See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/235699092

Calibration of nonautomatic weighing instruments

Article · January 2007

CITATIONS READS

2 4,873

1 author:

Adriana Vâlcu
Institutul National de Metrologie
51 PUBLICATIONS   66 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

PT in the field of verification of NAWI View project

All content following this page was uploaded by Adriana Vâlcu on 24 November 2014.

The user has requested enhancement of the downloaded file.

 XVIII IMEKO WORLD CONGRESS
Metrology for a Sustainable Development
September, 17 – 22, 2006, Rio de Janeiro, Brazil

Calibration of nonautomatic weighing instruments

Author 1: Adriana Vâlcu
1
National Institute of Metrology, Bucharest, Romania, adriana.valcu@inm.ro

Abstract: This paper describes the metrological - adjusting the error close to Max, using internal or
requirements and the measurements methods for calibrating adequate external weights, to allow compensation for
nonautomatic weighing instruments. It also may be a useful changing environmental factors such as temperature
guideline for operators working in calibration laboratories and, implicitly, air density.
accredited in various fields. This paper is also intended to Estimating the expanded uncertainty is based on:
enable metrology laboratories to prove a given expanded - repeatability
uncertainty using suitable procedures. Two examples of - resolution
calibration methods are given: one for electronic weighing - eccentricity
instruments and another for the mechanical instruments. The - the influence of temperature variations at the site of use
described methods include information regarding the - accuracy measurements
measurements standards used for calibration, environmental - the standards weights used in the accuracy
conditions, calibration procedures and estimation of the measurements
measurement uncertainty. - hysteresis
Keywords: electronic weighing instruments, mechanical
1. Repeatability
weighing instruments, multiple ranges and multi-interval
instruments. At least ten repeated measurements must be
performed. This test should be done at or near the nominal
Introduction maximum capacity of the weighing instrument or using the
largest load generally weighed in applications. In the case
Calibration methods and the evaluation of the
of a zero deviation between the weighings, the instrument
uncertainty described in the paper are in compliance with
shall be reset to zero, without determining the error of the
the OIML Recommendation OIML R 76-1 and ISO-GUM
zero indication [1]. The uncertainty due to repeatability of
[3] (Guide to the expression of uncertainty in
the weighing process, uw, is given by standard deviation s of
measurements-1995). The combined standard uncertainty
several weighing results obtained for the same load under
results from both the type A and type B evaluations of the
the same conditions. For multiple range instruments, this
measurements uncertainty.
test shall be carried out for each range used, thus for n
The environmental working conditions shall be suitable
measurements:
for the instrument to be calibrated. The room where the
n 2
balances are installed must be temperature and humidity 
∑  I − I  (1)
i
uw = s = i =1
 
controlled. It is not allowed to place balances near
equipment that generates vibration or in a room where dust n −1
may affect them. Also, heat transmission by solar radiation where Ii is the indication of the weighing instrument and n
through the windows shall be prevented. is the number of repeated weighings:
1 n
A. Calibration of electronic balances I = ∑ Ii (2)
n i =1
Before calibration, an electronic weighing
instrument should be checked on site to make sure it 2. Resolution
functions adequately for the intended application. The For balances having the resolution d (equal to the
preliminary operations are: scale interval), the uncertainty of the rounding error, ur, for
- identification of the weighing instrument (type, model, each reading I is [2]:
serial number, etc.); d /2 d
ur = = (3)
- checking the leveling; 3 12
- ensuring that electrically powered instruments have
The uncertainty of the rounding effect for I ≠ 0 is given
been switched on for a period of a least one hour
by [6]
(preferably overnight) and have reached room
temperature; d I2= 0 + d I2= L
ur = (4)
- ensuring that the pan of the weighing instrument is 12
clean and in good condition; Equation (4) is used for single and multiple range
- pre-loading for several times the weighing instruments instruments. For multi-interval instruments, for the different
to near maximum capacity; scale intervals di, the uncertainty due to the rounding effect
is:
 d12 + d i2 where: δi is the maximum permissible error of the “i”
ur = (5) applied weights
12 • When the indications of the instruments are corrected
where: d1 is the smallest scale interval for the errors of the weights (the calibration weights are
di is the scale interval of the appropriate partial introduced as conventional values) the standard
range. uncertainty from the calibration certificate (ucert) should
3. Eccentricity be combined with the uncertainty due to the instability
of the mass of the reference weight (ustab) as following
It is preferable to use large weights instead of [2]:
several small weights. The load L shall be applied on the
pan in the positions indicated in Fig. 1 in a sequence of 2
center, front, left, back, right, or equivalent. U 
2
ucert
uref = + u stab
2
=   + u stab
2
(9)
After the first measurement, tare setting may be done when  
k
the instrument is loaded. For instrument having no more When two or more weights are used for L, the equation
than four points (n ≤ 4) of support, the test load is 1/3 Max. becomes:
[1].
 ∑U i 
2

3 4 3 4
Fig. 1
1 1
2 5 2 5
The load is first placed in position 1 and is
subsequently placed in the other 4 positions in an arbitrary
order. Acceptable solution for uncertainty due to
eccentricity, uex is estimated as follows [2]:
∆ rectangular distribution. This component would be
uex = (6)
2 3
here, ∆ is the largest difference between off-center and
central loading indications;
The eccentric test is not carried out in the case of weighing
instruments with a suspended load receptor.
4. The effect of temperature variations, during the
calibration, uT is calculated from:
1
uT = (∆ t ⋅ TK ⋅ 10 ) ⋅
−6
L (7)
12
where: TK is the effect of temperature on the mean
gradient of the characteristic in ppm/K (estimated or taken
from data information sheet).
∆t = tmax – tmin is temperature variation during the
calibration, for load L.
5. Mass standards
The weights used as measurement standards shall
comply with the specifications in the OIML R111. The
traceability of the standards to the SI unit shall be ensured.
The standards shall be adequately acclimatized before the
calibration (to minimize the effect of convection). A
thermometer kept inside the box with standard weights may
be helpful to check the temperature difference.
• When the indication of the instrument is not corrected
for the errors of the weights (the calibration weights are
introduced as nominal values) the uncertainty of the
reference weights, uref, is estimated as follows:
δi
u ref = (8)
3 the indication error, EI , is obtained from the difference
or, when two or more weights are used,
 
 k  + u stab
2
uref = (9’)
 
Ui (k=2) is the uncertainty of the applied weights from the
calibration certificate.
The calculation of the uncertainty associated with
the stability of the standard (ustab) has to take into account a
change in value between calibrations, assumed that a
equivalent to the change between calibrations divided by
3:
Dmax
ustab = (10)
3
where Dmax represents the drift determined from the
previous calibrations.
If previous calibration values are not available, the
uncertainty from the calibrated certificate is considered to be
an uncertainty associated to the drift.
6. Accuracy measurements
Weighing instruments should be calibrated
throughout their range. When a weighing instrument is used
only over a part of its capacity, the calibration may be
restricted to this part of the measuring range. In this case,
the part range that has been calibrated has to be explicitly
mentioned in the calibration certificate and also a label with
this information should be fixed to the weighing instrument.
Measurements are made at about five equal steps across
the range of the balance (zero, 0.25Max, 0.50Max, 0.75Max
and Max). If the balance is typically used for a particular
load, the accuracy of the scale around this load should be
measured.
• When :
- the weighing instrument was adjusted before
measurement,
- the density of weights is close to 8000 kg/m3 and
- the air density is close to 1,2 kg/m3,
between the instrument reading I - upon application of a

u ref =
∑δ i
(8’)
load L - and the value of this load (conventional mass value
or nominal value).
3
EI = I - L (11)
 The weights are applied in increasing and Starting from eq. (16’) and (16’’), the relative standard
decreasing loads. The estimation of uncertainty associated uncertainty associated to the buoyancy correction (uBC) may
with the indication error taking into account the influences be calculated as:
of repeatability (uw), resolution (ur), reference weights (uref), 1 1 2 u ρ2 u ρ2
2
u BC = u ρ2 ⋅ ( − ) + ( ρ a − ρ 0 ) 2 ⋅ w + u 2ρ ⋅ w (17)
temperature (uT) and hysteresis (uH). a ρ w ρc ρ w4 a ρw 4

Hysteresis occurs when a balance displays a

u ρ2a u ρ2 w
different reading for the same load, when the load is applied 2
u BC = + (ρ a − ρ 0 ) 2 ⋅ (18)
increasing the weight and decreasing the weight. ρ 2
w ρ w4
If the difference is δx, the standard uncertainty due to Uncertainty of air density u ρa is determined
hysteresis is given by [3]: according to [2]. When the air density is not measured and
δx the average air density for the site is used instead, the
uH = = 0,29 δx (12) uncertainty associated to the air density is estimated
12 (according to chapter C.6.3.4 in [2]) as:
The expression of uncertainty associated with the
u (ρ a ) =
0,12
determining the indication error is: [kg/m3] (19)
3
(u w ) 2 + (u r ) 2 + (u ref ) + (uT ) 2 ⋅ L2 + (u H ) 2
2
u(EI)= (13)
The expression of uncertainty associated with
The uncertainty uEI should be calculated for each indication error (when the buoyancy correction is applied)
value of the load used. is:
From eq. (13), the relative standard uncertainty can u(EI)= (20)
(u w ) 2 + (u r ) 2 + (u ref ) + (u T ) 2 ⋅ L2 + (u H ) 2 + (u BC ) 2 ⋅ L2
2

be calculated as:
u(EI) rel= u(EI) / L (14) The uncertainty u(EI) should be calculated for each value of
the load used.
and the largest uncertainty u(EI) rel max is taken into account From eq. (20), the relative standard uncertainty can be
for further calculations. The indication error is not the same calculated as:
when the weighing instruments are calibrated using standard u(EI) rel= u(EI) / L (21)
weights under different conditions. In this case, the error of and the largest uncertainty u (EI) rel max is taken into account.
indication is:
E= I – (L+BC)= I – L – BC= Uncertainty of measurement for the weighing
instrument
  ρ a − ρ a 
I − L − − L ⋅ (ρ a − ρ 0 ) ⋅ (
1 1
= − )+  =
adj

 ρ ρ ρc The influences of the repeatability and of the

 w c  
rounding error are assumed to be independent from the load
 ρa − ρa  applied, while all the other components are proportional to
= I − L + L ⋅ (ρ a − ρ 0 )⋅ (
1 1
− )+  (15)
adj
the weight values. The standard uncertainties corresponding
 ρw ρc ρc  to the components that are proportional to the weight values
Where BC, the buoyancy correction is equal to [8]: are expressed as relative uncertainties. The combined
 ρa − ρa 
standard uncertainty uc is based on the parameters described
BC = − L ⋅ (ρ a − ρ 0 )⋅ (
1 1
− )+  (16) above (which can be grouped to obtain a simplified
adj

 ρw
where: ρw= density of the weight
ρa
= density of the air during the calibration
= 1,2 kg/m3 is the reference density of the air
ρ0
ρc
= reference (conventional) density of the
adjustment weight equal to 8000 kg/m3
ρa adj = air density at the time of adjustment.
Provided
- the instrument has been adjusted immediately
before the calibration , ρa adj = ρa , the buoyancy
correction may be calculated as:
BC = − L ⋅ (ρ a − ρ 0 )⋅ (
- the instrument has been adjusted independent of
calibration (ρa adj is unknown) the buoyancy
correction may be calculated as:
ρa − ρ0
BC = − L ⋅
ρw
with the next assumptions: ρa adj= ρ0 and ρw= ρc
ρc ρc  expression that would better reflect the fact that some of the
terms are independent from the applied load, while others
are proportional to the weight value) [5]:
uc = α + β ⋅ L (22)
 When corrections are applied to the error of indication
of the weighing instrument, the expression for
combined standard uncertainty uc is:
uc = u w 2 + u r 2 + L2 ⋅ (u exrel
2 2
+ uTrel + u (2EI ) rel max )
= u w 2 + u r 2 + L ⋅ u exrel
2 2
+ uTrel + u (2EI ) rel max (23)
1 1
− ) (16’)
ρw ρc
uTrel from eq. (23) is calculated by replacing the temperature
variation during calibration (from eq.7) with the actual
temperature variation recorded during the use of the balance.
 When no corrections are applied to the error of
indication of the weighing instrument, the largest
(16’’) relative indication error across the range that is
measured EI rel (Max) should be added to uc, in addition to
u(EI)rel(max), as follows:
uc= u w 2 + u r 2 + L2 ⋅ [u exrel
2 2 + (u
+ uTrel 2 L1and L2 are weights with nominal masses equal to the
( EI ) rel max + E I rel ( Max ) ) ]
minimum capacity of the balance.
L3 and L4 are weights with nominal masses equal to the
= u w 2 + u r 2 + L ⋅ u exrel
2 2
+ uTrel + (u ( EI ) + E I rel ( Max) ) 2 (24)
rel ( Max) maximum capacity of the balance.
The expanded uncertainty for k=2 is The following tests and calculations should be carried out on
a regular basis and are essential to the routine operation of
U = k·uc (25) the balance [7]:
B. Calibration of mechanical balances 1. Determining scale interval while the balance is loaded
with minimum capacity.
B1. Two pan balances: the balances with two pans dmin=m sw1/ |P4-P3| (28)
and three knife edges are also known as equal-arm balances 2. Determining scale interval while the balance is loaded
because the knife edges supporting the pans are nominally with maximum capacity:
equidistant from the central knife edge. The three knife- dmax=m sw2 / |P8-P7| (29)
edges are parallel and lie in the same horizontal plane. 3. Determining repeatability while the balance is no
Two-pan balances are generally undamped with a loaded and loaded with maximum capacity (by
“rest point” being calculated from a series of “turning determining the experimental standard deviations):
points”. Some balances incorporate a damping mechanism n n
(usually mechanical or magnetic) to allow the direct reading
of a “rest point”. In all cases, the reading in terms of scale
∑i =1
( Poimed − Poi ) 2 ∑ (∆
i =1
imed − ∆ i )
2

R0 = ; Rmax = (30)
units needs to be converted into a measured mass difference. n −1 n −1
A two-pan balance undamped is used less
where: P0i are the equilibrium positions while the balance is
frequently mainly due to the amount of time needed to make not loaded
a weighing compared with electronic balances. P0i med are the mean of equilibrium positions while the
There are two methods to calculate the ”rest point” (the balance is not loaded
equilibrium positions) “P” for balances (the accuracy of the ∆i are the differences between equilibrium positions
second is better than that of the first one): of the balance when it is not loaded and when it is
P = (e1+ 2e2+ e3)/4 or (26) loaded with maximum capacity
P = (e1+ 3e2+3e3+e4)/8 (27) ∆i med is the mean of ∆i differences
where e1…e4 are consecutive readings at the extremity of 4. Determining errors due to the fact that the two arms of
the swing of the pointer, i.e. where it changes its direction of the balance are not equal in length (this test is not
motion. applicable to balances with a single pan, case B2,
A calibration procedure, [7], is shown in the table 1. section to be discussed in).
Table 1. Calibration procedure for two pan balances P2 + P3 P1 + P5
Jmin = − (31)
No Loads applied Readings Equilibrium Difference
2 2
on receivers e1 e2 e3 position ∆i
div div div P6 + P7 P5 + P9
left right
Jmax = − (32)
2 2
1 0 0 P1
2 L1 L2 P2
3 L2 L1 P3 Uncertainty of measurement for the weighing
4 L2 L1+sw1 P4
5 0 0 P5
instrument
6 L3 L4 P6 The standard uncertainty is based on the described above
7 L4 L3 P7
8 L4+sw2 L3 P8
parameters as following:
9 0 0 P9 • uncertainty due to the sensitivity of the balance:
10 L3 L4 P10 ∆1= P10-P9 2 2
11 0 0 P11 us min = d min ⋅  u sw1  +  u( P 4 − P3 )  (33)
12 L3 L4 P12 ∆2= P12-P11 m   P −P 
 sw1   4 3 
13 0 0 P13
14 L3 L4 P14 ∆3= P14-P13 and
15 0 0 P15 2 2
  u )
16 L3 L4 P16 ∆4= P16-P15 us max = d max ⋅  u sw2  +  ( P8 − P7  (34)
m   P −P 
17 0 0 P17  sw2   8 7 
18 L3 L4 P18 ∆5= P18-P17
where: usw is the uncertainty of the additional small
19 0 0 P19
20 L3 L4 P20 ∆6= P20-P19
weights sw (sensitivity weights);
where: (P4 –P3) or (P8 –P7) is the change in the indication of the
sw1 and sw2 are additional small weights (sensitivity balance (due to the sensitivity weights) with the
weights) with mass msw, used to determine the scale interval uncertainties u(P4 –P3) or u(P8 –P7), respectively;
of the balance. The sensitivity weights should be calibrated d is the scale interval.
against suitable mass standards. From eq. (33) and (34), the relative standard uncertainty can
be calculated as:
 us rel = u s (min, max) / L (35) where: sw1... sw4 are sensitivity weights having nominal
and the largest uncertainty us rel max is taken into account mass equal to 1/4...4/4 from the maximum capacity of the
• the variance of repeatability : screen;
mswi is the mass of the sensitivity weight applied;
uw 2 = s2 = R2 (36)
Poi med are the average equilibrium positions (rest points)
• uncertainty due to the inequality of the two arms lengths while the balance is not loaded.
u =
u P2 6 u P2 7 u P2 5 u P2 9
+ + + +u2
(in scale divisions) (37) The indication error of the screen will be calculated as
J w
4 4 4 4 follows:
2
EI = ∆i – mswi – ρa·Vswi (41)
 0,2 ⋅ d 
uJ = u r2 + (u w ⋅ d ) = 
2
 + (u w ⋅ d ) 2 =

(in mg) (37’)
 2 3  where: ρa is the density of the air and Vswi is the volume of
the sensitivity weight (with uVsw uncertainty).
= 0.0033 ⋅ d + (u w ⋅ d )
2 2
3. Calibration of built-in weights: the built-in weights are
where: u w2 is the variance of the repeatability used in combination during the operation of balance. First of
2
u is the variance of limited resolution.
r
all, it is necessary to identify the built-in weights as nominal
The resolution is equal to 1/10 or 2/10 of the scale interval values. Ideally the weights built into the balance should be
d, the standard uncertainty being calculated as: removed and calibrated externally. If this is not possible
0,2d
they can be left in the balance and calibrated by dialing them
0,2d upon combinations. A standard weight S of mass ms and
ur = 2 = (38)
3 2 3 volume Vs is chosen for calibration, depending on the
Then, the combined standard uncertainty uc can be accuracy of the balance.
calculated as follows: The steps for calibration of built-in weights are [9]:
- record screen reading at no load indication I1;
uc = (u w ⋅ d ) 2 + u J2 + u s2 rel max
⋅ L2 = (u w ⋅ d ) 2 + u J2 + (u s rel max
⋅ L) (39) - record screen reading I2 when loading with
standard weight S;
The expanded uncertainty is reporting by multiplying uc - record screen reading I3 when loading with
with the coverage factor k=2 standard S (with volume Vs) and a sensitivity
U = k ·uc (40) weight of mass msw (with volume Vsw);
- record screen reading I4 when the standard S is
B2. Single pan balances
removed and on pan remains only the sensitivity
Displays on these balances tend to be of the optical weight.
variety, the sensitivity of the balance being usually adjusted To calculate the mass of the built in weight, the next
by a skilled person. formula can be applied:
In the case of single pan, direct reading analytical balances, I1 + I 4 I 2 + I 3
the following tests and calculations should be carried out: BW = mS − ρ a ⋅ VS + ρ a ⋅ VBW + K ⋅ ( − ) (42)
repeatability, calibration of the screen and calibration of 2 2
built-in weights. where K is a factor used to convert the reading in terms of
1. Repeatability: this test should be done at or near the scale units into a measured mass difference.
nominal maximum capacity of the weighing instrument or m − ρ a ⋅ Vsw
K= sw (43)
using the largest load generally weighed in applications. I3 − I 2
Repeatability is determined in the same way as was The formula (42) can be reduced to a simpler one if the
described in section B1, eq. (30). accuracy of the weighing allows it and it is known that K is
2. Calibration of the screen: on can determine the accuracy constant from the previous measurements:
measurements for the entire screen by the application
BW = mS − ρ a ⋅ VS + ρ a ⋅ VBW + K ⋅ ( I1 − I 2 ) (44)
standard weights at various points in the range of the screen
(1/4, 1/2, 3/4 and 4/4) according to the table 2: The values for ms, Vs, Vsw, msw , are given in the calibration
certificate.
Table 2. Calibration of the screen
Measurement uncertainty for the weighing instrument
N Load Equilibrium P0imed Diff Mass
o applied positions mg ∆i of To estimate measurement uncertainty for the
mg P1 P2 P3 Pmed mg sensitivity
weight
weighing instrument the following parameters need to be
mg mg mg mg
mg considered:
1 0 P1 1. Uncertainty due to the repeatability of the weighing
2 sw1 =1/4 P2 ∆1= P2-P01 msw1 instrument, uw, - given by standard deviation s of several
3 0 P3 P01 weighing results obtained for the same load under the same
4 sw2 =1/2 P4 ∆2= P4-P02 msw2
conditions, calculated as in eq. (36).
5 0 P5 P02
6 sw3 = 3/4 P6 ∆3= P6-P03 msw3
2. Uncertainty associated with the indication error of the
7 0 P7 P03 screen - calculated as follows:
∆4= P8-P04
8 sw4= 4/4 P8 msw4
u(EI)= u sw 2 + u ρ2 a ⋅ Vsw
2
+ ρ a2 ⋅ uVsw
2
+ u w2 + u r2 (45)
9 0 P9 P04
The above expression includes the parameters described: uw
(repeatability), ur (resolution), u ρa (uncertainty of the air
density), usw (uncertainty of the sensitivity weight), Vswi
(volume of the sensitivity weight with uVsw uncertainty).
3. Uncertainty of the built-in weights - calculated starting
from eq. (42) or (44):
2
u BW = u A + us2 + u ρ2 (VBW − VS ) 2 + ρa2 ⋅ (uVBW
2 2 I1 + I 4 I 2 + I 3 2 1 2 2
+ uV2 ) + uK ⋅(
a S
uBW= u2A +us2 +uρ2 (VBW−VS)2 + ρa2 ⋅ (uV2 +uV2 ) +uK
2
⋅ (I1 −I2)2 + K2 ⋅u2(I1−I2)
a BW
Since an experimental standard deviation cannot be
calculated for a single measurement to estimate uA, data
obtained from previous repeatability evaluations can be
used, thus resulting a pooled standard deviation.
Standard uncertainty of the reference weight, us is calculated
according to eq. (9) from above.
The combined standard uncertainty uc can be calculated as
follows:
 when corrections are applied to the error of indication,
the expression for combined standard uncertainty uc is:
results, but it should be remembered that the calibration
uncertainty represents only one part of the measurement
uncertainty stated in current applications of the laboratory.
Other contributions to the measurement uncertainty that
have to be taken into account are the influence of the
buoyancy correction, the influence of the properties of the
− ) + K ⋅ u [( I 2 + I 3 ) − ( I1 + I 4 )]
(46) product that is weighed (evaporation, hygroscopic behavior,
2 2 2
electrostatic charging, etc).
- Mechanical balances have generally been replaced by
(47) electronic balances, which often offer better resolution and
S
are easier to use. The recalibration period for all of them
(mechanical and electronic) can be different for each type,
being influenced by such factors as the usage of the
receives, operator skill and the environment in which the
balance is located. As a general guideline, balance should be
recalibrated yearly, until the stability of operation is
established.
References
[1] International Recommendation OIML R76 -1 “Non-
automatic weighing instruments”, Annex A. pp 68-73

uc = uw 2 + ur 2 + u BW
2
+ (u( EI ) ) 2 (48) [2] International Recommendation No111 “Weights of
 when no corrections are applied to the error of classes E1, E2, F1, F2, M1, M2,M3” ch. 6, pp. 70-74.
indication, the indication error across the partial screen
range that is measured EI should be added to uc, in [3] “Guide to the expression of uncertainty in
addition to u(EI), as follows: measurement” (ISO, Geneva, 1995), Annex F pp 54 .

uc = uw2 + ur 2 + u BW
2
+ (u( EI ) + EI ) 2 (49) [4] International Document OIML D 28 “Conventional
value of the result of weighing in air “pp 7-9, 2004
The expanded uncertainty for k=2 will be:
U = k·uc (50)
[5] Document 2089, Edition 00 – October 2000 “Specific
requirements relating to the calibration of non-
Conclusions
automatic weighing instruments”, ch.8.5 - 8.6,
This paper established metrological requirements pp 19-24.
for the calibration of nonautomatic weighing instruments
and provided useful information for operators working in [6] DKD –R-7-1, “Calibration of non-automatic weighing
accredited calibration laboratories in various fields, in order instruments”, issue 98, ch.5
to determine the mass of products.
[7] NTM 3-05-77 “Verificarea balanŃelor cl. 3, 4 şi 5” ch.5,
- A laboratory should not attempt to make measurements
pp 25-31
with an uncertainty of “x” using an instrument that has the
readability “x”. If the user wishes to apply no corrections, to
[8] EA -10 /18 EA “Guidelines on the calibration of non-
obtain an uncertainty of “x”, he should have a balance with a
automatic weighing instruments” , (draft) July 2004,
readability of “0.1x”, to be sure that are no gross errors
ch.4, pp 11 and 2.
present.
- The balance indications are closer to the conventional mass [9] Randall M.Schoonover: “Air Buoyancy Correction in
than to the true mass, on many occasions, the indication High-Accuracy Weighing on Analytical Balances”.
being directly used as the conventional mass. This is Anal. Chem. 1981, 53 , 900-902
normally not valid for mass (true mass). In current use, it is
necessary to convert the weighing result from the
conventional mass to the true mass (in section B2 described
above, the weighing result is transformed directly in true
mass).
- When a calibrated instrument is used, the calibration
uncertainty stated in the calibration certificate of that
instrument has to be taken into account when reporting the
measurement uncertainty associated with any measurement
View publication stats