evaluating systematic errors in quadrant scales
TRANSCRIPT
E V A L U A T I N G S Y S T E M A T I C ERRORS IN Q U A D R A N T S C A L E S
A. L. S h n e i d e r m a n UDC 681.26.088
In making the balancing and recording of weighed articles automatic and using remote-measur ing systems it becomes often necessary to evaluate systematic measurement errors.
Let us assume we are dealing with a weighing device (Fig. 1) which consists of a quadrant balancing mecha- nism I with lever system 2 at tached to it. An in te rmedia te lever carries inductive transducer 3 used for remote e l ec -
t r i ca l measurements with an output signal proportional to the displacement of the transducer plunger.
Taking into consideration that commerc i a l models of secondary remote-measur ing instruments have a uniform scale, it is obvious that they wil l produce a systematic measurement error.
Let us determine the maximum value of the error contributed to the system by remote measurements. By de- noting the current value of the measured quantity with P and the corresponding displacement of the plunger with y,
let us find relationship P = fly).
The quadrant 's condit ion of balance has the form of
Pl o cos r = ORo sin (qo + %). (1)
where ~ is the current va lue of the quadrant 's rotation angle, R 0 is the distance from the bearing of the quadrant to its center of gravity, ~o 0 is the angle of the quadrant 's in i t i a l position, G is the weight of the quadrant, l 0 is the weight-
carrying arm of the quadrant.
From (1) we obtain
6Ro P = -io - ( tan ~p cos % + sin %). (2)
By adopting notations
6Ro A ~ ~oo COS % ,
B = 6RQ sin %, lo
(3)
we obtain
sin q~
P = a V i - - sir,~ ~ ~ S . (4)
The current value of the plunger 's displacement in transducer 3 (Fig. 1) is represented by relationship
l~ (5) y =
where y~ is the displacement of the weight-bear ing lever 's knife edge for a rotation of the quadrant through angle ~o.
Taking into consideration that y~o = I0 sin 9 we obtain:
y -~- k sin r (6)
where
l_3_~ k = l0 l l �9
Translated from Izmer i t e l ' naya Tekhnika, No, i, pp, 24-26, January, 1967, Original ar t ic le submitted
October 15, 1965.
32
, ~ , ' : ~ !
<1 , / - 7-2 ~ '~-~ ,
' ~ D r a w bar of the weight-
bearing mechanism
Fig. 1.
H
i ma~
' ~Pma~
--252 f i +2. 2 -# ' r am
Fig. 2.
and for P = Pmax we have
By substituting in (4) for the sine its value from (6) we obtain:
A y P - - k .~ / y2 + B . (7)
V l--k- 7
The above relationship shows that the measured quan- t i ty has a nonlinear relationship to the displacement of the
transducer's plunger. Therefore, in observing or recording the readings on a uniform scale of a secondary instrument
we obtain a systematic error.
Let us i l h s t r a t e the method of evaluat ing this error on par t icular examples.
Let us first note the following. Coeff icient A and B in
equation (7) incorporate quadrant parameters G and R 0 which in prac t ice are not always avai lab le for c o m p u t a t i o n . There-
fore, it is advisable to express these values by the max imum effort P -- Pmax on the quadrant, which is normal ly known.
Let us derive a balance equation for the quadrant in
two of its positions, namely for P = I/2Pma x we obtain
I -~- Pmax lo = GRo sin q)o, (8)
Pmax lo ---- GRo {.tom qemaxCOS q~o q sin ~Po).
By comparing (8) and (9) and taking into consideration (3) it is easy to find that
1 1 I , 8 = P ax' A = ] - Pmax
tanqamax 2 l , *
(9)
(10)
By substituting in (7) for A and B their values thus found we obtain
P = -2- max ktanq)ma x " g=' " (11)
By assuming for a numer ica l computat ion that ~max = 23~ l0 = 23 ram, l , = 80 ram, g2 = 23.3 mm we find k = 6.7 and tan ~Oma x = 0.424.
By substituting in (11) for these quantities their values given above we have
_1 p ( 1 + 0 . 3 5 2 Y ) P = 2 max V l -o .0223y~ " (12)
The graph of this function is shown in Fig. 2 (curve 1). It is known that the mechanism of a quadrant with the a t tached lever system is assembled and adjusted so that in the middle position of the indicat ing instrument 's pointer the lever knife edges are in a horizontal plane. Thus, the measured weight value Pmax is made to correspond to a plunger displacement Ymax which according to (6) is equal to Ymax = 6 .7 .0 .391 m m = 2.62 ram.
In order to evaluate the nonlineari ty of the sys temat ic-er ror function (12) it is necessary to compare it with the l inear relationship which corresponds to straight l ine2 in Fig. 2 and has the form of:
1 Plin = 2 - Pmax -I ag. (13)
Coeff icient a,which represents the slope angle of the straight l ine ,can be evaluated from conditions y = Ymax = 2.62 mm and Plin = Pmax as a = 0.382. I/2 Pmax'
33
Finally (13) assumes the form of
I P l i n = ~ " Pmax (1 + 0.382 g). (14)
The difference between (12) and (14) represents the systematic error's current value 6 (y) = P -P l in , i .e. ,
I ( Y - - 0 . 3 8 2 y ) . (15) 8(y) ~- -~- Pmax 0 .352V1 _ 0.0223y 2
it is advisable to evaluate the systematic error in the middle of the interval, since it is difficult to find analyt- ically the extremal value of 5 (y) and the corresponding value of the independent variable y,
By substituting in (12) and (14) for y and Ymax their values 0.5 and 1,81 mm we obtain respectively:
1 1 P = -~- Pmax" 1,47U20, Plin = - 2 Pmax" 1.50042.
From the above we find the maximum measurement error on a linear scale of a secondary instrument to be equal to
A ~ 2 % .
34