Tg obviously has a deep impact on aging. You have to distinguish chemical and physical aging. Both are influenced by the temperature. Above Tg you have a rubbery or viscoelastic state where the diffusion controlled reactions (including thermo-oxidation) proceed faster. Surface reactions (e.g. photo-aging) are less influenced by Tg but the effect is not negligible, as the reactions following initiation are also diffusion-controlled. Physical aging (densification of the glassy structure) proceeds with appreciable rate below, but close to, the Tg. below Tg the rate processes can be more or less described by the Arrhenius relation, above Tg the Williams-Landel-Ferry or the Volgel-Fulcher-Tamman relation applies. Please specify your problem.

I think the question really is how one can distinguish? whether one has glass transition taking place from the appearance of the TTT diagram obtained by DSC plot as d H/dt (i.e., Heat Flow Rate) versus T where dT/dt=R (constant heating rate). One can easily show that d2 H/ dT dt = d CP /dt = R dCP /dT which is the slope of the DSC plot multiplied by the rate of heating. CP is heat capacity of test piece under the constant pressure.

Glass transition is an endothermic reaction which shows it self as a S- Shape transition on the DSC plot with a positive slope inflection point should be taken as Tg glass transition point where the second variations of Cp becomes zero. There is a plateau region well above Tg which ends up with the crysillation domain where DSC plot shows through or big diplet since crystallization transition is an exothermic reaction, where the second variations (or derivatives) of DSC plot shows positive sign (minima).

Note: Actually Tg transition is an autocatalytic reaction according to mathematical point of view where one can easily write the following expression that can be easily integrated to give proper S-shape kinetics:

d Cp /dt = Alpha Cp ( Cf - Cp)

The initial transient region may be represented by first order reaction (E. N. Yeremin '' The Foundation of Chemical Kinetics)

As Gyorgy wrote there are several impacts which are correlated with Tg.

As seals are typically used above their Tg some effects are not as likely to occure in application as others but i wanted to point out that typically your Tg is also not stable during ageing. Due to ageing your molecular structure is changed. Chain scission and crosslinking occure (typically both but one effect is predominant)which can have an influence on Tg.

We observed in our investigations some quite dramatic increase of Tg for HNBR and in a lesser extent for EPMD rubber. But these very high increases were observed for samples after ageing at rather high temperature were also DLO effects were observed.

Article Ageing of HNBR, EPDM and FKM O-rings

Article Effects of heterogeneous aging in compressed HNBR and EPDM O...

When you observed a change in Tg in the DSC measurement an increase would imply that the molecular mobility in your sample decreased which would be correlated with crosslinking and an decrease would indicate a predominant chain scission (if the effect/contact of possibly plasticizing species during ageing can be excluded).

If you provide more information on the material and the ageing conditions as well as the DSC result a better interpretation would be possible.

Thanks for your resourceful contributions

Under the constant pressure and the rate of temperature variations one can easily show that

the ordianate variations dH/dt is replaced by Y=R Cp and the absisa t by X=R T, where the heating rate R =dT/dt, Pressure P are constant. The transformation enthalpies of any events taking place in given temperature intervals Ti and Tf such as the glass, crystallization and melting may be calculate from the following area integration formulas:

R**2 Delta H = Integral { Y dX } from Xi to Xf - Y(initial) [ Xf - Xi]

explicit expression looks as:

R**2 Delta H = R**2 Integral {CpdT} from Ti to Tf - R**2 Cp(initial)[ Tf -Ti ].

Above final expression shows that the rate of heating doesn't enter into the calculation explicitly.

Note: Above areal connections are used in practice ad hoc fashion without proof

.

Best Regards

Since we are dealing with the reversible or quasi- equilibrium processes by assumption. At the transition temperature which may be defined as the point over which the sign of dCp/dT is changing (i.e., the inflection point of glass transition, the max and min points of melting and xtallization on Cp vs T plots) then one may write:

Delta G = Delta H- Ttrs Delta S = 0 at T= Ttrs (Reversible) that follows; Delta H = Trs Delta S

In most transition reaction one assumes that entropy of transition doesn't vary much and therefore the following approximation for Gibbs free energy of transformation may be used during the Boltzmann type kinetics formulations. The temperature dependence of the Gibbs free energy of transformation above and below the transition temperature may be given by:

Delta G = Delta H - T Delta H /Ttrs = Delta H [Ttrs - T]/Ttrs

Where [Ttrs - T] corresponds to degrees of supercooling or heating for given phase transition.

Note: According to my understanding glass transformation which may be follow up by the viscosity temperature measurements is an nonequilibrium dissipative transition (endothermic) takes place by local shear and rotation of macro molecules in cooperative way. It is autocatalytic in the sense that its rate dependence on not only the amount of transformed species but as well as what is left to be transformed. The kinetic equation for this reaction is given previously and the solution of which can be done very easily and exactly.

The formulated temperature dependent Gibbs free energy of transition may be used for the determination of the critical size of the nuclei at the super cooling temperatures as well as in the determination of the nucleation rate if the transition takes place by nucleation and growth process, which also involves surface free energy contribution. This mechanism does not operate for the glass transition but definitely it is operational for the crystallization and the melting phenomenon.

Don't forget that "Tg" is not an absolute number for the material. If you run the DSC curve at a different heating rate, you will get a different Tg. The faster the heating rate, the higher the Tg. The variation with heating rate will depend on the viscosity/temperature of the particular material you are measuring.