Experimental Procedure
We were given six different compositions of a Pb-Sn alloy:
- 0wt% Sn (pure Pb)
- 10wt% Sn
- 20wt% Sn
- Eutectic point: 61.9wt% Sn
- 80wt% Sn
- 100wt% Sn (pure Sn)
These compositions should give us enough data to draw a fairly accurate phase diagram. Each composition gives us another piece of information about important points on the phase diagram. For example, the eutectic compositions will allow us to determine the eutectic point. The two pure compositions, 100wt% Sn and 100wt% Pb, will give us the data we need to identify the melting points of the two metals. The other compositions should enable us to determine the liquidus, solvus and solidus lines on the phase diagram.
- Firstly, the alloys were heated up to between 350-400ºC in a crucible so they were in a molten state. A Bunsen burner was used to heat the alloy to its liquid form so safety glasses and lab coats were worn at all times. Heatproof gloves and metal tongs were used to handle the crucible.
- Once molten, an electronic thermometer was placed in the liquid alloy and the mixture was allowed to cool down to room temperature. The temperature was measured every 10 seconds because it will give us an accurate enough picture of the rate of cooling and it also a comfortable amount of time to write down the temperature before you take the next reading.
- The measurements were then plotted onto a graph to give us the cooling curves for each of the compositions.
Results
I collected the relevant data and used it to plot a cooling curve for each of the different alloy compositions. The temperatures at which phase changes occur can be seen on the graphs, I have also summarised them in the table below.
I have used the data above to plot a phase diagram which can be found on a separate page.
Discussion
The data collected has allowed me to plot a fairly accurate phase diagram similar to the one in the handout. Most points and lines are in the right place including the eutectic isotherm. There are some points which are significantly out of place, for example, the points marking the liquidus, solidus and solvus lines at 10wt% Sn are all slightly greater than they should be. Another point which is out of place is the point for 100wt% Sn. On the handout phase diagram the value is 232°C but the value I measured was 186°C, a difference of 46°C. I have marked on the phase diagram in red where the solidus and solvus lines should be.
The inaccuracies described above are due to sources of error incurred during the experiment. One way to avoid these errors would be to take more measurements by using more compositions such as every 5wt%, eg, 100, 95, 90, 85wt% Sn. That way we have several points to use to draw a line instead of just one or two.
Another way in which inaccuracies caused by sources of error may be minimised is by taking measurements at smaller intervals, eg, every 2 seconds. This would be hard to do by hand but if you were to setup a digital thermometer with a computer it could take readings as often as you like. This would produce a more accurate cooling curve and therefore a more accurate phase diagram. However, this is irrelevant when there are far more significant sources of error to be corrected.
To plot a more accurate cooling curve you could cool the alloy down as gradually as possible. This would make plateaus in the cooling curves more apparent and more accurate points for the phase diagram would be easier to determine.
Finally, when the molten alloy in the crucible is cooling the temperature of the metal nearer the outside is cooler than that in the core. It would be better to homogenise the mixture while it cools to give a more accurate picture.
In conclusion, the most significant source of error is the lack of points available to draw a line on the phase diagram. By taking measurements for more compositions the accuracy of the phase diagram could be greatly improved.
