November 17, 1933
INGERSOLL STEEL & DISC CO.
BUDD WHEEL CO.
Appeal from the District Court of the United States for the Northern District of Illinois, Eastern Division; George E. Q. Johnson, Judge.
Before Alschuler, Evans, and Sparks, Circuit Judges.
This is a patent infringement suit which was instituted by appellee. The bill charges appellant with infringement of United States Letters Patent No. 1,442,601, issued to Budd and Ibach, January 16, 1923. Appellant pleaded invalidity and non-infringement, and the court found the patent valid and infringed.
SPARKS, Circuit Judge.
The patent relates to the metal working art, and covers a process for making tapered disks, a machine therefor, and a process for placing a flange thereon. The claims in suit are 2 and 6 which are machine claims,*fn1 and 14, 17, 19, 20, and 25 which relate to process.*fn2
From patentees' application we gather the followings facts: One of the chief objects of the invention was to provide a machine by which tapered steel disks might be formed from a square blank of uniform thickness, with a minimum amount of waste and at a low cost. In general the invention consists in rolling rectangular metal blanks of substantial uniform thickness into disks such as are used in the manufacture of disk wheels, tapered outwardly from the center and having a substantially circular periphery, by rolling the metal blank along a plurality of different radii from the center outwardly so as to force the excess metal radially of the blank toward the periphery. This tapers the blank and also causes the periphery to approach a circle. The invention also comprises a machine for carrying out the method, including a pair of operating rolls which act upon the metal blank. One of those rolls has a substantially triangular working surface of cam formation, so that points on that working surface are at progressively greater distances from the axis of the roll.
The drawings accompanying the application illustrate a machine comprising upright frame members in which two rolls, one above the other, are journaled, said rolls being driven in opposite directions from a suitable source of power. A table extends forward from the frame, and is provided with ways on which the carriage is slidably mounted. An arm is pivoted at its forward end between the lugs on the carriage by means of a bolt. A pin projects downward from the rear end of the arm and is adapted to extend through a central opening formed in the metal blank. This arm and pin serve as a means for centering the metal blank with reference to the carriage and rolls. A cross bar extends between the uprights at the forward part of the machine, and a stop screw is threaded through the cross bar so that its forward end extends into a position to be engaged by the carriage and serves as an adjustable stop to limit the movement of the carriage toward the rolls.
The following figure correctly represents the rolls in patentees' machine:
A further description of their machine is as follows: A substantially triangular raised portion is formed on roll 4 and the height of the triangle is proportionate to the length of the square or rectangular metal blank upon which the work is to be performed. The narrow shoulder 19, to which we shall refer as the apex, is disposed midway between the ends of the roll, and the base of the triangle extends from one edge of the roll to the other. The surface of portion 18 is at all points equidistant from the axis of roll 4. A substantially triangular portion is formed on roll 3, and so positioned as to co-operate with 18 on roll 4. This portion extends from shoulder 21, which co-operates with shoulder 19 on roll 4. A notch 22 is formed centrally in 21 so as to accommodate the centering pin. At the apex of this triangular portion there is portion 23, all parts of which are equally distant from the axis of the roll 3. This portion is defined by the shoulder 21, the edges of the raised portion, the arc of a circle concentric with notch 22, and covering substantially ninety degrees, and lines drawn from the ends of said arc rearwardly to the edges of the raised portion, thus forming the horns 25 which are in the same cylindrical sufrace with portion 23. From the line 26 forming the inner boundary of portion 23, the cam surface 18' extends rearwardly to the base of the triangle, increasing in its distance from the axis of roll 3 continuously from line 26 to the base of the triangle. At the edges of the cam surface 18 there are formed the beveled surfaces 27 which are widest adjacent to shoulder 24, and grow narrower as they approach the parts 25 and the line 26.
Patentees describe the working of the machine substantially as follows: The square metal blank is heated and mounted on the carriage and is centered by means of the arm and pin hereinbefore referred to. While the rolls are in such position that the working surfaces are on opposite sides of the rolls from each other the carriage is moved inwardly until it engages the threaded stop screw, thus carrying one-half of the metal blank between the rolls. As they revolve in opposite directions the surfaces 18 and 18' move toward each other until the shoulders 19 and 21 are in line with each other and the notch 22 receives the centering pin. At this position of the rolls the central part of the blank is gripped firmly between the surface 18 of roll 4 and the part 23 of roll 3. The distance between these parts is such that there is no compression of the blank at this point. By this means the blank is firmly held at the beginning of the operation. As the rolls continue to revolve, the central part of the blank is given a continuous support by the corresponding surface 18 on roll 4. As the distance of the surface 18*fn1 from the axis of roll 3 increases continuously from line 26 to the base of the triangle, the pressure of the rolls increases continuously and the distance between the working surfaces on the rolls decreases continuously with the result that the metal of the blank is pressed and squeezed out giving the metal the tapered cross section, and elongating it somewhat. The fact that the parts 25 are in the same cylindrical surface as 23 prevents buckling or curling of the sheet during the rolling operation, as they tend to grip the blank on both sides of the point where the compression is taking place. The beveled edges 27 are provided in order to reduce the compression adjacent to the edges of the surface acted upon so as to eliminate the pronounced ridges which would otherwise be caused at those points, and prevent any material disturbance to the adjoining segments.
The machine described in patentees' application contemplates the rolling of the metal blank into an approximate circle with four passes through the rolls, that is to say the base of the triangular cam is the same length as one side of the metal blank, and the arc caused by each pass is ninety degrees. After the first pass through the rolls, the second pass is made 180 degrees from the first, or on the side opposite the first, and the third and fourth passes are made laterally to the first and second. Patentees state that the blank may then be trimmed and finished without further rolling, but that they find it desirable in many cases to give it four more passes along the radial lines in order to eliminate the ridges and cause the periphery of the blank more nearly to approach a circle. After the rolling operation is completed, the blank may be trimmed to bring its periphery to a true circle. The flange on the periphery is then pressed up from the body of the blank and a central opening punched in it thus forming the completed wheel disk. The flange and central opening, however, are not formed by the patented machine, but they are included in the process covered by the patent.
It is further stated by patentees in their application and specifications that, by the use of the machine and method, a wheel disk of tapered cross section may be produced very efficiently and at low cost; that the disk is formed by a simple rolling operation which can be quickly performed, and that there is practically no waste, as the square blank is little metal to be trimmed off.
Appellant's machine and process which are alleged to infringe, roll disks radially from the center and tapered on both sides, from equilateral octagonal blanks. Its method of cutting these blanks from strip stock involves an initial waste of material in excess of twenty per cent which is much in excess of the loss involved in appellee's method. Appellant's machine is provided with automatic indexing and feeding means which move the blank on 45 degrees after each pass, and eight passes are made in regular succession. These, however, are not closely confined to the octagonal sections. Two rounds of rolling are necessary in appellant's process. On the second round the indexing table is automatically set ahead 22 1/2 degrees, after which eight more rolling passes, again indexed at 45 degrees, are made in regular succession midway between the passes of the first round. The cam die which is used on the rolls is kidney shaped and has a double curvature contour. The working surface is ground to a double curvature, the central portion being cylindrically ground, and the end portions being ground tangent to the cylindrical portions, and relieved slightly toward the ends of the die. The working surfaces of these dies seem to us roughly to approximate a triangle, but not exactly comparable in size to the octagonal sector. The ends of the dies receding in height from the working surfaces constitute bevels on the dies of the machine, but they are not the same shape nor nearly so marked as patentees' bevels. Appellee's exhibit of appellant's device as applied to the octagon discloses that the beveled portions of appellant's device extend into each adjoining sector a distance of about one-third of the periphery of the sector. Indeed this exhibit discloses that a substantial portion of the working surface overlaps the left radial line of the sector, and by the same distance the working surface fails to meet the right radial line of the sector. This would seem to indicate that appellant's process, as applied to an equilateral octagon, does not require the exactness of application of the working surface to the octagonal sector, nor the refraining from disturbance to the adjoining sectors as contemplated by the patent when applied to a square or rectangle.
The first question presented is whether appellee's machine patent is valid. That question is raised by the record but appellant does not now deny its validity provided the metal blank used is confined to a square blank, which appellant claims is the only structure disclosed by the patent. Appellee starts with the presumption of validity in its favor which can only be rebutted by clear proof to the contrary. A patentee is entitled to a fair range of equivalents without specifically reserving that right. If he has substantially advanced the art he is proportionately entitled to a liberal construction of his patent in that respect, and the range of equivalents depends upon and varies with the degree of advance. He is required to set forth in his patent his novel principle and what he conceives to be the best mode in which he has contemplated applying that principle. However, he is not confined to that mode, and he need not point out all the advantages of his invention, but his claims may not be broader than his described invention.
With these fundamental principles in mind we shall examine the state of the art at the time patentees' application was filed, with a view to determining whether in their structure they have disclosed a blank other than a square from which the construction of the structure was contemplated by them, and if not, what is the range of equivalents to which they are entitled.
Rolled, tapered, and flanged disks for wheels were admittedly old in the art prior to the invention of the patent in suit.*fn3 Gap rolling mills with dies having an eccentric cam contour in which the distance from the axis of the roll increased continuously over all or part of the die were old in the art.*fn4 Eccentric gap roll dies of a more or less generally triangular shape were also old, but they were not of patentees' specific form.*fn5 The four pass operation of the eccentric dies of a gap rolling mill upon a rectangular blank was certainly taught in Millard, No. 80,083 (1868). Clement, No. 93,679 (1869), provided a gap roll mill adapted for taper rolling of a substantially rectangular blank in at least three radial directions. Rolling mill dies with working faces formed to the outline and cross section of a portion or the whole of the work to be finished were also old in the art as disclosed by the patents hereinbefore referred to, and also by McCune, No. 188,394 (1887), and Lathrop, No. 42,005 (1864). The prior art also discloses rolling mill dies with relieving bevels along the sides of their main rolling surfaces.*fn6 It also discloses dies adapted to the rolling of individual blanks and to the radial rolling of various sections of multiple blanks into symmetrical forms.*fn7
The gap roll mill of the Jepson patent more nearly approaches that of the patent in suit than any other of the patents disclosed by the record, and it was the only one cited as an interference to patentees' machine and process at the time of patentees' application. It was a machine for making tapered disk saws, and it rolled circular blanks of uniform thickness. The rolling was radial and from the center, and the working surfaces were rectangular in outline, but they were not designed to fit any definite sector of the blank. The blanks, being circular, wasted about twenty-two per cent. of the material at the source.
From a consideration of the prior art hereinbefore discussed we are convinced that the field in which patentees were laboring at the time of their discovery was a decidedly narrow one, and that both their machine and process must be held to contemplate the use of a square or an approximation thereof. This view we think is amply supported by the motivating object of the discovery, which was to construct a tapered disk with a minimum of material and labor. It is admitted by appellant that a square tapered by radial rolling with a ninety degree triangular die accomplishes both objects more perfectly than does the use of a blank in any other form. It is quite true that the disclosure specifies the use of any rectangular blank, but we cannot conceive of any reason why a rectangular blank other than a square would ever be used in the process, for such use would only increase labor and material, and that without reason. A rectangle other than a square, we assume, can never be rolled by any process into a circle whose diameter is greater than that of the circle rolled from a square whose sides are equal to the ends of the elongated rectangle. Furthermore, no rectangle, other than a square, can be divided into four triangles whose angles are equal at the apices, by drawing lines from the four corners of the rectangle to its center. The apical angles of the two longer triangles become less than ninety degrees, and those of the shorter triangles become greater than ninety degrees in proportion to the lengthening of the rectangle. It seems quite obvious therefore that according to patentees' process, one sized die could never roll the four triangles of a rectangle other than a square, nor could any number of triangular dies of different apical angles roll an elongated rectangle into a circle whose side is equal to the width of the elongated rectangle. Certainly the patent in suit discloses no method by which such a thing can be done.
It is not claimed, however, that appellant's machine and process infringe the patent by the use of a rectangular blank, or its equivalent, other than a square. What we have said in this respect is merely in support of our conclusion that patentees never contemplated the use of any blank other than a square or its equivalent. In view of the fact that appellant admits the validity of machine claims 2 and 6, we shall assume their validity without deciding the question.
The question next presented is the validity of the process claims. We are convinced that process claims 14 and 19 are too broad and for that reason are invalid. They do not limit the form of the blank except as to uniform thickness, and each of those claims reads readily upon Jepson.
Claim 20 includes a flange on the periphery of the finished product which is formed by upward pressure applied to the body of the blank after it has been rolled and trimmed to a true circle. Appellee's flange is not accomplished by the machine which is included in the patent, but by a separate die and press which is applied within the circumsference while the periphery is firmly held. This not only forms the flange, but the force when so applied also eliminates the waviness in the blank which is occasioned by the rolling process, and it is contended by appellee that appellant's flange is made in the same manner, and thereby infringes the patent in suit. To press up flanges on circular disks was quite old in the prior art.*fn8 That the flanging operation of Michelin is directly applicable to the flanging of the patent in suit can not well be denied. This was admitted by appellee's expert. That the function of the flanges is the same in Michelin as in the patent in suit, and that they are made in substantially the same manner, are facts admitted by Mr. Budd, one of the patentees of the patent in suit. If therefore appellee's flanging process is to be considered as a valid part of the patented process, it must be either by reason of some new and useful result occasioned by its combination with the other elements of appellee's process in making its tapered wheel disks, or else by reason of an old result accomplished in a more facile or more efficient manner. We are convinced that no such results are thus obtained, and that the flanging process can not be considered as a part of the patent. For this reason, claim 20 is too broad and is therefore invalid. Walker on Patents (6th Ed.) § 220; Robinson on Patents (1890) vol. II, § 537; Taggart v. Bremner (C.C.A.) 295 F. 506.
Claims 17 and 25 we think are valid. They are simple, it is true, but that fact is not sufficient to defeat the claims. The fact remains that for years the making of a tapered disk was a difficult and expensive process, and it remained so until outward radial rolling from the center was applied. It is true that Jepson made use of the latter process in the manufacture of circular saws as early as 1875, but he made them from circular disks with a much greater expenditure of labor and material than patentees. His dies were not of the same shape, nor were they designed to be applied to any specified sector as is contemplated by the patent in suit. There can be no doubt that patentees aggregated many old elements, but even if it be conceded that all their elements were old, yet we think they produced a new and useful result, or at least the same result in a more facile and efficient manner, and in such degree as to achieve invention, and we are convinced that they thereby advanced the art substantially.
We are content with not discussing further the validity of claims 17 and 25, because it is obvious that appellant has not infringed either of them, nor has it infringed the machine claims 2 and 6. When the patent in suit was pending, the only patent cited against it was that of Jepson who used a circular blank. To this citation patentees suggested that Jepson's machine could not be made to operate on patentees' square blank, and they thereby avoided the interference, whereupon the patent was issued to them. It necessarily follows, therefore, that the form of the blanks used is a matter of vital significance, and that patentees and the Department so considered it. In determining patentees' range of equivalents for the square, therefore, we must not lose sight of the fact that Jepson also was entitled to a fair range of equivalents for his circular blank, and although his patent has expired, yet the public, including appellant, is now entitled to all rights in that respect which Jepson had when his patent was in force and effect. Just where these ranges of equivalents meet, if at all, may be somewhat of a twilight zone. The record does not aid us in deciding with exactness where that zone lies, nor do we think that it should, because it seems to us that appellant's blank does not fall within it nor near it. The octagonal blank which appellant uses is equilateral and much more nearly approaches a circle than it does a square. If, after the issuance of the patent in suit, Jepson had used an equilateral octagonal blank, it could hardly be contended that he was outside the range of equivalents to which he was reasonably entitled. What Jepson could then have done can now be done by appellant without liability for infringement.
The decree of the District Court is reversed, and the cause remanded for further proceedings not inconsistent with this opinion.