LIBRARY COPY - UNT Digital Library

No. 1056
JUL 3 1196
By M. Kuhn
A.T,Z. Automobiltechnische Zeitschrift
Vol. 45, No, 3, February 10, 1942
December 1943
By 1 vi .
The discovery and introduction of the internal combustion engine has resulted in a very rapid development
Design has
in machines utilizing the action of a p iston.
been limited by the internal components of the engine,
which has been subjected to ever increasing thermal and
mechanical stresses.
Of these internal engine components,
the piston and piston rings are of p articular importance
and the momentary position of engine development is not
seldom dependent u p on the development of both of the components.
The piston ring is a well—known component and has
been used in its present shape in the steam engine of the
last century,
Corresponding to its importance, the piston
ring has been a rich field for creative activity and it
is noteworthy that in spite of this the ring has maintained
its shape through the many years. From the many and complicated desi g ns which have been suggested as a packing
between piston and cylinder wall hardly one suggestion
has remained which does not resemble the original design
of cast iron rectangular ring.
Material Development
From the metallurgical aspect it is known that the
gra p hite in the ring structure should be present in fine
filiform distribution, for only in this form of graphite
distribution maximum strength properties can be obtained,
The occurrence of eutectic grahite in the structure is
fundamentally undesirable because it is invariably associated with . the formation cf ferrite pockets, which, from
the point of view of wear, are bad. The remaining carbon
is bound as pearlite in iron and its appearance is identi—
fied as fine lamella sorbite.
The quantity of graphite
and structure formation, besides being dependent upon the
smelt charge and the composition of slag is dependent also
on the gas tem p erature, the mould material and the cooling
*"A,T.Z. Automobiltechnische Zeitschrift. 11Vol. 45, No. 3,
Feb. 10 3 1942, pp. 62-67.
NOTE: An English translation taken from "The Engineers'
Di est" Vol. III, No. 11, November 1942.
NACA Technical
Remorandum ' No . 1056
speed of the cast.
The speed of solubility- of the iron
in the surface is de p endent upon the alloying constituents, as y ell as .from the speed of heating •i., .anganese increases,
while silicon decreases, the speed of solubility.
This reaction of the alloying constituents necessitates
t'o uttormoc~t care in their uses because silicon expedites
the precipitation of =raphite in the structure while it
hinders the pearlite formation. A large addition of silicon
necessitates, therefore, an addition of manganese to n-revent excessive grap hite formation.
Better carbide formation results in improved thermal
and strength properties while better graphite formation
results in im p roved running _p roperties. The carbide formin g, elements are chromium, molybdenums, vanadium, tungsten,
arsenic, antimony, while the F'ra7) 'iite formin- elements are
silicon, nickel, titanium, copper, and aluminium.
As has already been pointed out, tie alloying of an
element of the first group usually necessitates the alloy—
in of an element from the second group, if neither carbide
formation nor graphite precipitation is to predominate in
the structure and cause the ring, to have either good running
,p roporti . s and bad thermal and strength properties or 'the
op p osite effects.
n addition, it is necessary that phosphorous be
alloyed. to the ring. Altheu j;h a phosphorous addition
affects unfavorably the strength properties of the ring,
the running p roperties are improved by its addition owing
to tale oossibilit ') of the formation of uniform phosphide
pockets, thereby improving the adherence properties of the
oil film.
Manufacturing Processes
In general, t,ie modern method. of manufacturing piston
rine's, is the individually cast method, whereby the blanks
are moulder singularly a.nd Il heaped if .cast,
Cylindrical pot
casting was p opular for a long' time, but its use,has declined
in spite of the fact th.FL.t a very fine uniform structure was
The elastic'prope.rties of the cylindrical pot
cast rings were not sufficient to withstand the conditions
in the modern engine;.
The individually cast ring showed
in this respect promise, in that it was possible to stress
the ring to a high degree. In Germany, the centrifugal
NACA Technical Memorandum No. 1056 3
method of casting rings is seldom employed, but in the
English—speaking countries the method is adopted,
Ring Stressing
Piston rings can be prestressed by three methods.
The first method is by hammering,a method which is
seldom used in Germany.
By this method, the ring is
-orestressed by stretching the inner fibres of the ring.
The process is applied by rolling or by notch—impact.
The second method of prestressing piston rings is the
heat formed method, which is being used extensively to-In this method a ring which is entirely free of
stresses is split and the gap thereby resulting is forced
apart by a wedge.
In this condition the ring is then
annealed whereby the stresses are nullified. An external
stress is then applied to the ring, and, in the closed
condition, is machined to finished size.
I ^
The third method of prestressing piston rings is the
out—of—round method.
In this method ring blanks are used
which are not circular but formed according to an involute.
The rin- is then cut by cutting away an appreciable piece
of the ring to form the gap.
The ring is then pressed
together and finish turned to a circle. After machining,
the ring will maintain its natural casting stresses. A
heat treatment follows which prevents any subsequent loss
in stress due to the temperatures occurring under engine
operation conditions.
Heat formed rings result in a uniform ring pressure
distribution throughout the ring circumference, that is,
a round radial pressure pattern. Out—of—round rings can
be liven any ring pressure pattern and it is possible to
make the rings with high radial point pressures especially
at the gap.
It will be shown later that this is also
possible with heat formed rings.
The Stress Distribution or Ring Pressure Pattern
The question now arises as to what type of ring pressure pattern is desirable. Originally, one had endeavored
to obtain a uniform ring pressure pattern; this type of
p ressure pattern is very favorable, as, with it, the ring
NACA Technical Memorandum No. 1056
is pressed at all -p oints with uniform pressure against
the cylinder.
It was found from experience, however,
that where the ring was subjected. to maximum load, that
is, at the gan, ring collapse took place.
It is desirable not to have a uniform pressure pattern
when the rint; is new, but after running the ring for some
time under operating conditions a uniform ring pressure
pattern should be attained.. This means that when the ring
is new the ga p should. have higher radial or point pressure
than the rest of the circumference. In addition to this,
with the continuous increase in engine speed, towards the
higher speed range, certain , p rocesses occur, which cause
sudden increases in piston ring blow—by and oil consume—
This process is termed piston ring flutter, a
p henomenon, which probably occurs when the ring is excited
by im-oulses in the range of its natural frequency, Reliable
data is, however, not yet available to analyse this condition
It can be shown experimentally, and the object of
this article is to show, that when the ring pressure at
the ring gap is high, the unfavorable condition of piston
rind flutter and its undesirable effects can be moved outside the range of actual engine operational speeds.
The mean specific ring pressure of the ring is considered as a. most useful criterion. The mean spac,if.ic
boring pressure is calculated from the tangential force
acting at the gap which is necessary to close the ring to
the normal gap cl •oarance. The measurement of the tangential
force is made by means of a "toledo rl balance.
Tho following relr.tion is-obtained:
2 Pt
pm =
kilograms per . squared centimeters
h D
in which
mean specific ring pressure, kilograms per squared
tangential force, kilograms.
rind; width, centimeter
ring diameter, centimeter
NACA Technical Memorandum No y 1056
The results of stress obtained by means of the toledo
method will agree with the actual mean surface pressure
only when the ring pressure is evenly distributed over
the whole ring circumference, which means that the
radial pressure distribution p attern must be round.
The toledo stress and the stress obtained from the
calculated mean specific p ressure are not in agreement
when the radial ring pressure is not evenly distributed
around the ring circumference. The ring with uneven
pressure patterns usually has high pressure points at
the ring gap and at other parts of the-ring circumference.
The remaining sections of the ring may have low or even
very low surface pressures. The whole ring cannot be
considered stronger than that of its weakest section,
and this section cannot be determined from the radial
pressure distribution. It follows therefore, that for
out--of--round rings, which have been stressed, the radial
pressure distribution is important.
Measurement of Pressure Distribution
To obtain actual measurements of radial pressures
is not a simple matter and a method free of error does
not exist. kany references are to be found in the
literature relating to radial pressure distribution and
the results obtained are often represented in a comprehensive manner.
It is, however, noteworthy that the manufacturers usually recede for one reason or another, directly measurements are asked for. Most of the measuring methods give such unreliable data that they do not
lend themselves to criticism.
It may be worth mentioning that the radial pressure
distribution can be measured by the steel band method,
by means of which the ring is stressed to its normal size
b y a thin steel band as shown in figure 1. With the aid
of calipers, the diameter of the ring is measured at various position of its diameter. The out—of—roundness obtained is ring ovality which can be as much as one percent
greater than the normal ring diameter. The positions of
low ring pressure cannot be determined by this method of
measuring ring pressures.
In spite of the fact that rings tested by this method
can give .pressure distributions which ap p ear to be entirely
NACA Technical l+emorandum No. 1056
satisfactory, they have failed when used in the engine.
The failures were more frequent, the higher the loading
of the combustion chamber. .After very short operations
in the engine, the faces of the rings at certain positions became discolored and there is an increase in
blow-by and oil consum p tion. Ring examination for high
duty engines should be carried out, therefore, by some
other method.
The obvious solution of the p roblem lies apparently
in the use of a ring gage, the inside surface of which
has iDrovision for a series of pressure gage sockets or
It is fully realized that the difficulty of
the solution may be the finding and mounting of such
pressure gage sockets, which in the radial direction,
permit the most accurate force indication s^rith the
smallest possible measurements, of the ring pressure.
A relativel..T simple solution of the problem is described
in the following arrangement and shown in figure 2.
The inner surface of a ring gage is arranged with a
number of steel rollers. Each roller can be loaded externally through ar. orifice. A pattern of the radial
ring pressure distribution is obtained by the resistance
of the roller to turning by manual operation. The latter
is very troublesome and, in addition, roller turning by
hand only gives subjective measurements.
A measuring method developed by the Ma.hle Company of
Germany is shown in figure 3. This method permits objective measurement to be made by means of pressure gages.
The ring to be tested is inserted in a ring gage with a
0.02 millimeter greater diameter than the nominal ring
diameter. The g^at;e points of sixteen clock gages are
in contact with points around the circumference of the
ring These pass through holes drilled radially through
the rir}g gage.
By suitable screw adjustment the gage
points are p ressed against the ring, so that all clock
;ages show identical readings a.nd, therefore, similar
stress. The gage points are then. sc:re ­ ed further in
against the ring, by merlins of a P m_ .nlweter' : to an amount
equal to 0.01 millimeter, The readi_Igb obtained from
the clock gages present a pattern of ttie radial pressure
It has been found that this complicated apparatus
cannot be used for practical purposes. The time required
to take measurements is too lengthy, for when setting the
NACA Technical kemorandum No. 1056
clock gages with equal pressure an alteration in the
setting of one clock necessitates variation and adjustIn addition, it is so comment of the-other clocks.
plicated that it is impossible to design an apparatus
suitable to test rings of different diameter.
A simpler apparatus for measuring pressure ring
distribution is that which has been introduced and de=w
veloped by Messrs. Alfred Teves in conjunction with the
Zeiss Co. This apparatus operates on the biezo—electric
princi p le and is shown in figure 4. An inducement to use
the piezo—electric effect of a ll ouartz" is that the quartz
shows only a small tendency to deform in the direction of
the a.pplied load or pressure. The movement of the measuring point is practically zero. But it is a disadvantage
that the p iezo—electric effect occurs not at constant
pressure but only during pressure changes.
The piston ring is inserted in a ring gage of the
apparatus. As the actual measuring procedure has been
described by C. Englisch 3 the author does not consider
it necessary to describe the apparatus at this stage.
As an essential component cf the equipment is a loop or
carthode ray tube, it follows that a well equipped laboratory is desirable.
Both the last two mentioned highly developed methods
of testing piston rings are so elaborate and lengthy in
time that they can only be used by specialist firms.
engineer who uses piston rings and. requires fur.ther knowledge concerning their characteristics is, therefore, forced
to use .simpler methods of measurements such as the methodwhich is about to be described.
The ap p aratus again consists mainly of a ring gage,
the nominal diameter of which corres p ond-s to that of the
piston ring to be tested`. The ring gage has a shoulder
against which the p iston ring is pressed. The gage points
of a clock gage, which is graduated in 1, / 100 or 11000
divisions, pass through a hole in the ring gage and rest
upon the circumference or bearing surface of the ring.
The ap p aratus is shown in figure 5. The front face of
the ring has a series of notches which are situated on'
the right and left of the holes drilled for the gage
These notches are used to locate a hook, upon
which weights can be suspended in order to load the ring
bearing surface.
XAC-A Technical N:emorandum No. 1056
The procedure of measuring piston ring pressure
distribution is as follows: After inserting the piston
ring in the ring gage, the ring gage is adjusted to give
P. zero reading. A simple turning of the ring gives an
indication of ring roundness, in the same way as is obtained with the light method. Everywhere, where the ring
diameter does not correspond with the nominal diameter of
the ring gage, tha clock indicates the occurring deviation.
A satisfactory ring will be within the limits of from 0.002
to 0.005 millimeter of its nominal dimension. It has been
found that many rings measured deviate from the nominal
diameter by as much as 0.03 millimeter. The weakest section of the ring is in the neighborhood of the ring gap,
while the ring ga-j itself retains the dimension of the
nominal diameter.
Measurement of r:^dial pressure distribution is carried
out in the simple manner of hanging weights on the suspen—
sion scale pan attached to the hook. The weight is increased
until. the clock shows a deflection of 0.01 millimeter. The
reduction in clearance at the gap is given by the ,product
of . 77
times th-e deflection and is of the order of 0.03
millimeter. This clearance reduction is unimportant, because even with rings of small dimensions the gap clearance is 0.15 millimeter, or greater..
In this way the ring
is measured at i , 11 sections and a pattern of the pressure
distribution is obtained of surprisin impressiveness.
The suspended weight shows, in addition, the bearing or
surf^- ce pressure occurring a.t each point with some approximation.
By loading the ring .until a reduction in diameter
of 0.01 millimeter occurs each point of the ring bearing
surface moves a small distance during the measurement in
the ring gage.
The frictional force, thereb , T set up, varies
according to the position of the gap,
Should the point at whicth the measurement be made be
near the gap, then with a reduction in diameter of 0.01
millimeter the ring gap will not be reduced but the loaded
rin g; ends will yield towards the center of the ring. Hence,
no friction between the ring and the gage exists. Should
the point at which the measurement is made be on the side
opposite the gan, then a reduction in ring diameter can
only occur with a corresponding reduction in ring gap. The
friction which is set u-o between the ring and the gage will
be measured.
It is for this reason, that the sum of all
the pressures ri.easured on the gap half of the ring are always smeller than those occurring on the opposite or other
side of the ring.
Cn the side opposite to the gap the
NACA Technical Memorandum No. 1056
measured pressures are higher than the actual ones by an
amount equal to the frictional forces which are of unknown
The occurring error in measurement
error is not small, but it can be easily
the position of the center point to the
bution curve, as the sum of the vectors
sures must be zero.
or experimental
estimated from:
pressure distriof all the pres-
Considered. as a whole the accuracy of measurement is
probably greater than that obtained with the other methods
of measuring ring pressure, described in this article,
because the latter only gives a pattern of the ring pressure distribution and not an actual value of the measurement,
With a certain amount of experience, measurements
with the apparatus can be made quite quickly. The time
required to measure the ring takes five minutes or less.
The radial pressure curves are so characteristic for a
ring, that from the curves it can not only be ascertained
whether the type of stress in the ring was obtained by
heat—forming or the out—of—round method, but also indicate
the.source from which the rings were obtained, as the tools
used in the manufacture give the ring product certain characteristics.
Figure 6 shows the radial pressure distribution or
pressure pattern of a heat formed ring, which had been
used for'some 8000 kilometers in an engine. The stress
distribution is very uniform. The reduction of the surface pressure at the gap, that is, the collapse, is not
The sum of all strosses on the side opposite
the gap is somewhat greater than those occurring on the
gap side of the ring. The cause of this is the error in
measurement which has already been fully discussed. This
error is a characteristic of the measuring method.
Figures 7 and 8 have been obtained from two rings
both of which have been run for some 2000 kilometers.
Both rings were heat formed. The first ring had a uniform
pressure distribution while the second had a high pressure
point at the gap. The stresses are partly typical. The
toledo stress of both rings are in agreement.
NACA Technical Memorandum No. 1056
It can be seen that high pressure points can be obtained with heat formed rings.
It is interesting to note
that, while with a ring having a uniform ring pressure
distribution, the piston blow—by increased above 3800 rpm
(fig. 9), the blow—by remained practically constant with
a heat formed ring which had uneven pressure distribution.
The opinion that the phenomenon of ring flutter can be
overcome-by the use of rings with high pressure points,
is confirmed by the above measurements..
Attention should be paid by the manufacturers that
high pressure points at the gap are not exaggerated,
otherwise rings are produced having ring pressure characteristics as shown in figure 10. The high pressure at
the gap is exaggerated and directly-next to it 'a position
of very low pressure is shown, so that the ring aan hardly
bear against the cylinder. The result is a break through
of the hot combustion gases at the dangerous section and
excessive heating of the rings, just at its weakest point.
In an engine of low output the progressive running—in can
remove at first the position of high pressure and eliminate
the position of low pressure.
In the high duty engine, the
ring is often disturbed at the point of its minimum stress
due tc overloading of thermal stresses, before the running—
in process can be of assistance.
This type of unsatisfactory stress distribution is
especially found in out—of—round rings. Figure 11 shows
a series of radial pressure distribution curves of piston
rings, which have been picked out from a batch of rings.
A number of these cut—of—round rings used in a high
duty engine had to, be replaced after short running periods,
because at different parts of the circumference the rings
showed burnt sections.
The measurements of the rings
showed ring pressure distribution curves as shown in figure 12. The position of the burnt sections occurring at
the circumference of the ring are indicated and prove that
the measurements of pressure distribution are correct.
Further measurements obtained, with the steel band method,
are shown in figure 12 by the pressure curve indicated by
the broken line. This shows that the ring ovality is about
6 percent, but does not show the characteristic of minimum
ring pressure which was apparent as indicated by the burnt
or discolored ring sections. That out—of—round rings with
satisfactory ring pressure distribution can be manufactured
is shown by the curves in figure 13. This type of ring,
NACA Technical Memorandum No. 1056 11
formed by turning, is especially of interest in that it
was manufactured from plain cast—iron. Its ackowledgeable quality is not due to highly developed materials,
but apparently due to careful balancing of the ring
pressure distribution, which was very uniformly maintained
with all rings.
The ring stress at the gap is not very
great, but the points of minimum pressure which occur
next to the gap are hardly noticeable.
Page intentionally left blank
NACA Technical Memorandum No. 1056
Figs. 112j31415
Fig. 1
Fig. 2
Fig. 3
Fig. 5
sr,» 9
Fig. 4
Page intentionally left blank
Figs. 619
NACA Technical Memorandum ivo. 1056
Fig. 6
^'rie a. orme evenly
Stressed ring8
He tfo
s^rQbse^. r^nga
3av 320 34Ox
W 3" O
Spsed r.p.m.
Fig. 9. Gas blow by with heat formed rings having uniform pressure
distribution and high pressure point or uneven pressure distribution.
Page intentionally left blank
NACA Technical Memorandum No. 1056
Figs. 758
Fig. 7. Heat formed rings from an Opel engine after 2,000 km., with
evenly distributed pressure. Mean specific pressure obtained from
the tangential force to close the ring, pm — 1.28 kg'cm".
90 •
I 7p
I/O •
f 70•
Fig. 8. Heat formed rings from an Opel engine after 2,000 km., with
high pressure point. Mean specific pressure determined from the
tangential force to close the ring, pm -- 1.35 kg cm2.
Page intentionally left blank
NACA Technical Memorandum No. 1056 Figs. 10111
r 0• iv,
tM e
Fig. 10. Heat formed high pressure point ring with exaggerated high
pressure point near the gap and sections of minimum pressure next
to it. New rings.
MG' 110'
Turned ring with exaggerated pressure point at gap,
Fig. 11.
followed by points of minimum pressure at other sections of the ring
circumference. Mean specific pressure determined from the tangential force to close the ring, pm = 1.42 kg/cm2-.
NACA T.M. No. 1056 Figs. 12 5 13
4e o' ,o•
eo• ^ r^
25 0 r
b oo
\ .Zj^
t'0,^ -V.
x, 4U
"°'^ X10
1W 1
Iv` .^w •'
Fig. 12. Turned high pressure point ring with
burnt sections. The radial pressure distribution is
exaggerated at the gap and is useless. The stress
determined by the steel band method is given by the
broken line diagram. Time of running not known.
0' 90•
.rip• M .
fV -
Turned high pressure point ring with
Fig. 13.
determined pressure distribution. No exaggerated
high pressure point at gap and absence of low pressure
points in the surface pressure. Mean specific pressure
determined from the tangential force to close the ring,
pm = 1.35 kg/cm z . New rings.