The sample plate width and length sound good. Take into account that it must be also thick enough in order to develop sufficient residual stress.
Make it of low carbon steel, 25 mm (one inch) thick (not less than 20 mm, or 3/4 inches), with a V groove, i.e., on only one side. In this way, it will develop a higher residual stress.
Weld it with SMAW (coated or stick eletrode), E-6010 or E-6012. I suggest E-6012 because the weld is easier to make and the residual stress will be the same.
Have you already found out someone who has an X-ray difractor to measure the residual stress? It's not an equipment easy to find.
To have a good idea of what the equipment looks like and how it works, enter www.rigaku.com
Once in there, go to X-ray difraction.
By the way, the correct word is X-ray difraction, not X-spray, as you've written.
I would like to receive a copy of your numerical analysis, if it's written in English. Can you send it to me?
Giovanni S. Crisi
Sao Paulo - Brazil