US2012796A

US2012796A - Method and apparatus for forming tubes

Info

Publication number: US2012796A
Application number: US560642A
Authority: US
Inventors: Morgan Arthur
Original assignee: REPUBLIE STEEL CORP
Current assignee: REPUBLIE STEEL Corp
Priority date: 1931-09-01
Filing date: 1931-09-01
Publication date: 1935-08-27
Anticipated expiration: 1952-08-27

Description

Aug. 27, 1935. R. c. F, KURTZE METHOD AND APPARATUS FOR FORMING TUBES Filed Sept. 1, 1931 4 Sheets-Sheet 1 MW m WlTNESSES 1935. R. c F. KURTZE METHOD AND APPARATUS FOR FORMING TUBES Filed Sept. 1, 1931 4 Sheets-Sheet 5 INVENTOR m 2. 5. A

Y-A X45 or PARABOLA c.

425:! sms 336=,592

m/wh -4v WETNE55ES Aug. 27, 1935. R. c. F. KURTZE METHOD AND APPARATUS FOR FORMING TUBES 4 Sheets-Sheet 4 Y q. I\ 3 Q Q m A 9 4 T m 9 w 0 m N H 0 v k 4 5 0 m h a 0 \v .0 K 0 a 2, I .I: wk? 3 .N ko & u m N Q o 2 6 9 9 A a QSNNM Q t z 5 8 6&6 wmslksmkafiz 35::

1M MZM Gag-MM z. WW

Patented Aug. 27, 1935 UNITED STATES PATENT-I OFFICE.

IIETHOD AND APPARATUS FOR FORMING TUBES Reimar C. F. Kurtze, Youngstown, Ohio; Arthur Morgan, administrator of said Kurtze, de-

ceased, assignor to Republic Steel Corporation, Youngstown, Ohio, a corporation of New Jersey Application September 1, 1931, Serial No. 560,642 .6 Claims. (01. 153-54) This invention relates to improvements in forming metal tubes by bending a flat strip of metal longitudinally and progressively to cylindrical shape.

In thus forming pipe and other metal tubes from an elongate, flat blank, it has been commercial practice to advance the metal strip through a series of shaping passes for the purpose of bringing the edges together gradually to complete the cylindrical shape of tubing. The object has been to form a seam that is suitable for welding. But heretofore each edge has tended to gather along the seam and crumple rather than meet the other edge uniformly, so that it has been necessary to apply considerable pressure to bring them into position and smooth them. With small, thin-wall tubing, the metal can be forced back into alignment, but not with blanks of heavy walls nor of larger dimensions. The edges of the larger or heavier tubing may buckle less but, in any event, considerable energy is consumed in undesirable movement of metal, andrthe metal is left in such conditions of strain as to impart to the tubing wall both mechanical weaknesses and increased susceptibility to corrosion.

The objects of this invention are to provide a method and apparatus for bending an elongate flat strip of metal longitudinally and progressively to continuous cylindrical form so as to bring the edges nicely into parallel alignment with'each other without distorting them, with a minimum expenditure of energy, and without impairing the walls in mechanical strength or in resistance to corrosion.

I have discovered that these objects, and others, are accomplished by preventing stretching of the metal longitudinally during progressive bending of a strip, and preferably by preventing stretching also with respect to any neutral circumferential axis. Necessarily, .the outer surface of a metal blank will be stretched circumierentially while the inner surface will be compressed, but the neutral circumferential axis of the blank should not be changed in length. According to this invention, in bending a flat strip of metal longitudinally and progressively,to a cylinder having parallel seam edges, stretching of the metal is prevented in part by causing the walls of the blank to follow certain smooth curves. This movement is such as to result in gradually shifting the middle of the blank away from the initial plane of the strip in such manner that the seam is drawn nearer that plane, and preferably until the completed seam lies between the initial plane of the strip and the middle of the blank. And in part, stretching of the metal is prevented further by maintaining the distance over which the edges are rolled together, from the flat strip to the finished cylinder, at least twenty-five times the cylinder diameter. Inthis way all neutral axial lengths of the strip are kept unchanged in extent during bending ofthe metal, and as a result of this absence of stretching less energy is consumed, the tubing is formed more readily, and residual strains from bending are prevented.

It is believed that in bending the strip the crumpling of the edges of the strip, referred to hereinabove, which form the seam edges of the tube, has been due to progressive extension of those edges resulting from progressive bending as practiced previously in this art. In the practice of this invention such crumpling, or longitudinal extension, is substantially avoided by performing the bending of the strip over a distance at least twenty-five times the tube diameter, as just stated, along an exponential parabolic curve. It is reasonable to suppose, although I do not limit myself to this, that bending in this manner prevents crumpling by causing each increment of the strip to be under compression, so to speak, in the region of the edge and that this compression resists and prevents the undesired crumpling of or longitudinal extension in the edge regions.

In further description of this invention, reference is made to the accompanying illustrative drawings, in which: Figs. 1 to 11, inclusive, show progressive contours of a strip, from. the flat plate in Fig. 1 through ten different cross sections of the strip during bending; Fig. 12 is a side elevation of a strip of metal as it is bent longitudinally from a. plane blank to bring its edges together progressively to form the seam of a continuous, cylinder, illustrative dimensions. and planes of roll passes being indicated; Fig. 13 is a plan view of a half width of the same strip, with dimensions indicated corresponding to those of Fig. 12; and Fig. 14 is an end view of a width of strip, showing a composite of sections of a half-width to outline the formation of the tubsame smooth contours as are obtained under similar conditions with 'a stripof flexible but sub-v stantially non-stretchable, non-upsettable mateguide. A flat strip of heavy paper is bent longitudinally until the edges are brought parallel to each other to form the seam of a cylinder. From the cylinder this is then permitted to unroll progressively until one end is flat. Now it is found that the unrolling of the edges naturally is accomplished over a distance greater than about twenty-five times the diameter of the cylinder. Assuming that the seam Eat the top of the blank, the bottom of the blank when no crumpling of the paper walls is apparent gradually curves upwardly from the cylinder to bring the plane of the fiat strip above that of the tubing seam. If the tubing is formed with its seam at the ,bottom, the middle of the blank then curves downwardly from the cylinder rather than upwardly, as will readily be understood since the positions simply are reversed. A side elevation of this paperblank appears as in Fig. 12, and a plan ,view appears as in Fig. 13. It is evident that there has been no longitudinal stretching of the paper nor any stretching of the neutral axis circumferentially, for there is no tearing nor any wrinkling of the material. Measurements may be taken directly across the top and up to the bottom of the blank for any sections, as illustrative measurestruct Fig. 14 without any necessity of further measurement, and thus to shape passes appropriately for forming tubing according to this invention.

Referring more in detail to the drawings, cross sections of a blank as it is shaped from a fiat strip S to a completed cylinder, are illustrated by Fig. 1 corresponding to a zero pass or pinch rolls, and by Figs. 2 to 11 corresponding to subsequent shaping passes. In these figures, the position of the passes is indicated with reference to the center line across the completed cylinder, showing the passes as being successively lower. I

Fig. 12 shows the blank in side elevation as it is shaped from the fiat strip -S at zero pass to form a completed cylinder. The lower line A represents the outline of the middle of the blank, which progressively is moved down from the initial plane of the strip until it forms the bottom of the cylinder. The upper line BB shows the outline of an edge of the blank as it is rolled over to the top of the cylinder. It found that the line A may be expressed mathematically as an exponential curve of parabolic type having the general formula:

Plotting specific measurements, from a model as taken tangentially along the bottom of the cylinder,'and the Y -axis is taken at the plane in which the cylinder is just completed. K is the usual constant of an exponential equation, and, as usual, its value is fixed during consideration of a given condition. With units of measurement expressed conveniently in terms of the diameter of the cylinder, and plotting on a logarithmic scale, the curve thus obtained is a straight line, as Equation (1) requires. From its slope the value of the exponent is obtained as 1.895: and from its y intercept with the total :r-distance taken as 50 diameters the value of the constant as 0.00069. Thus at any pass the height of the bottom of the blank being formed over the longitudinal distance illustrated is expressed as:

As shown, Fig. 12 contains numerical values for the height above the bottom of the cylinder of the middle of twelve passes when the passes are spaced equally at four diameters apart over a total distance of fifty diameters, with a cylinder diameter of twelve inches.

It appears from the nature of the curves that, independently of the total length of the passes or diameter of the cylinder, at the pinch rolls or zero pass, the plane of the strip is 1.16 diameters above the bottom of the cylinder, or in other words the seam of the cylinder is completed at 0.16 diameters below the initial plane of the strip.

The shaping of the blank may be conducted over any longitudinal distance, but to avoid undue stresses it should be a distance of at least 25 diameters. Any changes in this longitudinal distance are expressed readily by relating thechange to the unit-distance assumed in a standard graph. For calculations from Equation (1), increasing the total length of the passes from 50 diameters to diameters would alter the constant K1 of the mathematical expression, so that a new constant K2 would bear the relation:

El X.895 1.895 2 1( K, 100

Any increase or decrease in the diameter of the cylinder causes a simple proportional change in any given value of 3/. Thus if in a given longitudinal distance, a tube is formed of 24 inches diameter instead of 12 inches, the bottom of each pass is twice its former distance above the bottom of the finishedcylinder, the top or seam of the cylinder being located 0.16 of the new diameter below the initial plane of the strip. Referring to the plan view of a half width 0 strip, as shown in Fig. 13, the path of one edge is shown as that of a sine curve BB that is symmetrical about an X-axis taken lengthwise through the middle of the half-strip and also about a Y-axis taken halfway between the end of the flat strip plane and the completed cylinder. Thus the edge is brought smoothly to the seam and parallel to its original position. The width from the edge at any pass to the line of the seam is obtainable from the equation of its curve:

(4) y=K sin 0,

and extreme side positions by smooth circular curves, similarly to the showing of Fig. 14. Fig. 14 is an end elevation outlining the envelope of the curves intermediate A and BB1, or the wall of a half-width of a blank being shaped with numerical values indicated for passes corresponding to those of Figs. 12 and 13.

Thus in constructing Fig. 14, the bottom positions of the several passes, known from the 11 values of curve A, Fig. 12, are indicated on a vertical center line of the cylinder. Further, the extreme widths of the strip edge at the respective passes are known from curve BB1, Fig. 13; Further, it is known that the arcs defining the several passes all are equal to the original width of the strip, since the neutral axis across the strip is not stretched circumferentially. With the data thus given it is possible to calculate the radii of the several arcs. The first five passes in this illustration are simple semicircles; the higher numbered passes beyond the curve C are multiple arcs which for practical purposes are taken as those of two circles with the segment of smaller radius forming the middle of the pass.

For calculation ot-the radii of passes i to of Fig. 14, the length of the are, known as being equal to the width of the plate, may be designated as s; and the chord of that are, known as the dis-- tance between the edges at a given pass, may be designated as c. By calculation or from engineering tables this affords the value oi the subtended angle in radians, which may be designated as 0. Then the radius, r, readily is calculated from the appropriate equation r=s/0. At pass 2, for instance, r is 30.400 inches.

For establishing the passes beyond the curve C where the contour is a multiple arc, calculation of the radii is similar, but modified slightly in calculating the upper-segment. Referring for example to pass 8, the middle of the pass is established as a semi-circle of radius 7.560 inches, the radius being known from curve C of Fig. 13 as the distance from the center to the side of the blank at pass 8.- The length of the upper segment remaining to be established is known since the periphery of all passes is of constant length. Further, the height between this upper segment and a chord subtending it is known as the diiference between the widths to curves B and C of Fig. 13, which difference at pass 8 is 2.140 inches. Again from calculation or from engineering tables, the ratio thus known between an arc and the height to its chord afiords knowledge of the subtended angle in radians. And thus 1', the radius desired,

is calculated as before by the formula r=$/0. At pass I this radius of the upper arc is 11.000 inches, and its center is established on the line of the diameter of the lower arcof the same pass. Other values indicated on Fig. 14 are obtained similarly. The points of junction of these two arcs outlining the position of the side wall intdoubling to form the upper side of a blank, enable one to indicate curve C on Fig. 12 by way of supplemental information.

Thus the outline and the position of each pass is readily determinable in conformity with this invention. The result is that a metal blank that is shaped in conformity to these passes is not stretched longitudinally, nor stretched axially in any direction, but its edges are'brought parallel to each other without buckling, and the side walls are smooth and free from strain.

In operation, a flat strip of metal 'is moved into the series of shaping passes and while moving therethrough gradually is formed to cylindrical shape without stretching. A tube forming mill of this design is adaptable to substantially any range of wall thickness or of tube diameter. For example, heavy wall strips six feet or more in width or thin narrow blanks are equally well formed into tubes of accurate contour and of high strength.

It is within the scope of this invention to deviate somewhat from the exact values here given by way of illustration. Thus the values of exponents in the equations given may not be attained exactly and in practice may vary slightly with changes of material being formed, the values of the constants similarly may vary with practical requirements, and the methods of mathematical expression or of calculation may change likewise. Moreover imperfections of actual practice may entail considerable deviation from the curves .here described. However, according to the provisions of the patent statutes, I have explained the principle and mode of operation of my invention, and have illustrated and described what I now consider to represent its best embodiment, and I desire to have it understood that, within the scope of the appended claims, the invention may be practiced otherwise than as specifically illustrated and described,

I claim:

1. A process of forming metal tubing comprising bending a fiat strip of metal to a continuous cylinder having parallel seam edges by bringing the edges of the strip together progressively over a longitudinal distance of at least twenty-five times the diameter of the cylinder, while gradually bending the "bottom of the strip longitudinally away from its initial plane along an exponential parabolic curve having the form =Kz, where y is the vertical distance of the bottom of the fiat strip above the bottom of the completed cylinder, a has a value greater than 1, and a: is the horizontal distance in the line of the cylinder from a plane transverse to the cylinder at its point of completion, and thereby preventing substantial stretching of the metal axially.

2. A process of forming metal tubing comprising bending a flat strip of metal to a continuous cylinder having parallel seam edges by bending the strip transversely to bring the seam edges together progressively over a longitudinal distance at least twenty-five times the cylinder diameter and in curves maintaining the perimeter of the forming tube substantially equal to the original width of the strip, while gradually bending the bottom of the strip away from its initial plane along an exponential parabolic curve having the form y=K:c, where y is the vertical distance of the bottom of the flat strip above the bottom of the completed cylinder, a has a value greater than 1, and a: is the horizontal distance in the line of the cylinder from a plane transverse to the cylinder at its point of completion, and thereby preventing substantial stretching of the metal along its neutral longitudinal and circumferential axes.

3. A process of forming metal tubing comprising bending a flat strip of metal longitudinally and progressively below the initial plane of the fiat strip along an exponential parabolic curve having the form y=K:z:, where y is the vertical distance between the bottom of the flat strip and the bottom of the completed cylinder, a has a value greater than 1, and a: is the horizontal distance in the line of the cylinder from a plane transverse to the cylinder at its point of completion, while bending the strip, transversely and progressively to bring the edges together along 180 of a sine curve between planes transverse to the completed seam and the end of the flat strip, and in curves maintaining the periphery at all points between the end of the flat strip and the completed cylinder substantially equal to the width of the initial strip. and thereby continuously forming the cylinder while preventing substantial stretching of the metal with respect to its neutral longitudinal and circumferential axes.

, 4. A process of forming metal tubing comprising bending a fiat strip of metal longitudinally and progressively below the initial plane of the flat strip along an exponential parabolic curve having the form y=K:c, where y is the vertical a line connecting said mid-points forms an ex:

distance between the bottom of the flat strip and the bottom of the completed cylinder, a has a value greater than 1 and :r is the horizontal distance in the line of the cylinder from a plane transverse to the cylinder at its point of completion, while bending the strip transversely and progressively to bring the edges together along 180 of a sine curve between planes transverse to the completed seam and the end of the flat strip while maintaining the periphery at all points between the end of the flat strip and the completed cylinder substantially equal to the width of the initial strip, and bending the side walls of the forming cylinder along an exponential parabolic curve of higher order than that of the bottom of the strip between the plane transverse to the completed seam and the point where the edge of the strip passes beyond the forming side wall'in' moving along said sine curve, and

thereby continuously forming the cylinder while preventing substantial 'stretching of the metal with respect to its neutral longitudinal and circumferential axes.

5. In a tube-forming mill for progressively bending a flat strip of metal to a continuous cylinder having parallel seam edges, 9. series of bending passes having individually differing curved working faces, the sides of consecutive passes progressively approaching a vertical plane through the middle of the passes, the length of the are or working face of each pass being substantially equal to the original width of the flat strip, the mid-point of the bottom of successive passes being progressively removed below the initial plane of the flat strip supplied to the mill so that a line connecting said points forms an exponential parabolic curve having the form y=Ka:, where y is the vertical distance between the bottom of the flat strip and the mid-point of the pass, a has a value greater than 1 and a: is the horizontal distance in the line of the forming cylinder from the seam-completing pass, and the passes being disposed to bring the edges of the strip progressively together to form the com- 1 pleted cylinder over a distance at least equal to twenty-five times the cylinder diameter.

6. In a tube-forming mill for progressively bending a flat strip of metal to a continuous cylinder having parallel seam edges, a series of bending passes having individually differing curved working faces, the length of the arc or working face of each pass being substantially equal to the original width of the flat strip, the mid-point of the bottom of consecutive passes being progressively removed below the initial plane of the flat strip passed to the mill so that ponential parabolic curve of the form y=K:r where u is the vertical distance betwezn the bottom of the flat strip and the bottom mid-point of the pass, a has a value greater than 1 and a: is the horizontal distance in the line of the forming cylinder from the seam-completing pass, the sides of consecutive passes progressively approaching the middle plane of the passes with theirupper edgespositioned along 180 of a sine curve lying between the end of the flat strip and the completed seam, and thepasses being disposed tobring the edges of the strip progressively together to form the completed cylinder over a distance at least equal to twenty-five timzs the cylinder diameter.

REIMAR C. F. KURTZE.