A glass transition can be found by observing a endothermic reaction to a new value. In your curve between 330 and 340K. You might want to do the measurement at a lower speed to have a precise value, because a speed of 20K/min is pretty high. (1, 3 or 5 K/min)
Dear Aboothahir Afzal . These DSC curves are completely non-informative. There are several points that need to be improved in order to be able to determine the glass transition temperature.
1) Is your substance in an amorphous form? If not, then it is impossible to talk about the glass transition temperature
2) Does your substance melt with decomposition? If so, then the heating-cooling-heating cycle is incorrect to use.
3) I suggest, expand the temperature range and start at least from zero Celsius
4) The speed of 20K/min is very fast. You can skip the glass transition temperature. At least 10 K / min, and even slower is better.
5) Even if you take into account points 1-4, there will still be a possibility of error. Therefore, for greater certainty, I suggest doing Modulated DSC (if possible).
The DSC thermogram reveals some idea about phase transformation. However I like to suggest few points for mutual understanding.
1. Specify the type of material (alloy, polymer, organic matter) and preparation.
2. Once you know the material and processes, kindly refer to available literatures to find similarities with your data.
3. The cooling curve (2nd) does not show the peak, hinting at irreversible transformation. However, to identify glass transition temperature, you have to follow experimental parameters and ensure completely amorphous nature of starting material, as recommended by Nikita Vasilev.
I consulted the literature and found a mean value of 353 K, varying from 341-372 K depending on the degree of hydrolysis of the polymer sample.
There is no sign in your heating/cooling curves of any exo/endotherms around these temperatures. As already suggested by Nicolas, maybe a slower rate of heating or cooling would bring out more detail in the curves. Good luck. Mark.