Forge or quench crack? 8

[coreman73]

I received a couple of ratchet handles yesterday that show cracking in nearly the same location. Samples were hot forged from steel coil 4047 and heat treated to 42-46 HRC. Finish coating is Ni-Cr.

I am trying to determine the origin of this crack. Is there some criteria that differentiates forge cracking from quench cracking? I would assume that if this crack initiated during forging that there would be some internal surface decarburization along the crack path. I do not see this.

Please let me know any opinions regarding the orgin of this crack and how to reliably tell the difference between forge/quench cracking.

Thanks,
coreman

 

[Wrenchbender]

Only way I know to differentiate cracks after the fact is that no oxidation or decarb points to a quench crack - but you already knew this. I would have expected a crack to form in the head, where there was more hot-working (larger strain) and the cross sectional area is more irregular.

Assumptions:
a. Forging press came down perpendicular to the photograph, so material flow was in an up-down direction relative to the photo.
b. Armstrong has their process under control: forging temperature is high enough and strain rate is low enough.
c. Parts are tempered without a long delay after quenching.
d. This grade of steel does not air harden. (So ID markings, stamped into as-forged parts prior to heat treatment(?), are not causing the cracks.)

Were the two units from the same day's production? Perhaps one of the processing variables temporarily went out of spec.

Sorry, no answer, but forging temperature is my first suspicion. Also looks like it says USA on the handle. Shame on us. Hope it wasn't for export.

 

[TVP]

coreman73,

You have the right idea: if the crack occurred during/immediately after forging, then it should have some decarburization. Quench cracks are almost always straight, have some amount of scale within them, are propagate intergranularly. Your example seems to fit all three of these criteria.

 

[coreman73]

Thanks Bestwrench and TVP,

Yes, these two parts were from the same day's production. Stamping did occur prior to heat treatment.

I am leaning towards a quench crack as well for those three reasons.

There's one final issue that's bugging me though. The cracking occurred in EXACTLY the same area for both wrenches. What might be the reason for this and how could quench cracking be identical for two differnet parts? Could this localized area somehow simply contain more forming stresses? Very puzzling.

 

[TVP]

The answer is yes to your last question. Why should quench cracking necessarily be random?

 

[coreman73]

Thanks TVP. I guess I just found it a little odd that these cracks were so similar between different parts.

Since we've pretty much established that these are quench cracks, what might be some possible reasons for faulty quench? What variables regarding quenching could be manipulated to prevent the quench cracking? Fyi, the quench medium is oil. Also, as far as I know, the heat treating room at this facility is very cold right now.

 

[unclesyd]

I would question that the cracking is from quenching for several reasons.

The crack has too much oxide unless they lost control of their furnace.

I would expect quench cracks to the area of the hole.

One could also question the forging stock as to the quality of the bar or coil. I've investigated two case where there were defects in feed stock that were carried through the process. One was 3/4" bar stock and the bigger one was 3/8" coil from 4130 feedstock.
The coil was interesting as it was produced by the same company. the cause was traced to production people not trimming the ends of the coils.

 

[Compositepro]

I agree with unclesyd. The crack my have occurred in quenching but was probably due to a defect in the coil stock where defects would most likely be linear. Is there automated handling during forging? This may explain the cracks in the same spot, although forging stresses could also explain that.

 

[redpicker]

I would suspect quench crack, for the reasons noted above, but I don't think we can rule out the crack resulting from the forming operations.

Is the steel chemistry suspectable to quench cracking? (what is the carbon/alloy content?)

What forming operations are used? I would guess a close-die press forging, but it could be upset, roll formed, or even machined.

It looks as if, while both the cracks are similar in location, they are not exactly. In the upper ratchet, I can see a crack in the round handle, running through the middle of the numbers and the letters, but the end is all reflection, so I can't see where it is going. In the lower one, it looks as if there are actually two cracks; one in the round part, as above, and running through the numbers and the letter A. The seems to be another crack that can be seen at the top of the M runs through the longitudinally straight parts of the S, T, R, O and G and cutting through the top of the N. Are these two cracks in the second ratchet?

Is the heat treatment in a controlled atmosphere? I would guess it is becasue of the lack of decarb on any of the parts.

I would expect a quench crack to have formed along the straight part of the stamping, as I described for the letters S, T, R, O and G. I would not expect a quench crack to appear as shown in figrue b), becasue it is not following the straight part of the R.

So, I see features in the lower ratches that makes me believe it is a quench crack, but these seem to be absent in the upper one, or maybe they are just obsucred by the reflection.

One last question. You say the room in the heat treating department is cold. Is it possible these were allowed to sit at sub-freezing temperatures after quench and prior to tempering? That could be the simple answer, particularly if the cracks in both ratchets can be shown to be associated with the stamping since it would be difficult to get a forging crack to line up with stamping that wasn't performed until after the majority of forging reduction had been performed.

rp

 

[ornerynorsk]

I would try to rule out material defect as a first step, go through batch records, certs, etc. In the last 20 years, I have seen material quality go to heck in a hand basket. I have seen some of the strangest defects in ferrous and non-ferrous stock, and these problems seem to be showing up with greater frequency. Have you had opportunity to inspect raw coilstock?

Is quench medium brought to within acceptable temps prior to a start of production?

Has quench technique changed lately? Rate of entry or direction of entry, changing bath to spray, etc?

These are just a few obvious things that came to mind. Good luck on the detective work!

 

[coreman73]

I really appreciate all the excellent insight everyone. I will have to track down some of the unknowns regarding starting coil stock and forging specifics. This may be more difficult to do than not.

 

[coreman73]

Redpicker,

Carbon content is 0.47%. How can a steel's chemistry make it susceptible to quench cracking?

I apologize for poor photo quality. There is actually only ONE crack associated with each part. Both the location and orientation of these cracks are virtually identical for each ratchet.

Yes, heat treatment is in a controlled atmosphere.

It is likely that these parts were exposed to sub-freezing temperatures after quench and prior to temper. Unfortunately, there is really no way to prove this though.

 

[redpicker]

A 0.47% carbon steel can be very susceptable to quench cracking, particularly if allowed to sit for an extended period of time after quenching and prior to tempering, particularly if exposed to sub-freezing temperatures during that time.

The austenite to martensite transformation results in a volume increase that can create high internal residual stresses. Martensite formed in a steel with 0.47% C is going to be very hard and brittle and will crack due to these stresses. Conversely, the martensite formed in a steel with 0.25% C will not be as hard or brittle and will resist cracking in the same conditions. The lower temperatures futher lower the toughness of the material, increasing the likelihood of the formation of cracks.

A good heat treating procedure would limit the time the between the quench and temper operations (say two hours, but should be adjusted to meet shop conditions and experience). A limit on the minimum temperature the as-quenched parts could be exposed to may also be adviseable. If there the shop procedures do not limit the amount of time between the quench and temper operations, they need to be revised to include such a limit and the processing records should also record the time of these operations so the shop can document they followed the procedure.

rp

 

[arunmrao]

There are many valuable comments. The carbon is high 0.47% and certainly prone to cracking.

Have you missed out any form of stress relief after forging? Also,for such a high carbon material,a milder form of quenching will be advised.

No need to freeze after the quenching,if that is true, also the oil could be warmer to reduce the intensity of quench.

Finally, a flash tempering operation immediately after quenching,would have been helpful.

Just my 2 cents!

_____________________________________
"The richer we have become materially, the poorer we have become morally and spiritually." Martin Luther King Jr

 

[coreman73]

Yes, lots of excellent comments from everyone. I really appreciate it.

Thanks Redpicker. I will suggest to the heat treating area that they need to record and control how much time lapses between quench and temper.

Arunmrao,
I am not sure if any stress relief was done after forging. What could a milder for of quenching consist of? I guess I'd like to nkow which parameters would have to be adjusted to make it milder?

 

[arunmrao]

You can get slow queenching oils,which might reduce the severity of quenching,and perhaps they might suggest using a warm oil,rather than a cold one. Just a suggestion,can you try fast air cooling as the section thickness is not large and carbon content is high.

_____________________________________
"The richer we have become materially, the poorer we have become morally and spiritually." Martin Luther King Jr

 

[coreman73]

Thanks arunmrao. I will suggest to them to try fast air cooling and also to review their quenching protocol.

 

[salmon2]

redpicker, I think I read somewhere that tempering should begin 'immediately' after 'end of queching'. Any lapse could result in cracking. I agree end of quenching is sometime hard to define, especially for high carbon or alloy steel. Can you clarify your two hour limit further?

 

[redpicker]

salmon2
You are correct, any delay in tempering after the quenching (of low alloy steels) should be avoided. However, in the real world, unforseen things happen that can delay processing (quench elevators get stuck, forklifts break down, power failures knock-out furnace controls, etc...). The two hour limit I mentioned above was just intended to be an example of a pratical limit in heat treating procedures that I have experience with.

Typically, heat treating operations get the blame cracking parts. By having a having such a limit in their procedure, the HT department can defend their processing by saying they followed the procedure. To do this, they have to document the time the parts were quenched (start and stop) as well as when they were charged into the temper. With this information on file, if a cracking problem is discovered, the records can be reviewed to see if the quench delay on the suspect parts was any different than other loads processed. It also makes sure the HT personell understand that it is important to temper the parts soon after quenching and gives them a limit that they can use to process through delays without involving management. That is, if they have some equipment failure that results in the parts sitting around for 45 minutes, they can continue processing once the problem is fixed without needing to submit a descrepancy report.

Of course, if their plant experience shows that the two hour limit is too long and results in cracked parts, then this limit should be reduced.

rp

 

[salmon2]

redpicker, apprecite your clarification very much.

salmon2