Unexplained failure of 17-7PH sheet part 1

[mighoser]

Here's the story. This parts are made using 17-7ph sheet stock (AMS5528 or MIL-S-25043) which is cut into strips and then heat treated to TH1050 IAW Mil-H-6875. The parts are then formed using fine blanking process. The failure occured while theres parts are in opperation. They link together with one another and form a chain. We know that maximum normal opperating loads are 300 lbs and testing has shown these links failure at a minimum of 750lbs. This loading can't happen in reality because there is a weaker part which would fail first. These "links" were brand new so I can't imagine fatigue is the issue. Hardness was slightly high on the links @ Rockwell C 47. AMS 2759/3D says piece parts should be 38-44 HRC which covers heat treat of parts. AMS 5528 covers the raw material and its "responce to heat treatment" They indicate a range of 38-46 HRC. The failure surface is almost parallel with the sheet plane without necking. This is different from links that we tested. Any thoughts from the experts????

 

[TVP]

Incorrect heat treatment--> brittle microstructure--> high hardness, low ductility, low fracture toughness. You should have an appropriate metallurgical investigation performed to confirm the nature of the problem. Electron microscopy would be very useful for evaluating the fracture surfaces for the fracture mode (microvoid coalescence vs. cleavage or quasi-cleavage vs. intergranular.

 

[mighoser]

I agree that hardness could be the cause but our testing with other parts with similar hardness did not show premature failure. The failed part is already at a lab (no results) I just wanted to know if there was any obvious pitfalls common to this alloy and heat treat.

Thank you.

 

[swall]

Why are you forming after heat treat? Normal procedure would be to form "Condition A" material, then heat treat to TH1050. Consider the possibility of forming cracks.I agree fully with TVP to get a metallurgical failure analysis performed.

 

[mighoser]

The part is approximately 1.75" by .65" and .032" thickness. We have a .045" max free state requiement which historically can't be met with forming from condition A and then heat treat. Keep in mind, this is proably the millionth part that has been produced for us and these machines have been in service since 2000 (this unit was brand new and failed during acceptance testing = last step before selling to customer). Given the properties of this alloy and heat treat, do you think a hardness of 47 (one point out in terms of AMS 5528 and three points out in terms of AMS 2759)is sufficient to reduce the min failure load by over 50%. Does anyone know where to look for macro/micro surface failure images of this alloy (sheet) with various heat treats?

Thanks.

 

[CoryPad]

mighoser,

You can't use the AMS hardness ranges as guarantees that your part will meet its requirements. There isn't a direct cause:effect relationship for hardness and other properties. Based on the information you provided, it seems like TVP and swall are giving good advice.

For images, try these products from ASM International:

http://products.asminternational.org/matinfo/index.jsp

Regards,

Cory

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[EdStainless]

Hardness isn't the issue, ductility is.
When the parts are heat treated are they reannealed?
It may be that hte sheet was a little to hard to start with (from the temper roll pass). If this happened you would end up with CH material.

Are sample links from each lot destructivly tested for strength and elongation?

= = = = = = = = = = = = = = = = = = = =
Rust never sleeps
Neither should your protection

http://www.trent-tube.com/contact/Tech_Assist.cfm

 

[mighoser]

CoryPad,

Thanks for the link. I tried to look up some micrographs and I guess I need to have a subcription. Do you know of any free site that would have similar information?

Thanks.

 

[israelkk]

If the hardness is high there may be cracks formed during forming. Because the high hardness and maybe a problem with the base metal or the heat treatment the cracks may be critical from the fracture toughness point of view and the crack instantly expand through the whole part. You should check for cracks on the surface using an NDI test. Is the chain loaded under sustain constant load or a dynamic intermittent loads?

 

[CoryPad]

mighoser,

Yes, you need a subscription. You won't find this kind of information for free.

Regards,

Cory

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[mighoser]

Edstainless,

I'm not sure what reannealed means. I know the material is recieved in Condition A, cut into 24" pieces and then processed to condtion TH1050 per AMS 5528. This process is as follows: 1400F for 90 mins then cool to below 60F within 1 hr, hold below 60 for not less than 30 mins. Then Age at 1050F for 1.5 hrs. We do not do destructive testing on each lot. I'm not sure if this would help since we haven't been able to recreate the observed failure mode/load.

Thanks

 

[mighoser]

israelkk,

The chain is loaded under dynamic intermittent loads. A full analysis has been done and factor of saftey is above 2 given the material properties of this alloy and heat treat. I would estimate that this parts saw a max of 20-30 cycles of the maximum loading condition that I described in my initial post.

 

[mighoser]

Is the someplace I could put the photo so everyone can see it?

 

[israelkk]

mighoser

I am talking about KIc and fracture toughness theory which can explain the fail under such low stresses if there are critical (deep/wide enough as opposed to subcritical) cracks.

 

[mighoser]

Israelkk,

I follow what you are saying. Where would these types of cracks come from? Do you think the fine blanking process post heat treat could do this? One thing that I've observed on from macro-photos is a thin lighter line at the midplane of the sheet (almost like two sheets were laminated together). Failure appears to have propaged along this plane. Any thoughts? How would this phenomenon be generated? I'm working on getting some pictures posted somewhere.

Thanks.

 

[EdStainless]

My last post vanished, I'll try again.
You need to look at the original sheet material, both micros and properties. If it does not have a good elongation then you may need to reanneal at 1750F before you go into the 1400F.
I suspect under annealed sheet that cracked in blanking, but it will take some work to prove that.

= = = = = = = = = = = = = = = = = = = =
Rust never sleeps
Neither should your protection

http://www.trent-tube.com/contact/Tech_Assist.cfm

 

[CoryPad]

mighoser,

You will need a site other than Eng-Tips to host your image, then link to it.

Look in the Your Reply area of this page, Step 2 Options, click the Process TGML link. This will provide the code you need to have the image displayed in this thread.

If you don't have your own site to host images, there are free ones on the Internet. For example,

http://www.imageshack.us

Regards,

Cory

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[mighoser]

Thanks Cory,

I can access those types of website from my work. I've tried and I get blocked. I'll post some pictures when I get home tonight.

Thanks.

 

[mighoser]

Cory,

Typo Correction: Can't

 

[mcguire]

These alloys have a tendency to form persistent delta ferrite, especially at the centerline of cast slabs. These planes of ferrite have negligible toughness. Cracks are easily formed during shearing.
This condition is a consequence of sub-optimal casting conditions and insufficient soak time before hot rolling. There is little you can do at this stage to alleviate it. The homogenization annealing times would be too long.

Michael McGuire

http://stainlesssteelforengineers.blogspot.com/