COMPRESSION TESTS. 151 



They will be found to prevent the specimen from being 

 pulled through the wedges in the first instance, and in 

 this way help them to take hold of the bar. This 

 arrangement is shown on Fig. 69. In the three metals 

 which have been mentioned, as the metal is very ductile. 

 the percentages of extension and of reduction in area are 

 large. In these metals also the bar is often found to break 

 at a load less than the maximum load, as it is generally 

 possible, when local contraction is observed to be setting in, 

 to run back the load to such an extent as to balance the 

 pull on the specimen, and to continue diminishing the 

 load until fracture takes place. 



92. Low-grade Wrought Iron, the Harder Kinds of Steel, 

 Gun- metal, and Brass. These may be taken as represen- 

 tative of the less ductile metals. They are certainly 

 ductile, but to a smaller extent than in those already 

 mentioned. The holding is effected both by wedge grips 

 and screw dies. The extension and the reduction of area 

 are less than in the former cases, but are still considerable. 

 In these metals, fracture generally takes place at, or very 

 near, the maximum load. 



93. The Commercial Elastic Limit. The best way to 

 obtain the actual elastic limit with accuracy is to take careful 

 measurements with an extensometer, plot a diagram to a 

 large vertical scale, and notice where the curvature begins. 

 An inspection of the figures will also give it. 



The commercial limit, called also the " yield point " 

 and the " breaking down point," is best obtained with 

 certainty by setting a pair of dividers to as great a length as 

 is possible on the bar. One point is then held on the 

 lower gauge point, and the other point is made to scribe an 

 arc on the upper surface of the bar. When the bar has 

 definitely passed the limit the scribed line very perceptibly 

 and rapidly thickens. Some observers depend upon care- 

 fully watching the end of the weighing beam, which begins 

 to drop rapidly when the limit is reached. This is not 

 advisable, as it is not at all certain, and a considerable 

 error may be made by an unpractised observer. 



COMPRESSION TESTS. 



94- When a material is to be tested in compression there 

 are certain rules as to proportions and methods of 

 holding which it is important to observe. If the speci- 

 men is very short in proportion to its thickness, failure 



