Measure at Class: Using a DIC as an Videoextensometer  

Question: Can a DIC system be used as a video extensometer for standard norm testing?

Yes, a Digital Image Correlation (DIC) system can be used as a video extensometer for standard norm testing.

Question: What standards govern the use and implementation of video extensometer?

The accuracy, resolution, and precision requirements for video extensometer is governed by two fundamental norms;

• (International) ISO 9513: Metallic materials — Calibration of extensometers used in uniaxial testing.
• (USA) ASTM E83: Standard Practice for Verification and Classification of Extensometer Systems.

In standardized testing norms, the above-mentioned norms are referenced with a stipulate respective “Class” rating requirement for the used video extensometer.

Question: Who is allowed to calibrate a video extensometer to the ISO 9513 or ASTM E83?

To certify a video extensometer legally or formally to ASTM E83 or ISO 9513 norms, the calibration must be performed by a third-party calibration body or an internal metrology lab, that holds formal accreditation to ISO/IEC 17025. One aspect of this accreditation is that any hardware used for a calibration holds and carries an “unbroken chain” of traceability to national metrology institutes (like NIST – US, NPL – UK, or PTB – Germany).

Question: How is a calibration for a video extensometer typically performed?

Because video extensometers are optical and non-contact, the calibration focuses on verifying the software’s projection calibration model. Before the calibration, the projection calibration is performed using a (referenced) calibration target – this does NOT need to be a traceable reference. Then after, a traceable calibration fixture (that can be a highly precise micrometric screw, a laser interferometer, or a digital encoder) is placed in front of the video extensometer. The distance between a reference stationary marker and target moving marker is then measured at exactly, known displacement intervals. The difference between the known, referenced value and the measured value from the video extensometer is then cited as a maximum error, for which a “class” rating is then assigned and a certificate to ISO 9513 or ASTM E83 is issued. If the video extensometer has multiple fields-of-view (FOV), lenses, or calibrated software profiles, a classification must be provided for each gauge length range calibrated.

Question: Is the gauge length equal to the Field-of-View (FoV)?

No, the gauge length is a stipulated length for calibration of the video extensometer. It is a smaller subset distance of and within the Field-of-View (FoV). Both the gauge length and specimen travel distance should approximately equal the total FoV. The FoV should be equal to the nominal size of the calibration target used for the projection calibration. Assuming the “25% Target Field-of-View Rule-of-Thumb”, the gauge length can be considered to be 75% that of the calibration target nominal size i.e. 100 mm Calibration Target = 75 mm Guage Length.

Question: What is a class rating and what does mean for video extensometers?

A class rating (or accuracy class) is an official designation that defines the maximum allowable measurement error of a testing instrument. For example, ISO 9513:

• Class 0.2: maximum permissible relative bias error +/-0.2% of displacement value
• Class 0.5: maximum permissible relative bias error +/-0.5% of displacement value
• Class 1: maximum permissible relative bias error +/-1% of displacement value
• Class 2: maximum permissible relative bias error +/-2% of displacement value

The fixed threshold bias error (FTBE) in micron is 3x the respective Class.

• Class 0.2: 3x 0.2 = +/-0.6 um
• Class 0.5: 3x 0.5 = +/-1.5 um
• Class 1: 3x 1.0 = +/-3 um
• Class 2: 3x 2.0 = +/-6 um

In both the ISO 9513 and ASTM E83, this classification rating for video extensometers works on the so-called “2:1 ratio”. In the ASTM E83, the device’s required (in-plane) resolution (RR) must be at least twice as precise as the fixed threshold bias error (FTBE). In the ISO 9513, the same rule is achieved via writing the specific resolution limits directly into its classification tables.

A video extensometer must achieve the required (in-plane) resolution (RR) in order to be used for a specific “class” rating.

• ISO 9513 Class 0.2: FTBE +/-0.6 um, RR <0.3 um, equv. ASTM E83 Class A
• ISO 9513 Class 0.5: FTBE +/-1.5 um, RR <0.75 um, equv. ASTM E83 Class B-1
• ISO 9513 Class 1: FTBE +/-3 um, RR <1.5 um, equv. ASTM E83 Class B-2
• ISO 9513 Class 2: FTBE +/-6 um, RR <3 um, equv. ASTM E83 Class C
  
  

Question: How can one calculate the required (in-plane) resolution (RR) for a respective class rating?

Simple – as described in Part 4 of the DIC Beginner User Training Course, calculate the image scale using the factored x0.75 horizontal pixels used for the factored x0.75 FoV yielding the gauge length distance. Then multiply by a (video extensometer) conservative ORF of 0.015 px (instead of the typical DIC ORF in-plane of 0.01 px). This will yield the required (in-plane) resolution (RR).

• 2.3 Mpx Camera (1920 px) for a FoV of 200 mm / Gauge Length = 150 mm, RR = 1.56 um < 3 um (Class 2)
• 5.0 Mpx Camera (2448 px) for a FoV of 150 mm / Gauge Length = 112,5 mm, RR = 0.92 um <1.5 um (Class 1)
• 12.3 Mpx Camera (4096 px) for a FoV of 100 mm / Gauge Length = 75 mm, RR = 0.37 um < 0.75 um (Class 0.5)
  
  

Question: Can we (Dantec Dynamics) sell a DIC system for a specific class rating?

Yes, we (Dantec Dynamics) as a manufacturer can deliver a DIC system for a “stipulated” class rating based on setup design variables.
As a manufacturer, we do not need (and cannot) verify this “stipulated” class rating until the device has been calibrated to ISO 9513 / ASTM E83.
But we can state, for example:
“The TestDIC system of 3x 12.3 Mpx cameras (4096 px) supplied with a 300 mm calibration target = FoV = Nominal Guage Length of 225 mm is rated as Class 1 to ISO 9513 and Class B-2 to ASTM E83. This is achievable with the supplied lenses and assuming an operable working distance of 50 cm.”

Question: Are there any limitations in using a DIC system as a video extensometer?

Yes, the main limitation of using DIC as a video extensometer is the processing speed, and inherently, the larger data computation requirements associated with the selected processing algorithm.
A video extensometer can be used for both active (closed-loop) control and passive recording – this is because the algorithm only searches for a few discrete contrast features, making the data feedback relatively quick with minimal lag.
DIC, on the other hand, can only be used for passive recording – this is because the algorithm correlates an entire area, making data feedback slow with a time lag. As such, DIC cannot be used for active (closed-loop) control – however, the selection of Marker Tracking or Facet/Subset Tracking are quicker alternatives just as fast as a video extensometer algorithm – the downside is that neither method is available in Istra4D, in real-time (only via post-processing).
This means that DIC systems can only be used for passive recording (at the moment), which is permitted in most material testing norms, unless otherwise explicitly stated.
Another element that must be considered, is that the data acquisition speed must be captured at the required test speed, which is often cited in many material testing norms. This is often higher/faster throughout the complete test duration.