calibrate
surface
plates
in the plant
October 1955 The Tool Engineer
By J.C. Moody
1d Electrical Standarde Dept.
Albuquerque, Ne M.
Swce seasonsannrs ee no snore slab than
the surface pate on which they are referenced, ii
important 1s know exaciy the sora ofthe plate
Being wed. Surface plats are manufactred to
tccaraie varying from 0:02 to 0.005 inch of
Geviatin fom a tre plane, The user should chock
ech plats ater tiled 4 determine winter
it mes wpecfaions and rom tne tte there
ther to lea the eet of weet and envirnmnt
Tho dock monomer info. pstiol por
pes be done inthe work environment
predicable method of osurlely
calling surface platen in ovllble to ndutry.
‘Te methed used in the mavalogy labortory a
Sandia Corp. is highly accurate yo can be per
formed by semiilled personne uring istruments
set toany nde hod
is an application and extension of procedures de
Neloped by K. J. Hume (British metrologit) and
deal, th calibeton choo. ba. ecirm ka
2 room in which the temperature of the pate can
be kept in ogulibrium and from ‘which thermal
carrots can be exloded. Hower inary ue
jal laboratory. This 116
Fig. 2. Mirror mount, reflector mounts ar
‘surface plates under conditions that are less than
ideal. These plates can be satisfactorily calibrated
under the same conditions. Extremes of tempera-
ture changes, thermal currents, and vibration are
obviously to be avoided.
Calibration Method: Equipment needed for
this method is shown in Fig. 1. The autocollimator
jis exsentially an optical lens system from which
parallel rays are emitted. ‘These rays strike the
surface of a steel optical at and are reflected back
into the autocollimator. The reflected rays produce
image at the focal plane of the autocollimator
Ja angular displacements can be accurate:
ly determined.
‘The reflector is mounted on a bracket, the sup:
port pads of which are separated by # distance
Fig. 3. Positions of autocollimator daring readings
for the eight principal lines,
Eos!
optical las used for calibrating surface plates.
arbitrarily chosen so that it will divide evenly into
the dimensions of the surface plate. This distance
should be about 8 percent ofthe length of the short
side of the plate. Various reflector mounts needed
for different size plates are shown in Fig. 2. The
mirror mounting bracket is so designed that the
mirror mounting is normal to the surface plate.
Steel optical flats, the faces of which do not deviate
from a plane by more than 0,000003 inch, are
used for both the mirror and reflector. In addition,
4 straight edge, graduated in increments equal to
the distance between the support pads of the re
flector stand, is used,
‘A total of eight lines of readings are taken: four
perimeter lines, two diagonal Tines, and two center
lines, Fig. 3. The perimeter lines are laid out one
increment from the edge of the plate. The precise
stations at which readings along all eight lines are
taken are measured off in steps equal to the incre:
ments on the straight edge. Many more readings
could be taken, but « reasonable compromise be
tween accuracy and economy is achieved by this
method.
Detailed instructions for using an autocollimator
are supplied with the instrument and should be
studied carefully. The position of the autocoli-
rmator for each line of readings is shown in Fig. 3.
‘The readings along the north perimeter line, for
‘example, are taken with the autocollimator in the
southwest corner of the plate. The reflector is
moved along the line and readings are taken at
each station. These readings are entered directly
‘on the properly identified work sheet, Fig. 4 After
cach line is completed, the reflector is moved back
to the first station on that line and another reading
taken. If this does not agree within + 0.3 sec. of
‘The Tool EngineerFig. 4. Diagram showing relationship of
profile of northwest to southeast diagonal
{o line of first reading and datum plane.
Datum plane
3 7 Il 15 19 23 27 31 35 39 43 47 51 55 59 6367 71 75 79 83
are with the first reading taken at the same station,
the operation must be repeated.
Autocollimator readings entered on the work
sheet show only angular displacement in tenths of
a second of are in relation to the line of first read-
ing. To be readily meaningful, these must be con-
verted to linear deviations from a base plane. The
procedure for these conversions and their presenta-
tion will be discussed in the following paragraphs.
‘Though care must be exercised in each step, the
conversion is not a formidable task. Readings ean
be made in about two hours; an intelligent, proper:
ly instructed clerk can reduce the data to an ac:
curate profile of the surface plate in an equal time.
implified Data Conversion Procedure:
Directions for converting. autocellimator readings
into linear displacement in hundred-thousandths of
an inch are given without any attempt at theoretical
justification. One line of each class will be ex-
plained in detail. The reader who is interested in
the theoretical considerations should read K. J
Hume's Engineering Mesrology or the author's
paper, The Metrology of Surface Plates, copies of
‘which are available upon request
‘The person who is to reduce the data is given a
work sheet for each of the eight lines, TABLE 1
On these, the stations at which readings were taken
fare indicated in Column 1 in terms of inches from
the edge of the plate in the direction in which the
line was read, The autocollimator readings are
centered in Column 2. No reading is entered for the
firs station on the line,
Convension FoR DiscoNALs: To process the
Northwest to Southeast diagonal line data, the fol-
lowing steps should be taken in order:
October 1955,
1. Convert the sutocllimator readings into angular
displacement by determining the amount by which
tach value in Column 2 is greater or less than the
first value in Column 2. Do this at each station:
enter the result in Column 3, paying
the sign
2. New, determine the algebraic sum of the at
displacement at each station and enter this valve
jn Column 4. To do this, add the values in Column
3 down to and incleding each station.
Divide the last value in Column 4 by the total num:
her minus one of the stations on the line to deter-
mine the correction factor. (in the example shown
there are 21 stations. Hence, ~280/20 = —14.)
4 Set up an arithmetic progression in Column 5. Re-
verse the sign of the value in Column 4 opperit
the midstaton and enter it opposite the same sta
in Column 5. Working up Colun
dstation, add the correction factor cusmuatively
at each station. Return to the midstation and sub:
tact the correction factor cumulatively at each ta-
tion to the bottom of the column, The resulting
arithmetical progression is the cums
tion factor for each station,
ion, add algebraically the
5 from the
Columa 6. (The datum plane, Fig. is that plane
in which the center point of the rarface plate lee
parallel 10 the lines containing the end
of each diagonal.) Proceed
‘completed, enmpata
enter lines must be catried to this point. Do the
Perimeter lines next.
Converston ror Norrit Periwerer Ling:
1. Proceed exactly as with the diagonals through
Column 4 foreach of the perimeter lines
2 Prepare a chart of the surface plate ae illustrated
in Fig. 5. Enter the physical center, 0, and the
values found for the ends of the
"‘Table 1—Work Sheets for Calibrating a 48 x 78-Inch Surface Plate*
7 Z z z ¢
emg has at jaagael Tsai | mes
Se fie Sadie ce. aetna en
Commitee a = ae
Paes nei) cmea aetna ae Oo
Taga Noe San
= z 5
& j ‘
e 3 ‘
z 2 t
& fs A
5 4 3
3 - x
3 ‘
g s
: f
3 k
3 3
3 i
i =
> ‘
x = 3
5 = $
3 = 5 i
¢ 3 z ,
f 3 3 i
> 3 z :
5 3 z i
z s t :
‘ ey A 7
s = 3
= # : ‘
z te 3 %
» ‘ 5 «
3 = fe u
x at $ t
& 3 : 3
& = z :
a Faia iw Ra Sate
7 = = = a = z 7
: = = = ‘4 A
a i a 3 H i
‘ 3 eH = ; i
$ z = Ee i ‘
2 3 os —« 8 2
“ 8 =a a 2 6
‘The Tool Engineer‘Table 1 (Continued) *
0
7 2 3 + 3 . 7 °
font orlnne man ae
(i wolfe sein
rani iiae Tiron
ot, dk tate
South Prince Line Eat to West
a ‘s x
te ue o &
= = nal ™
x a =i ts oi 3
a & = Bote =
* ve ae tie (3 * *
® ‘s = im i x :
* a8 he a ‘ 3
3 8 ty Hee ey 3 3
= a ee 2 "
5 St ae ine 2 t
a =x an Oe z ‘
& 3 = iz 8 3 3
hw Sth
7 zn = e = me = A
2 fe a a = 3 2 ct
% $ a 3 aH = a ‘
» « a a a 3 ® :
x 5 = a an 3 a :
3 3 a =F 7 = : ;
2 > = = 4 = : '
x 3 : 3 th a :
: & : 3 te s :
% s = s ue a 1g
Cente Une East West
5 es == ae 7
Y +3 a 3 op <
‘ uM = =a. oe :
» tn ws tog "
x pe = =e %
3 Bes 3 ae 3
2 Be = og 3
3 aS ‘
3 OS u
& ad :
e 0 1 3 &
7 = >
: a z
= « °
x é &
a $ "
x
e
*
+
a
He
TAIT vateee exept Colne 1
October 1955
1920
Conversion For East ro West Cexten Live:
Column 6 in the work sheets a8 shown. This ch
ie important as without it there i danger of eon-
Sasing the figures
3. Enter the value for the NE ead of the NESW di
agonal in Colurnne 5 and 6 opposite the frst station.
Enter the value of the NW end of the NW-SE di
agonal opposite the last station in Column 6 on
44 Next in! the correction factor. Subtract the value
fopperite the last station in Column 4 from the value
‘opposite the same station in Column 6 [—36
199) = 163]. Enter this value opposite the
last station in Column 5, Subtract this value from
that opposite the frst station in Column 5 (—53
=I6d = —216) and divide the result by the total
number of stations on the Tine minus one (—216/
16 = —135}.. The result is the correction factor.
5. Beginning atthe lest station in Column 5, add the
re
tach station. (Since the correction factor in the
example is 135 to avoid decimals —13 and —14
ace used altermately.)
6. To find the anguler displacement from the datum
plane, algebraically add the values opposite each
Station in Columns 4 and 5 and enter the results in
Column 6.
Complete the conversion for each of the perimeter
Tines to this point and enter the values atthe mid>
points in Fig. 8. Now proceed with the center line
1. Carry the conversion through Column 4 ‘The proce
dre for the center lines is exactly the same as for
the diagonal and perimeter lines to this point,
2, From Fig. 5, enter the value for the midpoint of the
fast perimeter Tine opposite the fist station in
Columns 5 and 6. Enter the valve for the mid-
int of the west perimeter Tine opposite the last
Station in Column 6 ony.
4 Subtract the value opposite the last station in
Column 4 from the valie opposite the same sta:
tion in Column 6 and enter this value at the last
station in Column 5.
4, Subtract the last value ia Calumn 5 from the
and divide the result by th total number of stations
‘onthe line minus one, The result isthe correction
factor
5. Beginning at the last station in Column 5, add the
correction factor up the Column in an arithmetic
progression to find the cumulative correction factor
for each station.
6. At each sation, algebraically ad the values in
Columns 4 and 3 and enter the result jn Column 6.
‘This i the angular displacement frofh the datum
plane.
1, Change the sign ofthe value opposite the
{in Colin 6 aid add it to the value opposite each
sation in Column 6, Enter the sum in Columa 6a.
‘A word of explanation is necessary at this point.
‘The center line check is the criterion of accuracy
for the entire operation. The value at the point
at which the center lines and diagonals intersect is
physically zero, If everything were done perfectly,
the value opposite the center station would be zero.
But this is not possible because each slight error in
reading the instrument is reflected at the midstation
of the center lines. If the magnitude of 1
ation may be regarded as
‘satisfactory; job must be done over.
“Column 6a, which appears only in the work
sheets for the center lines, is used to move the error
away from the center, which is known to be zero,
‘out to the perimeter.
Final Steps in Conversion: The work sheets
for the eight lines are now completed through
Column 6, inckuding Column 6a for the center lines.
‘The procedure for Columns 7 and & is identical
for all work sheets and must be done together.
1. Seach through Column 6 for all work sheets, Column
Gn of the center lines, for the lowest value in all
fof the 8 enlomns. Add this value to the value
‘opposite each station in Column 6 (6a for the center
Tines) end enter the sums in Column 7. This
The Tool Engineer[rates Cater]
oo N24 Se
" 15
wo
se eoRremunne
is
wae
3 6 e138 7 7 Oe
wo7 63
a,
° 2
' Fig. 6. Linear deviations
from base plane indicated
° slong. the ight principal,
peace ata ines of x 48. x" T2inch
2
8
0
Banos
Figure is the displacement from the true base plane
This is defined as that plane containing the point of
lowest reading and parallel to the datum plane
2. Next, convert the values in Column 7 to linear
values in 000001 inch. To do this, multiply the
of the sine of 1 second (0.000005) times
enter lines of the mi
example) (0.000005 % 4
(0.00002) by the values in Column 7 at each ata
the values in Cal tenths of
4 second of are, itis necessary to divide the produet
at each station by 10 to got the decimal point in the
right place. Round out the anewer to the nearest
indredthousandth af an inch, drop out the des
mal point and enter the value in Calum 8.
Ociober 1955
Conclusion: ‘The data from Column 8, when
reported on the form shown in Fig. 6, allows the
user to see at a glance the features of the surface
plate he is using. It shows not only the extremes
of variation, but also the est areas on the plate,
Acknowledgments: ‘The author wishes to ac:
knowledge the technical assistance given to
work by J. M. Bunch, physical standards
Sandia Corp. and A. R. Darling, formerly of Sandia
Corp. Appreciation is also expressed for the e'
torial efforts of W. F. Carstens, reports and et
torial division, Sandia Corp.
a