Fracture face analysis 7

[coreman73]

I have a few fractured T-slot bolts that failed prematurely while in service. Simply based on what is visible from the attached fracture surface photos, what would be the general opinion on type of failure mode? I believe it's obviously fatigue but am unsure of anything else beyond that. Maybe by unidirectional bending? Any input would be appreciated.

Background information is that these bolts were formed from AISI 1045, not heat treated and surface finished with black oxide. The failures shown occurred in the threaded portion of shank. It is unknown how long these bolts were used before failure happened.

 

[metengr]

coreman73;
Based on what little information we have regarding prior service conditions of these T-slot bolts (dynamic loading, environment, etc.,) the fracture surfaces as shown appear to be subjected to undirectional, high stress conditions. The remainder of the fracture beyond the thumnail-shaped pre-cracked region could be brittle fracture induced very rapid (immpact) load conditions.

Just a guess until more information can be provided.

 

[swall]

Are there any cracks in the other thread roots?

 

[adamuk]

Really, you shoud get these bolts to the lab for analysis. Anything else is just a reasoned guesstimate.

Adam Potter MEng CEng MIMechE

http://www.ax-ea.co.uk

 

[arunmrao]

This is a general observation regarding alloys 1045,4140,etc,parts are quite often made and used in unheattreated condition. What benefit is achieved by specifying such alloys,which are supposed to perform best in heat treated condition. The stock reply I encounter is the drawing specifies it and then a blank stare.

 

[coreman73]

Thanks metengr. These bolts were said to be used for fixturing during machining applications. Unfortunately, that's all the info I have on their use. I do have some "competitor" bolts that were used in identical fashion but did not fail.

Swall, yes there is a crack in nearly every single thread root of the failed bolts. The competitor bolt was completely free of root cracks.

Microhardness results of the failed bolts show a surface hardness of ~27-30 HRC with core ~20-25 HRC. For comparison, the competitor bolt I tested has surface hardness ~35 HRC with core hardness ~30 HRC.

Chemical analysis showed the same steel grade was used for both the failed bolts and competitor bolts.

I've attached some photos illustrating the thread profiles for both the failed and competitor bolts. The failed bolts typically show inconsistent crest widths along with visibly different circular thickness between threads. The competitor bolt demonstrates consistent profiles throughout.

For the failed bolts, it appears that the damage pattern is different between threads. Some threads show crest deformation while others show deformation of the flank.

Could it be that the failed bolts did so simply due to poor quality of thread rolling?

 

[adamuk]

The bolt looks overloaded as you can see the crest of the tooth as moved to the left of the photo and the thread profile in cross section so the cracking and root deformation may be symptomatic of that. I.e. the bolt overloaded, cracked the root then subsequent vibration caused it to fatigue. The bottom (competitor) bolt has not seen deformation so either was not loaded as highly or is stronger.

Is one bolt machined thread and the other rolled? (etching the micro will see the reveal the microstructure and this can be easily seen). Rolled threads tend to be stronger than machined in most cases which my be why the 1st bolt failed prematurely.

Since it seems that you have met lab then are you a Metallugist?

Adam Potter MEng CEng MIMechE

http://www.ax-ea.co.uk

 

[boo1]

why do you think the theads were rolled?

 

[metengr]

coreman73;
Obviously, the failed bolts were without adequate heat treatment in comparison to the competitors bolts, in addition to the observed thread profiles. Poor thread profile will result in local stress concentration.
I would presume the competitors bolts are quenched and tempered, which is more desirable from the strength and notch toughness standpoint. Make sure that the bolts are adequately preloaded for the intended service, otherwise fatigue failures will occur in service.

 

[adamuk]

The uneven deformation would be due to the fact that bolt/nut contacts tend to be uneven. The highest stresses are int he first 2 or 3 threads then drops off dramatically after that. Also the machining varies hence the reason why the distortion'll occur in different areas.

The higher hardness in the competitor bolts'll indicative of higher yield and UTS so might be a stronger bolt.

Adam Potter MEng CEng MIMechE

http://www.ax-ea.co.uk

 

[coreman73]

Adamuk,
It was provided that the failed bolt MAY have been roll threaded. Otherwise, it's pure speculation as to how the threads were made for either of these bolts. I'm not really a metallurgist but have been trained to do some failure analysis. Otherwise, I do mostly material audits/evaluations. The lab I work in is set up to do both.

I've attached some photos illustrating the microstructure of the failed and competitor bolts. Both bolts show the same basic core microstructural constituents (ferrite and pearlite), which are also seen near the threaded surface of the failed specimen. In contrast, the competitor bolt transitions to a microstructure of what appears to be mostly tempered martensite throughout all threaded surface areas. Also, as expected the deformation is clearly more severe in the failed bolt versus the competitor.

So from these photos how could I tell how the threads from each sample were formed (rolled versus machined)?

Metengr,
Is it safe to say that the poorly formed threads could have contributed to failure?

 

[swall]

Coreman73---in the last set of micros you posted, the second to last set has two micros. The bottom is labled "failed bolt", but what about the top? It is captioned "near surface of thread" but it is not clear if this was the competitor bolt or the failed bolt.

 

[metengr]

coreman73;
Normally, evidence of thread rolling is indicated by local subsurface deformation of the microstructure, whereas, machined threads will reveal a cut surface with a very minor amount of subsurface deformation.

Yes, poorly formed thread profiles can result in local stress concentration and failure.

 

[dbooker630]

Macroetching a sample should reveal grain flow if you have a deep etch tank available. A strong HCl solution will work at ambient temps if you do not have a heated tank.

If you can take a photomicrograph at a lower mag (50x) that may help with the polished sample

 

[adamuk]

Normally, evidence of thread rolling is indicated by local subsurface deformation of the microstructure, whereas, machined threads will reveal a cut surface with a very minor amount of subsurface deformation.

Looks like I was beaten to it by metengr

Adam Potter MEng CEng MIMechE

http://www.ax-ea.co.uk

 

[coreman73]

Swall,
Sorry about that. The photo "near surface of threads" is related to the competitor bolt.

Metengr and Adamuk,
From what I can tell, it looks like there is little if any subsurface deformation so I'll go with machined threads for the failed bolts. I see much more subsurface deformation for the competitor bolt so will say the threads were formed by rolling.

 

[TVP]

coreman73,

Based on the information that you have provided so far, my assessment is as follows:

1. Bolt failure mode is by fatigue fracture, with relatively high nominal stress (as noted previously by metengr).

2. Failed bolts are overtightened which leads to thread deformation, root cracks, etc.

3. T-slot bolts should be quenched and tempered to at least Property Class 8.8 (ISO 898-1, ASTM F568, etc.) as required by product standards such as DIN 787. PC 8.8 requirement is for core hardness of 22-32 HRC (diameter < 16, otherwise 23-34 HRC).

4. T-slot bolts should also feature rolled threads, not cut.

FWIW, depending on the size of the bolts, 1045 may not provide sufficient hardenability to guarantee core hardness of 22 HRC minimum.

 

[coreman73]

TVP,
Thanks for the complete assessment. I just received confirmation from the manufacturer that the failed bolts feature cut threads. You stated that T-slot bolts should have rolled threads. Is there a standard that states this or what is the reason for this?

 

[stanweld]

coreman73,
Considering that the bolts were reported to be used for fixturing during machining operations, repetitive tightening and loosening would be expected. Tightening tension would be expected to vary considerably, from over tight to under tight. The failure mode should be expected, especially when manufacturing (machine cutting) threads from common, as rolled, 1045 bar stock.

 

[swall]

The near surface microstructure in the last image set appears to be martensitic. Either that, or it was only lightly etched. If the former, this suggests that the threads were hardened. This might make sense for a threaded item designed to be assembled/disassembled multiple times during its lifespan. Following up on TVP's comments on thread root cracks--these could also be additional fatigue cracks.