State of progress of dynamic calibration of force, torque and ...

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

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

EMRP Project IND 09 : « Traceable dynamic measurement of mechanical quantities »

2

For further information on the European Metrology Research Programme, contact EURAMET at their booth in the Metrology Village or at www.euramet.org

3

A few words to introduce the project …

What quantities are we concerned with? • Force • Pressure • Torque

Which countries are involved ? Czech Republic

Finland France

Germany Italy

Spain Sweden

Turkey United kingdom

What budget ? 3584 k€ or 242 PM

How long will the project last? 3 years, the planned end is September 2014

8 industrial partners : VSL; HBM; Spektra; Porsche; Volkswagen; AEP; Scandura; Simea

8th Workshop on Analysis of Dynamic Measurements / May 5-6, 2014

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


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


6

Organization of the project


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

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

11

2 primary calibration devices for shock forces :

• 20 kN and 250 kN shock force amplitude

• colliding mass bodies, interferometric acceleration measurement

• Pulse shaped force

WP 1 : Shock force facilities


12

On axis interferometric measurement

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

Shock Force Facilities : 20 kN


13

Interface sensor

Ringing behaviour

Behaviour : Pulse shape depends greatly on: Mechanical coupling Mounting conditions NOT on mounting torque at the base

Shock force signal response

HBM sensor

Conditions of measurements : Reaction mass for both sensors: 10 kg

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


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


Measured pressure peaks with different falling heights.

pist

on p

ositi

on



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

Frequency range : 10 to 1 000 Hz Torque range : up to 20 N.m

WP 3 : Torque facilities


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


19

Torsional stiffness

Moment of inertia

Rotational damping

Auxiliary measuring set-ups

20

Torsional stiffness

Moment of inertia

Auxiliary measuring set-ups

Rotational damping

Measured

Measured

21

Dynamic torque measurement

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

Next steps: 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


22

Frequency range : DC to 10 kHz

Charge amplifier Bridge amplifier

WP 4 : Measuring amplifiers


23

Dynamic Calibration for Strain Gauge Bridge Amplifiers

Principle of a dynamic bridge standard calibration device:

Measuring Amplifier

mV/V

Dynamic Bridge Standard System

D·K

D·K(t ) D K 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.


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.


25

DynBN Signal Generation

Goal of the DynBN calibration is to obtain the uncertainties for the ratiometric bridge voltage output amplitude (Uo /Ui) and the phase j between reference signal (Uref) and bridge output signal (Uo).

Illustration of the PTB dynamic bridge standard signal generation path.

• Signal generated by the MDACs represents the reference signal.

• The 1/200 resistive voltage divider reduces the amplitude to the bridge output voltage level.


26

Result of the PTB DynBN Calibration

Calibration of the 1mV/V signal:

10 100 1000 100000.9994

0.9996

0.9998

1.0000

1.0002

1.0004

1.0006

-0.12

-0.08

-0.04

0.00

0.04

0.08

0.12

Calib

ratio

n Fa

ctor

for 1

mV/

V (m

V/V)

Measurement Frequency (Hz)

Amplitude Cal. Factor for 1mV/V Range

Phase ϕ (Degree)

Phase Calibration Result

Uncertainty Border ±0.05% and ±0.1°

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°


27

Calibration of the 1mV/V range:

10 100 1000 100000.996

0.997

0.998

0.999

1.000

1.001

1.002

-16

-12

-8

-4

0

4

8

Calib

ratio

n Fa

ctor

(m

V/V)

Measurement Frequency (Hz)

Amplitude

Phase ϕ

Phase ϕ (Degree)

Dewetron Bridge AmplifierRange: 1mV/V

Result of the Dewetron Bridge Amplifier Calibration

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 -16°.


28

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

Conclusion: the next steps


* cf. presentation of B. Arendacká for extracting information from a reproducibility experiment

To go further during this workshop …

Acknowledgements : The EMRP is jointly funded by the EMRP participating countries within EURAMET and the European Union.

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

30

Thanks for your attention !


State of progress of dynamic calibration of force, torque and ...

Documents