Calibration and Field Checks

Comparison Calibration and Traceability

• Back to back with a reference sensor on a calibration exciter: the ratio times the reference's sensitivity
• ISO 16063-21 is the standard; traceability is the chain of calibrations back to a national standard
• Method 514.8 asks for traceable calibration; GEVS names the recognized programs
• A certificate that states no uncertainty and no traceability is a number with nothing behind it

The Uncertainty Is Spent Out of the Tolerance

• Commercial comparison services: roughly one to two percent across most of the shaker band
• Method 514.8: the whole control system within one-third of the sine tolerance
• The sensor's calibration is one part; the mount, cable, conditioner, converter and entry are the rest

What a Certificate Covers

• This serial number, at a reference frequency and the listed frequencies, at the amplitude used
• On the calibration mount, at the laboratory's temperature, on that date
• Not this mount, this cable, this conditioner, the sensitivity typed in, the chamber, a block, or anything since
• It says the sensor was right; it cannot say the channel is right today

The Hand-Held Calibrator

• A small battery-powered shaker with its own reference sensor and control loop
• The common class: 1 g rms at 159.2 Hz, chosen so velocity and displacement come out round
• Its own certificate, a tolerance of a few percent, and a heaviest sensor it can drive

Through the Whole Chain

• On the calibrator by its stud at its torque, on the cable, conditioner and channel it will use
• With the sensitivity, units and detector the test will run
• The screen should read 1 g rms, within the calibrator's tolerance and the chain's
• Logged with the date and the calibrator's serial: the first line of the channel's record

The Size of the Miss Names the Fault

• 1.41 g: a peak detector on the rms calibrator; 2.83 g: peak to peak against rms
• 9.81: meters per second squared read as g, or the reverse
• The ratio of two sensitivities: another sensor's number; the conditioner's gain: entered twice
• A few percent beyond the tolerances: the sensor itself, back for calibration

What the Check Does Not Prove

• One frequency at one level, with the sensor on the calibrator, not on the fixture
• Not the top of the band, the linearity at 50 g, the mounted resonance or the axis
• Those are proven in place: polarity, the tap, the noise floor, the last low-level run, the rigid-body check

A Dropped Sensor Is Judged, Not Condemned

• General-purpose shock limits of several thousand g; a hand-height drop onto concrete, typically 2,000 g
• The check: the bias, the calibrator against the certificate, the time history for ringing or a shifted baseline
• Reads as its certificate: back to work; moved, a hard fall from height, a miniature or shock type: back for calibration

TEDS

• A memory chip in the sensor, read over the same cable: model, serial, sensitivity, reference frequency, calibration date
• It removes the typing, and most of the wrong-serial errors
• It does not prove the calibration current, the chip rewritten at the last calibration, or this sensor at this point on this axis

Decide

• No current certificate, no article
• A dropped sensor checked against its rating and certificate; back for calibration if the check moved or the drop exceeded the rating
• Before every test, every channel's calibrator check through its whole chain, the control channel first, logged
• A miss read by its ratio and fixed before the sensor is mounted

Go to the quiz