US1826060A - Method of making helical springs - Google Patents

Description

Oct. 6, 1931. EATQN 1,826,060

METHOD OF MAKING HELICAL SPRINGS Filed May 19. 1927 WITNESS INVENTOR M j-w $101 17. 305-1 Patented Oct. 6, 1931 UNITED STATES GEORGE M. EATON, F IPITTSBUZRlGH, PENNSYLVANIA METHOD OF MAKING- HELICAL SPRINGS Application filed May 19,

This invention relates to improvements in the method of forming helical or coil springs and the articles produced thereby. Helical springs commonly employed as resilient supports loaded under compression or in tension are subject to failure resulting from initial causes other than true fatigue, which render them very unsatisfactory in service.

The present invention is based on my discovery that failure of helical springs made according to standard practice is of a characteristic form. The invention further deals with a method of making helical springs of a single rod in a manner to eliminate some of 5 the causes of the type of failure to which helical springs manufactured in accordance with methods heretofore practiced are subject. I

I have found thatfrom the focus of fail- 2o ure there is propagated a fatigue crack which in commercially sound steel is always located on the inside of the spring coil. The failure follows very closely the well known diagonal shear failure characteristic and if the steel is commercially sound the slope of the fatigue crack is always in the direction of the angle of the helix or coil (in a spring loaded in compression). The direction of forces of the couple producing torsion in the coil section makes it clear that the tendency is to tear this crack open. The fatigue crack after it is initiated progresses for a short distance in a shape approximating a semicircle before complete rupture occurs.

When a coil spring is under load, its section is subjected to direct shear applied as a uniform unit stress over any diameter perpendicular to the axis of the coil since the p section supports the entire applied load. The section is also subjected to torsional shear and as will be hereinafter demonstrated, the resultant stress at the outer fibre of the section is the difierence between thedirect and torsional shear, while at the inner fibre of the section the resultant stress is the sum of the direct and torsional shear. There is a further concentration of stress on the inner fibres due to-the fact that the inner fibre is shorter than the outer fibre and is therefore deflected through a greater angle, when the 1927. Serial No. 192,597.

spring is deflected. From a consideration of the application of forces on a spring section it is therefore obvious that the unit of stress is higher underservice conditions, in the region where the failure localizes, namely, on the inside of the coil, than in the regions which are immune from failure.

I have discovered by extensive observation that the fatigued portion of the great majority of helical springs which have failed in service has a slight step or off-set at the exact focus from'which the crack progresses. This off-set is substantially in parallelism with the axis of the rod of which the spring is made. From this it may be established as a positive fact that the initiation of failure is usually propagated from streaks in the steel. The term streak as employed throughout the specification and claims is intended to define any non-homogeneity of the structure lyingin the rod as the rod comes from the rolls.

uch streaks occur in the steel cast in ingot form as non-metallic foreign inclusions, oxides, segregations of metalloids, voids, etc., and during the process of rolling or drawing, these imperfections are elongated and attenuated. In the finished rod they appear as streaks substantially parallel with the axis of the rod. When the rod is coiled on a mandrel to form a helical spring, the axis of the rod remains at practically the same length as before coiling. The metal on the outside of the coil is lengthened axially while that on the inside of the coil is shortened axially. The axial shortening of. the metal can occur only through plastic flow, under the influence of viscous fluid compression and follows the ath of least resistance which on the surface of the inner diameter of the coil is in a direction substantially parallel with the axisof 90 the coil or spring. I g

I have found that a streak, when located at the smallest diameter of the coil, will exhibit an appreciable tendency to open in the coiling operation and as the location of the streak 95 is swung away from the inside diameter of the coil, the amount by which the streak opens decreases. The opening of the streak or even the tendency to open is a serious menace in heat treated-steel springs and is greatly 1 0 trationsat the radial, axial and tangential extremities of the streak.

If rupture does not occur during heat treatment, progressive failure and ultimate rupture are inevitable when streaked steel is subjected to service stress near the limit which sound steel will resist successfully.

This inherent failure characteristic of steel helical springs when subjected to heat treat- I ment, is especially pronounced in alloy steel and accounts for the fact that to date helical springs of alloy steel are quite generally inferior, to carbon steel springs. This probably is due to the fact that the difference between the characteristics of the streak and of the adjacent metal is more pronounced in alloy steel and consequently there exists more severe stress concentration.

It is among the objects of the present invention to provide a method of forming coil springs which shall relieve the material in the critical range of distress and which shall eliminate, or postpone the type of failures to which reference has been made.

The invention further contemplates the making of coil or helical springs of carbon or alloy steel which may be subjected to heat treatment without developing defects inherent in helical springs as formed by methods heretofore practiced. Another object of the invention is to provide a method of forming helical springs by subjecting the rod of which the spring is made to mechanical treatment whereby the physical characteristics of the steel are greatly improved and the spring section is rendered more capable of resisting failure in the region of greatest stress.

Another object of my invention is to pro vide a method of distorting the grain structure of the material by which the grains are elongated and orientated in a direction which will increase the resistance of the spring to the critical conditions which ordinarily produce premature failure.

' These and other objects will become more apparent from adescription of the drawings in which like reference characters designate like parts and in which Fig. 1 is a diagrammatic View of a coil spring illustrating the princi les of my invention; Fig. 2 a sectional elevation of a fracture taken along the line IIII, Fig. 1; Fig. 3 a cross sectional view of the spring rod taken on line III-III, Fig. 1; Fig. 4 an enlarged elevation of a portion of a spring illustrating a fracture; and Fig. 5

a an elevational View of a portion of a twisted bar utilized in the making of helical springs in accordance with the objects herein stated. The spring shown in Fig. 1 is formed of a roundbar or rod 10 coiled on a mandrel in the usual manner in theshape of a right hand helix. L designates the applied load. Ar-

rows P -P indicate the direction of forces of this crack open. The crack A after it is initiated, progresses fora short distance, in a shape approximately a semicircle, as shown 111 Flg. 2 before complete rupture occurs.

Theunit stress in the region of crack A.

where failures localize, namely, at the smallest diameterof the coil, is higher under service conditions than it isinthe regions which are immune from failure. This is clear from a consideration of the distribution of stresses on a section of rod 10 when the spring is under load as illustrated in Fig. 3.

Fig. 3 represents any section through rod 10 of the spring. lVhen the spring is under load the section is subjected to direct shear applied as a uniform stress over any diamcter of the rod perpendicular to the axis of the spring since the section supports the entire applied load. The unit stress Sd equals the load divided by the cross sectional area of the rod; I

where L is the applied load and D is the diameter of rod 10. 7

But this section is-also subjected to torsional shear St which at the surface of the wire is equal to the twisting moment divided by the resistance of the rod to twisting;

The above formula for St applies to the specific case illustrated in the drawings,namely where the radius of the helix is 1.7 times the diameter of the rod 10, Fig. 1. This case is quite closely representative so that the formula indicates roughly the relation of stresses on the inner and outer fibres.

It will be noted by reference to Fig. 1 that the resultant stress at the. outer fibre of the coil is the difference between the direct and torsional'shears, while at the inner fibre of the coil the resultant stress is the sum of the direct and torsional shears. As previously stated, there is still further concentration of stress at the inner fibre of the coil. This concentration of stress on the inner fibres of the coil accounts for the initiation of failures in this region. I a

As previously noted the fracture at the focus of failure is offset, this being illustrated in the enlarged view of Fig. 4 where 0 represents the streak in the bar lO which is disposed in parallelism with the axis of the rod, and the divergent lines designate the crack emanating from the streak which in- Sd D elusive of the streak O is the line of fracture along which failure occurs. In accord ance with the present invention the streak O in the spring rodis diverted or distorted from its original plane in which it was formed by rolling or drawing as previously stated. This is accomplished by twisting the rod 10 throu h an angle I in the manner shown in Fig. 5 efore coiling it in the form of a spring. The twisting of rod distorts the fibrous structure of the steel out of parallelism with the axisof the rod. The amount of twisting or the angle of twistI is dependent upon the dimensions of the rod 10 and the inner diameter of the coil, but should be of such a degree that the streak O of Fig. 4 is disposed substantially at right angles to the crack A of Fig. 1.

Referring to Fig. 1, A designates the line of fracture and the line H designates the direction of twist on the inner diameter of the coil, H and A being approximately at right angles to each other in which position there is the least likelihood of tearing apart the streak by the forces P -P as previously explained. As to the exact angle of twist,

- I have found that a twist of 360 degrees in a length equal to ten times the diameter of the rod. 10 was sufficient to prevent the streak from opening. I do not, however, wish to limit myself to this pitch for the helix of twist since more or'less twist may be necessary or suitable for the prevention of failures emanating from streaks.

The twist of the rod when coiled flattens appreciably on the outer diameter of the coil as illustrated by the line K of Fig. 1 and becomes materially steeper on the inner diameter of the rod as shown by the line 4 of Fig. 1. The angle of'twist may be determined by test, since the angle of fatigue in any given spring is known to be 45 degrees from the axis of the 'coiled rod (in sound material),

and from this angle the degree of angularitv and direction of helix of twist, which will place the streaks in the steel at approximately right angles to the direction of the fatigue crack, 1 can be established. The angle of twist, when established by test, will be more than the minimum angle which will definitely stabilize the streak against the tendency to open during the act of coiling. The direction of the torsion of load will then be the same as the direction of twist.

'The analytic determination of the angle of twist is dependent upon the plastic flow characteristic of the material, and is so involved that I prefer to determine this angle for any range of spring proportions and chemical analysis by the accumulation and tabulation of test data.

The twisting of the rod in addition to securing a structure in which the streaks are oriented in a direction where they have the least possible weakening effect, also secures strength of the rod in the finished spring.

During the twisting operation, which is preferably carried out at forging temperature, the grains will be elongated, their major axis in general following the helix of.

twist. This is the most favorable direction for the major axis of the grains, since there j is less tendency under severe servlce stressfor intergranular separation to start the initiation of a fatigue crack. 1

The reason for this is obvious, being in general analogous to the reasons for the directional disposition of streaks. During all high temperature cycles, subsequent to the twisting operation, there will be a tendency for grain growth to change the shape of the grains from the elongated to the equi-axed form butthere are a number of conditions combating this tendency toward grain reconstruction. If the surface of the wire is decarbonized the critical temperature of the decarbonized structure will be raised and also the temperature range of most rapid grain growth will be raised. Thus the direct effect of heat'treatment at a temperature adapted to the normal carbon content of the steel will tend to produce a comparatively inferior structure in the decarbonized material, but at the same time there will be less removal of the directional tendency of the grains, due to the comparatively slower grain growth.

The twisting of the rod aligns not only partlcularly at or near the surface. When due to decarbonization the surface during heating for treatment does not reach the critical temperature of the alpha-gamma change,

the desirable alignment of these crystal aggregates will persist in much the same condition as before heat treatment, and will benefit the performance of the springby resisting, in greater measure than would occur without this favorable alignment, the tendency of service stresses to start intergranular or -intercrystalline separation, with the later development of a fatigue crack.

In regions which are critical because of unfortunately located streaks, the grain distortion occurring during the twisting operation will have afurther reason for material persistance throughout subsequent high temperature. ranges on account of the obstruction of grain growth, due to the presence of the impurities resident in the streak.

The twisting of the rod may be effected during the coiling of the spring or may be handled as an entirely independent operation. If the operation is independent, the rod would preferably be twisted (in a commercial process) in mill lengths by special machinery which could be designed by any one familiar with the building of steelmill and spring coiling facilities. I prefer this independent operation to synchronous twisting and coiling procedure for the following reasons, although the resulting spring will be somewhat more expensive. Where the ends of the rod are forged down to a taper, gradually changing from a round to a rectangular section, it is very important that the twist of the rod be carried uniformly to a point well past that where the taper begins, because as this is quite a critical range, it is important to orient the streaks and the grain structure properly clear through this critical region.

It is entirely feasible, however, for any designer skilled in the art to produce facilities for coiling and twisting the rod synchronously but when this is done the tapered ends must be forged prior to the twisting and coiling operation and once the tapered ends are completed no subsequent twisting can be carried out in the-tapered portion excepting the small amount of twist required to overcome the tendency of the rod to rotate slowly as it wraps onto the coiling mandrel. (In this connection it is to be particularly noted that this coiling rotation of the rod for compression springs is always in a direction which, if resisted, produces a twist in the rod in a direction opposite to the twist which I produce in practicing my invention).

Thus, it is evident that in the case of synchronous twisting and coiling the twist must die out at the critical range in which the taper terminates, with material chances for undesirable stress concentrations unless the most exacting care is exercised during the operation. Therefore, this synchronous operation, while producing a spring which is materially better than can be produced by operations now in common use, will be less certain than independent twisting and coiling operations to secure the full improvement that is possible under'my inventioni It is of course theoretically possible to forge the tapered end as a reverse helix so that it can be twisted while coiling. But this is not a very practical process, as a large number of expensive dies would be required and it would be very difficult to avoid concentrat: ing excessive twist at the thin end of the taper.

I am aware of the fact that previous oonsideration has been given to the making of helical springs from a plurality of twisted rods in stranded form. Such structures, however, are utterly impracticable for springs which are heat treated after they are coiled, which is an essential and common practice, for the following reasons :If the rods are not welded together, the interstices between them and the alternate surface exposure of the individual rods and the burying of the rods in the body of the stranded structure or cable will inevitably result in widely varying exposure of the different parts of the rod to the quenching medium causing varying rates of cooling through the critical temperature range, with resultant variations of the physical characteristics over the different parts of the rod. A spring of this nature would be obviously unreliable and would also possess material internal friction or hysteresis. If on the other hand, an attempt were made to weld the rods into a homogeneous mass, the very purpose of the structure would be defeated because there is no known method for securing a perfect weld in such a complicated structure and a spring made from a plurality of twisted stranded rods welded after twisting would be far less reliable than untwisted helical springs made of a single rod. Furthermore, in any compound stranded cable structure, either the streaks between rods or the streaks in the rods themselves will of necessity lie outside of the direction at right angles to the direction in which fatigue cracks tend to form, and it is therefore impossible to attain with a stranded and twisted cable the desirable disposition of the streaks in the rod which I obtain in the use of a single twisted rod spring.

It is evident from the foregoing description of my invention that the method of forming helical or coil springs of a single twisted rod with the streaks in the rod disposed substantially at right angles to the plane in which fatigue cracks tend to form provides reliable means of manufacturing heat treated springs for severe service requirements and permits of successfully utilizing alloy steel 1n the manufacture of springs to obtain the advantages characteristic of these metals.

Iclaim:

1. The method of forming helical springs, which comprises, twisting a steel rod to a predetermined angle about its own axis and coiling said twisted rod.

2. The method of forming helical springs which comprises twisting a steel rod to an angle at which the streaks inherent in the steel are substantially at right angles to the direction in which torsion is imposed upon the rod, and coiling said twisted rod to the shape of a helix.

3. The method of forming helical springs which comprises twisting a steel rod so that the streaks inherent in the steel are disposed at a predetermined angle, and coiling said twisted rod to the shape of a helix.

4. The method of forming helical springs which comprises twisting a steel rod to definitely distort its fibrous structure out of parallelism with the axis of said rod, and coiling said twisted rod to the shape of a helix.

n testimony whereof, I have hereunto set In hand.

y GEORGE M. EATON.

Images