MEASUREMENT TRACEABILITY AND ERRORS RELATED TO FORCE MEASUREMENT

Transcription

1 MEASUREMENT TRACEABILITY AND ERRORS RELATED TO FORCE MEASUREMENT Henry Zumbrun II, President Morehouse Instrument Company 1742 Sixth Ave York, PA PH: web: info: The guy talking right now 1

2 Morehouse is committed to the following: 0 helping companies lower measurement uncertainties for force and torque. 0 offering products with the lowest measurement uncertainties available. 0 helping our customers make more accurate measurements, which save costs, reduce risk, and increase quality. 2

3 Abstract This session will cover the following: Metrological traceability in relation to force measurement Common force measurement errors and the importance of calibrating the instrument in the manner it is being used 3

4 Learning Objective By the end of the webinar the attendee should: Understand the metrological traceability hierarchy. Identify potential force measurement errors. Know the importance of using the proper calibration fixtures 4

5 Best Practices For Force Measurement Does anyone want to guess what best practice looks like? Morehouse believes best measurement practice is talking to the customer (understanding more than their requirements) and trying to replicate via calibration, how the equipment is being used. Best practice involves making sure your measurements are traceable. 5

6 Measurement Traceability The seven essential elements: 1. Reference to International System of Units (SI) 2. Unbroken chain of comparisons 3. Measurement Uncertainty 4. Competence 5. Documentation 6. Calibration Intervals 7. Measurement Assurance 6

7 1. SI Units The Seven Elements Per NIST SP811, Reference to SI units: the chain of comparisons must, where possible, end at primary standards for realization of the SI units. The SI base units are a choice of seven well-defined units which by convention are regarded as dimensionally independent: the metre the second the kilogram the ampere the kelvin the mole the candela. Note: Derived units are formed by combining the base units according to the algebraic relations linking the corresponding quantities. Force = mass (Kilogram) x Acceleration (metre per second squared) 7

8 2. Metrological Traceability Property of a measurement result whereby the result can be related to a reference through a documented unbroken chain of calibrations, each contributing to the measurement uncertainty. (VIM 2.41) (International Vocabulary of Metrology) 8

9 2. Metrological Traceability Metrological traceability requires an established calibration hierarchy Measurement Uncertainty Budget Worksheet Laboratory Parameter FORCE Range 2K-100K Morehouse Sub-Range Technician HZ Standards Date 4/18/2011 Used LOAD CELL S/N U-7660 CALIBRATED WITH DW MACHINE M-7471 Uncertainty Contributor Magnitude Type Distribution Divisor df Std. Uncert Variance % (Std. Contribution Uncert^2) u^4/df None Repeatability E-3 A Normal E E % 800.0E-6 NIST Uncertainty (k=2 100k lbf) E-3 B Expanded (95.45% k=2) E E % 6.8E-6 Resolution E-3 B Resolution E E % 3.5E-9 B None B None B None None None None Combined Uncertainty (u c ) = Effective Degrees of Freedom Coverage Factor (k) = Expanded Uncertainty (U) 100k = E E % 806.8E E % Slope Regression Worksheet Worksheet Applied Run 1 Run 2 Run 3 Average Std. Dev. Error Calculated Residuals #DIV/0! Example of Measurement Uncertainty Worksheet for Primary Standards. Calibration of Primary Standard (weights) was performed by N.I.S.T. 9

10 2. Metrological Traceability Calibration and Measurement Capability (CMC) is a per point analysis. It represents the measurement process with the best test instrument and a repeatability study done on a per point basis with at least three runs of data. CMC is required for ISO/IEC accreditation. On the Morehouse deadweight standards, the following was used to determine the CMC: Repeatability of a Test Instrument (TI) in the dead weight machine Resolution of the TI NIST Uncertainty for the weights used at the point tested (Includes air buoyancy correction, stability, wear, and local gravity correction). 10

11 Uncertainty 3. Measurement Uncertainty s Relation to Measurement Hierarchy BIPM/SI National Metrology Institute (NMI) Primary Reference Laboratory Morehouse Instrument Company Accredited Calibration Service Supplier Typical Uncertainties for Force Measurement k =1 N.I.S.T = % Morehouse = % Accredited Calibration Supplier = 0.02 % Working Standards = 0.1 % Field Measurement = 0.5 % Working Standards Instrument/Equipment 11

12 Uncertainty 3. Measurement Uncertainty s Relation to Measurement Hierarchy BIPM/SI National Metrology Institute (NMI) BIPM is the highest level on the chain of traceability and where 7 SI units are realized. From the BIPM, The international prototype of the kilogram, an artefact made of platinum-iridium, is kept at the BIPM under the conditions specified by the 1st CGPM in 1889 when it sanctioned the prototype and declared. Primary Reference Laboratory Morehouse Instrument Company In the United States the National Institute of Standards and Technology (N.I.S.T.) is the designated NMI Primary Reference Labs use dead weight Primary Accredited Calibration Service Supplier Working Standards Instrument/Equipment Standards, usually calibrated by N.I.S.T. as the reference standard. Morehouse uses Primary standards for Force measurements up to 533 kn (120,000 LBF). Accredited Calibration Suppliers use secondary standards (those calibrated by primary standards) Working standards and certified reference materials used in commerce and industry. Traceable to above These are typically the material testing machines used for tensile, compression, fatigue, impact, rheology, structural, materials or a range of other force applications. 12

13 Measurement Uncertainty CMC is defined as Calibration and Measurement Capability. It includes the following standard uncertainty contributors: Repeatability Resolution Reproducibility Reference Standard Uncertainty Reference Standard Stability Environmental Factors Note: I call these 5 R s and an E 13

14 3. Measurement Uncertainty Let s examine CMC (Calibration Measurement Capability) using a primary standard, as the reference and how it affects the Expanded Uncertainty. Morehouse Primary Standard as the Reference (CMC % for K=2 or K) Measurement Uncertainty Budget Worksheet Laboratory Parameter FORCE Range 10K Morehouse Primary Standards Sub-Range Technician HZ Standards Date Used Uncertainty Contributor Magnitude Type Distribution Divisor df Std. Uncert Variance (Std. Uncert^2) % Contribution Reproducibiliy E+0 A Normal E E % 000.0E+0 Repeatability E-3 A Normal E E % 2.2E-6 U-7643 LLF E-3 A Normal E E % 89.3E-9 Resolution of UUT E-3 B Resolution E E % 3.5E-9 Environmental Conditions E-3 B Rectangular E E % % 17.6E-9 Stability of Ref Standard E-3 B Rectangular E E % 3.8E-6 Contribution Ref Standard Resolution E-3 B Resolution E E % 11.5E-12 None Morehouse CMC E-3 B Expanded (95.45% k=2) E E % 204.8E-9 Combined Uncertainty (u c )= Effective Degrees of Freedom Coverage Factor (k) = Expanded Uncertainty (U) K = u^4/df E E % 6.4E % 14

15 3. Measurement Uncertainty Let s examine CMC (Calibration Measurement Capability) using a secondary standard, as the reference and how it affects the Expanded Uncertainty. Accredited Calibration Supplier with Secondary Standards as the Reference (CMC 0.04 % for K=2 or 4 LBF) Measurement Uncertainty Budget Worksheet Laboratory Parameter FORCE Range 10K Morehouse Primary Standards Sub-Range Technician HZ Standards Date Used Uncertainty Contributor Magnitude Type Distribution Divisor df Std. Uncert Variance (Std. Uncert^2) % Contribution Reproducibiliy E+0 A Normal E E % 000.0E+0 Repeatability E-3 A Normal E E % 4.1E-3 U-7643 LLF E-3 A Normal E E % 89.3E-9 Resolution of UUT E-3 B Resolution E E % 3.5E-9 Environmental Conditions E-3 B Rectangular E E % % 17.6E-9 Stability of Ref Standard E-3 B Rectangular E E % 3.8E-6 Ref Standard Resolution E-3 B Resolution E E-6Contribution 0.00% 11.5E-12 None Accredited Cal Supplier CMC E+0 B Expanded (95.45% k=2) E E % 80.0E-3 Combined Uncertainty (u c )= Effective Degrees of Freedom Coverage Factor (k) = Expanded Uncertainty (U) K = u^4/df 2.04E E % 84.1E % 15

16 3. Measurement Uncertainty Let s examine CMC (Calibration Expanded Uncertainty Measurement when Capability) calibrated and with what the Reference CMC does to the calibration Primary results. Standards Morehouse is approximately Versus Accredited 10 times Cal Supplier lower than using secondary standards Expanded 10K = 0.41 Expanded 10K = 4.03 LBF LBF Morehouse CMC = 0.16 LBF Accredited Cal Supplier CMC = 4.00 LBF Repeatability = LBF Repeatability = LBF 16

17 Calculating Force CMC s Guidance Documents 0 NCSLI RP-12 0 Lack of proper guidance document for non ASTM E74 0 ASTM E74 Appendix combined with A2LA R205 17

18 Force CMC for ASTM E74 Calibrations We will need the following: 1. Calibration Report for the Device. 2. The uncertainty of the instrument(s) that were used to perform the calibration 3. Calibration History (if available) 4. Manufacturer s Specification Sheet 5. Error Sources, if known The end user will then have to conduct the following tests: 1. Repeatability study 2. R & R between technicians 3. Complete Proficiency Testing Requirements 18

19 3. Measurement Uncertainty Morehouse has prepared a Measurement Uncertainty Calibration and Measurement Capability Excel Worksheet for anyone doing force measurements. Morehouse Measurement Uncertainty Calibration and Measurement Capability Worksheet START ON THIS SHEET AND FILL IN ONLY LIGHT GREY BOXES SECTION 1 DATA ENTRY NOTE: ONLY ENTER INFORMATION IN LIGHT GREY BOXES Laboratory Morehouse Ref Standard Stability Temperature Technician Initials HZ All information entered must converted to like units. FORCE Change From Interporlated Actual Effect Date: 2/26/2016 This spreadsheet is provided by Morehouse Instrument Company APPLIED Previous % 0 LBF Range 1K-5 K It is to be used as a guide to help calculate CMC % Standards Used Ref and UUT Ref S/N U-7644 UUT S/N Test % % Resolution UUT 0.1 LBF This is the resolution of the Unit Under Test you are Using for the Repeatability Study (What you are testing) % % REFERENCE STANDARD INFORMATION % ASTM E74 LLF * LBF * This is your ASTM E74 LLF Found on Your ASTM E74 Report. It will be converted to a pooled std dev (drop down for non ASTM) % Resolution of Reference LBF This should be found on your calibration report % Temperature Spec per degree C % % This is found on the load cell specification sheet. Temperature Effect on Sensitivity, % RDG/100 F % % Max Temperature Variation 11 per degree C of Environment 1 During a typical calibration in a tightly controlled the temperature varies by no more than 1 degree C. 12 Morehouse CMC % This is the CMC statement for the range calibrated found on the certificate of calibration. Leave blank if entering Eng. Units Miscellaneous Error % This can be creep, side load sensitivity or other known error sources. Enter and select Eng. Units or % Conv Repeatability Data To Eng. Units YES Repeatability of UUT Ref Laboratory Uncertainty Per Point MUST SELECT Applied Run1 Run2 Run3 Run4 Average Resolution STD DEV CONVERTED Force % Eng. Units Conv % Force % or Eng % % % % % % % % % % % % % % % % % % % % % % % % Avg Std Dev of Runs

20 3. Measurement Uncertainty SECTION 2 DATA ANALYSIS DISTRUBUTION IS THE ONLY COLUMN IN SECTION 2 THAT MAY NEED CHANGED Laboratory Section 2 Measurement Uncertainty Budget Worksheet Data Analysis (Nothing below should need filled out) Morehouse Miscellaneous error % Parameter FORCE Range 1K-5 K Sub-Range N/A Force % Reading Technician HZ % Date 2/26/2016 Standards Used Ref S/N U-7644 UUT S/N Test % Uncertainty Contributor Magnitude Type Distribution Divisor df Std. Uncert Variance (Std. % Uncert^2) Contribution u^4/df % Reproducibility (see R & R sheet) E+0 A Normal E E % 2.8E % Repeatability E-3 A Normal E E % 945.8E % Standard Deviation E-3 A Normal E E % 429.1E % Resolution of UUT E-3 B Resolution E E % 3.5E % Environmental Conditions E-3 B Rectangular E E % 18.5E % Stability of Ref Standard E-3 B Rectangular E E % 281.3E % Ref Standard Resolution E-3 B Resolution E E % 9.7E % Miscellaneous Error E-3 B Rectangular E E % 45.0E Morehouse CMC E-3 B Expanded (95.45% k=2) E E % 12 Combined Uncertainty (u c )= 2.31E E % 2.8E+0 NOTE: ONLY ENTER INFO IN GREY BOXES IN SECTION 1 UNLESS CHANGING DIST 4 Effective Degrees of Freedom 10 Formula uc ( y) Veff = N 4 4 Coverage Factor (k) = 2.23 ci u ( xi ) The grey column Ref CMC is what å Expanded Uncertainty (U) = % i= 1 vi populates individual sheets Slope Regression Worksheet Applied Run 1 Run 2 Run 3 Run 4 Average Std. Dev. Error Calculated Applied Ref CMC Repeatability (Of Error) E-14 Average Standard Deviation of Runs Slope 1 Intercept 0 Regression

21 3. Measurement Uncertainty Common Issues 1. CMC values are unrealistic 2. Measurement Uncertainties reported to end customer are less than the scope CMC. Morehouse provides calibration for X customer. Based on our knowledge the best X customer can obtain is 0.01 % of applied. It is more probable they are at 0.02 % Then X customer calibrates the device for another lab y. Y lab then claims values of above. 2 chain links away from primary standards and better CMC s than the secondary standards labs. 21

22 3. Measurement Uncertainty 0 Example where the Expanded Uncertainty is less than the Reference Standard Uncertainty % not 99 %

23 Question Is your calibration service provider meeting traceability requirements, which include properly reporting measurement uncertainty? a. Yes b. No c. Maybe 67 % 4 % 29 % Note: If Measurement Uncertainty is not being provided by your service provider, there is NO measurement traceability! 23

24 4. Competence Competence: The laboratories or bodies performing one or more steps in the chain must supply evidence for their technical competence (e.g. by documented proof, the service provider is accredited); Importance: If the laboratory staff performing calibrations do not know how to calibrate a device ie using improper adaptors, not aligning the Unit Under Test (UUT) properly, and improper loading procedures Are they competent? Can there be confidence in the reported results? How to provide evidence of competence; proof of training (records), OJT and evaluation by management. This is still questionable as it is self declaration. Accreditation to ISO/IEC is the best evidence available. Note: Calibrating an artifact with primary standards and then performing an ILC is the best way to ensure you are making measurements within your stated Calibration and Measurement Capability. 24

25 4. Competence Competence: The laboratories or bodies performing one or more steps in the chain must supply evidence for their technical competence (e.g. by documented proof, the service provider is accredited); In this section we are going to discuss the following: 0 Proper adapters 0 Common force measurement errors 25

26 Competence and Measurement Error Examples of Competency include: Using proper adaptors when calibrating force instruments. Improper adaptors can produce errors times that of manufacturer s stated accuracy. Proper alignment of UUT (Unit Under Test), adaptors, and proper methods for loading threads. Misalignment, different hardness of adaptors, and thread loading versus shoulder loading, contribute to a decrease in the repeatability of measurement results, resulting in additional measurement error. Repeatability and Reproducibility Tests, as well as, Proficiency Tests are good methods for detecting measurement errors. 26

27 ISO 376: 2011 ISO 376 recognizes the importance of adapters in reproducibility conditions of the measurement. Proper adaptor use in accordance with ISO 376 Annex A, helps ensure the reliability of reported measurements. Note: Annex A is not a requirement for labs to adhere to. A.4 Loading fittings A.4.1 General 0 Loading fittings should be designed in such a way that the line of force application is not distorted. As a rule, tensile force transducers should be fitted with two ball nuts, two ball cups and, if necessary, with two intermediate rings, while compressive force transducers should be fitted with one or two compression pads. 27

28 Tension Links 28

29 Tension Links Potential Error Source Loaded without the proper pin diameter to 50,000 LBF Sometimes referred to as load links, dynamometers, or crane scales 29

30 Tension Links Potential Error Source Loaded with the proper pin diameter to 50,000 LBF Manufacturer s Accuracy Specification is 0.1 % of full scale. What if you are using a device like this for a critical weighing application? 30

31 Tension Links Improper Vs Proper Pin Diameter Difference of 860 LBF or 1.72 % error at 50,000 LBF from not using the proper size load pins. Out of Tolerance Versus In Tolerance Note: Tension links of this design seem to exhibit similar problems. If you are unsure, TEST! 31

32 Tension Links Good measurement practice This following summary is from Dillon. Using correctly sized pins is critical. If links are damaged, highly used, or worn, decrease the time between recalibrations. The same size and style of shackle and pin used during operation should be used for calibration. Other factors have a larger effect on accuracy than pin rotation. Maintaining pin orientation may be best practice, but is not required to stay in tolerance. 32

33 Misalignment on S-beam Misalignment Video Demonstrating 0.75% error 33

34 Misalignment on S-beam Misalignment Demonstrating % error Output in mv/v Aligned in machine mv/v Output in mv/v Slightly misaligned in machine mv/v 34

35 Misalignment on 10,000 LBF S-beam Misalignment Demonstrating % Error Output in mv/v Aligned in machine mv/v Expanded Uncertainty 9.95 LBF Output in mv/v Slightly misaligned in machine mv/v Expanded Uncertainty 85.0 LBF 35

36 Misalignment on Shear Web Cell Misalignment Video Demonstrating % error 36

37 Different Hardness of Top Adaptors Example: A customer brought in a 1,000,000 LBF load cell for calibration. Morehouse performed a calibration. The output of the load cell was recorded as 1,500 LBF higher than the previous calibration for a force applied 1,000,000 LBF. Is this a stability issue, or an adaptor issue? After calling the customer, we were informed a new top loading block was supplied with this load cell for the current calibration. When we told them what was happening, they sent the original top loading block. When tested, the original block resulted in an output of 1,000,180 LBF when loaded to 1,000,000 LBF. When using the new adaptor and figuring the measurement error between the different top blocks (adaptors), Expanded Uncertainty would have increased from 269 LBF with original top adaptor to 1,490 LBF using the newly fabricated adaptor. 37

38 Loading Through Different Thread Depths Below is a test Morehouse did with two different types of adaptors and recorded the readings with 10,000 LBF applied. Output was 10,001.5 LB with 1.5 of engagement vs LBF with 0.5 engagement. There was a difference of 59.2 LBF on a 10,000 LBF cell. Column Type Cell Model RFG/F Different Type Adaptors. 1.5 versus 0.5 engagement The error on this measurement was over 5 % on a device expected to be better than 0.025%. How are you or how are your devices being calibrated? 38

39 Different Thread Depths The top light grey line shows the difference in output when loading a shear web load cell at different thread depths The darker bottom line shows the difference in output when loading against the load cell shoulder. 39

40 Different Thread Depths 40

41 Loading through the bottom threads 41

42 Loading through the bottom threads 42

43 Non Flat Base Error associated with installing a non flat base on a multi-column cell. This is an actual test result we observed on a Revere multi-column cell. Non Flat Base Flat Base Maximum Error Maximum Error Force Applied In Rotation In Rotation LBF LBF % error % error % 0.013% % 0.016% % 0.023% 43

44 Cable Length Error Load cells used with meters that have a 4-wire configuration are subject to additional error. This is because of voltage drop over cable lengths, and the effect on thermal span characteristics of the load cell, as temperature changes can alter cable resistance. 44

45 Cable Length Error Substitution of a 4-wire cable at a given length with another 4-wire cable of a different length or gauge will produce additional errors. (Recalibration will be required) 45

46 Cable Length Error What you need to know about 4 wire systems. 1. If you damage or replace your cable, the system may need to be calibrated immediately following replacement or repair. 2. Operating at different temperatures will change the resistance, which will cause a voltage drop, resulting in a change of measured output. 3. Cable substitution will result in additional error and should be avoided. 4. Cables used for 4-wire systems should have a S/N, or a way to make sure the same cable stays with the system, it was calibrated with. - This would be a Good Measurement Practice Technique Morehouse highly recommends. 46

47 Cable Length Error To eliminate the errors associated with a 4-wire system requires a 6-wire cable, which is run to the end of the load cell cable or connector, and is used with an indicator that has sense lead capability. With a 6-wire setup, the sense lines are separate from the excitation lines, thereby eliminating effects due to variations in lead resistance. This allows long cable runs in outdoor environments with extreme temperatures. 47

48 Cable Length Conclusion A 4-wire cable cannot be interchanged without requiring the system to be recalibrated. A 6-wire cable will yield similar readings, regardless of length and gauge. We performed a test with a load cell and a 6 foot 4 wire cable, output recorded was mv/v. We then the same test using a 13 foot 4 wire cable, and recorded the output at mv/v. The observed a difference from interchanging 4 wire cables a the same gauge wire was % 48

49 Using only part of the calibrated range Not exercising the load cell to full range may produce additional errors. 0 The load cell exhibited a decline in output, which correlated to the amount of time between the additional applications of forces. The potential error ranged from % to %. This error could be considerable when using the load cell as a secondary reference standard to calibrate other load cells. A Secondary Standard, as defined by ASTM E74-13a, is one that is calibrated by Primary Standards (deadweights) and has a test accuracy ratio of better than 0.05 %. A maximum difference of % was observed. 49

50 Decreasing loading If a load cell is to be used to make descending measurements, it must be calibrated with a descending range Ascending Descending The difference in output on an ascending curved versus a descending curve can be quite different. A very good 100K load cell had an output of on the ascending curve and on the descending curve. Using the ascending only curve would result in an additional error of %. 50

51 Measurement Error Information To learn more in depth knowledge about force measurement visit our website at:

52 Proper Fixtures Conclusion Communication with the customer is key to address these issues. Unfortunately, this does not always happen. Examples of this scenario are as follows: 0 3 rd party suppliers 0 Purchasing Departments 0 Management who does not care 0 Large Companies where it is difficult to reach the technician using the device. To minimize these errors, the Ideal solution would be to calibrate the device with the customer s adaptors or have the customer send the appropriate adaptors to the reference lab for calibration. 52

53 Common Torque Issues 53

54 5. Documentation Documentation: each step in the chain must be performed according to documented and generally acknowledged procedures; and the results must be recorded. ISO/IEC 17025, The laboratory shall use the appropriate methods and procedures for all tests and/or calibrations within its scope. Importance: The calibration must be performed using an acceptable and agreed upon calibration method or procedure (Contract Review). The results must be provided in a report with necessary information. What could go wrong?: Procedures not validated that do not cover all parameters and ranges (partial or limited calibration). This includes software. Results not reported accurately, clearly, unambiguously and in accordance with any specific instructions in the calibration method. Not following or properly understanding documented standards such as ASTM E74 54

55 5. Documentation In this section we are going to discuss the following: 0ASTM E74 Standard requirements that accredited labs often do not follow. 55

56 ASTM E74 Calibration 0 The Class A or Class AA loading range cannot be less than the first applied non zero force point (400 x = 52.8) 0 Per Section 8.6 of ASTM E74-13a The loading range shall not include forces outside the range of forces applied during the calibration 0 Per Section of ASTM E74-13a states In no case should the smallest force applied be below the lower limit of the instrument as defined by the values: 400 x resolution for Class A loading range & 2000 x resolution for Class AA loading range 56

57 ASTM E74 Calibration 0 It is recommended that the lower force limit be not less than2%( 1 50) of the capacity of the instrument. 0 Per Section If the lower limit of the loading range of the device (see 8.6.1) is anticipated to be less than one tenth of the maximum force applied during calibration, then forces should be applied at or below this lower limit 57

58 ASTM E74 Calibration Data Analysis 0 Deviations from the fitted curve 0 These are the differences between the fitted curve and the observed values 0 Standard Deviation is the square root of the sum of all the deviations squared/n-m-1 0 N = sample size, m = the degree of polynomial fit 0 Calibration equation Deflection or Response = A0+A1(load)+A2(load)^2+ A5(load) ^5 0 LLF is 2.4 times the standard deviation 0 Class A range is 400 times the LLF. Class AA range is 2000 times the LLF. 58

59 Test Accuracy Ratio ASTM E74 PRIMARY STANDARDS % SECONDARY STANDARDS CLASS AA 0.05 % WORKING STANDARDS CLASS A 0.25 % TESTING MACHINE 1 % Primary Standards are required to assign an ASTM Class AA loading range Those Secondary Standards with a Class AA range can only assign a Class A loading range Class A devices are then used to calibrate the testing machine to ASTM E4 Other types of calibrations performed may not adhere to any TAR, nor should they. The challenge is figuring out if they are calculating their CMC s properly. 59

60 ASTM E74 STANDARD 0 Remember Customer X Scope Example 60

61 Calibration In Accordance with ASTM E74 Secondary Force Standard an instrument or mechanism, the calibration of which has been established by comparison with primary force standards. Criteria for Lower Load Limit 0 LLF = 2.4 * STD DEV This corresponds to a 98.2 % Coverage Factor 0 Based on LLF or Resolution whichever is higher 0 Class A 400 times the LLF or resolution 0 Class AA 2000 times the LLF or resolution 61

62 ASTM E74 STANDARD 0 ASTM LLF is not the expanded uncertainty!! 62

63 ASTM E74 STANDARD 0 The Expanded Uncertainty is less than the Reference Standard Uncertainty % not 99 %

64 Scope Issues Example: 0 An AB acquires a new lab to assess. This lab is performing force calibration and is referencing the ASTM E74 standard using a set of load cells, 100K,50K,10K and 1K. This lab had their standards calibrated by another accredited lab using secondary standards. This lab becomes accredited and continues to perform Class A calibrations. What does go wrong! 0 The reference lab would have to be using primary standards to assign a Class AA loading range. 0 The Class AA loading range allows this lab to calibrate devices and assign a Class A loading range. 0 In the example above, this lab would only be able to use their load cells for ASTM E4 (Field Testing) and not ASTM E74 calibrations. 64

65 6. Calibration Intervals ( 90 Days? 1 Year? 180 Days? 3 Years.?) What is your risk level? Calibration intervals: calibrations must be repeated at appropriate intervals; the length of these intervals will depend on a number of variables (e.g. uncertainty required, frequency of use, condition used, stability of equipment). ISO/IEC ( ) A calibration certificate (or label) shall not contain any recommendation on the calibration interval except where this has been agreed with the customer. Importance: If inappropriate calibration intervals are used, the reliability of results reported are questionable. Profitability can be reduced if intervals are too short. What could go wrong? Inappropriate intervals can result in unknown drift, increased Measurement Uncertainty (MU) and incorrect claim of MU reported. Note: ASTM E74 and ISO 376 are standards that address recalibration intervals. For ASTM E74-13a See Section for appropriate interval criteria. 65

66 There are two certificates above. One is in 2004 and another one in

67 Calibration Intervals - Proving Rings % Diff % % % % % % % % % % 12 Year Change From Previous. Note: Morehouse does not recommend 12 year calibration intervals. 67

68 7.Measurement Assurance Measurement Assurance Practices put in place to monitor a testing or calibration process and to ensure the calibration status of equipment, reference standards or reference materials used in a measurement process. A measurement assurance program: needs to exists for each step in the measurement hierarchy to ensure the competency of measurement results. (Proficiency Testing (PT) - Inter-laboratory Comparisons (ILC) Intermediate checks ) SPC Statistical Process Control ISO/IEC 17025, Section When intermediate checks are needed to maintain the confidence in the calibration status of the equipment, these checks shall be carried out according to a defined procedure. 68

69 7.Measurement Assurance Measurement Assurance Practices put in place to monitor a testing or calibration process and to ensure the calibration status of equipment, reference standards or reference materials used in a measurement process. In this section we are going to discuss the following: 0 Force Verification Kits 0 ILC Tests 0 SPC 69

70 Verification Kits Measurement Assurance Practices put in place to monitor a testing or calibration process and to ensure the calibration status of equipment, reference standards or reference materials used in a measurement process. In this section we are going to discuss the following: 0 Force Verification Kits 0 SPC 0 ILC Tests 70

71 Verification Kits Monitoring your process by putting practices in place to ensure that your measurements are accurate is essential to limiting your risk and keeping your bottom line intact. A good stable force measurement system can be used to do the following: (1) Force Verification, (2) SPC Statistical Process Control, (3) ILC Intra- Laboratory Checks, (4) Proficiency Testing, and (5) A Test Standard to do repeatability and reproducibility tests used to calculate Calibration and Measurement Capability (CMC). 71

72 Verification Kits Statistical Process Control (SPC) This process is similar to verification, with the exception of a documented control process in which an artifact is used to monitor performance of the measurement process. A good load cell system can be used as a check standard to monitor that the process is in control. It can provide the objective evidence and reduce risk. If the process is continually monitored and an out-of-control situation is found, the root cause analysis can be performed to ensure proper corrective action before the machine or process actually goes out of tolerance. 72

73 Verification Kits Intra-Laboratory Checks (ILC) The force system can be used to compare machines, operators or processes. If you are using control charts and the process output is approaching control limits, the system can be used to test what the issue is and to determine which machine, operator or process needs to be corrected. 73

74 Verification Kits Repeatability & Reproducibility A device with very high resolution and low overall uncertainty will allow the end user to lower their Calibration and Measurement Capability (CMC). When calculating CMC, the resolution of the system being used must be figured into the calculations. The lab will need to perform repeatability studies. An artifact with low sensitivity to side loading, temperature compensation and stability will be a lab s best asset. A good system will often decrease the variation in output between multiple measurements. It will also allow the lab to test the true performance between technicians. R & R data may be derived from control charts if they are set up properly. 74

75 Consequences from lack of Measurement Traceability: 0 Without all seven elements in place, your measurements are not traceable. 0 What measurement errors do. They raise the risk level for product failures, false accepts, and false rejects. 0 If taken to court, are your measurements defensible? If you are missing any elements in the measurement traceability chain, No. 75

76 Consequences from lack of Measurement Traceability: 76

77 Celebration of Knowledge! The seven essential elements of traceability 1. Reference to International System of Units (SI) 2. An unbroken chain of comparisons 3. Measurement Uncertainty 4. Competence 5. Documentation 6. Calibration Intervals 7. Measurement assurance 77

78 Celebration of Knowledge! Metrology is not Meteorology! Even though Google Gmail tells me is not a word. If you do not know how the instrument should be calibrated what should you do? - Contact the user and find out how the MT&E (Measuring and Test Equipment) is being used! IE element # 5 Documentation Contract Review 78

79 Celebration of Knowledge! Thank You. If you want to learn more please visit read our Blog, and sign up for our newsletter 79