I have one trivial but important doubt in measuring tensile property at constant strain rate. When we tell constant strain rate to the instrument(UTM) it measures the stress corresponding to the strain. My doubt is how instrument keeps constant strain rate. We know constant strain rate (epsilon dot)=delL/L*(del t) . Hence to keep constant strain rate instrument has to keep (del L/del T) constant which means sampling time has to change with measurement since del L is also changing. Does equipment really measuring the sampling time. Do you have any clues about the actual measurement of constant strain rate by any standard equipment?
Strain rate is usually backward calculated. Machines can measure force and displacement. Since you have time, you can work out strain rate.
When you are doing room temperature test, strain rate does not have big influence.
In most of the universal testing machines, a constant crosshead speed is used as an input which corresponds to the constant strain rate at the macro level.
In actual practice, almost universally we use "constant' cross head speed instead of true constant strain rate. Since, strain rate =cross head speed/gauge length, the true strain rate decreases with time/deformation. It is possible to conduct a "true constant strain rate" tensile test with modern machine equipped with suitable software. Such softwares consider the real time increment in gauge length and adjust the cross head speed accordingly.
Thanks David, Sayed and Chandan da for your nice and insightful response.
To answer Chandan da's comment , I guess now -a days people use interferrometric and digital image correlation (DIC) technique to measure true strain rate measurement.
David thanks a lot for your answer. I understand room temperature strain rate does not matter for normal metals but viscoelastic material does show weird response at room temperature at different strain rate such as polymers.
Ankur, yes u are right on the money, but don't have one in reachable lab.... if u have DIC, please let me know......
Regarding RT tensile testing, be careful about stain rate sensitive materials, otherwise its fine.
As said in previous comments you need closed loop control on your tensile test machine. This essentially means that the crosshead displacement is adjusted by the information you read from the strain gauge or extensometer you use to measure elongation. On a first level, you can keep e_dot (engineering strain rate) relatively easily constant by controlling the machine by imposing a constant rate of opening of the extensometer. Most of the modern software packages can do that. To run the test with constant true strain rate epsilon_dot you need the feature of being able to update the rate of extensometer opening you want to achieve in your controlling software. This is by far not standard but would be fairly usefull for e.g. superplastic materials if you want to see the evolution of the flow stress with deformation.
From an experimental standpoint, working with extensometers to obtain fixed e_dot or epsilon_dot contains the substantial risk of sliding of the extensometer (the machine will go ballistic if the extensometer slides). Another important limitation is the relation between noise/resolution of your extensometer and the strain rate you want to achive on the one hand and the control loop frequency of your controller on the other. The uncertainty (noise or resolution) in the strain measurement divided by the strain rate gives you a characteristic time that should be shorter than the inverse of the control loop frequency. Otherwise your controler would try to correct on signal noise.
The Hopkinsons Pressure Bar provides the results by averaging out strain rates. The complete stress strain behaviour is generated for a particular strain rate.
Haa, there is no way to use DIC for PID control of strain rate guys. Plus with DIC you get what? Strain destribution over the surface.
I tend to use constant displacement rate term, though if necessary it is possible to put incerement in your displacement rate to have constant strain rate, but be careful it is rarely liniar dependency in fact, I have checked it so many times on so many materials :) also by using DIC
Yep, and most of the extensometers are limited to 15, 20, 30 or 40 % of strain, plus they have a bit different error at different strain levels, so if you'd like to use extensometer for feedback signal keep it in mind.
Though, there is pneumatic systems which can provide you constant strain with high accuracy, but they are preferable to use for very slow tests with strain rate of 10^-6 and slower
Dear Ankur. If you are extending at constant speed v=dL/dt, then the strain rate edot=dL/Lo*dt is constant. If you program the machine to extend at say x mm/min or as a percentage of original length Lo (%/min), then that is what it will do. It will turn the screw that displaces the clamp at the appropriate constant speed to achieve it. An extensiometer will help with accuracy and eliminate end effects, but in principle the rate will be constant.
Deon, that is true if we re talking about engineering strain rate but not if we look at true strain rate.
Quite right, thanks, Ludger. As the true strain is the natural log of the engineering strain, in theory it should be possible to program a tensile tester to speed up exponentially in order to achieve a constant true strain rate. I have not done so myself and would suppose it depends on the software that controls the machine.
where a close loop universal testing machine is employed, there is no doubt. Because the corresponding software can adjust the velocity response to keep the strain rate at constant value. But, in the case of typical universal machine which utilizes constant cross head speed the strain rate would be changed (epsilon=velocity / instantaneous length). However, you can suppose that the order of magnitude of the strain rate remains relatively constant. This can be verified through calculation of the mean strain rate values.
Dear Ankur use s tensile m/c have a velocity control as in the instron
- Badges
- Science topic
- Similar topics
- Engineering
- Materials Engineering