DMA question/ have you ever seen that?!?!

I assuming that you are talking about a temperature sweep (low --> high) at a fixed frequency? How is the sample mounted in the DMA?
Just a guess at this point, as some more experimental details would be helpful: If you are using circular plates (flate or conical, it doesn't matter for this point), you could be having poor contact and slippage with the sample. As the sample warms, it softens and the contact improves, thereby giving a stronger signal to the transducer and an apparent jump in E'.
John Aspen Research, - "Turning Questions into Answers"
Opinions expressed herein are my own and may not represent those of my employer.

One other last minute thought: what are you analyzing?
For simple polymers - single phase - this would be unexpected. But for multiphase or reactive systems, something like this is much easier to conceive of and explain.
John Aspen Research, - "Turning Questions into Answers"
Opinions expressed herein are my own and may not represent those of my employer.

You don't say much about the polymer and how it was prepared, so I am guessing about possible causes and I understand from your post, that a DMA-rerun of the very same sample doesn't show this steep increase in G' near the Tg anymore.
It looks like you are seeing a sudden stiffening of your sample close to the polymer Tg, which could be due to post-crystallization, in case the polymer sample was quenched from the molden state i.e. had not reached maximum crystallinity durinig its preparation. During the 1st DMA-run maximum crystallinity can occurs, which is why you don't see this "peak" anymore during the 2nd DMA-run.
I understand the following document talks about this kind of DMA-effect for PET-based polymers: Other potential causes for sudden increases in G' can be due to molecular interactions or due to crosslinking.
spiros wrote:

OOPS ... should have said COLD-crystallization (instead of POST-crystallization) ...
Found another document for you that talks about this DMA-effect with PET-fibers ... You don't say much about the polymer and how it was prepared, so I am

thanks for your help guys,
sorry if I wasnt clear, there was no second run
the only thing is that the effect was not only seen in pure polymer even if in that case it was more obvious.
the crystallization crossed my mind also but the thing is that the crystallisation (DSC results) during heating occurs way above the Tg while the effect i'm talking about takes places just below Tg
In addition the increase of E' during crystallization was noticed as expected as a small and broad peak after Tg,
in contrast the DMA before the Tg showed a plato at first, a sharp peak and finally a decline at Tg
thanks once again any other comments would be really helpful spiros

sorry i forgot
@_______@ / \ || || ||
pressing in fixed freq. in the centre of the sample

my comments are below
spiros wrote: Are we talking about a polymer blend and if so, only one of the blend polymers shows this effect also, but to a lesser degree? If so, again, then you need to "investigate" THIS polymer first.
Crystallation in DSC "way above Tg" means that your polymer (blend?) crystallizes relatively slowly? Right?
So you are seeing actually two peaks in E', one below Tg and one above? The peak above Tg ought to be due to cold-crystallization.
This effect could be due to molecular rearrangements or stress relaxations.
To verify my assumptions would it be possible for you to carry out a second run with one and the same sample and that you also run another new sample that was annealed at a temperature "way above Tg", before the DMA-run? Samples without any stress relaxations and/or any cold-crystallization should't show any peaks in E', but only typical plateaus at the various thermal transitions.
This goes beyond your original question, but I would be worried about a polymer (blend?), which, after a "typical molding process", shows a huge amount of cold-crystallization (and perhaps high stress levels), because molded enduse parts might be exposed to a significant change in physical properties (embrittlement?) and in dimensions (due to shrinkage), especially when those parts are used at higher temperatures ....

its a blend of polymer+ non-polymer but as I said the pure polymer samples show the effect more vividly
no it crystallises fast but at a temperature of 60oC above Tg
well the second one is not really a peak more like a stair-step
you think so?? wouldn't this be observed in DSC also, i dont see anything before the Tg in DSC
no, the samples after the first run or after annealing become brittle which makes it difficult or even impossible. they brake
you see the effect as I say above is more obvious in the pure polymer so the blending has nothing to do with this effect, I was just wondering what could cause this sharp peak (which people in the DMA business say they have never seen it again) if not crystallization.
thanks once again spiros

this is the E' of the DMA hope you can understand it
* * * * * * * ************* * * ************** * * * **********

sorry ignore the previous this is the E' of the DMA hope you can understand it
* * * * * * * ************* * Tg * ************** * * * **********

Yes, you said it ...! Non-polymer would be some kind of filler?
That's "as bad as" PET or semi-aromatic PA. Fast would be if the cold-crystallization peak of a quenched sample overlaps with the Tg-step in DSC, as you would see for PA6 or PA66.
Well, I am not sure, but if the DMA peak is not to do with a significant change in specific heat, you'd not see it in DSC, which doesn't pick up changes in modulus. Often you would not even see a Tg in DSC, but get a huge signal in DMA, for example.
That's bad, but probably good enough for academic studies ... ! Apart from brittleness during the DMA run, are there any other changes you observe on the tested sample? Like bad warpage, huge shrinkage, cracks etc... ? What happens to your sample if you heat it up to a temperature above the first DMA-peak temperature, but still below it's Tg and if you hold it at that temperature for some time, in a regular lab oven? Same brittleness (i.e.increase in stiffness?) as after the DMA run? Warpage? Cracks? Shrinkage? Color change?
That's for sure interesting and needs to be approached from the polymer side and it's preparation.

Not sure what you are playing with for a product but I experienced something similar to this in polypropelene blown film using the "double bubble process" specifically we were getting sheet curling in both the machine and transverse directions that was contrary to the natural curvature of the bubble (take a sample, layed it flat and it would roll up like a cigarette) we tried different cooling temps, die temps, cooling quicker by physically adjusting the distance of our cooling from the die face...nothing seemed to work. A year, several hundred thousand pounds of scrap, the machine manufacturers engineering team, two consultants and our own folks....
Turned out to be strain induced hardening (crystalization) described by visco/plastic/elastic theory put forth in Kanai and Cambell's book, FILM PROCESSING available from Hanser Gardner publications page 122..
if it gets worse on higher output rates, and less noticeable at lower rates...it might be what you have... the FIX involved a restriction in the entrance to the die (change in veloity throught the die)