Note: Descriptions are shown in the official language in which they were submitted.
10~i7~
BACKGROUMD OF THE INVENTION
Field of the Invention
.
This invention concerns pulse-echo type wall thick-
ness measuring in general. More speclfically, it relates
to a method and system for improving upon known pulse-
echo types of wall thickness measurement.
Descri~tion of the Prior Art
In the field of non destructived testing, or
measuring of thickness of the walls of elongated bodies,
use haæ been made of a longitudinally directed pulse
transducer with a 45 degree reflecting surface so as to
direct the energy from a parallel direction to one at
righ angles that is thus transverse to the walls of the
elongated body. Also, in connection with pipe line
inspection employing pulse-echo systems, it has been known
to make use of a plural transducer with rotator, which
employed a scanning or a physical rotation at a rate
proportional to the longitudinal travel in the pipe. How-
ever, neither of the prior arrangements conceived of
slmultaneously introducing another pulse directed at a
different angle relative to the surface of the pipe or
other wall. This invention makes use of the latter
method in con~unction with known elements and arrangements,
to provide ~or an improved method and system for deter-
mining a characterizat~on of a discontinuity so as to be
able to recognize a cavity or thin spot, as distinguished
from a welded ~oint or the like.
Thus, it is a object of this invention to provide an
improved method or system for measuring wall thickneæs
~067~
with a pulsc-echo type of system wllile employing an additional means for
characterizing a discontinuity in sa;d wall thickness that may be observed.
Briefly, the invention provides in a pulse-echo system for measur-
ing wall thickness of pipes or the like, the improvement comprising in
combination, means for generating a single pulse directed along a path
that is longitudinal relative to said pipe wall, means for splitting said
single pulse to form first and second named pulses, said splitting means
comprising first and second reflecting surfaces, said first reflecting
surface having an angle relative to said longitudinal path such that said
first named pulse is directed transversely relative to said pipe wall for
making said wall thickness measurement, said second reflecting surface
having an angle relative to said longitudinal path such that said second
named pulse is directed at an angle of incidence relative to said wall that
is greater than the critical angle of refraction in order to permit any
reflected energy from a discontinuity in said wall to be returned along the .
same path, said second reflecting surface being non-contiguous with said
first reflecting surface and having an offset along said longitudinal path .
for causing a time delay of said second pulse relative to said first pulse; :
and means for receiving any reflected pulses from both said first and second ~ : ;
named pulses whereby said wall thickness may be measured and any discontinuity
may be characterized.
1067~14
According to another aspect of the invention, there is provided
in a pulse-echo system for measuring wall thickness of a pipe, the improve-
ment comprising in combination a plurality of pulse generating means located
spaced peripherally about the axis and inside of said pipe, said means being
oriented for directing the pulses parallel to the axis of said pipe, a first
annular reflecting surface having a 45 angle relative to said axis and
spaced axially from said plurality of pulse generating means, said first
surface extending radially about half the width of said pulses for splitting
off a portion of each and directing said portion along a path transverse
to the wall of the pipe at that location, a second annular reflecting sur-
face having an angle of more than 45 relative to said axis and being spaced
axially at a greater distance from said pulse generating means than said
first annular surface, said second surface being non-contiguous with said
first surface and having an offset axially therefrom to provide said greater
distance and to cause a time delay of the remaining portions of said pulses,
and said second surface extending radially the other half of the width of
said pulses for directing another portion of each pulse along a path having
an angle of incidence relative to the wall of said pipe at that location
which is greater than the critical angle of refraction in order to permit
any reflected energy from a discontinuity in said wall to be returned along
said other portion path.
.,
~06761~
BRIEF DESCRIPTION OF THE DRAWINGS
The ~oregoing and other objects and bene~its of the
invention will be more fully set forth below in connection
with the best mode contemplated by the inventor~ for
carrying out the invention~ and in connection with which
there are illustrations provided in the drawings, wherein:
Figure 1 is a fragmentary schematic showing, partly
in elevation and partly in cross section, illustrating
one form of apparatus which may be employed with the
invention; : ~`
Figure 2 i8 an illustration representing oscillographs
which show various echo pulse signals which are the re-
flected pulses that are returned from the sur~aces of a
wall being measured in the manner indicated by Figure 1, :
Figure 3-4, 5-6, 7-8, and 9-10 are fragmentary
~chematic showings in elevation and crosæ section, and
plan views, respectively illustrating variouæ difications
as to different configurations of a reflecting surface
which may be used in order to control the characteristics
of an ultrasonic energ~ beamj
Figure 11 is a fragmentary schematic, showlng a
longitudinal cross-section taken along the lines 11-11
of Flgure 12. It illustrates another form of the in-
vention which is particularly for use in pipe wall thick-
nes~ measurement;
Figure 12 is a horizontal cros~-sectional view taken
along the lines 12-12 of figure 11; and
Figure 13 i8 schematic circuit diagram illustrating
an electrical circuit arrangement for use with the com-
bination of elements that are illustrated in Figures
11 and 12.
.,-
~067614
DESCRIPTION OF THE PRæFERRED EMBODIMENTS
The pul~e-echo technique ls well-known in connection
with ultrasonic energy, and has been considerably employed
with non-destructive testing or measuring of wall thick-
nesses of various types of material including plpes for
use in pipelines, and the like. However, in ~aking use
of this techni~ue heretofore it has been found that often
there is a difficulty in the ability to distinguish be-
tween thin or corroded spots and other anomalies such as
welded joints and irregular surfaces. This invention
provides a method and system which overcomes such ~ ;
difficulties. It makes it possible to have positive
identification of thin or corroded spots in the walls of
a pipe or the like.
Thus, with reference to Figure 1, there is shown ln
longitudinal cross-section, a fragemental portion of a
pipe 21. It will be appreciated that the wall 21 of the -
pipe mlght also be a wall of a tank or the like. Al~o,
it will be understood that inside the pipe 21 ~here is a
fluid 22 which will act as a good conductor for acoustlc
energy.
In order to measure the thickness of the wall of pipe
21, there is a pulse transducer 25 with a lens 26 for
creating a d$rected beam of energy when the crystal is
actived. It will be understood that in this type of
sy~tem a short time duration, unitary pulse of ultrasonic
acoustic energy is created by applylng a short time duration
electrical voltage to the crystal in a conventional manner.
The crystal material is preferably lead metaniobate and in
3 the lndicated transducer 25 it will be a flat disc shape
(not shown) with silvered faces as its electrodes (not
- 5 -
1067tii~
shown). Since the crystal is piezoelectric in nature, it
will deform and thus produce an acou~tic energy pulse.
Because of the orientation of the transducer 25 and
the focusing of the lens 26 the pulse will be directed
downward along a path 29 until it strikes a flat reflect-
ing surface 30 that is located at 45 degrees relative to ~ -
the axis of the pipe 21. The energy of that part of the
pulse which strikes the surface 30, will then be reflected
at right angles to the path 29 and so travel over a path
33 that is transverse to the surface of the pipe wall 21.
The wall 21 of the pipe has a inner surface 34 and
an outer surface 35, each of which is a boundary for the
material of the pipe 21 such that a reflection sf some
of the energy in the acoustic pulse will occur from each
of these surfaces. These reflected pulses wlll then travel
back along the same path 33 in the other direction and
then will be reflected from the surface 30 to travel up-
ward along path 29 until they strike the crystal in trans-
ducer 25, where an electrical signal corresponding to the
acoustic pulse will be generated. The technique for thus
first producing an acoustic pulse and after the short
time involved, producing the electrical pulse signals
created by the returning reflected acou~tic pulses is
well-known, as already indicated, and need not be describ-
ed in greater detail here.
Figure 2 illustrates four lines of corresponding
oscillograph records, showing signal amplitudes as a
function of time. The energy is ultrasonic in fre~uen~y
so that the time is short and distance 38 on the time
scale represents two micro seconds.
-- 6 --
~. .
- ' .
~ 6'761~
The uppermost trace on the Flgure 2 lllustration,
which is designated by the letter "a", shows a pair of
reflected pulses 41 and 42. These are the pul~e æignal6
generated by the transduc~r as it responds to the acou~tic
energy which would have been reflected from the surfaces
34 and 35 of pipe wall 21 if no cavity or other di~continuity
existed.
Referring back to Figure 1 it will be ob~erved that
there is another flat reflecting sur~ace 45 spaced ~ome- -
what farther away from the transducer 25. This reflecting
surface 45 i8 æituated at a angle-of more than 45 degrees
relative to the axis of the pipe 21. Con~equently, energy
from pulses travelling over half of the path, e.~. a path
46 that is indicated by a dashed line, will travel over
another acoustic path 47. Path 47 has an angle of
incidence relative to the inner æurface 34 that is greater
thQn the critical angle of refraction for the material o~
the pipe wall 21. Consequently, this pulse energy will
penetrate the surface 34 of the wall of the pipe 21 and
travel upward and out without returnlng unless there is
a discontinuity in the pipe wall 21, such as a cavity 51
illustrated. If there i8 such a discontinuity, there will
be a reflected pulse, e.g. from the cavity 51, which will
return over the same paths 47 and 46 to impinge upon the
transducer 25. It will then eenerate a reflected pulse
in the same manner as the reflected pulses of the thick-
ness measuring paths 33 and 29.
From the foregoing it will be understood that each
time the transducer 25 iæ activated to produce an acoustic
pulse, such pulse will travel downward in the fluid 22.
- 7 -
. . . . . .
10676~
Also, the pulse will have passed through the lenæ
26 so as to focus it to cover an area that is substantially
the same size a~ the crystal of the transducer 25. Aæ ~ ~
such pulses travel downward from the transducer 25, the ~ -
upper reflecting surface 30 wlll reflect half of each
pulse at right angles, so that thiæ portion goes trans-
ver~ly to the surfaces 34 and 35 and is reflected from
these walls of the pipe 21. The returning reflected
pulses thus indicate the thickness of the wall 21.
At a short time thereafter, the reflecting surface
45 will reflect the other half of each pulse so that thiæ
portion goeæ along the path 47 that tranæmits the energy
into the wall 21. Such energy travelling in the wall 21
will cause a reflected return pulse only in the event that
there iB some discontinuity in itæ path. Such a dis-
contlnuity will cauæe a reflected return pulse to appear.
It will be underætood that by having the reflecting
æurface 45 located farther away from the transducer 25
than the transverse energy reflecting æurface 30 at a
predetermined distance, it may be readily determined how
much time delay iæ to be expected between the first pair
of re Mected pulses and the single returning pulse~ each
along separate paths.
Of particular importance to this invention is the
fact that since conditions on the inner and outer wallæ of
the pipe 21 may include some discontinuities which will not
cause the delayed single reflected pulse to return, e.g.
weld joints or shallow irregularitieæ, such conditions can
be distingui6hed from others. In other wordæ some
conditions can cause loss of either or both of the first
pair of reflected pulæe~ which determlne the thicknesæ
- 8
.
~067f~1~
of the wall 21, i.e. pulses along path 33. ~owever, the
other slightly delayed pulse along path 47 will provide an
indication as to the presence of a discontinuity which
causes the single reflected pulse to return, e.g. a
cavity or similar thin spot.
Figure 2 illustrates that part of dif~erent
oscillographs, taken under some of the foregoing conditions
which shows the returning reflected pulses. Thus, already
indicated above, the line "a' of the oscillograph
illustrates conditions wherein the reflected returning
energy pulses 41 and 4~ are from the inner and outer
surfaces of the pipe wall 21 where there is no cavity or
other discontinuity. Consequently there is nothing to
return any reflected energy pulse along the other pulse
path 47. The absence of such later pulse may be noted
at a general location 54 on the trace "a".
On the other hand, the trace "b" of Figure 2
illustrates a first pair of reflected pulses 55 and 56
which are closer together in time and so indicate a
thinner wall condition of the pipe 210 In additionj
there is a third pulse 57 that is one which has returned
over the angled path 47 and thus positivity indicates the
presence of a discontinuity such as a cavity 51 which is
illustrated in Figure 1.
Traces "c" and "d" of ~igure 2 illustrate other
conditions similar to trace "bn, but with thinner wall
conditions. Consequently, the first pair of re~lected,
i.e. transverse pulses are closer together. However,
these traces also illustrat the present of a discontinuity
such as the cavity 51, since in each case there is a de-
layed single reflected pulse 60 and 61 respectively.
'1067~1~
Reîerring to Figures 3-10, there are various
modificatiolls shown which relate to the structure of
reflectors Ior ultrasonic energy beams~ These are
especially applicable to the type of wall thickness
measurement under consideration. It is quite feasible
to control the shape of the acoustic energy beams in-
volved, by means of determining the shape of the reflector
surface. mus different types of beam shaping may be
carried out. For example, in Figures 3 and 4 the beam
may be concentrated vertically while retaining the same
width in a horizontal plane.
Figures 3 and 4 illustrate side and plan views of a
transducer 65 that transmits pulse beams onto the angled
surface of a reflector 66, which has it reflecting surface
divided into two halves 67 and 68. These reflecting sur-
faces 67 and 68 have a small supplementary angle relative
to one another so that the beam of energy reflected will
be concentrated in the manner indicated, i e. toward a re-
flecting point 71 on a wall 72 of a pipe 73.
On the other hand it may be desilrable to slightly
disperse the beam of the acoustic pulse path vertically,
and such is illustrated in Figures 5 and 6. There is a
transducer 75 that sends its pulses downward toward a
reflector 76, which has its reflecting surface divided
into two halves making separate surfaces 77 and 78 as
indicated. In this case these surfaces have a slight
angular difference which creates the dispersal situation.
The transmitted beam of acoustic energy is indicated by
a dashed line 81, while the returning path is indicated
as a solid line 82. It will be observed that
the reflected energy from the wall 72 of the pipe
-- 10 -
10~7~
returns farther away from the axis of the transducer 75.
Figures 7 and 8 illustrate a reflector surface
structure for spreading the acoustic beam horizontally
and not vertically. In this case there is a trQnsducer
85 that has its beam directed onto a reflector 86. The
surface o~ reflector 86 has its face divided into two
halves 87 and 88. These are flat surfaces angled away
from one another in a convex manner a few degree~ in order
to spread the beam of acoustic energy horizontally as
indicated in figure 8.
Figures 9 and 10 illustrate one more embodiment
concerning the reflecting surface structure which may be
employed. Since the use of flat surfaces on the reflector
will cause a separation into two distinct ~eams, as
indicated in Figure 8, it may be preferable to employ a
curved convex surface. This i8 illustrated in ~igures
9 and 10 where there i8 illustrated a transducer 91 that
is directing its energy onto the reflecting surface of
a reflector 92 which like reflector 86 in Figures 7 and 8
is designed to spread the acoustic energy beam horizontally
but not vertically. In this case there is a curved sur-
face 93 that is continuously curved in a convex manner to
form the spreading affect. However, it will be u~derstood
that surface 93 is a oylindrical curved surface with a
straight center line situated at 45 degrees, as indicated
in the elevation view of Figure 9.
Figures 11, 12 and 13 illustrate a system for
providlng a multiple tranæducer arrangement that is
particularly useful in an instrument for determin~ng wall
thickness of a pipe Such an instrument is especially
useful for surveying a pipe line.
..
10676~
As schematically illustrated in Figures 11 and 12,
there is an instrument body 101 that has the diameter
designed for a proper size to pass through a pipe 102.
In order to make a continuous measurement, or survey
of the wall thickness of the pipe 102 there is a plurality
of pulse generating transduceræ 105 that are located
circumferentially situated on the body 101 of the in- .
strument. These transducers 105 have the axis of each
oriented parallel to the axis of the pipe 102.
The pulseæ generated by each of the transdueers 105
are directed down (as illustrated) parallel to the axis
of the pipe 102. Consequently, these pulses impinge
upon a first annular reflecting surface 106 that has an
angle of 45 degrees relative to the pipe axis. This sur-
face 106 extends radially about half of the width of the
pulse beams which are transmitted from the transducers
105~ so that a portion of each of the pulses are split
off and reflected by the surface 106 to travel over a
path 109 in each case. It will be understood that these
paths are transverse to the surfaces of the pipe wall 102.
Situated axially farther away from the transducers
105, there is another annular reflecting surface 112 that
haæ its surface at a angle which is greater than 45 degrees
relative to the axis of the pipe 102. This surface 112
extends radially far enough to encompaæs the other half
of the pulse ray paths of acoustic energy from transducers
105 so that this portion which is 3~1it off of the
original pulæes~ will be reflected along somewhat up-
wardly directed paths 113. Each of these paths 113 has
an angle of inciden~e relative to the wall of the pipe
102, that is greater than the critical angle of refraction
- 12 -
. I .
1067~1~
so that the pulses of acoustic energy will enter the
material of the wall of pipe 102. Then, as explained
previously in connection with Figure 1 and 2, the energy
from the pulses travelling o~er paths 113 will be reflected
back over the paths 113 if they encounter a dlscontinuity
such as a cavity 116, illustrated. Such reflected
cnergies will be returned to the transducers 105 where
electrical signals will be generated.
By having the instrument 101 constructed with a
plurality of transducers, in the manner illustrated in
Figure 11, 12, and 13 the instrument will be especially
suitable for being employed in the survey of a pipeline.
In order to employ the instrument 101 to survey
the walls of the pipe 102, the transducers 105 will be
pulsed sequently around the circumference of t~e
instrument 101. Consequently, as the instrument travels
through the pipe 102 there will be a continuous scanning
of the walls of the pipe which may, of course, be part
of a pipeline. A schematic circuit arrangement for
accomplishlng this iæ illustrated in Figure 13, where
it will be observed that each of the transducers 105
includes a piezo-electric crystal llg that has electrodes
120 and 121 associated therewith for applying a voltage
pulse which will generate the acoustic pulse. mereafter,
the crystal 119 and electrodes 120 and 121 will act
inversly to generate an electrical signal, in each case,
as the reflected return pulse is detected.
Still referring to Figure 13, it will be under-
stood that in each case the pulse generating circuit
includes a circuit connector 123 that leads from the
electrode 120 to a co~trolled electrode 124 of a
- 13 -
.
10f~7~
silicon controlled rectifier 25. The SCR 125 acts to
pass a voltage pulse from a charged capacitor 128 to
the crystal 119 whenever it is triggered by a signal
applied to a circuit 129 to trip the SCR into conduction.
The capacitor 128 is maintain charged by a relatively
high DC potenti&l, which is maintained at a terminal 132,
with a resistor 133 between the terminal 132 and the high
potential plate of capacitor 128.
It will be understood that throughout this
specification wherever the abbreviatlon SCR is employed
it stands for sillcon controlled rectifier. Such
abbreviation is well known to one skilled in the
electronic arts.
Referring to the pulse generating circuit of
Figure 13 again, it will be noted that there is a common
ground circuit 136 that is connected to the other
electrode 121 of the crystal 119. Also, the circuit
136 has one side of a variable inductor 137, as well as
one end of a resistor 138 connected thereto. In addltion,
there is a resistor 141 that is in the control circuit
129. The control circuit goes via a resistor 141 from
the output of an amplifier 142. And, the output of a
selector circuit 145 goes to the input of the amplifier
142 over a circuit connection 146.
It will be understood that after each acoustic
pulse is transmitted, a sufficient period of time is
allowed before the next one so as to permit the reflect-
ed pulses travelling over the transverse paths 109,
~n addition to those which may return over the angled
path 113 to reach the crystal 119 of the transducer.
Also, it will be understood that a circuit connection
- 14 -
10~7~
151, which goes to an amplifier (not shown), will carry
the electrical signals generated by the crystal 119 to
such amplifier, from which they may go to an oscilloscope
(not shown) or otherwise be used to develop oscillograph
signals like those illustrated in Figure 2.
It will appreciated that there is an individual
control and reflected-pulse amplifier circuit for each
of the transducers 105. This is indicated ln Figure 13
where there are rectangles 154 and 155 which represent
additional circuits like that described above in connect-
ion with the crystal 119. There will, of course, be one
such circuit for each of the transducers 105.
It will also be understood that the time elements
involved in sending and receiving individual acoustic
pulses and reflected return pulses, are relatively short
as indicated above in connection with Figure 2. Con
quently, a complete scan made be carried out around all
of the transducers 105 rapidly enough to provide adequate
testing of pipe wall conditions along a pipeline where a
normal speed of travel of the instru~lent through the
pipeline is maintained.
Method Steps
A method according to this invention may be
carried out by ~arious and different types of apparatus
which are not necessarily equivalent to one another. The
method relatesg in general~ to the field of pulse-echo
measurement for testing of wall thicknesses, and the
following steps should not be considered as limiting
the invention nor are they necessarily always carried
out in the order recited.
- 15 -
.
. .,: : , , . ... ' ' . . ~ -
10~;7~1~
A first step is that of transmitting a pulse
of energy along a path which is transverse to the wall
being measured. This is illustrated by the path 33
indicated in Figure 1. The pulse to be transmitted will
be one generated by the transducer 2~ half of which is
reflected from the surface 30 to change direction by 90
degrees and so travel toward the wall 21 in a direction
transverse thereto.
Another step is that of receiving reflected
pulses after reflection of the transmitted pulse from
the surfaces of the wall being measured, in order to
determine the thickness of the wall. mis is illustrated
in Figure 1 by the paths 33 and 29 over which the return-
ing reflected acoustic energy pulses will go ln travelling
back to impinge upon the crystal 25. The crystal generates
signals representative of suchrpulses. This is indic~ted
in Figure 2 where the time difference between the return-
ing or reflected pulses, e.g. pulses 41 and 42, will be
a measure of the thickness of the wall 21.
Another step is that of transmitting a separate
pul~e of energy along a path having a angle of incidence,
relative to the wall being measured, that is greater
than the critical angle of refraction for the material
of that wall. This is carried out by the splitting
of the acou~tic pulse which is transmitted from the
transducer 25. The half that is thus split off is re-
flected from the surface 45, and then travels over the
path 47. This energy pulse will penetrate the wall 21
of the pipe, and if it encounters a discontinuity such
as the cavity 51 illustrated, there will be a reflected
pulse returned along the same path 47 and back along path
- 16 -
~;7~
46 to the transducer 25.
Then a final step ~s that of receiving the :~
other pulse reflection, if it appears~ ln order to
characterize the presence of an anormaly. mis i8
lllustrated by the delayed reflected pulse e.g. pulse
57 of Figure 2, which returns along paths 47 and 46
to the transducer 25. The arrival will be at a time
spaced from the first reflected pulses since the length
of the paths of travel is greater.
The method may also be described in a more
comprehensi~e manner which relates to the multiple
transducer arrangement such as that illustrated in
Figureæ lla 12 and 13. The steps of such method include
the following.
First, the step of generating a plurality of
single pulses sequentially, as described in connection
with Figure 13. These are directed along paths that
are parallel to the axis of the pipe. Thus, in the
Figure 11 one of the transducers 105 is shown mounted
on the body of an instrument 101. It is oriented so
as to direct the pulses in a path parallel,to the axis
of the pipe 102. Generation of each pulse may be carried
out as described in connection with Figure 13. The
triggering of an SCR 125 will create a discharge of the
capacitor 128 across a p~th which includes the crystal
119. Consequently, an electrical pulse iæ applied, - `
which creates the piezoelectric af~ect so as to produce
an acoustic pulse output.
Then there are the steps of reflecting a
portion of each of the single pulses from a 45 degree
surface to direct them along paths transverse to the
- 17 -
. ~
..
. . . - . . . .
1067~
pipe wall. Figures 11 and 12 illustrate this step in
that the pulse energy travelling down from transducers 105
is reflected by surface 106 and then goes in a direction
transverse to the wall iO2 of the pipe, i.e. over the
path 109.
Another step Is that of reflecting another
portion o~ each single pulse from a surface more than 45
degree relative to said parallel path. This step is
illustrated in Figures 11 and 12 by the path 113, which is
the path of travel for that portion of the pulse travelling
down from transducers 105 that is reflected from the sur-
face 112.
Another step is that of receiving reflected
pulses which are returned from the transverse directed
energy, in order to determine the thickness of the pipe
wall. mis is carried out by the conventional circuit
controls for having the crystal 119 (of the transducer
which has ~ust transmitted its single pulse) connected
in a receiver circuit for amplifying generated electrical
signals that are caused when the crystal is deformed by
the reflected pulse energy returning to it. The earliest
reflected pulses returning wlll be those over the trans-
verse paths 109 and these provide an indication of the
thickness of the pipe wall 102.
Finally there is the step of receiving any
reflected pulses which return from the other portion of
the split single pulse that was transmitted from the
transducer in order to determine whether there is a
discontinuity of the pipe of the type which will reflect
and return some of this angular energy. This is carried
out by maintaining the crystal circuit available for
ampli~ying any reflected energy long enough to have a
- 18 -
1C)~7tj14
signal generated if the delayed returned reflected pulse
exists.
While the foregoing preferred emboiments
the invention have been described above in considerable
detail, in accordance with the applieable statutes, this
is not to be taken as in any way limiting the invention
but merely as being descriptive thereof.
- 19 - . ,
