Page 41

A New Standard
Elastic
Strength
                        for Ductile Metals
By J. B. KOMMERS
DESIGN engineers are always interested in knowing
D    what allowable stresses may be used for a given ma-
terial of construction. For a material like structural steel
it is common practice to make a tensile test to help in
answering this question. The tensile test determines, among
IJLIwtJJ.Lilili^s, tflC
UbL1C [l111 gS, Me
ultimate tensile
strength and the
yield point of the
material. An al-
lowable or safe
unit stress should
be well below the
ultimate, but it
should also be be-
low the yield point,
because when the
yield point is
reached the ma-
terial suffers per-
manent deforma-
tion.   In  other                J. B. KOMMERS
words, at the yield point and beyond, the material does not
behave in an elastic manner, and will not return to its
original dimensions when the load is removed. The above
remarks apply particularly to structures subjected to static
or quiescent loads. For machine parts subjected to millions
of repeated or reversed stresses, the allowable unit stresses
used in design must be still lower, in order to prevent the
formation of a crack due to fatigue action. Also for cases
involving impact, the allowable unit stresses must be lower
than for static loads.
  Under the heading "Methods of Tension Testing of
Metallic Materials," in the 1933 book of Standards of the
American Society for Testing Materials, the following defini-
tions are of interest in the present discussion. The four
values defined are all intended to be approximations of the
elastic strength.
  Elastic Limit-The greatest stress which a material is
capable of developing without a permanent deformation re-
maining upon complete release of the stress.
  Proportional Limit-The greatest stress which a material
is capable of developing without a deviation from the law
of proportionality of stress to strain (Hooke's Law).
  Yield Point-The stress in a material at which there oc-
curs a marked increase in strain without an increase in stress.
  Yield Strength -The stress at which a material exhibits
a specified limiting permanent set.
  The definitions of elastic limit and proportional limit re-
mind one of some of the political and economic plans which
are common topics of discussion today. They are examples
of things which seem very simple on paper, but which are
not simple when an attempt is made to get actual results.
   In determining the elastic limit it is necessary to apply
loads to the specimen, reduce the loads to zero or nearly
zero, and then determine from the extensometer measure-
ments whether any permanent deformation or set has oc-
curred.  This process is continued with increasing loads
until a permanent set is registered. A practical question
which arises is the amount of permanent deformation which
shall be used as a criterion. If the extensometer which is
used can be read to 0.000001 of an inch, there will be an
indication of permanent set at a much lower unit stress
than if the extensometer can be read to only 0.0001 of an
inch. Some experimenters have even reported that when
delicate apparatus is used there is evidence of inelastic ac-
tion at all stresses. The writer has also shown* that, at
least for certain brasses, the repeated loads used in determin-
ing the elastic limit may have the effect of improving the
elasticity, so that the value obtained for elastic limit does
not represent the material in its original condition. Besides
this, the test is tedious and expensive and therefore is not
a practical commercial test.
  In determining the proportional limit of a material it is
necessary to plot the unit stress against the unit deforma-
tion and then determine the lowest unit stress at which the
curve first deviates from a straight line. There are many
factors which may influence the result obtained from such
a test. In a paper published in the 1929 proceedings of
the American Society for Testing Materials, R. L. Templin,
of the Aluminum Company of America, discusses some of
these factors. He classifies them under five major headings:
(1) the material tested, (2) the test specimen used, (3) the
testing apparatus used, (4) the testing procedure followed,
and (5) the interpretation of the results obtained.
  Under (1), Templin mentions definite grain due to
method of manufacture, variations in structure due to voids,
seams, inclusions, grain size, segregation of constituents, and
uneven temper, any one of which may cause local yielding
and affect the proportional limit.
  Under (2), he mentions variation in dimensions, straight-
ness, and internal strain, resulting from the procedure used
in preparing specimens. He cites an example of an alumi-
num alloy, one specimen of which was slightly bent before
testing, with the result that its proportional limit was much
lower than that of a normal specimen. Machining of speci-
  *Engineering, London, June 22 and 29, 1934.
December, 1934
Page 4 1