"Nobleized - B" Ferralium 255 - low Copper

[cmcbain]

I had the ball and seats of a ball valve analyzed using X-Ray Fluorescence. They're supposed to be made of Ferralium Alloy 255 (UNS S32550) and treated by the "Nobleizing - B" process. What I found was that they all had much lower levels of Copper than they are supposed to.

ASTM A240 lists the copper content of Ferralium 255 as 1.5-2.5%. The parts that I have are coming in at 0.3-0.5% Copper.

Just FYI, the "Nobleizing - B" process does the following (according to the vendor):

"Nobleizing - B is a thermal diffusion process which diffuses Boron into the metal surface forming boride compound case in the metal surface. The case formed has a hardness level in excess of 1400 Knoop which exceeds the hardness capabilities of flame spraying, carburizing or hard chrome plating. To put the hardness into perspective, 950 Knoop is equivalent to 70 Rc. The case formed is from .0005" to .025" thick depending on the alloy treated."​

So, I've got a few questions:
1) Does anyone have any idea whether thermal diffusion of Boron could cause the copper content to drop? If I understand X-Ray Fluorescence right, it only detects the metal content of the surface alloy, where the Boron has been introduced.

2) How much would the lower copper content affect corrosion resistance? What I'm seeing is localized areas of bright orange rust where the ball and seats come into contact, and some very minor pitting in the same areas. The parts were in contact for probably 1 month in stagnant Deionized water (actually even less ions than most commercial DI water).

Any educated opinions would be appreciated.

 

[CoryPad]

1) There is no reason that thermal diffusion of boron would reduce copper concentration. XRF penetration depth depends on the energy of the excitation source and the resulting interaction volume on the sample, but it should be micrometers not millimeters.

2) Low copper should not affect the corrosion resistance, this alloy depends on the 25 % Cr and 3 % Mo for its corrosion resistance.

 

[cmcbain]

So if low copper doesn't affect corrosion resistance, I guess that raises another question: Is it normal to see bright surface rust and some shallow pitting on Ferralium parts after they've been wetted? All the rust areas are where the part was in contact with more Ferralium, and the patches are between 1/32" and 1/8" dia.

 

[cmcbain]

And I guess one more question:

If you asked a vendor for Ferallium 255, and you got something that doesn't meet the spec for Ferallium 255 because of low copper, do you shrug it off or do you ask the vendor to give you something that meets the spec?

 

[EdStainless]

The B treated surface will have lower corrosion resistance than the bulk alloy.
There will be CrB compounds so the 'effective' Cr at the surface will be reduced.
I would not trust XRF unless you know that you have ground through the B layer.
And then use a real XRF, not a hand held.

= = = = = = = = = = = = = = = = = = = =
Plymouth Tube

 

[cmcbain]

Ed,
Don't use the handheld because it is inherently inaccurate, or don't use it because it's not very accurate when your hand is moving around in relation to the test part? I only ask because we used a thermoscientific Niton handheld, but it was mounted in a fixture during the test so it wasn't moving around.

 

[EdStainless]

Given the power and detector technologies they are inherently less accurate.
They are better than they ever have been, and there are a lot of things that I use them for.
Determining the chemistry of a single sample isn't one of them.

= = = = = = = = = = = = = = = = = = = =
Plymouth Tube

 

[mrfailure]

I do not think portable XRF units are as inherently inaccurate as Ed thinks. However, surface preparation is very important. If you have not done a little grinding through the surface oxidees as Ed suggested, then you will get misleading results that may give you the low copper content you saw. See if you get a difference with surface grinding.

 

[MagBen]

The culprit might be the too low Cu.

Cu has a beneficil effect on corrosion rate in sulphuric acid environments, in seawater wher the formation of Cu sulphide can stifle the deblitating action of H2S. Due to its capability to chemcially stifle incipient pit growth, Cu is particularly imporatnt to prevent pitting resistance in chloride environments. If without taking Cu into consideration, PREN could not be a good gurantee for corrosion.

 

[EdStainless]

Ben,
I doubt that the Cu has much impact in this alloy.
There are a few other duplex grades with very similar Cr, Mo, and N, and no Cu.
The Cu will play a slight role in phase balance, but it is very weak.

MrF,
If you have multiple pieces of product, with identical surface condition then a hand held XRF will tell you if they are the same alloy or not. But if you don't have a reference of that alloy (with the same surface) to use as a type standard then I wouldn't trust the composition that you get.

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

 

[cmcbain]

The OP is still here and reading your replies, so thank you.

Question about XRF: I was under the impression that XRF broke composition down to the elemental level, so it wouldn't really matter if the surface layer had oxidized, because all the original elements would still be there, with a little extra Oxygen. True?

Wish I knew more about materials science so I could really understand what the Cu does in this alloy.

 

[EdStainless]

XRF is a whole study to itself.
The impact of composition and surface greatly depend on the energy levels of the X-ray beam.
The real problem is that changes in surface condition or compounds is that it changes the relative sensitivity of the various elements. This is also why you need to use type standards to make fine adjustments in calibration.
Most portable units use theoretical calculations which are not nearly as accurate as using real physical standards.

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

 

[mrfailure]

The metallic oxides that form on the surface will result in different concentrations than the substrate. These combine in the reading you obtain. This is why one first has to knock off the surface oxides using a portable grinder. This could be the real reading you are getting such low Cu.

 

[MagBen]

Since the agent in this case was DI water, the Cu content might not be critical to corrision. However, these two sources seemed to indicate the Cu had a fundamental impact on corrosion for this DSS in chemical and seawater environments.

Link
Link