Opinion
George E. Stebbins, Walter J. Blenko, and William H. Webb (of Stebbins, Blenko & Parmelee), all of Pittsburgh, Pa., and Hugh M. Morris, of Wilmington, Del., for plaintiffs.
Clarence D. Kerr and Charles H. Walker (of Fish, Richardson & Neave), both of New York City, and Richards, Layton & Finger, of Wilmington, Del., for defendant.
NIELDS, District Judge.
This is the usual patent infringement suit. Plaintiffs charge defendant with infringement of Durand patents No. 1,924,028 (claims 1-4, 7-9, 11 and 12), No. 1,918,089 (claim 3) and No. 1,918,090 (claims 1, 4 and 5). Defendant pleads that all the claims in suit are invalid for anticipation and lack of invention and are not infringed.
Plaintiff, Jean Baptiste Durand, is the patentee and owner of the patents in suit. Plaintiff, Societe D'Electro-Chimie D'Electro Metallurgie et des Acieries Electriques D'Ugine, a French corporation, holds an exclusive license under the patents. Birdsboro Steel Foundry & Machine Company is a licensee and agent of the French corporation in granting licenses in the United States. Defendant, Bethlehem Steel Company, is a Delaware corporation.
Patent No. 1,924,028 is for the 'manufacture of Foundry Molds'. The patent is directed to the use of a sand cement composition for casting. As will appear, Durand was not the first to use such compositions for castings.
For many years there have been two principal methods of molding; the green sand and dry sand methods. Green sand molds are used shortly after being made, while the dry sand molds are oven dried or heated to drive off the moisture. The constituents have been and still are sand, binder and water. In these two methods the practice has always been to hold the water content down as low as practical. Foundry Practice (1906), a standard work, states: 'The sand should be mixed evenly and to a dampness such that it will stick together when squeezed in the hand, but not so wet as to show moisture or dampen the hand.'
From early days it was the practice for the molder to determine whether the molding mixture had the proper amount of moisture by the feel of the sand which was ascertained by squeezing a handful in the hand. As the usual mix had clay in it, the 'feel' was due both to the clay and to the moisture. Such mixtures were friable and crumbly. Too much water in the sand clay mixes lowered the permeability of the mold. That was one of the reasons why water had to be held within a definite range. Green and dry sand molds are still used much more extensively than sand-cement molds. Thus Birdsboro, a licensee of the French corporation, does twice as much green sand molding as sand cement.
Patent No. 1,924,028.
Durand forms his molds and cores out of pure silicious or argillaceous foundry sand and a binder of cement or lime. The molding sand is mixed in the dry state with cement. The only example of the mix in the patent is 'five parts of sand to one part of cement', by volume. He states that 'water is added in such proportion that the mixture does not become pasty or plastic, but always stays subhydrated, i.e. the proportion of water is such that a part of the cement, mixed with the sand, does not set'.
The mixture is placed in dismountable wooden boxes for molding in earth or metallic frames and rammed or pressed down slightly. The molds are left to bind or set for several hours. The patentee states that the 5 sand to 1 cement mixture when molded, 'can be handled without special care in about six hours after use, this being the average time of setting'.
Durand makes it clear that he poured while wet. He states that most of the water is removed by chemical combination with the binder. He asserts as the result of his invention that ramming of the molds is unnecessary; the chemical removal of the water by hydration of the binding means in the preparation and setting of the mixture eliminates the stoving operation; the elimination of contraction and deterioration of the sand enables the accurate retention of the shape of the molds; unskilled employes are employed; foundry flasks are eliminated; the same sand-composition may be employed for cores and molds. Durand also states that 'an essential feature of this invention is the use of only so much water as will keep the mixture always in a sub-hydrated state and not permit it to become pasty or plastic, * * *. This calls for an upper limit of hydration considerably below the lower limit thereof prescribed according to the usual practice in making similar compositions, and below the limit commonly prescribed for the purpose of causing the hydraulic binder to set completely.'
In the second paragraph of his specification Durand describes what he says was the prior molding practice. His statement is misleading since it refers only to dry sand molding. He does not mention green sand molding which required no drying before use, or the prior sand cement molding disclosed in the Smith patent. The inference from this paragraph is that Durand uses no outside heat for his molds (yet all his licensees do) and that the sand had to be dried before use in the prior practice whereas usually there was no preliminary drying of the sand.
'Sub-hydrated' is a word coined by Durand or his attorneys for use in his application. The experts had never seen it before reading it in the '028 patent. As a direction to foundrymen as to how much water to put in the sand cement mix it is meaningless. No quantities or proportions are indicated in this patent.
Durand's Patents '089 and '090.
The '089 application was filed April 4, 1930; the '090 on March 10, 1933 as a division of the earlier patent. The two specifications are practically identical. The '089 relates to a process and the '090 to a composition and to foundry molds. Two objects are stated. The main object is 'to utilize in the process of my aforesaid application ('028), in place of the natural sands therein described, artificial sands of any composition which have the property of becoming set when mixed with a hydraulic cement, and have also the refractory property necessary for resisting the cast molten metal'.
The second object is to use as 'artificial' sands material obtained by crushing old molds made by the '028 sand cement process. The proofs show that no one ever used any of the artificial sand mentioned by Durand for mold surfaces which come in contact with the molten metal. The supposed invention in the '090 patent is stated to be for a 'Composition for Foundry Molds'. The composition is the mixture of artificial sands and an hydraulic binder.
Bethlehem does not use artificial sand. It mixes with the cement a new almost pure silicious sand. The patent '089 can not cover the use of silicious sand with cement since such sand has been used by Bethlehem since 1901. The sand in each instance is used in the same way and for the same purpose so there can be no invention in thus using it. As will appear hereafter all the claims of the '089 and '090 patents are anticipated by the Smith patent.
Regardless of the question of the validity or invalidity of the Durand patents they undoubtedly enjoyed a marked commercial success in Europe and in the United States. There is a large number of licensees paying substantial royalties. At one time defendant negotiated for a license. Not being able to agree upon the amount of royalty to be paid defendant declined to take a license. Thereupon defendant instituted a thorough search into the prior art and as a result thereof determined that the Durand patents were void and that licenses thereunder were worthless.
The Durand patents are instances of the familiar stuffed shirt patents. The simple idea of severely restricting the water content is obscured behind the use of the formidable word 'sub-hydrated'. Thousands of children from generation to generation have made mud pies. The child instinctively realizes that the use of too much water will make his pie slop down. The use of less water was never claimed as an invention. A little infantile reflection, call it mechanical skill if you choose, teaches the use of the least amount of water to make a friable ball of cement and sand. No amount of theorizing concerning the thickness of a water film about the particles of cement and concerning water that is 'adsorbed' about the particles of sand obscure the simple teaching of limiting the amount of water in the mix. Limiting water produces a half soaked mix. Unlimited water produces a fully soaked mix. The distinction between soaked and half soaked is too common to require chemical explanation.
Prior Art.
The disclosures and practices of the prior art were presented to show invalidity of the Durand patents. Plaintiffs distinguished this evidence from Durand on the ground that all mixes in the prior art were plastic, i.e., 'fully watered' while Durand's mixes were not plastic but subhydrated or under watered mixes. It is interesting to note that the original French and German applications, offered in evidence in this case as being for the identical inventions covered by the patents in suit, describe Durand's sand cement mix as 'plastic'.
Smith Patent. John Smith patent No. 575,074 for 'Molders Core' issued on January 12, 1897. A witness whose foundry experience began in 1900 testified: 'Any part of a mold that is made in a core box is a core '. Smith states his object is to provide a strong durable core which does not need to be baked. That his cores are all of accurate and uniform dimensions and do not emit gases to injure the metal. He used sand with a binder of Portland cement or other hydraulic cement. He taught that for ordinary iron castings the proper proportions were approximately one pound of cement to ten pounds of sand. As to the water content, Smith says 'the sand and cement may be mixed together and sprinkled with water until the composition is moist enough for molding'.
Smith, as a Pennsylvania foundry man, was familiar with the green and dry sand molding practice. The usual water content for green sand was 3.6% and for dry sand 5.8%. As one of plaintiffs' witnesses said: 'The lower you can work your moisture the better it would be. The moisture has always been an enemy to the foundryman.'
Moldenke Publication.
Doctor Moldenke is a recognized authority in the molder's art. He was author of publications in 1908, 1910 and 1917. In a pamphlet entitled 'The Principles of Iron Founding,' by Richard Moldenke, published in 1917, he states on page 316: 'Cement.-- Mention has been made of cement used as a core binder. The author at one time devoted six months to testing out this binder for small and very large cores in the production of perhaps fifty tons of castings daily. One part of cement was used to nine parts of sharp sand, with just enough water to dampen the material and allow the cement to set. The cores after being rammed up were not baked but set aside for 24 hours to harden and then-- if not used at once-- stored in a warm and dry place. Excellent results were obtained in the foundry and the cost was exceedingly low. The cores could be kept for months as they did not deteriorate-- being practically cement blocks of great porosity. The disadvantage of the process, however, was found to be the effect on the hands of the operatives, the caustic properties of the cement being serious unless rubber gloves were made use of. There is no reason, however, why such a cement binder core-making system should not work out well with the core-machine, as the material need not then pass through the bare hands of the workmen. Cement is cheap and the absence of a core-oven a desirable economy. The idea of using cement as a core binder came from John Smith, one of Pittsburgh's pioneer builders, with whom the author worked out the problem to the perfect satisfaction of everything but the effect on the hands of the girls and men making the cores in question'.
Further disclosures in anticipation of Durand are contained in patents to Conard, No. 1,053,787; Chapin, No. 1,547,787; Gathmann, No. 1,630,612; Messler, No. 1,737,665.
Prior Use at Raritan Copper Works.
Raritan Copper Works at Perth Amboy, New Jersey, is a subsidiary of Anaconda. Anaconda is the second largest copper refinery on earth having a capacity of 46,000,000 pounds of copper per month. It has used cement cores for a number of articles over 25 years. Sand cement cores of the character described by Durand in patent 1,924,028 were in commercial use at Raritan at least since 1926. Certain documentary evidence relating to 'wire bars' being available as evidence, defendant has singled out wire bar molds. These bars are made in copper molds. The molds are made by pressing a hollow water cooled copper core into a bed of molten copper. The hollow water cooled copper core is made by pouring molten copper into a mother mold in which a sand cement core is suspended to form the hollow interior.
In 1922 one Miller, in charge of the Anaconda plant in Montana, visited the Raritan plant. The general foreman of the furnace department at Raritan, an employe since 1899, showed Miller the sand cement cores then in use at Raritan. Thereafter Miller used sand cement cores in the manufacture of wire bar molds in Montana.
In 1926 the manager at Raritan wrote to Miller asking about his method of making sand cement cores for wire bars. Miller replied with instructions and sketches. Miller said: 'All the cement cores are made of a 3-1 sand cement mixed with enough water so it will pack solid. This mix should not be mushy.'
In making these sand cement cores the general foreman of the furnace department at Raritan did not follow Miller's instructions about a 3-1 mixture. He used the standard mix he had been using for other cores which was a 7-1 mixture.
The general foreman at Raritan testified:
'Q40 * * * Will the sand and cement mixture make a ball in his (the molder's) hand when it is at the proper consistency? A. He can make a ball, but I am very doubtful if it will hold to do anything with it, and he can make a ball if he opens his hand very cautiously, but his sand and cement you are sure of pretty nearly all the time. 'Q41. Would you explain to the Court how this mixture of sand and cement that you have used, compared with ordinary concrete, such as is used for sidewalks and building foundations and things of that sort? A. We do not use anything near as much water, nothing near as much water. If we did, we would never dig it out of the molds or cores. It would be just too hard. 'Q42. Can you tell the Court whether this mixture of yours was mushy? A. No, not mushy. 'Q43. Was it pasty? A. No, sir, it is not pasty or mushy. You can, with caution, make a ball on the floor, but not very easily, but then that in turn is rammed in hard, tight, into the core of whatever we are making, rammed in tight by hand. 'Q44. Did the men who handle this sand-cement mixture in your plant ever complain of sore hands? A. Not on this mixture.'
The sand cement cores so made were used.
Prior Use of the National Tube Company.
In 1907 welding rolls, casting cars and ingot stools were cast at the plant of the National Tube Company at McKeesport, Pennsylvania. The castings were large and the heat intense. Ordinary green molding sand with a loam bond was used with a certain amount of fire clay for the mold facings. In 1907 a molder's helper took some cement from the supply shed instead of fireclay. The castings made from this mixture were superior to the old mixture because the cement prevented the burning down of the cope facing. Thereafter the use of cement was continued for twenty years and was adopted by the other molders in the foundry. Its use was extended to drag facings. Of course the sand cement was a layer two or three inches thick against the pattern and was not used for the entire mold.
The proportion of sand to cement varied from 20-1 to 14-1 depending upon the different molders. All agreed the water content was low. One said water was added according to the feel of the mixture and he would make a ball that could be crumbled. Another witness said that the ball could be readily crumbled but was not mushy or pasty. On an average the mold could be poured the day after it was set up. The cement was used to prevent burning down but it also made the mold stiffer.
This prior use also involved a practical anticipation of Durand's artificial sand patents. In cross-questioning a National Tube witness plaintiff brought out that the practice at the National Tube plant was to break up the old molds and to re-use the broken mass to make new molds.
Prior Use of Attalla.
The prior use at Attalla was fully discussed at the trial and a demonstration was made to the court at the Wilmington plant of defendant.
For some years prior to 1918, National Pipe and Foundry Company at Attalla, Alabama, manufactured pipe and steam fittings. In 1918 a fire destroyed the entire steam fitting department. In 1921 in an effort to restore the manufacture of steam fittings, without core ovens, experiments were made with sand cement cores. By recourse to Moldenke's book an adequate practice was developed. The smaller size cores were blown on Demmler machines. Five thousand cores was an ordinary day's run for one machine.
In the demonstration at the Harlan plant of defendant the molder made a mixture of sand and cement and from the mixture made cores on one of the Demmler machines which had been brought to Wilmington from Attalla. The molder mixed sand and cement on the floor with a shovel, splashed water on it from a bucket until it felt right to him and finally riddled the moistened mix through a hand sieve. After each core had been blown on the machine the molder took out the core box, hit it a blow with an iron bar and separated the two halves of the core box. Finally he set the cores out on trays.
In an attempt to discredit this demonstration plaintiffs' witness Phillips testified that the air pressure used in the demonstration was only about 80 pounds whereas at Attalla it was around 100 pounds. On cross-examination Phillips was shown a photograph taken at Harlans in which the air gauge could be seen registering almost exactly 100 pounds. Phillips had not seen the gauge but had based his testimony on what 'two or three of the boys' staying at the hotel had told him.
In 1923 Morehead, vice president in charge of engineering of the Walworth Company, inspected the plant at Attalla and made a report more than five years before Durand's alleged invention. There he said: 'The National Pipe and Fittings Company, located in Attalla produce soil pipe and fittings in quantity and also are producing incomplete lines of screwed and flanged cast iron fittings. Their equipment, while limited as to extent, was in general of the latest design. Pottstown tapping machines and Demmler core machines were used. * * * Cores for screwed fittings are made of eight parts sand to one of Portland cement, blown and allowed to dry. They are ready for use in less than twenty four hours.'
Shortly thereafter the Walworth Company purchased the Attalla plant.
A description of the Attalla practice, as it existed long prior to Durand's earliest date, was published in 'The Foundry' for March 1, 1931.
'A dry sand core department without ovens and where cement is the only core binder employed, occupies a space close to the south end and extends along the west side of the foundry. '* * * Most of the cores employed in this section of the foundry are made of sharp sand bonded with cement and air dried. 'At the Walworth, Attalla, plant dry sand cores are made on three machines made by Wm. Demmler & Brothers, Kewanee, Ill., and the employes are not forced to handle the sand to any extent. Cement in the proportion of 1 part cement to 10 parts sand, by volume, is added to a species of sharp sand that carries a small amount of natural bond.'
After a full opportunity to examine the cores, particularly defendant's Exhibit 68 (as representative of the regular Attalla cores) and plaintiffs' Exhibit 57 (one of defendant's molds), plaintiffs' operator Chastain testified:
'XQ190 Plaintiffs' Exhibit 57 is just about the same degree of hardness as Defendant's Exhibit 68? A. I believe it is. 'XQ191 Do you see any difference between these two exhibits? A. There is a difference in color. 'XQ192 Anything else? A. Difference in the shape of it. 'XQ193 Is that all? A. I think so.'
When the practice was being established at Attalla, experiments were made to determine just what moisture was best for a given product. The criterion was to have the cores hard enough to give satisfaction to the foundry and, at the same time, soft enough to come out easily in the rumblers. As to this moisture content of the mixture, Wilkinson, the general utility man at Attalla, testified:
'A. In experimenting with this cement-sand mixture we tried to make it the same temper as ordinary molding sand or possibly a little wetter, that is what we would try to do. 'Q60 How would it behave in your hand? A. Just as ordinary molding sand. 'Q61 Would it make a ball? A. Yes, sir. 'Q62 Could the ball be crumbled readily? A. Yes, sir. 'Q63 Was it a mushy substance? A. Oh, no. 'Q64 Was it pasty? A. No, sir. 'Q65 Was it anything like structural concrete mix? A. Absolutely not.'
The artificial sand Durand patents '089 and '090 are also directly anticipated by the prior use at Attalla. Some of plaintiffs' witnesses testified that occasionally the butts remaining from used sand cement cores were broken up in the muller and the resulting material included in the next batch of sand cement mixture. The granular material produced by grinding up these old butts is 'artificial sand' as defined by Durand in patents '089 and '090. One example of what was meant by artificial sand is stated as 'The material obtained by crushing old molds.'
Defendant's Practice.
The Bethlehem practice is to dump 9 parts of sand and 1 part of cement in a muller or sand mixer and thoroughly mix the sand and cement therein. The sand contains water which varies from 2.33 to 4.84%, and water is added to bring the water in the total mix up to 9 1/2% by weight. In the commercial operation the water normally varies from 9 to 10% and sometimes under very moist atmospheric conditions, which make drying difficult, the water is reduced to 8 1/2%. The water is added in the mixer, in which the ingredients are intimately mixed, and the water is distributed thoroughly over the sand grains and particles of cement.
Using cement in the proportion of 1 part to 9 parts of sand of course makes the material much cheaper than Durand's 1 to 5 mixture, and at the same time does not make the mold as stiff or as rigid as molds containing a larger amount of cement. The 9 1/2% of water makes the mixture more coherent and develops the strength of the cement to a large degree, but does not make the mold so rigid that it would be detrimental to the casting. This coherence improves the molding qualities and the mixture molds very readily, much more so than a drier mixture. To ensure the castings against failure because of too great rigidity of the molds, the molds are made with soft centers of new sand without any bond, which permits the molds to yield as the castings shrink and also assists in the ease of breaking up the molds after casting.
Reinforcing bars, rods and nails are used for reinforcing the mold parts, and the larger molds are placed in flasks and the smaller molds are very heavily bolted or clamped together. Ramming of the Bethlehem sand-cement molds is necessary. The sand-cement is rammed into the molds by a jolt machine, an air rammer, peen rammers and by tamping by the feet of the men.
After the molding operation is finished, the mold parts are allowed to dry, depending on the size, from 48 hours to a week. To ensure drying when atmospheric conditions are unfavorable, salamanders (coke heaters) are placed near the molds, or they are torched with gasoline torches, or heated by heated air. These drying operations strengthen the mold and also permit evaporation of the moisture. If poured within a few hours after completion, the molds would blow and ruin the casting, even though the mold might be strong enough to hold the metal. Within three minutes after pouring the mold sends up clouds of steam and water runs out of the mold. The steaming lasts for a half hour or more. The Bethlehem molders are decidedly skilled. They have had a service of from ten to twenty years. Its molds are substantially the same as the prior art molds.
The court finds the claims in suit of the Durand patents 1,924,028, 1,918,089 and 1,918,090 invalid and not infringed.
This opinion contains a statement of the essential facts and of the law applicable thereto in conformity with rule 52 of the Rules of Civil Procedure, 28 U.S.C.A.following section 723c.
The bill of complaint should be dismissed.