March 4, 1904.] 



SCIENCE. 



363 



Professor J. L. Van Ornum, of Wash- 

 ington University, St. Louis, Mo., dpseribed 

 the results of his experiments on 'The Fa- 

 tigue of Cement Products. ' In his experi- 

 ments he made tests of cubes and prisms of 

 neat Portland cement, and of concrete, ap- 

 plying different loads at the rate of about 

 four times per minute until rupture en- 

 sued, and plotting a curve of results show- 

 ing the number of repetitions necessary to 

 cause rupture when the load was a given 

 per cent, of the ultimate static strength. 

 The results of the tests of cubes of neat 

 cement show that repeated loads, less in 

 intensity than the ultimate strength of the 

 material, will cause failure. The number 

 of repetitions necessary to produce this ef- 

 fect increases very rapidly for loads less 

 than 65 per cent, of the ultimate strength, 

 and seems to become infinite at about 50 

 per cent, value. For example, 180 repeti- 

 tions of the load of 80 per cent, of the ulti- 

 mate static strength are sufficient to cause 

 rupture. Four hundred repetitions of the 

 70 per cent, load, or 1,000 repetitions of the 

 62 per cent, load, or 1,700 repetitions of the 

 60 per cent, load, or 4,000 repetitions of the 

 56 per cent, load, or 5,000 repetitions of 

 the 55^ per cent, load, will do the same. 

 The same general law applies equally to 

 concrete. The above results with cement 

 and concrete are, therefore, similar to those 

 obtained by "Woehler on iron and steel. 

 The modulus of elasticity of cement and 

 concrete is greatly reduced in value under 

 the infltience of repeated loads of the in- 

 tensities indicated. Prisms 5" X 5" X 12" 

 high were used in this work. The paper 

 will be printed in The Transactions of the 

 American Society of Civil Engineers. 



A paper on ' The Design of Steel Concrete 

 Arches,' by Professor E. J. McCaustland, 

 of Cornell University, Ithaca, N. Y., was 

 read in his absence by his colleague, Pro- 

 fessor H. S. Jacoby. The author calls at- 

 tention to the lack of clean-cut, definite 



knowledge as to the action of steel com- 

 bined with concrete under stress, and par- 

 ticularly in an arch ring subject to moving 

 loads, and states that arches are built with 

 factors of safety ranging probably all the 

 way from 3 to 150. He is of the opinion 

 that we do not so much need new theories 

 as we do an extension of our practical 

 knowledge of the mechanical properties of 

 concrete. He gives an abstract and dis- 

 cusses a graduating thesis on the subject 

 by Mr. W. S. Edge. After briefly stating 

 the theory which formed the basis of the 

 investigation and describing the details 

 thereof, he summarizes Mr. Edge's con- 

 elusions as follows: (1) that the graphic 

 method of solution is as accurate as is 

 justified by our knowledge of safe unit 

 stresses in concrete; (2) that an arch 

 ring designed for thrusts due to uniform 

 live loading will be too thin at the 

 haunches to resist stresses due to eccentric 

 loads; (3) that in large spans it is more 

 accurate to use Cain's method of sub- 

 dividing the arch ring, since it gives, in 

 general, results for thrusts which are about 

 two per cent, greater than will be given by 

 dividing the arch into equal horizontal sec- 

 tions; (4) it is better not to try to use a 

 modified semi-ellipse for an earth-filled arch 

 when the rise is less than one sixth, or pos- 

 sibly one eighth of the span, as the equi- 

 librium curve flattens it too much at the 

 haunches. It is better to take full advan- 

 tage of the rise by making the arch linear 

 from crown to springing, thereby reducing 

 the crown thrust. (5) The maximum bend- 

 ing moment is not produced by a live load 

 covering one half of the bridge. For the 

 crown section the bending moment is the 

 greatest when the load is about three fifths 

 on the bridge. The greatest positive mo- 

 ment, however, occurs with the arch prac- 

 tically one half loaded. The greatest nega- 

 tive moment occurs when the arch is three 

 fifths loaded. As the result of designing 



