Tan delta, viscoelasticity, DMA, and specimen size

[tetmesh]

Hi All,
I'm trying to wrap my head around viscoelastic testing. I would expect greater energy loss (higher tan delta) at lower frequencies/strain rates for creep and strain relaxation reasons, and greater energy loss at higher frequencies as bonds in polymer chains are more pulled and broken, rather than a slower unfolding of the chains. I see this in data from two methods of viscoelastic testing we're benchmarking but when considering the two tests, I confused myself about a potential influence of specimen size. A small specimen on a DMA specific machine and a larger specimen on a larger dynamic load frame will effectively have different 'spring rates' because of different widths. With different spring rates, there's a different natural frequency for each, and I'd expect the energy loss (or efficiency?) to depend on the natural frequency - much like pumping legs too slow or too fast won't get you anywhere on a trampoline. There should be a 'most efficient' frequency. Is the trampoline analogy too simple? I'm trying to account for small differences in the two measurement results and specimen size is one of the differences in the measurements. Thanks in advance for any ideas, thoughts, or references!

 

[Demon3]

I am not an expert in DMA but I expect there to be no resonances during your measurement. Small and large specimens should give the same results. No bonds will be pulled or broken during the measurement, the chains just change configuration by rotating around bonds.

What polymer is being tested? You may see small differences due to temperature or humidity.

Chris DeArmitt

Expert consulting & training

http://www.phantomplastics.com

 

[tetmesh]

Thanks! Temperature and humidity are controlled. It is a latex-based polymer. I agree that dimension shouldn't influence the results and thanks for making the point about twisting bonds; breaking bonds gets into plastic deformation. I still need to convince myself (for sanity's sake) that the dynamic properties of a material don't interfere with a dynamic analysis. The loading phase (tension) will always load the load cell, but if the unloading phase happens at a certain speed, as fast as or faster than the natural rate of the material returning to an unstrained state, the load cell could read zero or negative values. In other words, unloading a tensile sample so fast that it it doesn't have time to respond so the measured load is zero or less.

If this phenomenon happens, despite the dimensional independence of stress and strain, a wider sample will encounter problems at a higher frequency than a narrow sample. The dimension could have an effect on dynamic testing that isn't apparent in quasi-static tests. Does this happen? Thanks in advance for any suggestions/thoughts.