ASTM E517

What the test does

A strip of sheet metal is pulled in tension to a specified plastic strain and stopped, deliberately short of fracture. What matters is not the force reached but where the stretch came from: how much of it the strip took out of its width, and how much out of its thickness. The ratio of those two strains is the plastic strain ratio, r. Because thickness is awkward to measure accurately on thin sheet, the thickness strain is normally derived from the length and width strains on the assumption that the volume of the deforming metal does not change. Specimens are cut at several angles to the rolling direction, because rolled sheet rarely behaves the same way in every direction in its own plane.

What it measures, and why it matters

Resistance to thinning. A metal with a high r-value draws material in from the flange rather than stretching it out of the thickness, which is exactly what a deep-drawn shape needs — a flat-bottomed cylindrical cup being the standard illustration. ASTM E517 describes r as a measure of plastic anisotropy tied to the preferred crystallographic orientations in the metal, and therefore as a measure of drawability, particularly useful where a substantial part of the blank is drawn from beneath the blank holder into the die opening. In a press shop that translates directly: a low r-value shows up as thinning and splitting on the drawn wall, on a die that has run for years without trouble.

The standard is careful about how the number is labelled. Where r varies with strain, a superscript records the percent strain at which it was measured, so a value taken at 20 % elongation is reported as r20. The angle of sampling is a subscript, so a specimen cut parallel to the rolling direction gives r0. That notation is not decoration — an r-value quoted bare is not comparable with anything.

Specimen and sampling

A sheet tension coupon with a reduced parallel section, but the measurement it serves is a ratio of two small strains, so the dimensions before the test matter as much as the ones after it.

Material

Sheet metal intended for deep-drawing applicationsThat is the scope in the standard's own words. The value is read as a measure of drawability.

Geometry

Rectangular test piece with a reduced parallel sectionGripped by its wide ends, with the original gauge length marked before the test.

Thickness

Uniform through the gauge section; normally the full sheet thickness

Orientation

Conventionally 0°, 45° and 90° to the rolling directionPracticeThe standard requires the angle to be recorded as a subscript; sampling at three angles is what makes the anisotropy averages available.

Original width

Measured, not assumedDakAn error in the original width goes straight into r at full weight. Measure it on the coupon, at more than one position along the gauge length.

Surface

Free of scratches and edge damage in the parallel lengthPracticeA nick localises strain and the width reduction stops being uniform, which is exactly what the calculation assumes it is.

Sheared edges left rough from the blanking press are the usual cause of scatter between nominally identical coupons. The rolling direction has to be marked on the blank before it is cut, not reconstructed afterwards.

Strain level and rate

The r-value has a strain attached to it. Which strain is a decision, and it belongs in the specification rather than to the operator.

Strain at which r is taken

A specified plastic strain, short of maximum forceFor many materials r is essentially constant over a range of plastic strains up to maximum applied force; where it is not, the strain has to be recorded.

How the strain is recorded

As a superscript, so 20 % elongation gives r20

How the orientation is recorded

As a subscript, so a coupon cut along the rolling direction gives r0

Discontinuous yielding

Continue the test beyond the yield-point elongationAccuracy and reproducibility are reduced unless the test runs past it. Coarse grain size has the same effect at low strains.

Never evaluate past maximum force

Necking ends the uniform regionPracticeThe constant-volume relation used to get the thickness strain holds only while deformation is uniform.

Two laboratories reporting r at two different strains on the same coil will disagree and both be right. Agree the strain and the orientations with the customer before the first result is issued.

Calculations

Thickness strain is almost never measured directly on thin sheet. It is derived from the length and width strains on the assumption that the volume of the uniformly deforming metal does not change.

How the test runs

1. 01Mark the rolling direction on the blank and cut coupons at 0°, 45° and 90° to it.
2. 02Measure the original width at several positions along the gauge length and record the average.
3. 03Mark the original gauge length, or set the extensometer gauge to it.
4. 04Fit the specimen squarely in the grips, aligned on the load axis, and zero the force before loading.
5. 05Fit the axial extensometer, and the transverse device if width is being followed under load.
6. 06Pull at a steady rate to the specified plastic strain, running past any yield-point elongation before evaluating.
7. 07Stop short of maximum force and unload.
8. 08Measure the width and the gauge length again, in the same positions as before.
9. 09Compute the true plastic length and width strains, and the thickness strain from constancy of volume.
10. 10Report r for each coupon with its orientation subscript and strain superscript.
11. 11Form the weighted average and the degree of planar anisotropy across the three orientations.

The single most damaging shortcut is measuring width in one place. Sheet does not thin evenly along a gauge length, and a width taken at whichever point the operator reached first is a different measurement from one averaged along it.

Watch the test

Grips and fixtures for this method

Self-tighteningTJ-15

Universal Parallel Wedge Grips

Parallel-closing wedges hold a flat sheet coupon square to the load axis, which is what keeps the width reduction symmetrical across the gauge length.

Specifications

Axial & transverse

Clip-On Extensometers

The transverse variant reads width directly across the coupon while the axial device follows the gauge length, which removes the manual width measurement that dominates the scatter in this test.

Specifications

Non-contact

Advanced Video Extensometer

A non-contact device follows both axes at once with nothing touching the specimen, and covers the full gauge length rather than one point across it.

Specifications

What the report has to contain

• Full designation and edition, and the units the method was run in
• Material identification — alloy or grade, temper or condition, coil or lot, and nominal thickness
• Orientation of every coupon to the rolling direction
• The plastic strain at which r was determined, and how it was established
• Original and final gauge length and width for each coupon, and how each was measured
• The strain measurement route: manual, a transverse extensometer, or an optical device
• r for each coupon, written with its orientation subscript and strain superscript
• The weighted average and the degree of planar anisotropy where they are reported
• Number of coupons per orientation, with the mean and the scatter
• Any coupon discarded, and why

What the machine must be capable of

Very little in the way of force, and a great deal in the way of strain measurement. Sheet coupons of this kind rarely need more than a few kilonewtons, so this is ordinary universal-testing-machine work. Extensometers used for the strain measurement have to conform to ASTM E83, and both strains have to be captured — axial extension along the gauge length and the reduction in width across it. A transverse extensometer or an optical device reading both axes at once removes the largest source of scatter in the whole test, which is a manual width measurement taken on an unloaded specimen.

The method must also be run beyond the awkward part of the curve. A material with an upper yield strength followed by discontinuous yielding stretches unevenly while that yielding runs — Lüders bands propagating across the surface in steels — and the standard states plainly that the accuracy and reproducibility of r are reduced unless the test is continued beyond the yield-point elongation. Coarse-grained material is affected the same way at low strains.

What goes wrong in practice

Stopping the test inside the yield-point elongation, the failure the standard warns about most directly and the one that looks least like a mistake. Reporting r without the strain and orientation it was measured at. Taking width on a specimen that has sprung back unevenly. Measuring near a shoulder rather than in the uniformly strained middle. And running past maximum force, where necking begins and the constant-volume assumption quietly stops holding.

How it differs from the standards nearest to it

All four are tension tests on sheet. Only two of them report a formability parameter, and they report different ones.

r and n are not substitutes for one another. A panel that splits in a stretched region needs n, and one that splits on a drawn wall needs r; testing only the convenient one leaves half the question unanswered.

Questions we are asked about this test