STAINLESS STEEL 316 - DEMAGNETISATION

Dear Rangasamy,

The crystal structure may be will be changed in cold worked S316. then the S316 is partially magnetic .

If you want to eliminate the magnetic, you could use the degausser -demagnetizer (TVP45 said); heat treatment or continuously throw the parts.

Please you could select these method according as your parts status.

Henry

TVP45 has it right.

I'd like to expand the idea of the difference between "being attracted to the magnet and being magnetic."

get a gas meter to tell you if the material is in fact magnetic.

a one to ten guess should be sufficiently discernible.

If the parts are not magnetic, yourefforts to demagnetize may in fact be "charging" the work hardened layer, making the problem worse.

To solve this problem, first take measurements to adequately define it. is it residual magnetism inherent in the part, or is it attraction to magnet?

Gauss meter will tell you.

If you don't havedata that describes the problem, you haven't adequately defined the problem.

Bring DATA!

milo

Hi Rangasamy,

there is no condition: completely or totally nonmagnetic.

You can improve your situation by:

A. get a higher current in your demagnetiser, most are much too low in flux, so more current will help.

B. Build your own demagnetiser if your existing one is not tolerating higher current.

C. Make a careful heat treatment below protective gas: heating to above the Curie-pint (the temperature where "magnetisation" vanishes) will remove most of the magnetic field. But be sure not to oxidise your parts: external and internal oxides are pretty magnetic! That's why specialised magnetic materials are heat treated (1100°C) in hydrogen atmosphere for hours to remove internal oxides by reduction and diffusion.

D. Change your material to a better one (magnetically spoken): iron-free alloys of Al, Cu, Ti will do a good job.

Tell us for what purpose you want to have low magnetic flux density, so a useful estimate of needed performance can be done.

RHABE

Hydrogen causes embrittlement in ferrous metals, so the high-temperature soak in hydrogen may introduce new difficulties. If your housing needs to be both non-magnetic and metallic, a non-ferrous alloy is probably indicated. Keep in mind that some other metals, like cobalt, also exhibit magnetic susceptibility.

The CRC Handbook is a good beginning place for your further investigation. It's a standard reference, available in most libraries, if you don't already own a copy.

Just a stupid comment. Stainless steel work hardens. There may be some residual steel from the drill bit used adhering to the inside of the bore? perhaps a light acid treatment? As I said probably stupid. The other comments are valid indeed.

Not stupid at all. I hadn't even thought of that, but it's a real possibility. Good catch.

Stainless steel needs argon arc welding (see local exhaust fitter)

Might help

The person that suggested work hardening is correct. Work hardening increases the magnetic permeability of 316 stainless steel. The only way to eliminate the magnetism is to fully anneal it. Heat it to 1600°F and allow it to cool slowly. To keep it from oxidizing and ruining your finish wrap it in stainless steel foil to keep the oxygen away or do it in vacuum. A better alloy would be 20Cb3 which has a magnetic permeability one tenth that of 316.

Good luck.

Try heating up the pc. to 450-500 degrees F. and let it slow cool. Not sure why it works but I have had success with this procedure in the past on Stainless, and Chrome. Moly. steels after becoming magnetized from cold working.

pipewelder

We experience similar problems with residual magnetism & are currently trying to resolve this. Our parts are made from 316LN EFRB which we use because of it's good properties under vacuum. The parts are vacuum stoved at over 900°C after machining & the finished assemblies are degaussed in a strong coil.

The level of magnetism is quite low, we believe something in the order of 0.5 gauss, but it is enough to affect the flow of electrons in the adjacent vacuum envelope.

We believe that we have narrowed down the affected area to the weld joints, we are aware that st.steel tends to become more magnetic when welded but this seems to persist after degaussing.

We are also looking at using copper/nickel alloy in place of the stainless but this will give rise to problems in assembly so we are still trying to resolve the stainless magnetism problem.

Try welding the stainless steel with inconel 625. We build components for superconducting magnets that require very uniform magnetic fields so any residual magnetism is a problem. As mentioned above 20Cb3 is an alloy with very low magnetic permeability as is Inconel 625. Of course the Inconel 625 is several times more expensive than stainless steel. 310 stainles steel weld wire is also low magnetic permeability but very crack sensitive. Learned that the hard way!

Another filler metal that is used in superconducting magnet cryostats is ER385L (a GTAW electrode)

Thanks for the suggestion, I'll pass that one on. at present we TIG weld without using any filler. The cost of the Inconel is not an issue, it will be small compared to the cost of solving this problem.

Just an update, we have now obtained a gauss-meter & have measured the residual magnetism at around 150-160milligauss. This does seem to be localised at the weld joints.

Our experience has shown, if this magnetic permeability is a significant problem got you, there is a way to select heats of 316L stainless steel that has the best chance of having the lowest magnetic permeability. There are two ratios that can must checked using the nickel and chromium ratios. If this is a big problem for you and you have an interest in the information, I can dig out that information from an old job and can provide the formulae and the ratios to select.

Thanks for the offer, I think we'll try the other routes first because our use of the 316 is pretty low & selecting material batches would be quite a large job.

We will probably try your suggestion of using Inconel filler. Using a filler rod would be quite difficult with our welding set up but we think we could make a thin ring of Inconel & sandwich it in the weld joint.

Hi Nigh,

you have two different problems:

if you weld there is severe plastic flow in the vicinity of the seams: this is changing the magnetic properties to have more coercive force Hc.

Degaussing with AC of amplitude going down from maximum to zero in a strong coil is ok, but be aware of the decay time of the eddy currents that will prevent complete magnetisation and thus complete demagnetisation.

As we had similar problems and cost did dictate a more economic solution we switched to aluminum plate constructions: better than ever estimated. (Photo available on request, size is 40 x 40 x 170 cm with 50cm pre-chamber and 120 cm 3 stage sputtering chamber.

RHABE

What is the problem with copper-nickel?

What is the problem with copper-nickel?

Expansion, we are brazing to ceramic so the expansion rates are very different & we have to allow for this in the design. The Cu/Ni will expand a lot more than the ceramic then become locked to the ceramic as the braze material solidifies, then contract at the rate of the ceramic until it reaches ambient. We can calculate the allowance for this but with complex shapes it becomes more difficult & the inside of this Cu/Ni part has to be shaped to focus the flow of electrons within our assembly.

Not impossible to overcome but it generally involves a bit of trial & error with parts that are very expensive to scrap.

Hi Nigh,

you are absolutely right with these problems.

But CuNi is better suited than stainless steel as the expansion is nearly equal

but elastic modulus is much lower for CuNi, so stress is lower.

Did you try zirconium-oxide? Y-stabilised ZrO2 has more than twice the TCE than alumina.

I can design these joints either with elastic or with plastic deformation of the metallic parts to overcome difference in TCE. If you have a need please let me know.

I am using vacuum brazing too - mostly diamond to carbide - smaller size but same experience.

RHABE

We've had some success brazing Cu/Ni 'Ferry metal' to alumina ceramic & will probably continue down this route, partly because we have the alumina components in stock but also because we have to best match the expansions of optical components as well.

The expansion of the alumina is 6.8 µm/mm/°C, which is the same as the Kovar Ni alloy that we use in weld joints elsewhere. The Ferry is 14.7 & stainless about 18. The optical parts are either 6.8 or 8.

The Ferry also seems to weld to Kovar ok. We're still learning about using Ferry but our experience with copper is that, after brazing, the material has softened such that residual stress is not a problem.