Grain structure of forged 440C

[CR100]

We have tested two different 440C shafts in a simulated failure mode to anaylys the strength or torque of whent he shaft will break.

Upon completion of the test we have found some very odd results.

Shaft #1 had the highest strength and the most refined grain structure. This part is forged from 440C and then heat treat to a hardness of 56 HRC.

Shaft #2 had roughly 65% of the strength of shaft#1. It is also forged and heat treated to a hardness of 56 HRC. But if you look at the picutre the grain structure is completely different. I am unsure of what to make of it.

Any ideas? I am thinking that it is a heat treating process but I am unsure.

 

[remetaper]

Structure and its coarse grains indicates a overheating.
Verify parameters of forging process and hardening too.
Just a look to temperature, strain rate and soaking.
Finally, Hrc is the same for both shafts ?
Do you know Hrc after forging process ?

 

[CR100]

remetaper, I will look for the data in regards to the temperatures, I know there was mention that the pieces were heated multiple times due to the pieces shifting while being heat treated.

Yes both shafts are HRC 56.

 

[dbooker630]

I would check out the forging process. The billet temperatures may have been excessive to the point of partial incipient melting that a subsequent heat treat cannot correct.

 

[remetaper]

dbooker630,
at first look , the reason why material failed depends on forging temperature, strain rate and soaking. In other words, higher temperature!
Coarse grain is the result of something of wrong in this process. For this reason, I asked for Hrc after forging process before slow cooling after this one. This value could be a good info about exagerated forging temperature or exagerated strain rate because you could have an idea about the influence of retained austenite whose amount depends on above parameters.
The higher retained austenite and the higher forging temperature, the lower Hrc after cooling.
In case of only one tempering after forging ( just to avoid cooling cracks ) HRc cannot reach Hrc = 56 if the structure had been overheating or "burned".
For this reason I agree about "incipient melting "but an easy optical microscope testing of grain coarsened piece would point up some grain boundaries with a evidence of melting (dentrides together decoesion between the grain boundaries)
However,the main problem of a hot forging /rolling process of this high carbon stainless steel grades is the habit to consider these ones as an austenic or low C martensitic grade. A lot of people forget that higher Carbon means lower liquidus/solidus borderline.
Finally, I think that if CR100 will collect other info about process,it will be easy to find the reason why of that "bad" structure without our help.

 

[redpicker]

This illustrates the problem with only specifying a hardness of a heat treated part. It is possible to achieve 56 HRC with 440C upon air cooling from forging temperatures. To obtain the optimum combination of hardness and toughness, however, you would need a full anneal followed by a harden and temper operation. If you only specify a hardness, you don't know what you'll get.

rp

 

[stanweld]

Often times, reforging billets can be 'lost' in the furnace and held at the forging initiation temperature for substantial time periods, which promote very large grain growth. Hot work by forging will not fully restore a finer grain structure.

 

[remetaper]

redpicker,
probably depends on my poor english, but I am talking about Hrc after hot forging and cooling. If temperature of forging were ( or had been ) highest ( close to Liquidus and decoesion ) the structure is the most part of retained austenite with some zone of fresh martensite. Therefore, Hrc will lower ( large amount of austenite ) compared to a right temperature of hot forging and cooling ( large amount of fresh martensite ). In this last case Hrc will surely be higher. Retained austenite is "soft" and , then, gives us an information about the temperature of forging. The higher temperature, the softer will be the structure due to retained austenite.
At the end of day, I presume that all people understood the reason why material failed and it could continue as academic discussion between metallurgists.
Finally, I know that after annealing ( or slow cooling after forging ) and H+T process , hardeness will be 56 because whole ( or part) of retained austenite is transformed in tempered martensite whose hardenss depends on tempering temperature. Even if we are facing to large grain !
Thanks for your notes on this subject.

 

[redpicker]

remetaper,
I think I understand your comments and I agree that knowing hardness after forging process can give valuable information. My comment was directed more at a common practice of only specifying a hardness as a heat treat condition. If these shafts were produced by two different vendors, Vendor A may be knowledgable of the intended use and provided a grain refining step in the processing. Vendor B, without this knowledge, only provided tempering after forging and checked to make sure he met the requirement. It is possible to have the coarse-grained structure without overheating if a grain-refining treatment was not employed after forging.

 

[CR100]

redpicker, if indeed this was the case where only the hardness was specified. Would UTS, YS and elongation % properties on the drawing cure this problem?

 

[Wrenchbender]

If I might chime in…, yes, I think specifying UTS, YS, EL (and RA) might go a long way in curing your problem. Also consider the following drawing requirements. They should address many of the previous concerns.

Definitely an HT spec such as AMS-H-6875 (Table 1B).
Definitely ASTM E8 (tensile).
Optional, ASTM E23 (CVN), or even E299 if it’s a supercritical application.
Optional, ASTM E381 (Macroetch).
and maybe throw in some NDT, e.g. ASTM E709 (mag particle) for surface cracks.

I have no suggestion for forging temperature, but I'm sure its documented somewhere for 440C. Any respectable forging shop should know the proper range of temperatures. (Generally, more hot work reduction gives smaller grains.)

Throw away the hardness requirement, or just keep it for reference. As you can see, different paths to the same hardness will not necessarily produce the same ductility. (Search Hall-Petch for more on the influence of grain size.)

BTW, what's the application for this shaft?

 

[Wrenchbender]

PS - Sorry, typo, that was E399 for Frax toughness.

 

[redpicker]

I agree that specifying tensile properties would be an improvement, but I am not sure it is really necessary.
A heat treat specification is really what is needed. I do not know if AMS-H-6875 is appropiate for 440C, but if it is, it would not be a bad choice. Depending on the specific applicaiton of the part, calling out an appropiate heat treat specification and a hardness may be adequate.

My original point is that only specifying a hardness can be insufficient if other properties are important. If tensile properties are important, then a tensile test should be specified. If toughness is needed, then CVN or other toughness test should be specified. For many steels and many applications, specifying a heat treatment and hardness is adequate to insure minimal properties are met. For more critical applications, addtional testing is justified

rp