Raw material defect? 4

[coreman73]

I've got an extension bar that's showing some prominent surface defects. This extension was fabricated by friction welding two pieces together (drive/female end is material 5140 and tang/male end is material 4142). The sample I have is in heat treated condition. The defects are found only on the 4142 steel (tang end) portion of the part and almost exclusively present in areas unaffected by the weld.

The attached file shows three photos. The first is an overview, the second shows "spider crack" defects and the third shows a longitudinal defect (probably a seam).

I'd like to see if anyone could give some possibilities on what these defects might be and how I could further prove it. I've got some experience dealing with seams so am confident in that assessment. What I have no idea on is the "spider crack" defects and what could have caused them.

Any help is most appreciated.

 

[redpicker]

You are probably correct on th elongitudinal defect. A tranverse section polished and etched with nital should reveal heavy decarb along the sides with oxides mixed in the decarb layer, suggesting the defect existing when rolling/forging.

The spider cracks are not so definative. They could be quench cracks, straightening cracks, or come from some other cause. Again, a microstructure examination can tell you a lot; whether or not there is decarb associated with the crack, is the crack filled with scale, and if it is, is it a temper scale or a high temperature scale. If the cracks blunt or sharp-tipped. Could be contamination (copper comes to mind) during processing. I am assuming these were found during production of the part, that is, they cannot be service-induced defects. If this extension has seen use, then there is a whole 'nuther set of mechanisms that could have caused the spider-cracks

rp

 

[metengr]

coreman73;
It is nearly impossible to just visually confirm the origin of these defects without a proper metallurgical analysis. For example, the star burst appearance of the defects could be local chemical segregation or banding with the material, improper metal working operation, or local surface contamination during heat treatment, etc. You will need to know how the bar was originally supplied - forged or rolled.

Get this piece to a metallurgical lab for proper analysis.

 

[arunmrao]

4142 steel can throw many surprises even to the most experienced ones. As suggested,please come back with a complete report and there could be meaningful responses.

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

 

[swall]

I have been doing failure analysis for 30 years, including 5 years with axle shafts, and have never seen defects that look like that. Just a guess--is the 4142 steel perhaps LaSalle etd 150 or similar product? If so, the "elevated temperature drawing" processing could have introduced some sort of tensile cracks of a somewhat ductile nature.

 

[coreman73]

Thanks for all replies so far. Just to clarify, these are not service-induced defects. Our QC guys found these defects following heat treating. Unfortunately, I have no other details as to how this extension was made/processed.

The 4142 material was supplied in rolled condition.

I made a transverse and longitudinal section through the defective areas and did a quick metallographic review (see attached file). Findings are:

1. Multiple internal voids were found around the "spider crack" areas. A thin layer of oxide outlines the inner perimeter of each void. I refer to these as voids since I can actually focus in and out of them.

2. There is absolutely no evidence of decarburization detected for either the "spider crack" or seam defect areas.

3. There is a very large internal crack present in the core. It does not make any contact with the surface in any way. I found oxide along the crack path.

4. The steel structure looks solid. There's no segregation and only light banding through the core.

With this information, my best guess root cause for "spider crack" defects would be that the steel making process formed the voids (oxide inclusions maybe impregnated with gas?). These stress risers were pushed to the surface where they burst/propagated in transverse/longitudinal directions during heat treatment resulting in the spider-like orientation.

Could this be at all possible or does this sound like a load of bs? It's all I can come up with.

 

[metengr]

Just speculation on my part given limited information it looks to me that the spider crack cross section shown in Photo 8 is an OD initiated quench crack based on appearance.

The internal crack you mentioned implies a serious processing problem during forming where the material in the core may not have been at the proper temperature for working, similar to an internal burst during extrusion. Either way, there was a serious process problem with this steel.

Just curious, what kind of material purchase specification are you enforcing? This material should have never left the mill.

 

[metengr]

It just occurred to me after I hit the send button that the internal cracking could also be from dissolved hydrogen during steel making - seen this before as flaking or internal bursts during cooling from rolling or forging in steel rounds.

 

[mrfailure]

I wonder what the microhardness of the spider crack area is on the surface compared with core interior hardness. Are they the same? If not, that could provide a clue.

Aaron Tanzer

http://www.lehightesting.com

 

[emxhps06]

Could it be the weld laps from friction welding that has been left in situ? Mind you, if they aren't removed, they can act as stress concentrator.

 

[swall]

Yes, figures 7 and 9 are evidence of hot cracking during steel processing. I have no hypothesis for the spherical features, though.

 

[TVP]

Wow, that is some bad steel. Spherical voids are usually due to entrapped gas, but are usually deformed during hot working, and turn into seams. My guess is that steel experienced very inhomogeneous deformation conditions during rolling, causing the internal cracks. The voids were never fully deformed, which is why there are so many clustered together.

 

[coreman73]

Thanks again for the great replies everyone. I think it's safe to say this is poor quality steel.

metengr,
Excellent point. I honestly do not know if we have a specificiation for this particular grade of steel. We purchase many other grades from this supplier and have specs for them. I need to see what QC has available for 4142. But yes, this steel should have never been shipped to us in such condition.

swall,
Could you please confirm you meant photos 7 and 9? Photo 7 showed the possible seam defect and not a crack.

 

[swall]

Yes, photo 5, not photo 7. I did not enlarge the images enough to read the mags the first time.

 

[redpicker]

Is this a "one-off" defective part, or characteristic of a lot of defective parts?

If it is a "one-off" unusual defective part, I'd chalk it up to it being the end of the original bar from the steel mill that had problems during processing (cold straightening cracks, extreme overheating, etc...) that should have been cropped, but for some reason, wasn't.

If it is part of a larger lot, then the steel mill has a problem and your best action is to make a claim against the mill. For some mills, this will not do any good, they just ignore any claims. Others take them seriously, but probaly will not tell you what went wrong. I've even had some tell me exactly what went wrong, then tell me they aren't going to do anything about it. But, it certianly appears to be a steel mill issue, so your first action would be to work with them.

rp

 

[coreman73]

swall,
Thanks for confirming photo 5.

rp,
These defects are showing up on the entire lot of defective parts. I've been in contact with the mill and they will be visiting the plant next week to take a look at the defects in person. I sent them my photos and they agree that the defects came from them. They've been really good in the past reimbursing us for the occasional bad coil so should do the same again. This is the first time I've ever run across such type of "spider crack" defect though. I'll see if they can pinpoint the root cause and then post back here.

 

[deadrange]

41XX series grades are highly suspectable to hydrogen flaking, especially when the bar cross sections gets larger. Vaccuum degassing is a must for this grade and in many cases a slow cooling to help in the removal of hydrogen. You should find out what kind of degassing cycle was run on this material as well as a hydrogen content of the material. Please also be aware where they are taking the hydrogen reading. If the hydrogen is being measured in the ladle, versus the tundish the values may be significantly different. There is a great deal of moisture in the tundish especially if it is recently relined tundish.

My previous experience indicates that you are going to need at least 2-3 hours of continuous operation on a tundish before the moisture is removed enough to not greatly change the hydrogen content of the steel.