Howard C. Ludwig, Appellant, v. Jesse S. Sohn and Harold Robinson, Appellees, 324 F.2d 1004 (C.C.P.A. 1963)

This appeal is from the decision of the Patent Office Board of Patent Interferences awarding priority of invention to the senior party, Sohn et al., in Interference No. 88,770. The interference involves application serial No. 204,722, filed January 6, 1951, by Sohn et al., and an application, serial No. 286,348, filed May 6, 1952, by the junior party, Ludwig.¹ 

The interference has a single count which reads:

"1. The method of joining material by are welding said material with a consumable electrode, which comprises the steps of impressing a potential to produce and maintain an are sufficient to melt said electrode between said electrode and said material, striking an arc between said electrode and said material while said potential is impressed therebetween, maintaining said arc sheathed in an atmosphere of commercially pure inert gas, and adding sufficient oxygen to the atmosphere to stabilize the arc, the said method being characterized by the fact that the spacing between said electrode and said material and the magnitude of the arc current are maintained such that the arc burns in a nonturbulent atmosphere so that substantially the only gas in addition to the inert gas present in said shield is said oxygen and by the further fact that the quantity of said oxygen is maintained just sufficient to stabilize the arc without oxidizing the weld."

The invention at issue is directed to a method of joining material by arc welding with a consumable electrode. The method involves (1) maintaining the magnitude of the arc current as well as the spacing between the consumable electrode and the material to be joined such that the arc burns in a non-turbulent atmosphere, and (2) maintaining the arc sheathed in an atmosphere of oxygen and a commercially pure inert gas, e. g. argon, such that the oxygen is substantially the only gas present in addition to the inert gas and the quantity of oxygen is maintained in an amount just sufficient to stabilize the arc without oxidizing the weld.

Pursuant to a stipulation by the parties, evidence on their behalf was submitted in the form of affidavits supported by documentary and physical exhibits. Ludwig, the junior party, submitted evidence allegedly showing actual reduction to practice of the invention of the count in December 1947 and in August 1949, both of which dates are prior to the filing date of the application of Sohn et al., the senior party. Sohn et al. regard their evidence as showing actual reduction to practice of the invention in November 1948. Neither party argues that he should prevail on the basis of an earlier conception followed by diligence extending to a later reduction to practice.

Ludwig thus will prevail irrespective of the proofs of Sohn et al. if he is found to have proved an actual reduction to practice of the invention in December 1947. He also can prevail on the basis of proof of an actual reduction to practice in August 1949, if Sohn et al. have not proved actual reduction to practice as contended in November 1948.

The Board of Patent Interferences held that Ludwig's evidence failed to prove actual reduction to practice prior to the filing date of the application of Sohn et al. The board further stated that if it "be regarded" that Ludwig proved an actual reduction to practice in August 1949, then the proof submitted by Sohn et al. as to the work done in November 1948 "should also be regarded as a reduction to practice." Thus it concluded that Sohn et al. are entitled to prevail on the basis of their earlier filing date on the one hand, or, on the other hand, on the basis of an earlier actual reduction to practice.

The activities Ludwig relies on for reduction to practice involve a first series of welds designated 1AO through 9AO made no later than December 8, 1947 and a second series designated 65AO through 67AO made in August 1949.

Ludwig urges that one of the welds of the first series, weld 6AO, was a sound weld and was recognized as a sound weld by him when made. He contends that he and his supervisor knew from the comparison of weld 6AO with the others in that series that when less oxygen than employed with weld 6AO was used, as in welds 1AO, 4AO and 9AO, the arc was unstable as shown by Esterline-Angus²  meter traces, and that when more oxygen was used, as in welds 2AO, 5AO, 7AO, and 8AO, the weld was porous as shown by lower density measurements.

With reference to the second series of welds, Ludwig contends that his Exhibit J, a physical cut of weld 65AO, shows a sound weld produced by maintaining the oxygen just sufficient to stabilize the are without oxidizing the weld, while his Exhibits K and L, physical cuts of welds 66AO and 67AO, respectively, show that with "higher percentages of oxygen" unsound welds resulted "by reason of oxidation."

Ludwig states that welds 65AO, 66AO and 67AO were made in protective atmospheres of argon with different quantities of oxygen and that he "elaborately studied" the effect of the oxygen on the welds with respect to stabilization of the arc and oxidation.

Sohn et al., on the other hand, contend that Ludwig's welds 6AO and 65AO did not involve an actual reduction to practice of the invention defined by the count. It is argued that welds 6AO and 65AO were but isolated single experiments which did not represent a completion of the invention. Sohn et al. contend that both welds 6AO and 65AO were followed by long periods of further experimentation "before anything that could be called a useful or practical welding process was developed by Ludwig."

Our consideration of whether the board erred in finding that Ludwig failed to show by a preponderance of evidence an actual reduction to practice prior to the filing date of the Sohn et al. application requires careful analysis of Ludwig's alleged reductions to practice in December 1947 and in August 1949.

In this connection the count itself should be considered. The count relates to the use of an inert gas as a shield surrounding an electric welding arc and more specifically involves the addition of small amounts of oxygen to such inert gas, the quantity of said oxygen in the shield being maintained "just sufficient to stabilize the arc without oxidizing the weld." We regard that limitation to the quantity of oxygen as a critical element of the invention at issue. It is firmly established that express limitations in an interference count are deemed material and cannot be ignored. Liebscher v. Boothroyd, 258 F.2d 948, 46 CCPA 701. Moreover, the record shows that the Board of Appeals, in allowing the claim corresponding to the count in Ludwig's application, held that patentability could not be predicated merely on the use of oxygen in an amount which would stabilize the arc. Rather the board based its finding of patentability on the ground that the prior art on which the claim was rejected did not teach a correlation between the proportion of oxygen to be added and the factors of arc stabilization and weld oxidation³  as expressed in the following language of the count:

* * * the further fact that the quantity of said oxygen is maintained just sufficient to stabilize the arc without oxidizing the weld.

Although the Board of Patent Interferences in its decision stated that neither the Ludwig application nor the Sohn et al. application have "an express teaching" of the above quoted limitation, it evidently meant only that neither application used the words of the count for it did not indicate any doubt that the limitation is supported in both applications.⁴ 

The Board of Interferences found, however, that the record failed to show any evidence that Ludwig or any of his associates recognized at the time of Ludwig's alleged reductions to practice in December 1947 and in August 1949, that "to produce a good weld the added oxygen should be an amount just sufficient to stabilize the arc but insufficient to oxidize the weld as required by the count," and accordingly that the work in December 1947 and in August 1949 fell short of actual reductions to practice. A careful analysis of the record reveals no error in that conclusion of the board.

The evidence indicates that in early March 1947 Ludwig was assigned the problem of studying the behavior of the arc in an arc weld in different gases and combinations of gases. It was contemplated that he would study the behavior of the arc in inert gases and later in the combination of inert gases and other gases such as oxygen, carbon dioxide and nitrogen. Ludwig testified by affidavit as follows:

"5. In the latter part of 1947 I made a number of welds shielded in combinations of argon and oxygen in different percentages by volume in this chamber. I produced traces for the current and voltage of the arc with which these welds were produced with Esterline-Angus current and voltage recorders and to my surprise, these traces were substantially flat indicating that the arc was very stable. Some of the welds which were produced in this way in a shield of argon and oxygen appeared sound on the surface. I also made photomicrographs of some of these welds and they appeared sound.

"6. I was so impressed by the flat current and voltage traces and by the apparent soundness of some of the welds described in paragraph 5 that I decided to determine whether or not these welds were porous throughout. This porosity would not be readily visible on the surface of the weld nor would it show up when a cut is taken through a weld for the purpose of producing a photomicrograph. I conceived the idea of evaluating the welds for porosity by measuring the specific gravity of the weld metal. These welds were made with mild steel having a specific gravity of about 7.84, if the welds were sound they should have a specific gravity approaching 7.84. The specific gravity of the weld metal made in a shield of oxygen and argon which appeared sound on the surface was found to approach, and in one case even exceeded, 7.84.

"7. At this point I had been studying welds in various gases for almost a year. Many of these welds particularly those made in pure argon or helium had been very porous. I concluded that such extensive porosity could not be caused by the oxygen alone but must be caused by another factor. Since nitrogen is present in the air and in fact constitutes about eighty percent of the atmosphere as compared to twenty percent for oxygen, I suspected that nitrogen was producing the gross porosity. I then embarked on a program of determining the extent to which nitrogen affects welds made in a shield of gas. * * *

* * * * * *

"9. Having determined that the nitrogen was an important, and probably the most important, contributor towards porosity, I was faced with a problem of providing some way of preventing the nitrogen from penetrating into the weld by means of a shield. I knew that nitrogen from the air tends to penetrate into the shield if the shield is turbulent and that an important contributing factor towards this penetration is the instability of the arc itself. I had found that when a short circuit occurs in the arc, the arc and the surrounding medium is quickly cooled, reducing the pressure in the arc and in the medium appreciably. In welds accessible to nitrogen; for example those made in air, this would cause the nitrogen to penetrate. It then became important to maintain the arc so stable that short circuits and other disturbances are avoided. I had discovered in December of 1947 that small additions of oxygen in an argon shield produced stability of the arc and that sound welds can be made with oxygen in the shield. I now proceeded to make a large number of welds with shields having different mixtures of argon and oxygen to achieve the desired stability of the arc.

"10. I started to prepare the instrumentation for this work during the early part of July of 1948 and the actual making of welds for this work started in early September of 1948 and continued for two years until late September of 1950. During this period I made welds in more than one hundred shields of different compositions of argon and oxygen. I analyzed the weld metal of a large number of these welds in many cases producing photomicrographs. I found that while the oxygen stabilized the arc, excess oxygen would tend to oxidize the weld and produce porosity. This was revealed by the photomicrographs and corroborated by analyses of samples of the atmosphere enclosing a weld before the arc was fired and during the welding operation. I found that where the percent of oxygen in the argon was high, the content of oxygen before the arc was fired was substantially higher than the content of oxygen during the weld. This indicated that some of the oxygen in the shield was being absorbed in the arc, that is, was oxidizing the weld.

"11. By late September of 1950 I had realized that for satisfactory welding the amount of oxygen in the shielding atmosphere must be maintained such as to stabilize the arc without oxidizing the weld. This required that before a welding operation is carried out, consideration must be given to the manner of welding and to the circumstances surrounding the welding operation. For example, in welding at reverse polarity less oxygen is required to stabilize the arc than in welding at straight polarity. The quantity of oxygen which would in many cases be required to stabilize the arc at reverse polarity would be so small as not to oxidize the weld. On the other hand, in welding at straight polarity substantially more oxygen is required in the shield to stabilize the arc. In this case the oxidation must be suppressed by some means and I conceived the idea of adding a deoxidizing material such as the metals aluminum or maganese. Whether or not a deoxidizer is to be added is also determined by the character of the metal being welded and of the electrode. Where the electrode or the metal has a high content of the oxidizer, for example, is a high manganese (2% manganese) steel, it is not necessary to add a deoxidizer either at reverse polarity or at straight polarity. But where the de-oxidizing content of the work to be welded is relatively low as is the case, for example, of rimmed or semi-killed steel, a deoxidizer must be added."

Ludwig further testified that during the early part of 1951 he devised various ways of applying deoxidizers in welding in an atmosphere of argon and oxygen. He testified that on about April 16, 1951 he made a study of welds with steel electrodes and steel work in pure argon, which showed that high-level voltage incidences increase⁵  as the welding current increases. This study caused Ludwig "to recollect that at the same time; that is, during the latter part of April, 1951, I also made similar welds in argon-oxygen atmospheres to verify the effect of the oxygen on the high-level voltage incidences. I found that when welding argon-oxygen atmospheres the high-level voltage incidences disappear and do not occur with any large frequency at very high welding currents."

We think Ludwig's testimony supports the board's decision that Ludwig's alleged reductions to practice in 1947 and 1949 fell short of actual reductions to practice. After making weld 6AO, Ludwig embarked on a program of determining the extent to which nitrogen affects welds. Determining that nitrogen was probably the most important contributor towards porosity, he proceeded to make a large number of welds with shields having different mixtures of argon and oxygen to achieve the desired stability of the arc. That work lasted until late September 1950. While he testified that by late September 1950, he "had realized that for satisfactory welding the amount of oxygen in the shielding atmosphere must be maintained such as to stabilize the arc without oxidizing the weld," he proceeded to make further tests to verify the effect of the oxygen on the high-level voltage incidences and that on April, 1951, which date is subsequent to the filing date of the Sohn et al. application, he found that "when welding argon-oxygen atmospheres the high-level voltage incidences disappear and do not occur with any large frequency at very high welding currents." We do not think that Ludwig's activities demonstrate that, at the time of his alleged reductions to practice, he had a conviction of success of the present process utilizing an amount of oxygen "just sufficient to stabilize the arc without oxidizing the weld."

In his brief Ludwig sets forth the following as an error of the board:

"Failing to recognize that Ludwig's statement in Exhibit N⁶  to the effect that the oxidation of the weld should be considered in determining the per cent of stabilizing oxygen to be added to the inert gas is as much a teaching of maintaining the oxygen `just sufficient to stabilize the arc without oxidizing the weld' as the respective portions of the applications of Sohn et al and Breymeier on which it relied in holding that the Sohn et al and Breymeier applications support said count."

However, we find a very significant difference in the teaching of Exhibit N and the teaching of the Sohn et al. application. As the board stated:

"The Sohn et al. disclosure teaches the use of oxygen in amounts ranging from ½% to 2%. The stabilizing effect of the oxygen addition is described on page 7 of the specification, particularly in the sentence beginning at line 19. Also, on page 8 of the specification as originally filed the following appears:

"Analysis of weld metal produced by this invention reveals no practically measurable quantities of metallic oxides due to the oxygen addition to the inert shielding gas when the oxygen does not exceed 1.5% by volume of the total. More oxygen than this begins to form oxide inclusions which, when the oxygen exceeds about 2%, causes the weld to be metallurgically unacceptable and outweighs the advantages gained by the addition of oxygen. Slightly less than 0.5% oxygen by volume of the total is insufficient to produce appreciable beneficial results.

"This appears to us to disclose an appreciation that the amount of oxygen must be sufficient to stabilize the arc and that too much oxygen will cause undesirable oxide inclusions which should obviously be avoided."

We agree that the excerpt from the Sohn et al. application shows an "appreciation" that the amount of oxygen must be just sufficient to stabilize the arc without oxidizing the weld. We do not think that the teaching in Exhibit N "to the effect that the oxidation of the weld should be considered in determining the per cent of stabilizing oxygen to be added to the inert gas" demonstrates that Ludwig had the same "appreciation" by August 1949 when he completed Exhibit N.

From the foregoing, it is apparent that the welds made in December 1947 and August 1949, the activities on which Ludwig bases his claim of an earlier reduction to practice, did not actually constitute a reduction to practice of the claimed process wherein the oxygen in the inert gas is maintained "just sufficient to stabilize the arc without oxidizing the weld." Ludwig thus is limited to his filing date, which is subsequent to that of Sohn et al., for reduction to practice since he does not contend he should prevail on the basis of a prior conception followed by diligence.

Appellant has moved that the costs of preparing so much of the printed record as includes the added material be taxed against appellees. Upon agreement of the parties, action on the motion was deferred until final decision. We find that certain of the papers accounting for one-fourth the costs of printing the added material were properly included in the record while the remainder was unnecessary. The printing costs being the only costs handled through the court, the motion is granted to the extent that three-fourths of the costs of printing the added material is assessed against appellees and the remainder of said costs is assessed against appellant.

The decision of the board is affirmed.

Affirmed