State of progress of dynamic calibration of force, torque and pressure sensors including conditioners

Transcription

1 State of progress of dynamic calibration of force, torque and pressure sensors including conditioners EMRP Project IND 09 : «Traceable dynamic measurement of mechanical quantities» Claire Bartoli, M.Florian Beug, Thomas Bruns, Loic Coquelin, Clemens Elster, Trevor Esward, Leonard Klaus, Andy Knott, Michael Kobusch, Anti Lakka, Anne Francoise Obaton and Christian Schlegel

2 For further information on the European Metrology Research Programme, contact EURAMET at their booth in the Metrology Village or at 2

3 A few words to introduce the project What quantities are we concerned with? Force Pressure Torque What budget? 3584 k or 242 PM How long will the project last? 3 years, the planned end is September 2014 Which countries are involved? Czech Republic Finland France Germany Italy Spain Sweden Turkey United kingdom 8 industrial partners : VSL; HBM; Spektra; Porsche; Volkswagen; AEP; Scandura; Simea 3

4 The challenges and the problems to be solved The transducer has a distinctive dynamic response to dynamic signals The structure has a dynamic response to excitation The desired measurand is not located at the sensor Provide primary procedures and/or devices for calibration of force torque and pressure Define the methods and procedures, then establish them in documentary standards Develop methods for dissemination into the field Validate by comparison at a higher scale 4

5 Our proposal : Set up a parameterized model Measure / identify the (model) parameters of the response Isolate and document the parameters of the transducer including uncertainties Calculate the measurand from the measured response using the new model (input prediction) Calculate the uncertainty from parameter uncertainties and additional uncertainty sources 5

6 Organization of the project 6

7 WP 1 : Sine force measurements at CEM System : Shaker with 6500 N and freq range 10 to 2400 Hz 3 sensors evaluated 5 masses from 0,4 to 12 kg Evaluation of : Influence of acceleration distribution on the top of the mass (rocking motion) Reproducibility tests Mounting torque * cf. presentation of Nieves Medina

8 WP 1 : Sine force measurements at LNE * cf. presentation of Loic Coquelin 8

9 DATA AVAILABLE Mass 400g 27 data sets Variation due to the mounting and to the position of the accelerometer on the shaker table Mass 1 kg 27 data sets Variation due to the mounting and to the position of the accelerometer on the shaker table Measurements done by Loic Coquelin and Anne Francoise Obaton from LNE

10 Current steps in sine force - A design of experiments was realized to evaluate influence of : Mounting/unmounting sensors (applied torque on shaker/force sensor and force sensor/top mass) Weight of the top mass Position of the base accelerometer on the shaker table - Comparison of the results from the 3 participants : Evaluation of the influence of theses parameters through a statistical model Integrate them as an uncertainty components for the estimation of the 3 parameters linked to the sensor and the 2 linked to its coupling. - Dissemination of the calibration results and parameter identification to the secondary level * cf. presentation of Christian Schlegel on primary and secondary periodic calibration of force transducers 10

11 WP 1 : Shock force facilities 2 primary calibration devices for shock forces : 20 kn and 250 kn shock force amplitude colliding mass bodies, interferometric acceleration measurement Pulse shaped force 11

12 Shock Force Facilities : 20 kn Sensors : 3 strain gage force transducers * different designs * different coupling Possibility: Use of a load button to complete and improve datas set for parameter identification On axis interferometric measurement 12

13 Shock force signal response Interface sensor HBM sensor Ringing behaviour Conditions of measurements : Reaction mass for both sensors: 10 kg Behaviour : Pulse shape depends greatly on: Mechanical coupling Mounting conditions NOT on mounting torque at the base Current work: Model-based parameter identification in function of the mechanical set-ups including coupling, mounting torques, (additional) load buttons * cf. presentation of Michael Kobusch 13

14 WP 2 : Pressure facilities at MIKES Device manufactured by AVL Operating principle based on drop weight system Generated pressure pulse measured using up to 3 pressure transducers simultaneously Good repetability : up to 300 pressure impacts without any notable leakage

15 piston position WP 2 : Pressure facilities at MIKES Measured pressure peaks with different falling heights.

16 WP 2 : Pressure facilities at MIKES Accelerometer measurements Interferometer measurements * cf. presentation of Anti Lakki * cf. presentation of G. S. Sariyerli for pressure measurements at UME * cf. presentation of C. Matthews for application to industry-level measurement

17 WP 3 : Torque facilities Frequency range : 10 to Hz Torque range : up to 20 N.m 17

18 Actual torque facilities Improvements of the set-up: More powerful rotational exciter Reinforced air bearing Improved air supply All complementary set-ups commissioned and tested 18

19 Auxiliary measuring set-ups Rotational damping Torsional stiffness Moment of inertia 19

20 Auxiliary measuring set-ups Rotational damping Torsional stiffness Measured Moment of inertia Measured 20

21 Dynamic torque measurement Next steps: Yet done : Measurements have been carried out Software for data acquisition and processing was developed Raw output of rotational vibrometer is acquired and demodulated Analog output of the angular accelerometer at the bottom of the DUT is also acquired Model-based parameter identification is currently under development Derived frequency responses will be used for parameter identification by means of a non linear least squares approximation Unknown parameters will be identified through acquired datas and known model parameters 21

22 WP 4 : Measuring amplifiers Frequency range : DC to 10 khz Charge amplifier Bridge amplifier 22

23 Dynamic Calibration for Strain Gauge Bridge Amplifiers Principle of a dynamic bridge standard calibration device: D K D K(t ) Dynamic Bridge Standard System K D Measuring Amplifier mv/v 2,000.0 mv/v The dynamic bridge standard provides the ability to do a frequency dependent amplitude and phase calibration of measuring amplifiers. The measuring amplifier needs to be calibrated before transducers can be calibrated. 23

24 Possible DynBN Calibration Strategies Schematic operation principle of the PTB DynBN. Connected to a bridge amplifier, the DynBN simulates a dynamic strain gauge transducer output voltage. Both components of the signal generation path (MDAC + resistive voltage divider) were calibration separately. Since the DynBN phase is defined with respect to the MDAC generated reference signal, only the resistive voltage divider needs to be investigated for phase calibration. 24

25 DynBN Signal Generation Signal generated by the MDACs represents the reference signal. The 1/200 resistive voltage divider reduces the amplitude to the bridge output voltage level. Illustration of the PTB dynamic bridge standard signal generation path. Goal of the DynBN calibration is to obtain the uncertainties for the ratiometric bridge voltage output amplitude (U o /U i ) and the phase j between reference signal (U ref ) and bridge output signal (U o ). 25

26 Result of the PTB DynBN Calibration Calibration of the 1mV/V signal: Calibration Factor for 1 mv/v (mv/v) Uncertainty Border ±0.05% and ±0.1 Phase Calibration Result Amplitude Cal. Factor for 1mV/V Range Measurement Frequency (Hz) Phase ϕ (Degree) The amplitude and phase calibration uncertainty borders can be defined including all correction values and its uncertainties. Uncertainties (k = 2) of the DynBN set up to 10 khz : Amplitude < 0.05%; Phase <0.1 26

27 Result of the Dewetron Bridge Amplifier Calibration Calibration of the 1mV/V range: Calibration Factor (mv/v) Amplitude Dewetron Bridge Amplifier Range: 1mV/V Phase ϕ Measurement Frequency (Hz) Phase ϕ (Degree) The traceable calibration of a bridge amplifier has, up to 10 khz, a similar behaviour to a low pass filter. The amplitude deviation is within 0.02% of the nominal value and the phase presents deviations up to

28 Conclusion: the next steps Force : Evaluation of the uncertainty components on the estimated parameters and dissemniation to the secondary level Torque: Determination of the rotational damping and improvement of the quality of measurements Amplifiers: Nearly ended, next step could be to publish a guideline for traceable calibration of amplifiers Math: continuation of the development of modelling and uncertainty analysis on estimated parameters * cf. presentation of B. Arendacká for extracting information from a reproducibility experiment 28

29 To go further during this workshop B. Arendacká Extracting information from a reproducibility experiment L. Coquelin Parameter identification for dynamic calibration of force transducers using sinusoidal excitations and assessment of the associated uncertainty M. Kobusch Analysis of shock force measurements for the model-based dynamic calibration A. Lakka Drop-weight system for dynamic pressure calibration C. Matthews Dynamic pressure calibration: Application to industry-level measurement N. Medina Dynamic calibration of force transducers at CEM G. S. Sariyerli Measurement method for dynamic pressure sensors in Tubitak UME C. Schlegel Primary and secondary periodic calibration of force transducers Acknowledgements : The EMRP is jointly funded by the EMRP participating countries within EURAMET and the European Union.

30 Thanks for your attention! 30