XRF Metal identification: uncertainty?

[kingnero]

Could anybody briefly explain (or link me to a reliable source) whether or not XRF PMI accurately shows the composition of metals? What are the uncertainties on the different elements?
In particular the carbon content? Does it matter if you are testing a ferritic or austenitic steel?

Thanks in advance...

 

[arunmrao]

I have been using PMI from Electro Nite. They are simple and fast to use, covering most of the alloy ranges . The battery lasts long and instrument does not demand much attention.

However as X rays are produced adequate precautions must be taken.

Surface cleanliness is important and the results taken on rough surfaces are not correct. On round surface need a minimum contact

Also elements having atomic number 12 and below cannot be measured by XRF. Thus Carbon content is not displayed.

Silicon is one of the lighter elements that can be measured accurately.

http://www.nitonuk.co.uk/pdf/periodictable.pdf

"Even,if you are a minority of one, truth is the truth."

Mahatma Gandhi.

 

[kingnero]

Thanks for this valuable information.

However, I have received XRF analysis displaying C up to 1/1000%.
This was not done using a hand-held machine, but rather (I saw the machine but I did not witness the testing) using a machine the size of a large welding post (lets say 80 - 100 kg) on wheels.
The place where they tested it was ground to clean, shiny material, and had one or few black spots, I suspect this is from the spark emission.

I have heard that this might not be accurate for carbon (or perhaps not even possible at all) - I will ask clarification where they got that C% number.

 

[mrfailure]

XRF, like EDS, is not suitable for measuring carbon even if the machine gives you a number. Too many people take this as gospel but the nature of the test does not support this.

 

[EdStainless]

If they are doing C then it is OE, either spark or laser ablation.
We can reliably do Al, Si, S and up with our lab scale XRF (vacuum path and huge tube).
If you want a real C number pay to have it done by combustion (Leco).

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[bobjustbob]

As others have said the XRF will not give you a carbon content. It seems as if it were done with a portable OES. OES will give a carbon content, however, it is not as accurate as a combustion analyzer such as a LECO. A stationary OES will give a decent carbon content if standardized for it and with good surface preparation. I would trust a portable OES measurement done on a whole part less so due to difficulty achieving a good flat surface. I would trust it to tell me the difference between a 1010 and 1030 material but not between CF8 and CF3.

If it were in dispute between the customer and the company I represent over the carbon content I would require an area sectioned out to run a LECO carbon. Be sure to look up and understand what is allowable by the ASTM specification for "product analysis."