangent vectors at any point along the surface to be machined can be determined readily by the expedient of determining a partial derivative of the equation of a line of intersection between the computed surface contour and a plane parallel to one of the coordinate planes. Another tangent vector containing that same point can be determined with respect to another coordinate plane. Having these two tangent vectors, it is possible to obtain the unit normal vector. The center of the cutting tool can be adjusted accordingly along the unit normal vector so that the cutting tool itself always machines a tangent plane at any given point on the computed surface.

The machine control may include an automatic parabolic interpolator that requires the simultaneous definition of the three coordinates of the points. Having rethe three coordinates of selected points in its control systern, the milling cutter, several of which are known in the industry, will direct the cutting tool to machine the surface 32 through the three points along a parabolic arc rather than along straight line segments between the points.

The information necessary to obtain this unit normal vector can be used also with modified computer subroutines to determine the normals for the incremental end surfaces of the rod which in turn determine the angle of the cutting tool that cuts the individual rod segments and the angular position of the rods during the cutting operation.

In FIGS. 9A, 9B and 9C we have shown a process flow chart for various machining operations and the assembly procedures necessary to convert hexagonal bar stock into a finished pair of matching die sections for working sheet metal. The chart should be read from left to right. The various process steps have been indicated in the upper portion of the view of FIGS. 9A, 9B and 9C; and the corresponding functions have been illustrated in the sketches in lower portion. The sketches and the method steps of the block diagram have been related by corresponding numbers.

The assembly procedure has been divided into 13 stages, each stage being identified by a Roman numeral character. Between each of the stages the nal product of the previous stage is transfeired and reloaded for entry into the next stage, which may take place at a different physical location.

At the outset, a turret capable of accommodating hexagonal rods is loaded. If the particular die under consideration requires areas of varying hardness, rods otv varying hardness are entered at this Stage. Rods of differing alloy content also may be used. Automatic feeders can be provided for feeding the hexagonal rod stock to the turret. A preprogrammed control tape is provided with information which will cause the numerical control system to command the exact angular adjustment of the turret that will permit proper material selection as designated in the sketch of step I. After the turret position is Stich that a proper material is conditioned for advancement through the turret, the rod is advanced against a xed stop. As the stop is withdrawn, a cutting tool moves across the plane of the raw bar stock as indicated in step number 4 for stage I. A tail stock, which is situated in alignment with the turret, then is adjusted to a proper X- axis position. The amount of the adjustment depends on the preprogrammed instruction that is provided to the control system by the programmed tape. Each rod, of course, will have its discrete position determined at step 5. It is different for each rod, unless, of course, the die section required rods of constant length. After the X-axis position is determined for the tail stock, the bar cut in step 4 is fed into the tail stock as indicated in step 6. At that time the rod is cut to length by a cut-off tool as indicated in step 7. The cut rod now becomes an inprocess hex rod whose length represents the adjusted, computed length resulting from the computer program.

The cut rod is transferred and loaded into a fixture to permit milling cutters, operating in tandem, to form a flat on one side of the hex rod. The rod is clamped and fed into a stop to establish a suitable position for one of the milling cutters. The other milling cutter is adjustable in 'the direction of the X-axis indicated in stage II, the amount of the adjustment depending upon theilength of the rod involved.

The milling cutter adjustment is necessary to accommodate varying lengths of rods, and it is made in response to computed data. The machined flats provide a convenient area for impression stamping an identification number at each end of the rod such that the metal raised around the characters will be below the original surface of the material where it will not cause interference preventing surface-to-surface contact of the rods after assembly.

The machined rod then is transferred again and loaded into another fixture as indicated in stage III. Identifying binary codes are punched on the flats that are machined in stage II. This code identifies a rod so that it can be assembled in its proper registry with the adjacent rods in the final assembly.

The bar is automatically positioned against an end stop and clamped. One marking head is fixed and the other is adjustable in response to the Command of the numerical control system to accommodate the spread between the flats previously machined. The identifying numbers are consecutive from end to end and progressive from rod to rod.

After the identifying marks have been impressed on the milled flats at each end of the rod, the ends of the rods are chamfered by chamfering tools that are advanced into the rod along the rod axis as indicated in step 2 of stage III.

The chamfering operation can be automated. It does not require instructions from the numerical control tape. Further, each of the transfer and loading operations can be fully automated independently of the numerical data in the control tape.

The selection of the material in stage I and the positioning of the tail stock to a discrete length in stage I also requires instructions from numerical control tape, although the other functions in stage I can be fully automated independently of the numerical data. It is considered practical operation to establish groups of similar length bars during the previous computation to simplify this operation. Alternatively, maximum yield from standand bar lengths could be obtained by optimized grouping.

The chamfered and milled rod then is transferred and loaded into a preoriented collet indicated in stage IV. The collet is oriented about the A axis is response to numerical control data since it must be capable of receiving a rod in its proper angular position with reference to the position orienting reference angle that accounts for the position of the milled flat ends.

If the particular rod involved requires a fiat side, the proper flat can be machined at this stage. This is done by feeding the rod through the preoriented collet into the path of motion of a milling cutter, the X-axis position of which is determined by numerical data in addition to the X-axis position of the cutter, it is necessary to properly orient the angular position of the rod prior to the machining. This is indicated in the step 2 in stage IV. This adjustment also occurs in response to numerical control instructions provided by the control tape.

If the rod does not need milling, of course, this stage can be omitted. Milling may be needed, for example, when it is desired to provide an axial passage in the finished die to permit entry of a temperature controlling media or lubricant, or to provide desired discontinuities in the surface.

The machined rod then is transferred to an automated transfer and load device and received by a preoriented collet in stage V. The angular position of the collet is determined with numerical data.

A cutter blade in the cut-off machine is adjusted as indicated in step 2 in stage V to a proper angle 0, with respect to the center line of the rod. The rod then is adjusted in the X-axis direction as indicated in step 3 of stage V. This stage requires feeding of the rod through the preoriented collet shown in step 1 of stage V until it engages a stop, and then determining the angular position of the blade and the X-axis position of the rod. The rod is cut as indicated in step 4. This forms two rod segments, one of which will form an element of one die section and the other of which will form a complementary element of a companion die section, The cutting itself can be fully automated, although the angular positioning of the blade, the X-axis adjustment of the rod and the angular orientation of the collet each require instructions in the form of numerical data.

The cut pieces are transferred through an automated transfer and loading device and received by appropriate mounting fixtures at stage VI where an automated wire brushing operation removes flash and burrs from the cut edges of the rods. After the rods are wire brushed they may be transferred to an automated conveyor. Selected rods are received then by a degreasing apparatus if this stage is desired, which immerses the rods into a degreasing liquid bath. The rods also may be grit blasted if required.

The grit blast and the degreasing steps can be fully automated steps, but the copper spray step requires numerical data instructions since under some circumstances it might be desired to apply stop-off material instead of the spray copper. This is desired if a copper braze for that rod is to be avoided. For example, if an opening is desired in the nished surface contour at the location normally occupied by a particular rod, that rod may be withdrawn if it is not bonded to its adjacent rods.

The rods then are transferred to a station which will permit orientation of the rods with a scanner operation whereby the identification numbers on the rod ends can be read by a reading head. This occurs, as indicated, in stage X. The rods then are separated into odd and even numbers and are arranged in numerical sequence. The odd numbered rods are arranged in an assembly that defines one die section, and the even numbered rods are assembled to form the other die section. The assembled rods are indicated in stage XIII. The assembled rods of FIG. 9C correspond to the assembly indicated in FIGS. 2, 3, 4 and 5.

This overall operating procedure is illustrated schematically in summary fashion in FIG. 1. A numerical representation of the die surface is one ingredient of the numerical data required by the numerical system that controls the various process steps. In addition to the numerical representation of the die surface, the numerical system also requires process data necessary to adapt the numerical system for an incremental die program. That includes information derived from the surface normals described with reference to the points in space that define the die surface,l which information is used to establish the cutter angle in stage V as well as the orientation of the collet in stage V which holds the rod during the cutting operation. In addition to this, the added process data must be sufficient to enable the cut-off machine to locate the proper X-axis adjustment for the rod parts.

The output data for the system is derived after the two process data inputs are integrated. The output data is used to perform the various functions indicated by the legends in FIG. l.

Having described a preferred form of our invention, what we claim and desire to secure b-y U.S. Letters Patent 1. A method for making a rough machined casting equivalent comprising steps of assembling one or more rods into a work piece handling fixture, adjusting the axial position of each rod with respect to a stop, cutting said rod to a predetermined length, identifying each rod by code notations, mounting said rods in a work holding fix* ture, adjusting each rod with respect to a fixed reference point in the direction of its axis, adjusting the angular position of said rod about its axis, cutting said rod into two segments to form two complementary end surfaces having la common direction for their respective normal vectors, repeating the foregoing steps with a plurality of other pieces of bar stock of predetermined length and material, separating the rod sections cut from stock into separate groups, and assembling together the rod sections of each group into registry with a coded side of each segment oriented in a common direction whereby the cut ends of said rods are contiguous with respect to each other and define an approximation of a predetermined die surface contour.

2. The method as set forth in claim 1 wherein each assembly of rod segments following their positioning into registry are bonded together to form an integral body.

3. The combination as set forth in claim l wherein the loading of said work piece handling fixture prior to the identification of each rod segment includes the step of selecting rods of proper material to satisfy predetermined hardness and metallurgical specifications.

4. The combination as set forth in claim 2 wherein the loading of said work piece handling fixture prior to the identification of each rod segment includes the step of selecting rods of proper materiai to satisfy predetermined hardness and metallurgical specifications.

5. The combination as set forth in claim l wherein the assembled separate rod groups define complementary die surfaces, the rod segments that are cut from a common bar segment being co-axial when one die section registers with the other die section, the angular orientation of both with respect to their common axis being identical, and finish machining of complementary surfaces to form finished die surface contours.

6. The combination as set forth in claim 2 wherein the assembled separate rod groups define complementary die surfaces, the rod segments that are cut from a common bar segment being co-axial when one die section registers with the other die section, the angular orientation of both with respect to their common axis being identical, and finish machining of complementary surfaces to form finished die surface contours.

7. The combination as set forth in claim 3 wherein the assembled separate rod groups define complementary die surfaces, the rod segments that are cut from a common bar segment being co-aXial when one die section registers with the other die section, the angular orientation of both with respect to their common axis being identical, and :finish machining of complementary surfaces to form finished die surface contours.

8. The combination as set forth in claim 4 wherein the assembled separate rod groups define complementary die surfaces, the rod segments that are cut from a common bar segment being co-axial when one die section registers with the other die section, the angular orientation of both with respect to their common axis being identical, and finish machining of complementary surfaces to form finished die surface contours.

9. The combination as set forth in claim S wherein the step of cutting each rod segment into sections is preceded by a numerical representation of the finished die surface including the development of a process drawing of the surface of the part to be formed, preparing a numerical representation of the surface on a two-dimension draft, locating surface definition lines, outlining surface characteristics and cutter Iflow lines as well as edge lines and superimposing said lines on said draft, converting coordinate data from said lines into a set of numerical readings and using said readings |as raw input data for developing a mathematical surface comprising computed surface points, projecting points of said mathematical surface onto the numerical die surface to provide threedimensional data and surface normal data for preparing an output numerical data medium to effect proper numerical control responses during the aforementioned rod processing steps.

10. The combination as set forth in claim 6 wherein the step of cutting each rod segment into sections is preceded by a numerical representation of the finished die surface including the development of a process drawing of the surface of the part to be formed, preparing a numerical representation of the surface on a two-dimension draft, locating surface definition lines, outlining surface characteristics and cutter -ow lines as well as edge lines and superimposing said lines on said draft, converting coordinate data from said lines into a set of numerical readings and using said readings as raw input data for developing a mathematical surface comprising computed surface points, projecting points of said mathematical surf-ace onto the numerical die surface to provide threedimensional data and surface normal data for preparing an output numerical data medium to effect proper numerical control responses during the laforementioned rod processing steps.

11. The combination as set forth in claim 7 wherein the step of cutting each rod segment into sections is preceded by a numerical representation of the finished die surface including the development of a process drawing of the surface of the part t0 be formed, preparing a numerical representation of the surface on a two-dimensional draft, locating surface definition lines, outlining surface characteristics and cutter flow lines Ias well as edge lines and superimposing said lines on said draft, converting coordinate data from said lines into a set of numerical readings and using said readings as raw input data for developing a mathematical surface comprising computed surface points, projecting points of said mathematical surf-ace onto the numerical die surface to provide threedimensional data and surface normal data for preparing an output numerical data medium to effect proper numerical control responses during the aforementioned rod processing steps.

12. The combination as set forth in claim 8 wherein the step of cutting each rod segment into sections is preceded by a numerical representation of the finished die surface including the development of a process drawing of the' surface of the part to be formed, preparing a numerical representation of the surface on a two-dimensional draft, locating surface definition lines, outlining surface characteristics and cutter -floW lines -as well as edge lines and superimposing said lines on said draft, converting coordinate data from said lines into a set of numerical readings and using said readings as raw input data for developing a mathematical surface comprising computed surface points, projecting points of said mathematical surface onto the numerical die surface to provide threedimensional data and surface normal data for preparing an output numerical data medium to effect proper numerical control responses during the aforementioned rod processing steps.

References Cited UNITED STATES PATENTS 881,912 3/1908 Emrick 76-107 1,591,572 7/1926 Stimson 76-107 1,826,783 10/1931 Hess 76-107 2,332,360 10/ 1943 Wakefield 76-107 2,691,905 10/ 1954 Onksen 76--107 CHARLES W. LANHAM, Primary Examiner G. P. CROSBY, Assistant Examiner y U.S. Cl. X.R. 72414, 475

Images