Problem in Nitriding

Is it the developmental job? has this been earlier tried out with other vendors?

We have a lot of our components nitrided, carburized, carbo-nitrided etc. However we have never come across components where no dimensional variation has taken place during any of the above operations.

Based on the requirement of the tolerances after the necessary hardening operation we do the following.

a) Stress relieve before finish machining with very little allowance, finish machine and then send for the treatment(s)

b) Finish machine with allowance. Treat (carburize,...) , Finish machine, Harden (Quench)

c) Finish machine with allowance, Treat (carburize,..), harden(Quench) , Finish (hard cut or grind)

If it is precision component, except for some very costly treatment processes (like ion exchange etc, never resorted to though) i don't see how you can maintain dimensions through the heat treatments.

BTW even solid forgings we grind OD after carburizing, simply because of the volumetric expansion of transformation of austenite to martensite.

just googled and a paper may be of interest

http://www.hardening.com/index/Aufsatz2-e.pdf

http://www.finishing.com/483/40.shtml

Thanks sb and all others for valuable guidance and suggestions.

I will come back with additional information, as few have asked shortly.

Regards,

gsuhas

Consider using a QPQ (Kolene's tradename before they sold it off) or salt bath ferritic carbonitriding process. I've found that since the temps are lower and more even, the distortion due to stress relieving is lower.

You could try doing a rough machining operation, stress relieving the parts prior to final machining then nitriding.

If this had just started occurring in a process that had run for some time prior, then look for what has changed, new nitriding vendor, new materials, new supplier, new machine shop, new manufacturing processes, new employees at any of the process steps, even different seasons might be the culprit by changing storage temps or shop conditions.

As a example of seasonal effects, where I used to work we used to machine (or rather our outside vendors did) close tolerance parts out of a super duplex stainless material. This material had so much residual stresses in it, even after solution annealing, that it would never ever stop moving on you. The material had every element, including the kitchen sink in it and it was simply not dimensionally or metallurgically stable. Our vendors worked out a system in which the parts were rough machined, then put on a pallet out in the parking lot where the hot texas summer sun beat down on it for 4-6 weeks before the part was chucked back up and finish machined. The day-night heating and cooling cycles did a very good job of stress relieving the material. but that didn't work in the winter. Vendors had to resort to using kerosene fired heaters pointed at the pallets in the back shop bay during the day and allowed to cool during the evening. If they didn't, the parts had to be routinely scrapped.

you may try low temperature 'TUFTRIDING'( Compounding of Nitriding with toughning). Before this treatment, you should anneal the machined parts to relieve machining stresses.

We too have faced this problem while manufacturing precision components for steam turbines. However, in our case the required case depth was 0.3 mm. Initially we had to rework on these components but finally in consultation with Nitriding Vendor we kept a further clearance of 0.015mm on bore as well as outside dia. This worked well for us.

Hi,

"are machined to close tolerance"

are you looking for 100µm, 10µm, 1µm or what?

Which material, which previous heat tretament, which nitriding procedure and temperature?

How are the parts (dimensions?) put into the nitriding (bath, gas, plasma)?

More information will give better results to your questions.

RHABE

Finally, a pertinent posting...

Need more facts as has been requested.

from the limited information that has been presented, I would suggest that if the parts are put into process in a large mass (piled in a basket) then both absorption and quench efficienecy will be affected by the location in the mass/ basket/pile.

a description of the nature of the variability, CAse thickness on parts, between parts, frequency of too thick vs too thin, whether hardness is consistent on all parts all locations or frequency distribution of any of the above would be quite helpful.

Their is certainly a volume pickup from nitriding. but in a controlled process, that volume change will be consistent for all parts.

A better description of the nature of the variation would be more than helpful, but right now, I'm betting on parts going in in a large heap.

As would knowlwedge of what residue from manufacturing may still be on the parts that might inhibit nitriding...

milo

You don't say what alloy you're using or how much case you're specifying. If the case is more than a .002-.004 in, I generally see a brittle layer (shows white after a 2% nital etch on a metallograph) that must be ground off, so a small dimension change, .001 in or so, should not be a problem.

For higher carbon steels, deep cryogenic treating will stress relieve and strengthen, but not harden, the material making it significantly more resistant to abrasive and corrosive wear and does not result in dimensional change. And unlike coatings the entire structure of the material is improved. Dave, dkimmel@300below.com

Hi gsuhas, I suggest your trying TOUGHNITE / GAS SOFT NITRIDING PROCESS. This process was developed by DOWA MINING CO. LTD., Japan as a substitute for salt- bath based nitriding systems. This process requires a temperature of only about 570 degree C, well below the 730 degree C A1 transformation temperature & therefore generates minimal distortion & dimensional change. When stress relieving treatment precedes the toughnite process, very high precision can be obtained. In India, HIGH TEMP FURNACE is undertaking such heat-treatment jobs. You may contact them- their Web site : www.hightemp-furnaces.com. Good luck. Rangasamy

Further thought.

ask to see your vendors process control charts for temperature and for temperature "evenness" through out the furnace.

Without Statistical Process control evidence, The words that they say are just his wishful thinking.

How many thermocouples does he have in eachj zone?

How many are working? Same with burners etc.?

If they have to crank up the temperature (or increase the time ) (or need to but don't!!!) for all zones because of a 'slack zone' then that may be the Root cause of the dimensional differences.

Your furnace may not be a zone type (my experience is with continuous roller hearth with exothermic gas atmosphere) but my thought s may guide you.

Heres how I always evaluate these kind of Vendor problems in my professional practice:

1) Is there a process?

2) Was it followed?

3) Was it effective?

In the absence of SPC DATA, one cannot claim the answer is "YES" to any of the three questions. Thus, unwanted variability.

milo