I received the following two questions:
Do we need to put error bar on the curve? Because there are many points on the graph, the error bars are impossible to be seen. Also when drawing calibration curve, should we use all the four data sets, or should we only use the data set obtained in our experiment?
Should we get yield stress and tensile stress from nominal stress strain curve or true stress strain curve?
Also for the true stress strain curve, do we need to draw the necking part? Because the equation relating nominal stress, strain with the true stress, strain is invalid when necking happens, so we can't get true stress, strain from the experimental data.
To answer the first question, I would say first off you probably won't need to use so many data points to make the strain vs. displacement curve. I think the only error you need to quantify here is how well your correlation fits the data (how close the data points are to the best-fit curve). It should be fine to just use the data collected for your section to develop that relationship.
For the second question, if you refer back to your lecture slides, you'll see from page 3 that you will get all your material properties (elastic modulus, yield strength and ultimate tensile strength) from your nominal stress-nominal strain curve.