Spectral analysis is widely applied in metallurgical, casting and metal industries. However we as a newbie may some time feel unsure of the result though the spectrometer is in a good and stable performance. This is largely due to your poor understanding and usage of data selection principles. What affects the accuracy of the result? How do you select the right data? The following approach is crucial:

1) RSD (Relative Standard Deviation) is the key. Its value is represented in percentage (0.5 means 0.5%) and calculated automatically by computer. The small value of RSD means the small difference among several measurements. As a result, the average result is closer to the result as it is .

2) Effective data should fall into the range of RSD. Different elements relate to different RSD. Take steel as an example:

The above data derives from GB/T222-2006.

1) The way that you prepare samples has a weight on the result.

2) Sometimes, machine’s malfunction can happen, rendering some data ineffective.