Appeal from the District Court of the United States for the Eastern District of Pennsylvania; W. H. Kirkpatrick, Judge.
Before BUFFINGTON, DAVIS, and THOMPSON, Circuit Judges.
BUFFINGTON, Circuit Judge.
This patent case concerns machines for ascertaining and recording the breaking point of, for example, concrete under compression and steel under tension. As stated by the court below: "Testing machines have to deal with tremendous forces but they must measure them with the highest degree of precision. The crushing of a block of cement or the breaking of a bar of steel requires a powerful and rugged machine, but unless it can register with exactness the breaking point it will be of little value in modern engineering practice."
In such machines the breaking force, known as the "load force," is exerted against the article to be tested, and such "load force" is a great one. For example, the proof in reference to a testing machine in the National Bureau of Standards at Washington is: "This machine is for testing specimens in either tension or compression and has a capacity in tension of 1,150,000 lb. on specimens of any length up to 34 ft. 4 in. after straining, and a capacity in compression of 2,300,000 lb. on specimens of any length up to 33 ft. 1-3/4 in. between platforms 36 in. in diameter." And machines made by the defendant "for slab and long column tests" are described as adapted to tests up to 10,000,000 pounds.
Such being the conditions which confront a testing machine, it has been found that it is impracticable to balance millions of "load force" pounds against millions of specimen pounds, and the art resorted to using instrumentalities located between the "load force" and the ascertaining and recording mechanism, whereby a lesser force than the actual "load force," but proportionate thereto, could be exerted on the ascertaining and recording measuring mechanism. In accord therewith it was customary, roughly speaking, in the graduated beam or lever and counterweight in weigh scales, to use levers, to one end of which the "load force" is applied and at the other a counter balance weight, movable by hand.
In his handbook on the art (published in 1899), Professor Adolph Martens, director of the Royal Testing Laboratories of Berlin and at Charlottenburg, says: "The same principles which apply to scales, gauges, etc., also apply to the load-indicators of testing-machines. There are some additional special features which have individually been adopted more or less generally in testing-machines. The scale, especially the beam-scale, is frequently such a predominant feature in the design of the testing-machine, that it becomes apparent at the first glance in the modern machine."
Indeed, that the lever type machine was the prevalent one in use, is shown by defendants' catalogue, where, referring to the accompanying picture, in which will be seen the graduated lever weight beam and the movable counter, the defendant says:
"The above illustration is of our 100,000, 150,000 and 200,000 pounds capacity Olsen New Automatic and Autographic Universal three-screw type motor-driven Testing Machine as used by all up-to-date testing laboratories throughout this country and abroad.* * *
"These testing machines are the recognized standard for high-grade testing throughout the world. They excel in accuracy and sensitiveness, are designed for maximum strength, durability, and ease in operation."
Incident to the operation of a lever, and necessarily so, a "knife edge" was required, and this resulted in inaccuracy. The testimony introduced by defendant shows this. Its Exhibit 5, Yale and Towne Bulletin, says:
"The corner-stone of all existing systems of weighing machines is the 'knifeedge'. This consists usually of a trianglar piece of hardened steel, resting on a flat plate of the same material, and forms the fulcrum on which the scale beam or lever oscillates."
"Theoretically, the knife-edge rests and oscillates upon a mathematical line, and the more nearly practice conforms to theory in this respect, the more accurate will be the scale. In practice, however, and particularly in large scales, the knifeedge is required to support heavy loads, and its bearing surface must thus have a sensible area to prevent the crushing of the material. This is sought by increasing the length of the knife-edge, the usual rule being to limit the pressure to a maximum of 12,000 lbs. per inch of length. As a result the knife-edges in large scales have considerable length, and this fact introduces another element of error, viz., the difficulty of fixing the knife edge in absolute parallelism with the axis of rotation of the beam or lever, and of preventing flexure of the knifeedge under pressure, so that it does not bear uniformly on the whole of its length. Obviously the least want of coincidence in either of these respects would introduce a large and variable element of error. Even assuming a knife-edge to be originally true in all respects, its bearing surface brought to a perfect and sharp edge, its axis exactly normal to the plane in which the beam vibrates, and the two plates on which the ends of the knifeedge rest to be true planes and perfectly aligned, how long is it probable that all of these conditions can be maintained? Oxidation tends ...