The Challenges of Strain Control
In 2009, one of the primary metals testing standards was updated to include a method based around controlling the strain rate on the specimen. The implementation of these updates, along with how they could benefit your laboratories, is commonly misunderstood.
Some mechanical properties of metals will be affected by the speed of the test and are therefore ‘strain-rate sensitive’. In Stress Control or Crosshead Speed control the overall machine stiffness will affect the speed on the specimen, which can cause differences in results.
When testing strain-sensitive materials, the allowable test speeds could cause more than a 10% difference in proof stress results from testing at the slowest and fastest rate allowed in ASTM E8/E8M and ISO 6892-1
The graph below shows two plots of the same material, the upper plot is at the Maximum allowable stress rate per ISO 6892-1 Method B and the lower plot is the lowest allowable stress rate per Method B.
“Method A is intended to minimize the variation of the test rates during the moment when strain rate sensitive parameters are determined and to minimize the measurement uncertainty of the test results.” – (ISO 6892-1:2009)
Requirements to Achieve Closed-Loop Strain ControlExtensometer
High-precision device with stable feedback
Grips
Need to securely grip the specimen during the test without slippage, preferably with high stiffness. The graph shows how different types of gripping devices can effect the system stiffness, and how a machine in strain control would need to compensate.
Lab Environment
No vibration or shock can be transmitted onto the testing system and specimen
Load Frame
Needs a precise and stable drive system with high stiffness. Below shows a stress strain curve on nominally similar materials, one tested on a high stiffness frame and the other on a lower stiffness frame. Using an estimated strain method calculation, both tests are run at a constant crosshead speed of 2.25 mm/min. There was a 21% difference in ‘specimen speed’ (expressed in mm/min), which lead to a 5% difference in the yield result.
The graph below shows the specimen speed difference between a ‘stiff’ system and a less stiff system. At the start of the test it is shown that the crosshead movement is transferred into strain on the specimen very quickly, whereas on the weaker system it takes longer. This is due to the machine/load cell/grips deflection, which causes that movement to not be transferred to the specimen. If both machines where using strain control then the results would be much more comparable, but the control will likely be more challenging on the weaker system.
Software and Controller
Responsive control loop to maintain close tolerances on strain rate, with the ability to tune the gain settings as the stiffness changes when the specimen yields
Specimen
A proportional specimen and a proportional gauge length extensometer are ideal. In reality, a specimen with good gauge length to parallel length ratio is well suited to minimize the strain seen outside of the gauge length, allowing the control to be more stable.
If your specimens vary from discontinuous yielding to continuous yielding it is important to change control methods for each type. As local yielding can occur outside of the gauge length on discontinuous yielding material, it is impossible to control from the strain feedback and should be in crosshead speed control during yield point elongation (YPE/Ae).