bro-test equipment in a quality system-e

7/26/2019 BRO-Test Equipment in a Quality System-e

1/28

Handbook of Weighing ApplicationsBalances and Scales Used as Measuringand Test Equipment in a Quality System

turning science into solutions


2/28

Preliminary Considerations

In many areas of applications, the balance,scale or weight value is only a means to anend. The quantity that is actually of inter-est is derived from the weight value or

mass. For this reason, each booklet of theHandbook of Weighing Applications thor-oughly treats a specific topic. For everysubject, the individual booklets include anexplanation of the general and theoreticalprinciples of the application concerned this is not always possible withoutdiscussing equations according to the lawsof physics or mathematical formulas.Part 3, which is now available, discusses thesubject of Balances and Scales Used as TestEquipment in a Quality System.

An important part of all quality systemsis the area covering inspection, measuringand test equipment and its monitoring foraccuracy. The quality element control of

inspection, measuring and test equipmentrequires that the supplier of a product orservice develop and maintain StandardOperating Procedures (SOPs) for inspecting,calibrating and servicing test and measur-ing equipment. The purpose of these SOPsis to ensure that the suppliers productsconform to defined quality standards. Whenreferring to the control of inspection,measuring and test equipment, we mean anorderly sequence that ensures that theequipment is inspected in a timely fashionand, if necessary, appropriate measures are

taken so that the equipment correspondsto the given requirements.

Using the laboratory balance as an exam-ple, this chapter explains how one canestablish an acceptable level of confidencein the test and measuring equipment beingused. Suitability of the equipment is theinitial requirement for obtaining reliableresults.

Marketing, Mechatronics Division

July 2008

2


3/28

Contents

4 Motivation4 Quality

5 Overview of Quality Systems

5 Universal Quality Systems5 ISO 9000 Series5 EN 45000 Series5 Legally Regulated Quality Systems5 GLP (Good Laboratory Practice)5 GMP (Good Manufacturing Practice)

6 Selection of Suitable Test andMeasuring Equipment

6 Equipment Qualification6 Design Qualification (DQ)6 Installation Qualification (IQ)6 Operational Qualification (OQ)

6 Performance Qualification (PQ)6 Device Qualification | Final Report

6 Test Methods

7 Determination of the Uncertaintyof Measurement

7 Weighing Range7 Repeatability7 Standard Deviation8 Linearity Error

9 Influence Quantities

9 Sensitivity9 Temperature Coefficient9 Zero Point Drift9 Off-Center Load Error

10 Operator

10 Weighing Location10 Leveling10 Gravitational Acceleration11 Mechanical Disturbances11 Humidity11 Barometric Pressure

11 Air Buoyancy12 Electromagnetic Disturbances

13 The Sample13 Static Electricity13 Magnetic or Magnetizable Samples14 Hygroscopic Samples14 Sample Temperature

15 Traceability of a Measurement

15 Calibration and Adjustment15 Calibration15 Adjustment15 External Calibration and Adjustment15 Internal Calibration and Adjustment

3

17 Mass and Weights

19 Documentation19 Description and Identification of

the Test and Measuring Equipment19 Calibration Equipment and Results19 Defined Maximum Permissible Errors19 Ambient Conditions and Correspon-

ding Adjustments20 Maintenance Procedures20 Modification of the Weighing

Instrument(s)20 Appointment and Identification of

Personnel Responsible for MonitoringTest Equipment

20 Restrictions on the Suitability of Testand Measuring Equipment

20 Defining the Interval ofConfirmation

20 If Non-Conforming Test andMeasuring Equipment CausesConsequential Damage

21 Manufacturers Recommendation21 Tendency Toward Component Wear

and Drift21 Environmental Influences21 Demands of Customers, Standards

or Laws21 Experience with Similar Test and

Measuring Equipment

21 Summary

22 Error Calculation22 Systematic Errors22 Random Errors24 Deriving the Uncertainty

of Measurement from theStandard Deviation


4/28

ISO 10012 provides a more extensive andconcrete explanation of the requirementsfor test and measuring equipment. Accord-ingly, a series of measures for using the test

and measuring equipment can be summa-rized as a few general, basic requirements.

The objective of each quality system is toprovide a product or service with theappropriate quality. But what is quality?The term quality is defined in the EN ISO8402 standard as follows:

Totality of characteristics of an entity thatbear on its ability to satisfy stated andimplied needs

Motivation

In the meantime, extreme ranges of resolu-tion have been attained in the field ofanalytical weighing technology. Reachingthese new limits, however, has opened up

discussion about the competence of indi-vidual laboratories. For this reason, mostlaboratories keep certificates, accreditationdocuments and written attestations on file.These credientals provide objective evidenceof the laboratorys performance and assurethose using the laboratorys services thatanalytical questions will be answered by anexpert.

In addition, the flood of analytical data hasconfronted laboratory employees with aproblem. Namely, they must test and vali-

date many measured values for plausibilityand accuracy. Here again, quality assurancemeasures are essential for correct, compara-ble and verifiable results, and are funda-mental to long-term success.

Regulations and standards of the mostprominent quality systems that relate tothe control of inspection, measuring andtest equipment are:

GLP (Good Laboratory Practice)

GMP (Good Manufacturing Practice)

ISO 9000 series

EN 45000 series

They have been generalized to cover a largenumber of devices and procedures and,therefore, must be interpreted accordingly.

4

QualityDefinitionResponsibilities

Ensured byquality systems

GLP/GMP

Quality elements: apparatus, test equipment

ISO 9000 EN45000/DIN17025

TraceabilityTolerancerequirements

Uncertainty ofmeasurement

Documen-tation

Confirmationsystem

Mathematicalmethods

Technicalspecifications

Influence factorsDisturbances


5/28

Overview of Quality Systems

The following descriptions of the qualitymanagement systems intend to highlightthe key features and application areas. Onething that all quality systems have in com-

mon is in their requirements placed on testequipment. These common requirementsdescribe how equipment qualification iscarried out for all test equipment beforeinitial operation and how this equipmentthat is used daily is tested and calibratedon a regular basis.

Quality management systems aresubdivided into various categories. Wedistinguish between universal andindustry-specific quality systems.

Universal Quality Systems

ISO 9000 SeriesThe ISO 9000 series is a set of widely usedinternational standards applicable toproduction and the service industry.Considering its general applicability, ISO9000 does not contain requirementsspecifically related to laboratories, but isa suitable approach for assuring quality inlaboratories. The ISO 9000 series coversvoluntary requirements for all areas ofproduction and service. A management

representative for quality oversees theintegrated quality management andassurance entities. Internal audits arecarried out continually, and recertificationis done every three years.

The focal points of this quality system lieon the following:

Internal and external interfaces

Purchaser-supplier relations

Corrective action

EN 45000 SeriesThis quality system involves European-widerecognition of testing laboratories. A test-ing laboratory accredited for compliance

with European Standards obtains the statusof an institution qualified for specifictasks. This laboratory is accredited for adefined scope of validity and is re-accredit-ed every five years. A typical example of alaboratory accredited for compliance withthe EN 45000 standards is a testing labora-tory commissioned to perform analysesrelating to environmental protection. Theparticular focal points of a quality systembased on European Standards are thefollowing:

Employee qualification Qualification of the processes used

Accuracy of the results

Device testing

Calibration and validation of themethod used

Legally Regulated Quality Systems

GLP (Good Laboratory Practice)GLP is a system of standards applied world-

wide and is legally regulated for data usedto assess products for safety approval inorder to protect people and the environ-ment from hazards. The requirementsimposed by GLP refer to the organizationand to personnel. An audit for compliancewith GLP requirements is performed everyfour years. A typical example of a GLP-compliant unit is a toxicological or analyti-cal laboratory in a chemicals company thatconducts research, or a testing laboratorythat is commissioned to perform tests. Thefocal points of GLP are the following:

Organizational rules and formalrequirements

Documentation

Independence of the quality assuranceunit

GMP (Good Manufacturing Practice)This system is prescribed for the pharma-ceutical industry and medical device manu-facturers. The scope of application for GMP

lies in the manufacture and analysis ofpharmaceuticals. The focal points of GMPare the following:

Defined and validated manufacturingprocesses

Release of each product lot

Self-audits

The most important prerequisites forimplementing GLP and GMP are listedas follows:

Organizational structure of the testingfacility

Qualification of personnel

Quality assurance program

Testing facilities

Equipment, materials and reagents

Test and reference materials

Standard operating procedures (SOPs)

Study plans, raw data and test reports

Filing and preservation of records andmaterials

If we compare all quality systems with oneanother, we discover that many areas over-lap. These systems differ from one anotherin their focal points because each systemhas different objectives. For instance, GLPis a system of documentation that con-tributes towards improving quality. Bycontrast, accreditation according to theEN 45000 series entails less work for docu-mentation. For the latter quality system,

the focus is on the competence ofpersonnel and the quality of results.

5


6/28

Selection of Suitable Test and Measuring Equipment Test Methods

Equipment QualificationThe use of inspection, measuring and testequipment in a quality managementsystem requires a detailed description and

documentation of the results of measure-ments and of confirmation. Processes andstandard procedures must be traceablydocumented and these documents filed.

Many leading quality systems, such asGLP/GMP and the ISO9000 and the EN45000 series, explain how to comply withthe standards.

Equipment qualification provides docu-mented evidence that an instrument isappropriate for its intended use to ensure

that it will operate on demand, under spec-ified service conditions, to meet systemperformance and accuracy requirements.

Equipment Qualification is subdividedinto 4 sections:

1. Design Qualification (DQ)

2. Installation Qualification (IQ)

3. Operational Qualification (OQ)

4. Performance Qualification (PQ)

Design Qualification (DQ)In design qualification, the user defines hisor her requirements on the test or measur-ing equipment. Parameters, such as accura-cy, method of measurement, and require-ments on the supplier that relate to designvalidation or services, must be defined anddocumented before purchasing (procure-ment). The purpose of design qualificationis to ensure that the measuring equipment in this case, the balance, scale or weigh-ing system is suitable for the particularapplication.

The data generated using the test equip-ment are merely observed values of a qual-ity characteristic, for example, the weightvalues generated in a laboratory. Systemat-ic and random errors that occur during theweighing process and result from theweighing equipment itself affect the accu-racy of these values. Therefore, the resultdetermined by the weighing instrumenthas a degree of uncertainty, which is calleduncertainty of measurement and must beindicated as a matter of principle for eachweighing process. The factors that play arole in this uncertainty of measurement areexplained in the following.

The selection of a suitable measuringinstrument must be based on answeringthe question of how great the uncertaintyof measurement may be to allow reliable

compliance with the required tolerances. Agood approach to answering this questionis to apply the golden rule of metrologythat says that the measurement uncertain-ty of a measuring device may only be 1/10of the tolerance of the measured values.

For example, lets suppose that a 10-mgsample is to be weighed to an accuracy of1 percent, which corresponds to 0.1 mg.According to the golden rule, the totaluncertainty of the balance may not exceed0.01 mg.

Especially if a cost-intensive process isused, it is important that this criterion bemet under economically feasible condi-tions. Under certain circumstances, a ratioof 1/3 is acceptable if these tolerances aremet through suitable measures, such as thefrequency of testing to ensure that the testequipment is appropriate.

The basis for the selection of a measuringinstrument or test equipment is providedby the manufacturers technical specifica-

tions, such as repeatability, linearity ortemperature coefficient. Besides theseinstrument parameters, additional factorsthat may affect the results of a measure-ment must be considered. These include theambient conditions at the place of meas-urement, qualification of the operator, testobject and test procedure.

Installation Qualification (IQ)Installation qualification describes startupand the detailed sequence of setting up themeasuring equipment. Special attention

must be paid to the completeness and cor-rect installation of the equipment supplied.To operate high-resolution analytical andmicrobalances, you should essentially con-sider using specially designed anti-vibra-tion balance tables. In addition, the climateconditions (particularly the temperature)should be kept as constant as possible.

Operational Qualification (OQ)Operational qualification describes themetrological testing of a weighing instru-ment at the place of installation. Ade-

quately trained personnel must test weigh-ing instruments using the correspondingauxiliary equipment and weights that havethe appropriate accuracy. In addition, thetest results must be documented in a cali-bration certificate or test report of theweighing instrument. This testing must beperformed at established intervals (knownas intervals of confirmation).

Performance Qualification (PQ)All manufacturers specifications refer tonearly ideal measurement conditions as re-

commended in the installation and operat-ing instructions. In practice, however, oper-ators frequently operate weighinginstruments under conditions that differfrom these. Therefore, performance qualifi-cation requires verification that the meas-uring equipment functions as intended inits normal operating environment (e.g.,weighing a sample under a laboratory fumehood).

Device Qualification|Final ReportOnce all qualification procedures described

above have been successfully performedand the adequate performance of themeasuring equipment has been verified,equipment qualification along with a finalreport is completed.

All manufacturer specifications are basedon idealized weighing conditions. Other-wise, comparisons could not be madebetween different instruments. But themethods actually used in the field oftendiffer from those used by the manufactur-er. Variations in the methods used should

be documented appropriately in the SOP,and allowances should be made for devia-tions in the weighing accuracy that mayresult. For example, if a hanger for below-balance weighing is used to weigh a mag-netic sample, the manufacturer specifica-tions, which were determined under thebest weighing conditions, cannot be main-tained. In this case, preliminary tests mustbe run using reference samples to verifythe attainable degree of accuracy.

6


7/28

Determination of the Uncertainty of Measurement

Manufacturer specifications, which are thebasis for selecting test and measuringequipment, are explained and interpretedin the following sections.

The limits of a weight measurement, i.e.,the range within which the definedcertainty of measurement is maintained,is called weighing range.

Repeatability describes the ability to dis-play corresponding results under constanttesting conditions when the same load isrepeatedly placed on the weighing pan inthe same manner. Repeatability is essen-tially independent of the load on the bal-ance or scale. It can be designated as the

most important metrological featurebecause its influence on the uncertaintyof measurement especially with lighterloads becomes the dominant factor.

Either the standard deviation or the differ-ence between the highest and the lowestresult for a defined number of measure-ments is used to specify this quantity.

Example of a Weighing Series:

Weight No. Weighing Series

1 9.997g

2 10.002g

3 9.998g

4 10.002g

5 10.001g

6 10.002g

7 10.001g

8 10.000g

9 9.998g10 10.002g

11 9.997g

For evaluating the quality of a weighinginstrument on the basis of its technicalspecifications, both values (lowest andhighest result) are approximately compara-ble with each other if the minimum|maxi-mum specification is compared with threetimes the standard deviation. Within thestandard deviation times three, you willfind 99.7% of all representative values ofa weighing series.

The standard deviation corresponds to thespread of the bell curve on either side fromits point of inflection. Sixty-eight pointthree percent (68.3%) of the individualvalues will be located within this area or, toput it differently, the individual values willfall within the range of

__ s with a confi-

dence interval of 68.3%. In practice, theuse of the standard deviation times two hasbecome the norm. This interval has aprobability of 95.5%.

This means that 95.5% or 99.7% of all

values will be distributed with respect tothe mean value within the range definedby the standard deviation times two orthree.

7

The mean value is calculated from the sum of the individual values W1 to Wn, divided bythe number n of individual values; hence_ = 1 . s

n

i__n

i =1

Using our example, this means:

_ =

9.997+10.002 + 9.998 +10.002+10.001+10.002+10.001+10.000+ 9.998 +10.002 + 9.997

11_ = 10.000 g

The difference between the highest and lowest result in the weighing series iscalculated as follows:

10.002 g 9.997 g = 0.005 g

The standard deviation is computed using the following equation:

s =

____________________

_

The standard deviation of our example is:

s =____________________________________________________________________________________

1. [(9,997-10,000)2 +(9,997-10,000)2] 0,0020976 g 2mg

1 n____ s ( i - )2

n-1i=1

i = individual value measured in theweighing series

n = total number of weight measurements mean value of the individual resultsmeasured


8/28

The following Figures show a weighingseries listed in a chart and plotted as agraph; the eleven individual weights aremarked as points.

Frequently, the relative standard deviationis also given in percent

[ ].In our example, the standard deviation is:

The values of our weighing series given asan example yield the following results:

Weighing Series

Number of individualvalues measured "n" 11

Sum of the individualvalues measured 110.000 g

Mean value 10.000 g

Standard deviation 0.002 g

Approximate (relative)standard deviation 0.00167 g

Repeatability acc. to OIML R76 0.005 g

0.002 g= ____________ 100 %= 0.02%

10.000 g

8

9.990 g 9.995 g 10.000 g 10.005 g 10.010 g

Numberofindividualvaluesn

0

1

2

3

4Standard deviations = 0.0020976 g

The linearity error (usually referred to aslinearity) indicates how much a balance ora scale deviates from the theoretically lin-ear slope of the characteristic calibration

curve. In the case of an ideal characteristiccurve, the mass on the weighing pan willalways equal the weight displayed on thebalance or scale. If the zero point is correctand the weighing instrument has been cor-rectly calibrated and adjusted at maximumcapacity, the linearity can be determinedby the positive or negative deviation of thevalue displayed from the actual load on thepan. Linearity is caused by the specificinherent properties of a weighing instru-ment and is therefore unavoidable. Two ofthe most frequent curves are slopes of the

2nd order (convex or concave curve) and ofthe 3rd order (S-shaped curve).

The maximum deviation between theactual characteristic curve and the linearslope of the two interdependent values the zero point and the maximum capacity

is defined as linearity. The maximumlinearity is given in the data sheets ofbalances and scales. In some cases (such asan analytical balance), a limited range isspecified, for instance, 200 g = 50 gwithin 2 g = 10 g.

Value displayed

Characteristic curve ofthe 2nd order

Ideal characteristiccurve

Characteristic curve ofthe 3rd order

Mass on theweighing pan100 g

100 g

Linearityerror

s__ 100 %__


9/28

Influence Quantities

The measured result, or weight, can beaffected by so-called influence quantities,such as temperature, barometric pressureand humidity. In general, a distinction is

made between the temperature coefficientof the zero point and of the sensitivity.Each of these parameters shows a moreor less considerable impact on themeasured result, affecting both theelectronic components and weighingsystem to an equal extent.

The sensitivity is the change in a displayedvalue divided by the change in the loadsignal generated by the mass on the pan.If a balance or scale with a digital displayhas been correctly adjusted, the sensitivity

must always be exactly 1.

The equation for the sensitivity is asfollows:

DS=

m

where D is the number of scale intervalsthat correspond to the change in load m.

A sensitivity error S is caused by usinginappropriate calibration weights to adjust

a balance or scale. The sensitivity error isalways indicated as a relative number, e.g.,20 ppm per K (1ppm = one part per million= 10-6).

If the value of the zero point or of thesensitivity changes because the tempera-ture fluctuates, the temperature coeffi-cient is used to characterize this change.If a weight is divided by the change intemperature, this will yield the value of thetemperature coefficient.

Example:Temperature coefficient: 2 10-6 K-1

Initial sample weight : 10 gChange in temperature : 5 K

Systematic error due to the temperaturecoefficient:

2 10-6 K-1 10 g 5 K = 0.1 mg

The value of the temperature coefficient isthe major criterion for judging whether ornot the weight readout has stabilized when

a balance or scale is exposed to fluctua-tions in the ambient temperature.

If a light load is left on the balance or scale,over time you will see a drift in the zeropoint ZP on the display. Zero point driftis only important for long-term measure-ments involving a constant load, as in ther-mogravimetric and sorption measurements.

The off-center load error, also calledcorner load error, means the change inreadout when the same load is placed invarious positions on the weighing pan orload plate.

The off-center load error is officially calledeccentric loading error. To verify the

error, a weight is placed exactly in themiddle of the weighing pan and thebalance or scale is tared. Then the weight isplaced in 3 to 4 different locations on theedges of the weighing pan; if the pan isrectangular, the weight is placed in thecorners. The off-center load error can thenbe directly read off the display. This valuecan be negative or positive and usuallyranges from 1 to 10 digits or scale intervals.Therefore, especially when you use bal-ances with high resolution, the sample tobe weighed should always be placed exact-

ly in the middle of the weighing pan. Inaddition, other factors that can substan-tially influence the weighing results mustbe taken into account: operator, weighinglocation, sample and weighing procedure.For this reason, it is recommended that theeffects of these factors be minimizedwhenever possible.

In the following, these factors will be dealtwith in more detail.

9

Measured value displayed

Characteristic curve ofthe 2nd order

Ideal characteristiccurve

Mass on theweighing pan100 g

100 g

~ S


10/28

Because of the earths rotation andgeographical features, the gravitationalacceleration varies depending on wherethe balance or scale is set up. We thereforerecommend that the weighing instrumentbe adjusted each time it is set up in a new

location and before initial startup. Duringthis procedure, a known mass is loaded onthe weighing instrument and the adjust-ment factor is determined from the weightvalue displayed. Another effect that oftengoes unnoticed is a change in altitude andhow it can influence the gravitationalacceleration when, for example, thebalance is moved to a higher location.Moving the balance will affect theaccuracy of the weight displayed!

The following is obtained in the relationshown below for a difference in altitudeof only 4m:

Leveling foot

Level indicator

Operator Weighing Location

Today, leading manufacturers offerbalances with a readability of up to 0.1 gand a resolution of up to 21 million digits.It almost goes without saying that theoperator must receive proper training inorder to capitalize on the accuracy and

precision of these instruments.

For instance, the operator must be awareof and strictly comply with basic rules,such as:

Placing the sample in the middle of theweighing pan (to avoid off-center loaderrors);

Attempting to work as consistentlyas possible (to maintain the specifiedrepeatability);

Making sure that the balance is setup on a level surface (to prevent asystematic sensitivity error)

A scale or balance is adjusted in a manu-facturing process so that the force trans-mitted to the weigh or load cell when thescale is loaded is parallel to the direction ofthe gravitational acceleration and perpen-dicular to the cell. A level indicator (small

spirit level) attached to the scale enablesthe operator to pinpoint this positionexactly, allowing it to be reproduced at alltimes. This step is called leveling.

The importance of leveling a balance orscale will be explained using the followingexample. Suppose a laboratory bench withan edge length of 1,000 mm is raised atone end by 5 mm. Then the followingapplies to the angle of inclination:a= arctan 5/1,000 = 0.2865

Moreover, the following applies to theforce generated by a load in the directionof the weighing axis:A = W cos = W 0.9999875,i.e., the weight measured by the tiltedbalance is 2.5 mg too low for a samplewith a mass of 200 g.

10

g = Gravitational acclerationRE = Earth's radiush = Difference in height (altitude)

RE

h

RE hg (R+h) g

K (________ ) gK 1 - 2 (____)R

E+ h RE

g ( R+4 m) g 1-2 ______________ ) =g 0.9999987(

6370000m

4 m


11/28

This means that a semi-microbalance,which measures a mass accuratelyto 200.00000 g, will only measure199.99974 g for the same mass whenset up 4m higher. This underscores thenecessity of calibrating and adjusting

a balance or scale each time it is movedto a different location.

As a result of the moment of inertia,mechanical disturbances register on thebalance or scale as periodic or stochasticweight changes depending on theirattributes. A digital filtering feature on theweighing instrument, which can be activat-ed by selecting a suitable integration time,can reduce these disturbances.

Low-frequency interference, however, is

less likely to be filtered out because thefilter can no longer differentiate betweenmechanical interference and a slowlychanging weight readout (for example,during filling). We generally recommendthat specially designed weighing tables beused for balances that have extremely highresolution. If vibrations in the buildingcause the disturbances, we recommendthat the balance be set up on a lower floor.If this is not possible, the balance should beused with a specially designed wall console.

Mechanical disturbances can be caused bypumps, laboratory shakers, turbulenceunder laboratory fume hoods, and so forth.

Under normal circumstances, humidityas an ambient quantity affecting theweighing procedure can be neglected.However, for balances of older designs andscales with a strain-gauge load cell, thechange in humidity must be kept as low aspossible as damage caused by corrosion ofthe connections can occur at high humidi-ty. The humidity also affects the long-term

stability of such load cells.

For standard weighing procedures,barometric pressure is a negligible sourceof error.

However, for precision weight measure-ments (urel =< 5 10

-4), the air buoyancymust be taken into account as it is ofconsiderable importance for assessing theaccuracy of the value measured by thebalance or scale.

If an object is in a medium, this liftingforce opposes the weight of this object.Buoyancy reduces the weight of the massto be measured by the amount that equalsthe weight of the displaced medium.

If you consider two materials of the sameweight but of a different volume, such asan aluminum cylinder with a density of2.7 g/cm3 and a weight standard with adensity of 8.000 g/cm3, both of these are inequilibrium when weighed under vacuum.

mSTD = Mass of the weight standardm

S= Mass of the sample

g = Gravitational acceleration

If you consider the same setup weighedin air, both samples are no longer inequilibrium.

A

= Density of the airVSTD = Volume of the weight standardVS = Volume of the sample

This is caused by the different buoyanciesresulting from the different materialdensities and volumes.

11

Weighing under vacuum:

mSTD g mS g

Standard Sample

Weighing in air:

mSTD g mS g

A VSTDA VS


12/28

Because air has mass (density understandard atmospheric conditions:

A= 1.2 mg/cm3), the weight W of a

sample depends on the density of its mate-

rials and thus on the volume it takes up.

Let m be the difference measuredbetween two masses mSTD and mS. You willobtain the actual difference m using thefollowing general equation:

where VSTD and VS are the volumes of the

objects of the masses mSTD and mS , and Ais the density of the air according to theconditions prevailing during the weighingprocedure.

The mass of the sample is determinedaccording to the density values available,

A

1 _____STD

ms= ____________A1 ____

s

where STD is the density of the standardand

Sthe density of the sample.

The graph shows how the weight readoutof a mass is corrected for air buoyancy asa function of the material density for a fewselected density values given in g/cm3.

Electromagnetic disturbances consistmainly of electromagnetic radiation inthe range of a few kHz up to several GHz,which is frequently used for wirelesscommunication:

Radio communications

Mobile or closed-circuit radiocommunications

Transmission of weights

Telecommunications through remotecontrol

Radar transmissions or measurement ofnoise in electric circuits

Every measuring instrument, in other wordsa balance or scale, must be able to functionproperly when exposed to the effects ofthese electromagnetic disturbances, gener-ally referred to as radio frequency interfer-

ence. Every balance or scale that is suppliedwith a Declaration of Conformity (CE mark)has passed the test prescribed by the ECCouncil Directive 89/336/EEC Electromag-netic Compatibility (EMC). This means thatthe balance or scale has a defined immuni-ty to emissions in residential, commercialand industrial areas including light indus-trial environments. Based on the results ofthe EMC test, electromagnetic disturbanceshave no effect on the weighing results.

12

m =mS mSTD =m + A (VSVSTD) or

mSTDA VSTD =mSA VS 1 2 3 4 50

2

4

6

0

0.8

1.4

2.0

8.0

m [mg]

Weight readout [mg]


13/28

The Sample

In the majority of cases, the properties ofthe sample itself are the cause of inadmis-sible results. The most important factorsthat influence weighing accuracy are:

Electrostatic charges

Magnetic or magnetizable materials

Hygroscopic materials

Sample temperatures that deviate toomuch from the ambient conditions inthe laboratory

Static electricity or electrostaticcharges, which are particularly noticeablewhen the humidity is low is characterizedby a weight readout that drifts consider-

ably and by poor readability. This phenom-enon primarily affects substances that havea low electrical conductivity and can there-fore pass on charges (caused by frictionwith air, internal friction or direct transfer)to their environment only slowly. Examplesof these substances are plastics, glass andfilter materials as well as powders andliquids.

Depending on the polarity of the chargedparticles involved, this force either attractsor repels, so a weighing result may deviate

in either direction. This effect is based onthe interaction of electrical charges thathave built up on the sample weighed andon the fixed parts of the balance that arenot connected to the weighing pan.

This problem can be eliminated by:

Shielding the sample (using a metalcontainer)

Increasing the surface conductivity ofthe sample by raising the level of humid-ity inside the draft shield of an analyticalbalance

Directly neutralizing the surface chargesusing so-called static eliminators

If a sample is magnetic or magnetizable,i.e., contains a percentage of iron, nickelor cobalt, forces of a different origin aregenerated, which also have a significantinfluence on the weighing result. If the

sample is magnetized, as is the stirring barof a magnetic stirrer, the forces of attrac-tion that this magnet exerts on the magne-tizable parts of the balance will overridethe weight of the sample. Vice versa, theinfluence that the residual magnetic fieldof the electromagnetic-force compensatingweighing system has on a sample cannot beruled out. Magnetic forces manifest them-selves as a loss of repeatability of theweighing result because they depend onthe orientation of the sample within thefield of interference. Unlike electrostatic

interference, magnetic interference is stableover time.

13

Static electricity eliminator withintegrated high-voltage source

Microbalance for weighingfilters; with a metallic pan cover

Semi-microbalance with a staticelectricity eliminator integratedas a standard feature

--

-

-

-

-

-

-

-

--

-

-

-

-

--

-

-

-

- -

+

-

--

-

-

-

+

+

+

+

+

+

+

+

+

+

+

+

Example of the pattern of fieldforces generated by a magnetic ormagnetizable sample

Interaction of electrostatic charges thatattract one another; the sample appears tobe lighter

Interaction of electric charges that repel oneanother; the sample appears to be heavier


14/28

To eliminate problems with magneticforces, one of the following approachescan be taken:

Increase the distance between thesample and the weighing pan

Use a hanger for below-balanceweighing (under-scale weigh kit)

Use a shield made of a soft magneticmaterial

Use a special anti-magnetic weighingpan

Hygroscopic samples cannot be preciselyanalyzed because they absorb moisture,which causes a constant increase in weight.If appropriate steps cannot be taken tokeep the humidity to a minimum at theweighing location, we recommend that the

sample be weighed in an enclosedcontainer that is suitable for its size.

The sample temperature is an influencequantity that is often underestimated.Especially during very precise weighingprocedures, it is imperative that the samplebe adapted to the ambient temperature.Otherwise, convection currents on thesurface of the sample can lead to majorerrors in measurement. Research has shownthat when beakers with a large surface areaare used during weighing, temperature

differences of a few degrees [C] can causethe readout to differ in the gram [g] range.

14

Hanger for below-balance weighing

Special anti-magnetic weighing pan


15/28

Mass on theweighing pan10.000 g

Weight readout

9.009 gActual value

10.000 gNominal value

0.991gDeviation

Traceability of a Measurement

Calibration and AdjustmentThe previous sections covered a series ofinfluence quantities that can adverselyaffect the accuracy of test and measuring

equipment in a variety of ways. Therefore,it is hardly surprising that the test andmeasuring equipment standards used in allquality systems require that errors inmeasurement be quantified. In addition,measures for eliminating such errors mustbe specified. This is done throughcalibration and adjustment.

Calibration checks the deviation betweenthe weight readout on the balance and areference weight (in the field of weighingtechnology, this is a weight whose value is

indicated on an accompanying certificate).Calibration is the most important source ofinformation for checking a balances orscales uncertainty of measurement underactual installation and operating condi-tions. Therefore, it plays a central role incontrolling the accuracy of inspection,measuring and test equipment.

Adjustment always entails corrective inter-vention in the balance or scale to eliminatethe existing error as far as possible. Duringadjustment, the weight readout is compa-

red to the correct value of the calibrationweight, and the resulting correction factoris stored in the balances or scales proces-sor until the next adjustment. Weighingprocedures performed after adjustmentare corrected accordingly.

How frequently a balance or scale needsto be adjusted depends significantly on thefollowing parameters:

The frequency of weighing procedures

The ambient conditions

The effects of an incorrect result

A variety of instruments and methodsexist for performing both of theseprocedures. In general, a distinction ismade between internal and externalcalibration and adjustment.

The external calibration and adjustmentprocedure is used mainly on older-modelbalances and scales or those with highcapacities. Comparison and correction are

accomplished using one or more weightswhose value and uncertainty must beknown and documented.

National testing laboratories, calibrationlaboratories and qualified manufacturersprovide appropriate certificates for thispurpose.

For internal calibration and adjustment,a reference weight that is built into thebalance or scale is used. The exact value ofthis weight was previously determined

during manufacture and stored as a fixedvalue in the electronically programmableread-only memory (EPROM) of the weighinginstruments processor. On the simplestmodels, the user places a weight on thebalances or scales weighing system withthe help of a mechanical device. The moto-rized calibration weight feature, which isoperated at the touch of a button, hasrecently become the standard. The mostadvanced balances and scales are equippedwith a fully automatic calibration andadjustment device that initiates calibration

after a preprogrammed or user-definedamount of time has elapsed.

15

External calibration and adjustment of a precisionbalance


16/28

In addition, an internal sensor continuouslymonitors the balance or scale temperature(as a parameter for determining accuracy)and triggers automatic calibration once

a certain temperature difference has beenexceeded. This ensures the continuousaccuracy of the balance or scale withoutrequiring the user to intervene.

The figure below shows the sequenceof functions that take place during fullyautomatic calibration.

Besides the advantages offered by thisconvenience feature, internal calibrationis generally considered preferable overexternal calibration.

The internal weights are better protectedfrom dirt and damage and are always atthe same temperature as the balance orscale, per se. Moreover, the motorizedcalibration feature ensures that the weightis placed on the balance or scale in themost reproducible manner possible.The fully automatic mode ultimatelyensures that one of the most importantrequirements of the test and measuringequipment is fulfilled.

The question is often asked about how thetraceability of a balances or scales built-incalibration weight can be ensured. This canbe accomplished by tracing the internal

calibration weight to an extremely precisereference weight from the manufacturer.With regard to its materials and surfaceproperties, the internal weight must possessall of the features of a classified weight. Asis the case with all external weights, inter-nal weights must also be tested at certainintervals to ensure that they are withintolerance limits. This is usually done whenthe balance or scale is serviced.

16

99.991 g 100.000 g

100 g 100 g

isoCAL

+

-

Motorized calibration weights of a micro-balance that are spherically shaped toimprove the area-to- volume ratio

Built-in calibration weights


17/28

Mass and Weights

To enable comparison of the results obtai-ned with various balances and scales, wemust be able to trace these results to adefined standard. A balances weighing

results are traced and monitored bycomparing them to a standard that repre-sents the value of the measurand (quantitysubject to measurement) that is required tobe correct. This standard is also traced tothe international prototype through anuninterrupted chain of such standards forcomparison.

Relation to the Base Unit

Nano- ng 1ng =gram 0.000,000,000,001 kg

Mikro- g 1g =gram 0.000,000,001 kg

Milli- mg 1mg =gram 0.000,001 kg

Gram g 1g =0.001 kg

Kilo- kg 1 kggram base unit

Ton t 1 t = 1000 kg

The necessity of tracing other units to the

kilogram by mass comparison has given riseto the hierarchical structure of mass stan-dards. In this hierarchy, the uncertainty of

17

Germanys national kilogram prototype

Mass standards Mass comparison

At the BIPMwhen necessary

At the BIPMwhen necessary,e.g., every 5 years

At the BIPM whennecessary, e.g.,every 12 years

At the placeof use,


18/28

measurement at a certain level depends on the number of previous mass comparisons.

18

+ / - in mg

Nominal value E1 E2 F1 F2 M1 M2 M3

1 mg 0.002 0.006 0.020 0.06 0.20

2 mg 0.002 0.006 0.020 0.06 0.20

5 mg 0.002 0.006 0.020 0.06 0.20

10 mg 0.002 0.008 0.025 0.08 0.25

20 mg 0.002 0.010 0.03 0.10 0,3

50 mg 0.004 0.012 0.04 0.12 0.4

100 mg 0.005 0.015 0.05 0.15 0.5 1.5

200 mg 0.006 0.020 0.06 0.20 0.6 2.0

500 mg 0.008 0.025 0.08 0.25 0.8 2.5

1 g 0.010 0.030 0.10 0.3 1.0 3 10

2 g 0.012 0.040 0.12 0.4 1.2 4 12

5 g 0.015 0.050 0.15 0.5 1.5 5 15

10 g 0.020 0.060 0.20 0.6 2 6 2020 g 0.025 0.080 0.25 0.8 2.5 8 25

50 g 0.030 0.10 0.30 1.0 3.0 10 30

100 g 0.05 0.15 0.5 1.5 5 15 50

200 g 0.10 0.3 1.0 3 10 30 100

500 g 0.25 0.75 2.5 7.5 25 75 250

1 kg 0.5 1.5 5 15 50 150 500

2 kg 1.0 3,0 10 30 100 300 1000

5 kg 2.5 7.5 25 75 250 750 2500

10 kg 5 15 50 150 500 1500 5000

20 kg 10 30 100 300 1000 3000 1000050 kg 25 75 250 750 2500 7500 25000

M1 E2F1F2

5,000

10,000

50,000

100,000

500,000

Digits

ClassE1*

1,000,000

n=max: d n = resolution of the weighing instrument (digits)max = max. weighing capacity of the weighing instrumentd = readability of the weighing instrument

* or E2, DKD calibrated


19/28

Documentation

A characteristic element of all quantitysystems is the requirement of documen-tation. Requirements as to the extent anddepth of the documentation vary signifi-

cantly depending on the system being used.In any case, it is helpful to use the rule offive Ws as a guide when developing a setof instructions that must be followed.This rule states that procedures must bedocumented in such a way as to answerthe question:

Who Did What, with What, When andWhy?In the area of management of test andmeasuring equipment, experience hasshown that this requirement is best met

by introducing and maintaining an SOPand a logbook for the weighing instrument.While all aspects of operation are laid outin the SOP, the logbook contains entriesabout the maintenance, service and repairprocedures for the particular balance orscale.

Practical examples of an SOP and alogbook are given in the Appendix of thisHandbook.

In particular, the following must be

recorded:

Description and identification of the testand measuring equipment

Calibration equipment and results

Defined maximum permissible errors

Ambient conditions and correspondingadjustments

Maintenance procedures

Modification of the weighinginstrument(s)

Identification of the personnelresponsible

Restrictions on the suitability of test andmeasuring equipment

These items are discussed in more detail inthe following sections.

Description and Identification of theTest and Measuring Equipment: Thisincludes general information about thetype of weighing instrument (e.g., analyti-

cal balance with a motorized draft shield);the most important manufacturer specifi-cations; and the model, serial number orinventory number at the weighing location.

Calibration Equipment and Results: Thesetwo factors are decisive for maintainingthe desired degree of weighing accuracy.Depending on the resolution of the balanceor scale and its construction features(motorized placement of the weight on theweighing pan, fully automatic calibrationfunction), determinations must be made

about the nominal value, the maximumpermissible errors and how the weights areto be used. The weights or sets of weightsemployed are also considered test andmeasuring equipment and must be labeledand identified accordingly. Intervals forrecalibration of the weights must also bedefined. Especially when there are largedeviations in the calibration results, controllimits must be defined, and a proceduremust be developed for reporting suchdeviations.

For defined maximum permissible errors,the overall uncertainty of measurement,which was determined using the test andmeasuring equipment described above,

must be traceable. On the basis of thisvalue, the user can determine whethera balance or scale is suitable for thetolerance indicated in the SOP (e.g., theanalysis.)

Ambient Conditions and CorrespondingAdjustmentsThe specifications that characterize thebalance or scale are determined by themanufacturer under well-defined standardconditions. In reality, however, certainusually unfavorable conditions often

cannot be avoided. For example, if thebalance or scale is located under a fumehood in the laboratory or in a place wherethere are great fluctuations in temperature,the analysis can be adversely affected.Modern balances and scales can be adaptedto the ambient conditions at the weighinglocation by varying the set of parametersin the operating system so that thebalance or scale may be used in thatlocation. However, this usually resultsin the accuracy being reduced.

19


20/28

Example:For example, if the stability range para-meter is increased, the balance or scale candeliver accurate results even when it is

subjected to a field of interference of agreat amplitude. The attainable repeat-ability however, is sacrificed in the process.In this case, the change in the parametersof the balance or scale operating systemand the influence on the uncertainty ofmeasurement must be documented.

Maintenance ProceduresDeterminations must be made about

when the balance or scale should becleaned,

who should service it and at whatintervals,

and how to proceed if a repair isnecessary.

The results of regularly performed mainte-nance procedures can also be useful foranalyzing the trend of certain deviations.This facilities appropriate definition of theinterval of confirmation.

Modification of the Weighing

Instrument(s)A variety of technical applications requirethat a standard-equipped balance or scalebe modified. For example, a hanger forbelow-balance weighing might be used ifeither the size of the sample or specialambient conditions (such as magneticfields, temperature, humidity and so forth)dictate the manner in which the analysisshould be conducted. Weighing pans ofmodified shapes and sizes and analyticalbalances with specially designed draftshields are also often used. Today, leading

manufacturers are in a position to offertheir customers application-specific solu-tions with respect to digital filters or otherweighing parameters. Dynamic weighingprocedures constitute one of the mainapplication areas for which this typeof modification is necessary.

Appointment and Identificationof Personnel Responsible for MonitoringTest EquipmentThe laboratory manager appoints a person

to oversee the test and measuring equip-ment. This person is responsible for theappropriate use of the balances and scales.

Restrictions on the Suitability ofTest and Measuring EquipmentIf a confirmation or calibration proceduredetermines that the test and measuringequipment can no longer operate withinthe defined maximum permissible errors,even if corrective intervention is taken, thebalance or scale should no longer be usedfor the intended purpose. Of course, it is

possible to use the balance or scale foranalyses that do not require such a highlevel of accuracy. In this case, the limitedapplication range must be clearly denotedon the instrument and indicated in theSOP.

Defining the Interval of ConfirmationWe use the term confirmation tosummarize all activities that ensure thatthe predefined properties of the test andmeasuring equipment are maintained.Therefore, the interval of confirmation

corresponds to the time interval or numberof analyses performed with the test andmeasuring equipment between two succes-sive inspections. From an economic stand-point, testing should be optimized so thatit is performed before a balance or scaleexceeds the maximum permissible errors.This is also closely connected to the previ-ously mentioned rule, which states that theuncertainty of measurement of the testand measuring equipment should be muchlower than that required by a particularweight measurement application. The

following should be taken into accountwhen first defining the interval ofconfirmation:

The extent of possible adverse effects onthe analysis due to non-conforming testand measuring equipment

Manufacturers recommendation

Tendency toward component wearand drift

Environmental influences

Demands of customers, standardsor laws

Experiences with similar test andmeasuring equipment

The following questions should betaken into account if non-conformingtest and measuring equipment causesconsequential damage:

1. When should data obtained witha nonconforming instrument berejected?

2. What additional expenses can resultfrom overfilling expensive substances?

3. Can the customer assert product liabilityclaims in such case?

20


21/28

Summary

Manufacturers RecommendationLaboratory balance manufacturers if theyprovide service and maintenance for theirproducts have an extensive amount of data

at their disposal with respect to all importantfeatures of the balance. This is especially truegiven various areas of use and ranges ofapplication of lab balances.

Tendency Toward Component Wear and Drift:

On advanced laboratory balances and scales, thistendency can be neglected because theseweighing instruments are designed andconstructed to keep component wear to aminimum when they are operated accordingto the manufacturers instructions. The readoutmight drift in individual cases and after

prolonged use of the balance or scale due tothe electronic components.

Environmental Influences:The range of uses for balances and scales arespecified according to temperature and humidi-ty classes. If a weighing instrument is mainly orconstantly subjected to temperatures or levelsof humidity that border on the allowable limitsof these classes, the specifications will likely beaffected and must be taken into accountaccordingly.

Demands of Customers, Standards or LawsIf the equipment is to be used in sensitive areaswith very high security standards (e.g., in theaerospace industry, for medical technology, forpharmaceutical production and so forth), thecustomer will place high demands on thesuppliers quality system. These demands cango far beyond the standard requirements and,therefore, can have an influence on the controlof inspection, test and measuring equipment.

Experience with Similar Test and

Measuring Equipment

Because of the multitude of factors that mustbe considered when defining the interval ofconfirmation, a general recommendation onhow to do so cannot be made. It makes moresense to follow your technical intuition andconsider the relevant factors to determine asuitable interval. Statistical data from thecurrent inspection can be used to checkcalibration and optimize the interval that isinitially selected. For example, the interval ofconfirmation can be gradually adjusted bycutting the test interval in half, if the maximumpermissible errors are exceeded, or doubling it ifthe requirements are met satisfactorily. From aneconomical standpoint and to ensure the trace-ability of test results, it may be useful to com-bine extensive inspections at longer intervalswith additional short-term tests or calibrationprocedures using suitable working standards.

The control of inspection, measuring andtest equipment is an element of functionalquality management. It is a prerequisite forobjectively demonstrating the performanceof a laboratory as well as for introducingand maintaining processes that can becontrolled.

This starts with the selection of a suitabletest or measuring device based on thetolerances to be tested, which, for instance,are indicated in the laboratorys SOPs.Measuring equipment suitable for this

purpose has an overall uncertainty ofmeasurement that is much lower than thesample with respect to the specifications ofthe equipment and all factors that have aninfluence on the measurement.

Suitable SOPs should be indicated inwriting to ensure that the test require-ments are always met, and all related datashould be documented.

21

Protect people andthe environment

Internationalrelationsbetween suppliers

Ensure marketabilityand competitiveness

Optimize costsBoost productivity

In-house cooperationand motivation

Intensify supplier/customer relations

Protection againstdisputes and claimsfor damages

QUALITY


22/28

Error Calculation

A deviation in the displayed value from thetrue value is commonly known as an erroror deviation; the standardized term iserror of measurement. In the following,

we will use the simpler form error. Wedistinguish between two types of error:systematic and random errors.

Systematic ErrorsThe cause of the error is known, perhapseven the value of this error, or at least anupper limit of error.

Examples:

1. A scale of lengths is not exactly accuratein length; all measurements are made

with the same scale of lengths.2. The same holds true for a weighing

instrument; e.g., a balance with anincorrectly adjusted sensitivity.

3. A measuring instrument is adjusted to20C, but the measurement is carried outat 25C (this is important, e.g., in thecase of a volumeter.)

Random ErrorsThe cause of a deviation is either unknown,or this deviation is caused by varyinginfluence factors.

Examples:

1. Friction in a measuring instrument thathas mobile components

2. Random fluctuations in the zero pointof a weighing instrument

3. Statistical influence of the operator(e.g., parallax errors when the operatorreads off the measuring instrumentdisplay that has a pointer; or a changeon the mass of the object being weighed

when the operator touches it with his orher hands)

Note:There is no hard-set difference betweensystematic and random errors. By means ofadditional measurements or information,many random errors can be transformedinto correctable systematic errors.

22


23/28

The standard deviation s is given as the quantity for repeatability:

s =___________________

n = total number of measurements

i = individual results measured__ = mean value of the individual results measured

_ =

Gaussian distribution (normal distribution)__ = average weight; s = standard deviation68.3% of the weighing results lie within the range of

_ s95.5% of the weighing results lie within the range of

_ 2s99.7% of the weighing results lie within the range of

_ 3

ESum =

_______________

F1, F2 = individual errors

Example:The gross weight mG of 210.213 gAnd the tare weight mT of 205.171 g

Yield the mass mNet of 5.042 g

The individual errors of mG and mT, respectively, are each 1 mgHence, the absolute error of m

Netis :

ENet =_______________________

and the relative error is:

EResult______ =

_______________________

Result

Example: Density determination in accordance with the equation:

m = mass = 150.27 g 0.01 gV = volume = 173.4 cm3 0.1 cm3

= density

150.27 g g = _____________= 0.866609 ____173.4 cm3 cm3

E___ =

_________________

E___ =

__________________________

5.80 10-4

g gE= 5.8 10-4 = 0.8666 ___ 5.8 10-4 = 0.5 10 -3 ___cm3 cm3

gFinal result: = (0.8666 0.0005) ___cm3

23

There are mathematical rules for randomerrors:

Rule 1:

If a measurement is repeated a sufficientnumber of times and the frequencydistribution of the individual valuesmeasured are plotted, you will obtaina characteristic curve, the so-calledGaussian curve.

Rule 2:

Law of error propagation for sums anddifferences: In sums or differences, thesquares of the absolute individual errors(E) are added and the square root of thissum is taken:

Rule 3:Law of error propagation for productsand quotients:In products or quotients, the square ofthe relative individual errors are addedand the square root of this sum is taken:

1 n____ s (i -__)2n-1

i=1

1 n____ s in i=1

(E1)2 + (E2)

2

(1 mg)2 + (1 mg)2= 1.4 mg

ENet

1.4 mg___ = _______ = 0.028% = 2.8 10-4mNet 5.042 g

E1 2 E2 2(_______) + (________)Value1

Value2

m = _____

V

Fm 2 Fv 2(____)+ (___ )m v

0.01 g 2 0.1 cm3 2

(__________) + (__________ __)150.27 173.4 cm3 =


24/28

Deriving the Uncertainty of Measure-ment from the Standard DeviationOnly approximately 68% of the measuredresults lie within the range of s from the

mean value. Therefore, for practical purpo-ses, the (maximum) uncertainty of measu-rement u is frequently defined as 2 s(95% of the measured results lie within therange of 2 s from the mean value).

We will use: u = 2 s

Example for Calculating the Uncertaintyof Measurement for Samples of Approx.10 g:Small amounts (approx. 10g) are to beweighed on a GENIUS ME2545 semi-

microbalance with a resolution of 0.1 mg.Ambient conditions are good (no tilting;temperature difference of 5C max.; noneof the containers or objects is electro-statically charged, nor is there anyelectromagnetic interference.)

The containers are small and must becorrectly centered, as directed in thestandard operating instructions.Therefore, the off-center load errorcan be neglected for 10 g.

The repeatability/standard deviation is:< 0.07 mg

The temperature coefficient for thesensitivity is 1ppm/K => < 110-6 /C,as stated in the technical specifications.Hence, the error for 10 g and T = 5C is< 10 g 1 10-6 /C 5C =< 0.05 mg

The max. linearity error is as stated inthe technical specifications:< 0.15 mg

The balance has been calibrated andadjusted with a standard E2 class weightof 200 g (maximum error of 0.3 mg).

In relation to a 10-g load, the error is:< 0.015 mg

The samples density is 2.0g/cm3, with anuncertainty of 20%; the differencebetween air buoyancy of the samples andthat of the standard weights used to adjust

the balance is thus 2.25 mg with anuncertainty of 20% 0.45 g.

The uncertainty of this air buoyancy cor-rection value due to fluctuations in the airdensity of 10% is considerably less thanthat of density fluctuations.

Deriving the Uncertainty of Measure-ment from the Standard DeviationWith the exception of the repeatability|standard deviation, all values are maximum

errors. If the equation of u=2s is used toexpress the maximum uncertainty of therepeatability and if the air buoyancy hasbeen corrected, the uncertainty ofmeasurement will be as follows:

However, if no correction is made for airbuoyancy, a systematic error of 2.25 mg isadded to the uncertainty of measurementu so that the total deviation can be asmuch as 2.75 mg.

The uncertainty of measurement of aweighing instrument can be exactlydetermined over its entire weighing rangeby calibration in a DKD*-accreditedlaboratory, which Sartorius has.

*DKD = German Calibration Serviceofficially recognized throughoutEurope

24

u=(2 0.07 mg)2 + (0.05 mg)2 + (0.15 mg)2 + (0.015 mg)2 + (0.45 mg)2

u= 0.50 g


25/28

Adjustment 15Air buoyancy 11Ambient conditions 19Appointment and identification 20

of personnel responsible formonitoring test equipment

Barometric pressure 11

Calibration 15Calibration results 19Consequential damage 20

Defined maximum permissible errors 19Demands of customers 21Description of the test and 19measuring equipment

Design qualification (DQ) 6Determination of the uncertainty 7of measurementDocumentation 19Drift in the zero point 9

EN 45000 series 5Environmental influences 21Equipment qualification 6Error calculation 22Experience with similar test 21and measuring equipmentExternal calibration|adjustment 15

GLP (Good Laboratory Practice) 5GMP (Good Manufacturing Practice) 5Gravitational acceleration 10

Humidity 11Hygroscopic samples 14

Influence quantities 9Installation qualification (IQ) 6Internal calibration|adjustment 15Interval of confirmation 20ISO 9000 series 5

Legally regulated quality systems 5Leveling 10Linearity error, linearity 8

Magnetic and magnetizable samples 13Maintenance procedures 20Manufacturers recommendation 21Mass and weights 17Mechanical disturbances 11Modification of the weighing 20instruments

Non-conforming test and 20measuring equipment

Off-center load error 9

Operational qualification (OQ) 6Operator 10Overall uncertainty of measurement 19

Performance qualification (PQ) 6

Quality 4Quality systems 5

Random errors 22

Sample 13Selection of suitable test and 6

measuring equipmentSensitivity 9Sensitivity error 9Standard deviation 7Static electricity 13Structure of mass standards 17Systematic errors 22

Temperature 14Temperature coefficient 9Tendency toward component wear 21Tendency towards drift 21Test methods 7

Traceability of a measurement 15

Uncertainty of measurement, 24deriving from the standard deviationUncertainty of measurement, 24example for calculatingUniversal quality systems 5

Weighing location 10Weighing range 7

Index

25


26/28

26

References

(German titles have been translated intoEnglish in parentheses for convenience.)

Christ, G.A., Harston, S.J.; Hembeck,

H.-W., Opfer, K.-H.1998. GLP-Handbuchfr Praktiker (GLP Handbook forExperienced Professionals). Darmstadt,Germany: Gt Verlag GmbH.

Deutsches Institut fr Normung e.V.(German Institute for Standardization).1995. Leitfaden zur Angabe der Unsi-cherheit beim Messen (Guidelines forIndicating the Uncertainty duringMeasurement). Berlin, Germany.

Deutsches Institut fr Qualitt e.V.(German Society for Quality). 1998.

Prfmittelmanagement (Managementof Inspection, Test, and MeasuringEquipment). Frankfurt, Germany.

DIN ISO 10012. 1996. Forderungen andie Qualittssicherung fr Messmittel,Messunsicherheit und Fhigkeit, Qualittund Zuverlssigkeit (Quality AssuranceRequirements for Measuring Equipment,Uncertainty of Measurement, andCapability, Quality and Reliability).Geneva, Switzerland: InternationalOrganization for Standardization.

Verein deutscher Ingenieure (Associationof German Engineers). 1998. Prfmittel-management und Prfmittelber-wachung (Management and Controlof Inspection, Test, and MeasuringEquipment). Dsseldorf, Germany.

Weyhe, S. 1997. Wgetechnik im Labor(Weighing Technology in theLaboratory). Landsberg/Lech, Germany:

Verlag Moderne Industrie.


27/28

27


28/28

Sartorius AGWeender Landstrasse 9410837075 Goettingen, Germany

Phone +49.551.308.0Fax +49.551.308.3289

www.sartorius-mechatronics.com

Specifications subject to change without notice.Printed in Germany on paper that has beenbleached without any use of chlorine.

bro-test equipment in a quality system-e

Documents