Possible TME Failure?? 2

[nlj]

I have just read Thread 330-248132 on Temper Embrittlement and have some questions regarding a failure that I am currently analyzing.

The component is supposed to be 4147 but is 4140 deoxidized with silicon and grain refined with aluminum. The mechanical properties of one failure are TS 184 ksi, YS 121 ksi and Elongation 12%. The second failure has TS 216 ksi, YS 137 ksi and Elongation 2%. The mechanical values are well below the typical. Hardness readings however were quite high (52-55 HRC) I have submitted samples for Charpy testing but do not have the results yet. Microstructure is nonhomogeneous with tempered martensite, untempered martensite, bainite and pearlite. The fracture morphology is intergranular across the majority, microvoid coalesence across the remainder. What phenomenon could be causing the low mechanical properties? Is it possible TME? Poor steel quality? A mistake in heat treatment? The print states a minimum tempering temperature of 600F, hardness range 47-53 HRC.

Any thoughts would be appreciated.

 

[metengr]

nlj;
Did you confirm the chemical composition and presence of tramp elements?

 

[nlj]

The chemical compositions are as follows:

Carbon 0.388 0.395
Manganese 0.82 0.86
Phosphorus 0.0062 0.012
Sulfur 0.0036 0.0084
Silicon 0.283 0.243
Chromium 0.86 0.96
Nickel 0.129 0.112
Molybdenum 0.193 0.178
Copper 0.135 0.234
Aluminum 0.023 0.029
Niobium <0.0040 <0.0040
Vanadium 0.0064 0.0074
Titanium <0.0010 <0.0010
Tin 0.0083 0.0079
Boron 0.0007 0.0010

 

[metengr]

One additional comment - your statement above

Microstructure is nonhomogeneous with tempered martensite, untempered martensite, bainite and pearlite.

Was the microstructure you observed and mentioned above throughout the failed part? You should be observing a consistent tempered martensite microstructure with other phase constituents introduced further away from the quenched surface because of hardenability limitations with this alloy. What was the thickness of the failed part and heat treatment practice?

 

[nlj]

The non-homogeneous structure is throughout the entire part. It is machined from a 2.25" bar, machined to approximetaly 1.5" I do not know the details of the heat treatment other than oil quench and temper. Temper set to a minimum of 600F.

 

[nlj]

I am going to re-run chemistry to find antimony, arsenic, and tin content. I will re-post.

 

[swall]

While the subject matter is 5160 hot coiled, quenched and tempered springs, a passage from the 1993 edition of the SMI spring design manual may hold some relevance for your situation:
"It should be noted that as hardness increases above Rockwell C 48, ductility and toughness of the steel fall off rapidly..." "...springs made to a hardness of Rockwell C 53, for example, have been known to shatter under a constant load."

You may just be dealing with a brittle component because of the hardness range you are at.

 

[TVP]

(1) Non-homogeneous microstructure is a definite problem. It appears that the heat treating process was significantly off, especially the quenching phase.

(2) Yes, tempered martensite embrittlement is a possibility. Because the steel was presumably tempered at 600 F, this is right in the TME range (400-600 F, 205-370 C). Tempering in this range causes films of cementite to cover the prior austenite grain boundaries, which reduces the tensile strength, fracture strain (elongation), and fracture toughness.

(3) Low yield strength could be due to decarburization or other defects at the surface.

 

[redpicker]

I don't know why people want to use Temper Embrittlement as a failure mechanism. It is not. Perhaps the failure is due to low toughness (probably so), but the failure would have occured regardless of whether the low toughness was due to temper embrittlement or just plain poor heat treatment (as seems the case here).

Look at the yield to tensile ratios, 66% and 63%. Any highly loaded part should have a Y:T ratio of at least 80%.

The substitue of 4140 for the 4147 is a poor choice. 4147 has much higher carbon and manganese, both contribute strongly to the hardenability. With the reported mixed microstructure, the hardenability is most certianly an issue, as a higher hardenability steel would have developed more martensite on quenching, improved the yeild to tensile ratio, and would have much improved mechanical properties.

Temper embrittlement is a red herring. Much more important is the failure occured because of a poor transformation during heat treatment (quenching), and a lower-hardenability steel was substituted for what was originally called for.

rp