Problems regarding the Heat Treatment in Bearing steel 2

[oka137]

Last week my company had some problems regarding the hardness of bearing steel after heat treatment. Bearing ring , AISI 52100, (inner and outer) was found after heat treatment process, having low hardness.

Before HT : 8 - 13 HRc
After HT : 61 - 64 HRc
Problem Ring : 18 - 22 HRc

After searching the whole day, we cannot find where the problem came from ; is it from the HT machine that have some trouble, or human error that makes the ring does not proceed do HT process?

We conducted three methods of investigations : using Back-Scatter to find the retained austenite, etching with nital to find the microstructure, and simulation on line.

The assumption was, if using EBSD we find significant peaks of Fe-C, than we can conclude that the ring follows the HT process, but if not, then it must be something else. Our first suspect were the grain consisting of Fe - Cr - Ni, the compound of stainless steel formed on 300 degree celcius.
We did not found any Fe - Cr - Ni compound, but we found a high ammount of Fe - Cr - O compound, so we suspect the ring had been heat treated under the eutectoid temperature, say 700 degree celcius, and the cooling rate is not fast enough to make tempered martensite.

Due to our limited optical microscop, we cannot determine the grain and the phase. The etchant available is only Picral and Nital, no Zephiran Chloride available.

I would like to ask any opinion about this. Is there something wrong in our investigation methods? or is the etchant not significant enough (i really think it is)? or our conclusion is wrong?

Thanks =D

 

[mw1st]

There's probably nothing wrong with your investigation.
1. What final hardness you looking for?
If you temper 52100, at 800F you should get about 45HRC.
2. How you conduct your heat treatment process. (see attached recommended heat treating practice for 52100).
3. And your assumption is correct, if your part was austenize at to low temp.(700C that's to low), and/or cooling rate was to slow, you will end up with nonmartensite products and further tempering will not correct the problem. Now, I'm not 100% sure, but when 52100 is quenched from austenitizing temp.,cooling rate equal or greater then 100F/s (measured at 1300F) is needed for avoiding the nose of TTT.
4. Also, I assume that the chemistry is OK. But even then, if we take specific analysis of 52100 we noticed potential variances in the spec.
C .98 - 1.1 Mn .25 - .45 Si .15 - .45 Cr 1.3 - 1.6

The gross variances will be:
C 11% Mn 44% Cr 19%
Although carbon potential is within the specification, it doesn't take great deal of carbon to influence as quench hardness condition.

Mark

 

[redpicker]

There are two possibilities I can think of.
1) The Problem Ring is not 52100 material.
2) The Problem Ring did not receive the proper heat treatment.

You can verify both by haveing the Problem Ring reprocessed. If it comes out fine the 2nd time, then you're probably #2.

If it remains soft, you are most likely looking at #1. A microstructure examination might help, but you really need a chemical analysis.

And, yes, Picral and/or Nital will both etch 52100 material. You generally don't need Zephiran Chloride. If you are having a wetting issue, you can use a drop or two of dishwashing detergent, but, particularly with Nital, it isn't needed. Longer eteching times may be needed.

rp

 

[sodeen]

I know this pretty much just echoes redpicker's synopsis, but it's so rare that I actually get to contribute to one of these because I actually have experience in the field.

How are the parts processed through heat treat? Are they furnace hardened or induction hardened? Are they hardened one at a time or in batches? If they're done in batches yet only one part failed to harden, it makes heat treatment a less likely culprit.

And yeah, if you run different materials in your facility other than 52100, make sure you know the true identity of the material. Thrice now I've been called to look at a batch of bearings that were soft after the normal Stainless Steel heat treatment only to discover that the parts were made from Precipitation Hardening Steel.

What size of part are we dealing with?

 

[arunmrao]

I echo the comments of redpicker. Also was there any instance of surface decarb . In my experience, bearing heat treaters are very well equipped and have tight process controls,such laxity is surprising.

_____________________________________
"It's better to die standing than live your whole life on the knees" by Peter Mayle in his book A Good Year

 

[oka137]

@mw1st : the final hardness is 61 - 64 for bearing products. We conduct our heat treatment process in a batch furnace and a conveyor furnace. Thanks for your advice, i'll look into it.

@redpicker : well, the element analysis shows that the ring is still 52100. I'll check again for your advice, red.

@sooden : the diameter is 22mm, the width size is 7mm.

@arunmrao : i suspect the same thing too, but i'm not quite understand the mechanism of decarburization, and also is there any reagents that initiates the decarburization? i suspect the conveyor furnace, because it is not vaccum, so the carbon can react with oxygen, but not quite sure about the parameter too....
do you have any knowledge about this, arun?

Thanks for all your advice, i'll report later tonight if i have made progress =D

 

[arunmrao]

Leakages in the furnaces,door curtains are the potential sources of air leakage. also you must have a slightly reducing atmosphere in the furnace,or any water leakage or dampness may cause problem. .

_____________________________________
"It's better to die standing than live your whole life on the knees" by Peter Mayle in his book A Good Year

 

[mw1st]

When parts are austenized under condition that doesn't provide some form of surface protection, your parts may get coated with rust (oxidized), and/or decarburized. And to put it very simply, the mechanism of decarburization is that the carbon in steel (under high temp.)reacts with the atmosphare so that the amount of carbon is reduced.

 

[redpicker]

oka137
I am not aware that Backscatter analysis reveals quantitative chemical analysis. Maybe you mean some other analysis method (or maybe I am not up with the latest innovations of electron microscopy). I am also confused about your use of the word "compound", such as "Fe - Cr - Ni compound" or "Fe - Cr - O compound". Did you clean the surface of the sample before you performed the analysis? You are obviously picking up contamination from somewhere since you should not be seeing any measurable amount of oxygen in the steel.

rp

 

[oka137]

@arun and @mw1st, i have done another back up investigation, and i found that the problem ring is on this three conditions : some ring have some black dust on the appearance (i am sure this is the product of decarburization), some other have a black-sticky layer on its raceway (which reveal the elements of aluminium-silicon-oxygen after SEM/EDAX), or having some slightly shiny coating from some molten metals (i think it is aluminium). I attached one of my SEM/EDAX analysis on the black sticky layer.

my analysis is : there are foreign material that cover the bearing steel, might be the aluminium that melt after contact with high temperature bearing steel or the sludge covering it, so it never reach the eutectoid temperature, resulting the low hardness of metals. Now the problem are ; where are the foreign material come from? and if it is prooved that aluminium is contamined the furnace, how can it be? =D

@rp, yes, EBSD cannot do the quantitative analysis. I use it to proove the existance of retained austenite phase (which is contained a high ammount of Carbon and Iron). If there is any retained austenite found, so i guess the bearing steel somehow reach the temperature more than eutectoid line.
PS: what i mean by compound is actually a phase, sorry for the misuse of the word ^^'

 

[rizzy77boy]

If a conveyor furnace was used, it is possible for individual parts to "hangup" as the fall from the conveyor into the quench. They slow cool and then eventually "unhang" and drop in with other good parts. The hanging points come from poor furnace design, warped furnace parts that form a ledge, or a buildup of carbon from oil quench fumes. This last one is very difficult to prevent. So much so that we open the back of each of our conveyor furnaces every three months to inspect and clean the drop chute into the oil quench.