Carburizing Surface hardness 4

[mfgenggear]

We manufacture many aircraft gears which material is AMS6265, AISI 9310 Steel, We do not have in house capability
for Carburizing and Hardening, al of it is farmed out.

We are having a recent problem with a part which
requires .022 efective case depth, 90,5 HR15N at .005 depth
however the verification is done with Vickers,via a cut Met Lab sample Mounted.

The parts are verifing 15N 89 at .002 depth. however it is meeting the hardness at .005 after is ground with 15N superficial.

The procedure as specified per AMS 2759/7 effective hardness
at 50 HRc cutoff.

I have been advised by the supplier that the problem occured because of loss of carbon potential.
This job is run in Endothermic Carburizing Furnace.

what I am aware of is air leak, dew point adjustment, adjustment of gas mixture(methane)

can you please asvise me some finer points in the matter above.

Thanks in advance

 

[redpicker]

I am not sure what finer points you want to be advised of.

Yes, an low carbon potential will result in a low case hardness. It would be possible for a loss in carbon potential would give you a low case hardness.

And, yes, an air leak would cause the loss of carbon potential, which would affect the dew point. At first, the heat treater could try to adjust the DP at the Endo Generator and, if unsuccessful, would make natural gas addition.

A possible result, particularly with 9310, is that the gas addition raised the carbon potential too high, resulting in retained austenite in the case. If this occured near the end of the cycle, it might only be a couple of thousanths deep. If this is the problem, they probably could be salvaged by a low temperature (cyrogenic) treatment. If the low surface hardness is due to a low carbon content, the cyro treatment won't help.

rp

 

[mfgenggear]

Redpicker thanks for your reply.

I guess I am trying to verify that what the
Supplier Informed me of, "is correct". I am trying to prevent further such Instance by getting more Knowledgeable in this area.

Please excuse my Inexperienced in these matters
what is "adjust the DP at the Endo Generator"

what was puzzling me was at .005 depth the case hardness was correct? should I be asking for pictures of the microstructure to verify the correct martensitic transformation?

Thanks

 

[dbooker630]

In this case "adjusting the dewpoint" would be to lower the air/gas ratio to produce a richer gas mixture exiting the endothermic atmosphere generator. The endo generator creates a protective atmosphere that is then enriched with methane or propane to provide a source of carbon for the parts to absorb.

Yes, absolutely request photomicrographs if not the specific samples examined, as well as the hardness survey so that you can examine the case profile.

Low carbon potential is certainly a possibility along with retatined austenite at the surface, but with a report and/or test sample, you will have a better idea of what happened. You can post the report on this thread and perhaps one of us can explain it if necessary.

 

[redpicker]

Yes, to add to what dbooker630 said;

The type of carburizing you are talking about, Endothermic Carburizing, is performed by heating the parts in a controlled atmosphere furnace. The atmosphere inside the furnace is composed of, among many different gases, carbon monoxide, which is produced by the "Endothermic Generator". The "quality" of the atmosphere depends on the ratio of carbon monoxide to carbon dioxide. It is difficult, however, to measure these concentrations accurately. In practice, the H2O content of the atmosphere is measured as the H2O is one of the "many different gases" that is produced by the endothermic generator and the H2O content can be easily measured by measuring the Dew Point (DP).

So, when running a Endo carburizing system, if you start to lose carbon potential in the furnace, the first step would be to adjust the output of the generator to produce gas with a lower Dew Point. If that does not correct the situation, then a natural gas (methane or propane, generally) is made to the furnace. This can be tricky, since it may result in the furnace going "BOOM" (I know from personal experience).

And, yes, the microstructure of the case needs to be examined to determine the cause of the low hardness.

rp

 

[tripleZ]

Does the heat treater have the ability to check the dewpoint both within the furnace case and at the endo generator? If your gas going in has a low DP, but your DP is running high in the case no matter how much gas you pump in, you're likely running an air leak somewhere, leading to the possible loss of carbon potential.

Second, I might question what flows they are running in the furnace now, and what they have historically run. That's not normally a huge proprietary issue with a trusted supplier. Our sintering process uses hydrogen (purchased) & nitrogen (self-generated). With th escalating cost of process gases such as hydrogen and propane, there is an industry-wide effort to cut processing gasses as close to the line as possible and increase consumption of the gases we generate.

 

[Helicopterjunky]

My opinion only..........

On relatively short case hardening jobs like yours in 9310, in order to get the hardness level at depth such as you mention the heat treater must flood the furnace with enriching gas throughout the cycle. The reason for this is that the short case depth job is all non-equilibrium carburizing, meaning that the surface of the material can only buildup a limited carbon potential. The surface material potential never reaches the potential of the surrounding furnace atmosphere. Even in a perfectly tight furnace, there must be a large available supply of enriching gas. Most likely this is line natural gas. Boost-diffuse cycles do not work well for short carburizing jobs. Stay with a straight carburizing cycle flooded with natural gas (high furnace carbon potential) . You may then tend to get some retained austenite at the surface which must be deep frozen out. Some people freeze right after quench and before temper, some snap temper first and then freeze. There is some risk of cracking intricate parts by freezing right after quench. Either way an additional temper must be perfomed after deep freeze to transform any converted untempered martensite to tempered martensite. Keep in mind that high furnace carbon potentials will over time tend to soot up a furnace and the furnace will need to be regularly burned out. This is running the furnace up to temperature and holding for lengthy periods with only air in the furnace. I would recommend a micro be made and reviewed and more discussion of carburizing process details with the heat treater. Modern heat treats have automatic dew point control at the endo generator and carbon probe potential control at the furnace.

 

[Helicopterjunky]

The case hardness can be correct after grinding .005" and checking direct with 15N but not meet on the mounted sample
transverse from the surface at .005 by Vickers. Which method actually buys off the parts. Most likely the direct 15N on finish ground part. The check on the sample is to give an idea of whether at grind stock removal the hardness will be sufficient. It should be possible to pass both in your example with 9310. I prefer using knoop for these case hardness at depth checks.

 

[mfgenggear]

Gentlemen

wanted to report the Independent lab report the
following results

Microhardness test

Knoop hardness @ 500 grams

Location OD gear

.005 90.7 HR15n 90.7HR15n
.022 51.5 HRc 50.0 HRc
.025 50.0 HRc

Requirement .005 90.5 HR15N Minimum

core 35 HRc 38.0 HRc

effective
case depth .025" .022"

requirement .022-.030 .022-.030

It conforms

The supplier results
at .005 57 HRc 58 HRc

does not conform

however after grinding gear

at surface 90.7 HRc

I am waiting to receive microcraphs

Thank you all

 

[mfgenggear]

Typo error
last line should read
photomicrographs

Thanks

 

[dbooker630]

I can comment more once you receive the photomicrographs, but are the 57-58 HRC @ .005" supplier results converted from HR15N? Also, I assume that the 90.7 result is a HR15N reading.

As far as the test lab results, if your 15N hardness is allowed to be converted from Knoop or Vickers then I agree that you are OK. I would confirm the allowable conversion. All parties have to agree on the scale to be used.

If the near surface case is homogeneous then it might be permissible to convert from the lighter load. The 15N takes more of the structure into account. The micros should tell the rest of the story.

 

[mfgenggear]

Dbooker630

Thank You for your reply.

The hardness at surface is my mistake should be 90.7 HR15N
and not HRc

Quote "but are the 57-58 HRC @ .005" supplier results converted from HR15N"

The 57-58 HRc was converted from Knoop/Vickers

It is allowed.

Take Care

 

[Helicopterjunky]

It is great that the gears conform per independent lab analysis. Your question now becomes why your supplier test results are differing from the independent test lab. Maybe faulty microhardness test or lack of clarity on buyoff being test sample, ground part or both. I can't say I really understand what the buyoff is. How are supplier results at .005" depth determined and/or converted? Does supplier grind .005 and then check direct with R/15N? If so could supplier have inadvertantly ground off more than .005". Is supplier also using Vickers at .005" depth on lab sample to verify? Still alot of questions in my mind.

 

[mfgenggear]

Helicopterjunky

Thank you for your reply

I am to meet with the supplier on this friday
about this matter an other issues.

The supplier is is using knoop/vickers microhardness on a
cut & mounted, polished & etched sample from the carburized gear teeth & journals . it is visually verified & micro harness tested per AMS2759/7, it is converted from knoop/vickers to HRc, It did not conform.

an actual gear was returned to us and it was precision ground then returned to the supplier for HR15N hardness test. and this is what I am trying to avoid in the future.
this did conform to the hardness requirement.

I never question their converting method.
but I shall find out.

the Independent lab used a cut, mounted and polished test sample and tested per the AMS 2759/7 and it conforms.


I do believe there is a problem with the suppliers micro hardness testing proceedure. This is going to be one of the Items disscussed in our meeting.
still waiting for the Micrographs.

Take Care

 

[mfgenggear]

here is the photomicrographs

.engineering

http://files.com/getfile.aspx?folder=cf4477bd-b9b7-4211-83f8-19f18ce0ae9e&file=micrographs1921.pdf

Please review and advise

Thanks

 

[mfgenggear]

if this doesnt work can some one delete these post
thanks

 

[mfgenggear]

Any one have a chance to look at the photomicrographs I uploaded for everyone review and comments,?

Thanks

 

[redpicker]

I took a look at them. It can be hard to evaluate merely looking at scanned images, so don't put too much into this evaluation.

That said, it looks like there may be some retained austenite in the near-surface area. That is not uncommon for 9310. I can't be sure just looking at the scanned image as what I am seeing might just be some lightly etching martensite. As mentioned above, if the low hardness is a result of retained austenite, a cyrogenic treatment will fix it.

Was a cyrogenic treatment used in the processing? It is often used as a matter of course when carburizing 9310. Because of the alloy content of 9310, you can get retained austenite with case carbon contents over around 0.7 %, which can make achieving 60+ HRC difficult without a cyrogenic treatment.

rp

 

[TVP]

I agree with redpicker on all points. The images showed some potential retained austenite, which can be converted to martensite with subzero processing. Aerospace gears using 9310 are almost always processed using subzero treatment, so it is a little surprising if this wasn't done for your parts.

 

[Helicopterjunky]

I agree also. May have some retained austenite but hard to tell at the 500X magnification. Can you get some higher magnification micrographs? Have your microhardness test inconsistencies from heat treater and test lab been resolved?