Carburizing Question 2

[metalman8357]

Hey guys,

I've got a quick question about carburizing and I can't seem to find a good answer in my references. I'm running a carburizing process on a custom alloy with similar chemistry to 8620 at 0.95% constant carbon potential and 1700F for 17.3 hours. The part is a rectangular bar measuring 1" x 1" x 4". I've ran 6 runs so far with identical process parameters and so far I'm seeing average effective case depths (depth at 50HRC) vary from 0.065" up to 0.073". For testing, I section the bar in half and run a traverse from all four sides. I'm also seeing a range between the hardness on sides up to 0.010" difference (never seen this before). The process I run is sand blast part, carburize then furnace cool, sand blast, coat with stop-off compound, austenitize at 1550F with endo-gas for 8hr, quench, and blast to remove stop off compound.

1.) One question I have is does the austenitizing process continue to drive carbon into the material even though I am using stop-off compound? Should I be having a better control on the time that these are austenitized in order to have smaller variation between carb runs. For example, if I were to carb and then immediately quench and temper would the depth on this bar be less than if I were to carb, furnace cool, austenitize for 8hr, and then quench and temper?

2.) What could be causing the range of up to 0.010" seen on case depth between two sides of the same block. Usually three of the four sides will be within 0.0005" in depth and one side is an outlier and off by up to 0.010".

3.) I'm using these test rectangular bars to dial in the process for a much larger part (20" in diameter). Is there any reason to believe that the case depths I'm shooting for on these test bars will be different on the larger part?

Any input is GREATLY appreciated. I've been scratching my head over these issues for weeks...

 

[redpicker]

Your stop-off paint may be part of your problem. You don't need it if you are running endo for your austenitize. Why are you running an 8 hour austenitize? That could be part of the issue as well. If you have a 1" square, why not just austenitize for 1 hour? The carbon that is in the part will diffuse during the austenitize cycle, although much slower at 1550F than at 1700F.

The difference in observed CD could be a function of the quench.

You could expect a slightly lower CD on the larger part, due to both the differences in quenching rate of the two parts and the fact that the larger surface area is going to consume the carburizing atmosphere faster and you may find it more difficult to control the carbon potential. How are you controlling the carbon potential?

Also, I would also consider carburize at 1700F, cut back to neutral atmosphere at the end of the carburizing cycle and furnace cool to 1550F to soak for an hour, then oil quench. Unless you are doing some post-carburize machining on the case, I don't see a big advantage to the second quenching operation unless you have trouble with controlling the carbon potential and/or you need to do some post-carburize machining.

 

[Lyrl]

1. Being at a high temperature enables carbon diffusion. While the stop-off will prevent the surface carbon from interacting with the carbon potential of the endo, basic diffusion will cause the carbon profile to flatten, meaning the surface carbon concentration will drop and the total depth of carbon diffusion will increase. Long reheat and quench times will tend to increase the case depth and lower the surface hardness. (Side note that if you quenched directly from a 0.95% carbon potential you might have excess retained austenite, so the reheat cycle's effect of lowering the surface carbon is helping you to minimize retained austenite; if you wanted to direct quench the normal procedure is to run between .9 to 1.2% for the "boost" and then when the temperature is dropped to 1550°F also drop the carbon potential to .75 to .85% for a few hours of "diffuse" before quenching.)

2. One factor to investigate is surface roughness, which can have a surprisingly large impact on how the surface accepts carbon. If the sandblasting isn't completely uniform (due to lack of appropriate tumbling or turning, for example) the surface texture difference may contribute to differences in case depth. A second factor that comes to mind is circulation of the endo gas around all the part surfaces. How is the part racked? Could one side have reduced or enhanced access to the carburizing atmosphere?

3. The hardenability of 8620 material is sensitive to section size in larger diameters; assuming oil quench the core hardness starts to drop off in section sizes over 2". I haven't worked with any 8620 parts of the size you're talking about but would estimate you'd start to see case depth affected around section sizes of 5-6". Your test piece will accurately represent the depth of carbon diffusion (as long as you stay away from the corners, which of course have deeper depths due to the higher surface area to volume ratio). The problem comes in when parts with identical carbon profiles quench out to different hardnesses depending on cooling rate, with thin parts cooling faster and getting harder compared to thick parts. The result is that the effective case depth to X hardness level is shallower in the thicker part.

 

[metalman8357]

Thank you both for the detailed responses. This gives me some great sources of variation to look into. I am going to run a few more tests and I will report back my findings. Thanks again.

 

[Maui]

You mentioned that,

What could be causing the range of up to 0.010" seen on case depth between two sides of the same block. Usually three of the four sides will be within 0.0005" in depth and one side is an outlier and off by up to 0.010".

Is the case depth shallow by 0.010" on that one side? If it is, then that may be because that one side is laying face down on the tray inside the carburizing furnace. If the gas has difficulty reaching that surface, it may not carburize to the same depth as the other sides do.

Maui

 

[dbooker630]

I would also check out the furnace to make sure the circulating fans are working properly; that could also be a source of variation. I also agree that you only need about an hour of soak time for your test coupon, although you will need a longer cycle for all of your actual part to get up to heat. Loading vertical vs. horizontal is another possible source of side variation; with the vertical load all four sides enter the quench at the same time, so that it would not matter which side you analyze as long as your furnace is working properly.