Most important are eutectic mixtures (type and volume fraction), in as-cast condition alloy may have different melting behaviour in comparison to heat treated state. In most measurements heating rates of 20-30 K/min are used.

After first heating, when you will acquire homogenous chemical "composition" you will be measuring "exact reaction temperature" according to the given chemical composition. In my experience, NI based superalloys are not often present in real application in "as cast" state, they proceed several types of different heat treatments. In these the chemical composition is more less homogeneous.

To Mr.Grzegorz Cios,

If the eutectic mixtures are most important factor for measurement, in as cast condition, where should I choose the specimen for measurement to obtain exact reaction temperature?(In chill zone or columnar zone or equiaxed zone?)

To Mr.Peter Oslanec,

My alloy is applied in as cast condition for automobile turbo charger part.

Therefore, I should measure the reaction temperature in as cast condition. 

For the alloy used in as cast condition, where should I choose the specimen for measurement to obtain exact reaction temperature?(In chill zone or columnar zone or equiaxed zone?) as mentioned above answer.

It would be interesting to perform all three measurments. That means from chill zone, columnar zone, equiaxed zone. Then you will see if there is any difference or not. Actually, I guess you are misleading yourself. You are writing, there are so many differences in microstructure between different parts of the casting. So, why do you think there is one solidification temperature of the alloy? At least width of the solidification interval could be different.

Thanks for answer.

That is what I am saying.

Because microstructures are different in chill, columnar and equiaxed zone, solidification temperature should be different.

If so, if I want to know the entire solidification temperature of alloy such as solidus and liquidus, how can I make the specimen?

Hello, 

I am not sure whether I undertand your further question. I am not DTA or DSC expert, but some devices use very small specimens, with weight in miligrams. You can consult the person who will perform the measurment, how big sample is needed. From my point of view, there might be conflict with the grain size and the size of the specimen. In case of large grains from the casting, your sample might be, depending on the choice of sampling, with or without grain boundaries which will affect the DSC measurment. Grain boundaries are the last part to solidify, therefore first to melt. Incipient melting is taking place at grain boundaries. I assume there might be also differences in incipient melting temperatures due to different degrees of chemical heterogenity  resulting from different cooling rate during solidification in different parts of the casting.

Which alloy do you study, IN713(L)C?

Thanks for answer.

My alloy is IN713C.

I think your point is same as my opinion that measurement results must be different depending on the sample position (sample cut from chill, columnar and equi-axed), due to different chemical composition and grain size.

I understood exactly about that, but my question is, because the measurement results can be changed depending on the grain structure, if I want to obtain the solidification information of entire alloy, how can I prepare the sample and measure?

I want to avoid the experimental error which can come from the different grain structure.

Do you have any idea?

Hi, sorry for being late. If you have the opportunity, I suggest to perform preliminary measurement e.g. of these 3 structures at one heating rate, and you will see how big are the differences in temperatures of phase transormation (dissolution of g') and melting, if there are any. Then you will see. From the whole alloy, you probably obtain large interval. Due to small size of specimens for DSC, you will have to perform some statistics. Or lower and upper estimate by measurment of chosen microstructures.

Or as Peter Oslanec proposed, melt the specimen and then you will have destroyed the effect of different microstructure. Note for precise measurment, usually different heating and cooling rate experience are performed, since the transormation temperatures can vary with them.

Have you found something useful in a literature regarding your question? From people I know, in Ostrava they performed some DTA analysis, but don't know how they choose the experimental material.

https://www.researchgate.net/publication/257615729_Determination_of_thermophysical_properties_of_high_temperature_alloy_IN713LC_by_thermal_analysis

Article Determination of thermophysical properties of high temperatu...

I think that you should take into account that there might be some undercooling so there is no exact reaction temperature it depends on you're composition (also the minor elements) and process.  Generally 3 steps are considered in a Solidification process, germination, growth and coalescence and each of them could be affected not only by composition but also by  you're process.

I'm not an expert on DSC or DTA but while using those technics changing the heating or the cooling rate will give you different transformation temperature for the same sample. If you want to analyse the solidification you should analyse the cooling curve but it might be affected by the heating rate you used. Also DSC and DTA analysis will give you a peak over a range of temperature to analyse and it's not that easy to give one absolute temperature for the reaction.