University of Wisconsin Digital Collections
Link to University of Wisconsin Digital Collections
Link to University of Wisconsin Digital Collections
The University of Wisconsin Collection

Page View

Niles, Donald E. (ed.) / The Wisconsin engineer
Volume 48, Number 7 (March 1944)

In this issue,   pp. 4-[6]


Page [6]


Shrinkl
Fitting
        by Harold May, me'44
of
Metals
C  OOLING     the male or internal member of mating
    metal machine parts so as to obtain sufficient shrink
to allow assembly with little or no pressure application.
This might easily be termed as industries jump out of
the frying pan into the refrigerator. Heretofore, the im-
portant method of assembling such parts was by expand-
ing the external or female part by the application of heat
to produce sufficient expansion for easy assembly. Upon
cooling the two parts are firmly held together as one.
Perhaps one of the earliest applications of this method
was the shrinking of the rim on a railroad wheel. An-
other method of assembly was that of applying pressure
,to force parts together for the required assembly. In
either case a negative clearance on the parts before assem-
bly gave a firm friction grip afterwards.
  Neither of these methods have been suddenly dropped
by industry; as a matter of fact they are probably the
more important methods in use today, but the stimulus
of war and the searching for faster and better methods
of assembly has brought out the use of refrigeration in
mating parts with negative clearance. This method was
not brought into use overnight, but it has had its greatest
development in the last year or so. Perhaps the first rea-
son for the search for new methods was the tendency for
change in metallic structure and properties, if heated to
sufficient temperature to produce the expansion desired.
At the same time such temperatures often caused warp-
ing and scaling, giving an uneven bearing surface when
assembled. In force or press fits such troubles were not
encountered, but it was found difficult to make extremely
tight fits without tapering the pin, and even under these
conditions distortion and scoring are almost inevitable.
  It was such difficulties as these which led to the work in
freezing or shrinking before assembly. With a room t-m-
perature of about 700 F. and assuming reasonable nega-
tive temperatures, the amount of shrink could be easily
calculated by means of temperature coefficients. "Dry
ice" was perhaps the first freezing method used. The
temperature of "dry ice" is approximately -1100 F., but
as it was difficult to obtain results when in solid form, it
was necessary to immerse in kerosene or alcohol to pro-
duce a liquid form. Under such conditions a covering of
CO gas often formed around the dry particles, thus re-
ducing their effective working temperature to about
-900 F. With a room temperature of 700 this gives a
temperature differential of about 1600 F., which except
in the case of small parts was not sufficient to produce
the required contraction, so a combination of heating the
external and freezing the internal part was used. In this
way a much lower heat was necessary, greatly reducing
many of the previous heating difficulties involved, but at
the same time the necessity of the two operations intro-
duced new problems, both in cost and time, so experi-
menting continued. Liquid air at a temperature of
-297f F. gave sufficient temperature differential in most
cases, provided the metal was actually placed in the liquid,
which again brought out new problems. The liquid air
caused chemical action on certain metals and when the
metal was removed from the liquid into air, it immedi-
ate-y produced a frost coating, undesirable for assembly.
Keeping the metal from actual contact with the liquid
eliminated most of these difficulties, but the required low
temperature was not obtainable under these circumstances.
Liquid nitrogen with a temperature of -3200 F. seemed
to be the most adaptable for varying conditions, and
therefore the logical freezant for continued experimenta-
tion. Keeping the metal from direct contact eliminated
chemical action somewhat and yet produced sufficient
temperature differential for most conditions, so apparent-
ly this was the answer to the problem.
  This liquid, or one of the others, has been and is still
used extensively in such operations, but it did not mean
the end of experimentation with other methods. The
chief difficulty with these methods seemed to be in cost
and speed. It was found difficult to find containers of
sufficient quantity to work on a mass production basis
without too great a loss by radiation and evaporation.
Small parts could be forced through tubes which are
immersed in the coolant, coming out on the other end
with proper shrink, but this was hardly possible with
larger parts.
  Such difficulties led to experiments with, and final de-
velopment of, mechanical deep-freeze units which are in
popular use today. These mechanical units, which are
yet in their experimental stage, are capable of producing
temperatures of --150f F. or perhaps below with unbe-
lievable efficiency. It is this greater efficiency, and thus
reduced operating cost, as well as its convenience, ease
                 (continued on page 18)
THE WISCONSIN ENGINEER
7


Go up to Top of Page