Another casting failure 6

[emonje]

Continuing from this thread:

http://www.eng-tips.com/viewthread.cfm?qid=418786

Taking the sound advice & opening a new thread as it's a different part.

This casting failed over the holiday break.

Here are the shots of fracture surface:

All these ridges and facets, does this look like "rock candy" fracture? I'm thinking of asking for Al and Boron content tests.

Why is there this different layer on the outside?? Brittle fracture inside then ductile failure on this shinier layer??

Other side survived.

The casting will be in town from the site next week, trying to get my head around & see if there's any obvious hint what happened here.

Thanks for looking & your comments highly appreciated.

 

[mfgenggear]

just an unofficial observation.
it appears to me that with out knowing the the drawing requirements, in other words the detail of the material type, the manufacturing requirements, and any process that involves such as heat treating, chemical processing, the specification of the material. and the non destructive testing requirements. it is a guess. there should be the starting points. what are the allowable defects, what are the allowable porosity requirements. what are the non destructive and destructive test results. basic under standing of the casting requirements. next should be stress and strain analysis. make sure the load requirements are adequate.
I can not help to notice strange oxide over the failed areas. I would expect it to be a clean bright surface. I admit I am not a casting specialist but these are common sense approach ideas.

 

[weldstan]

.appears to be ductile overload failure. There appears to be cold deformation

 

[mrfailure]

The fracture morphology looks more to me like fracture began at the weld repair. The origin you identified could be a second origin but more likely it was an extension of the crack that started at the weld. Without knowing the base material of the casting, I want to ask if 309 is an acceptable material for such a repair. Does it have compatible thermal expansion/contraction properties with the casting material? Why did a repair have to be performed there in the first place? If a crack was present as the reason for the repair, are you sure it was completely removed at the time of repair?

 

[metengr]

309 stainless steel has been used for local weld repairs to castings and even wrought components over the years with the thinking that PWHT can be avoided because of the inherent ductility associated with the filler metal.

I do not agree with this approach because you have a locally hardened heat affected zone to contend with that can be susceptible to cracking during welding or in service.

 

[emonje]

Material is cast steel (Avg 0.2% C, 1% Mn, 1.8% Ni, 1% Cr, 0.4% Mo) with yield strength of minimum 700 MPa.

Repair weld was done by the foundry without our knowledge, we found out only after it failed.
Foundry is supposed to tell us before doing any kind of repair. We are using this foundry for the first time & they seem to have some serious issues.
We usually do not use stainless wire for our weld repairs.

On a side note we have asked a third party to do detailed failure analysis, they will be starting work from next week. I'll update with their finding as soon as I get them.

Any hint what these big steps in fracture face indicate?

Thanks.

 

[metengr]

The steps are called ridges and are formed from multiple fracture planes.

 

[kingnero]

309 stainless steel has been used for local weld repairs to castings and even wrought components over the years with the thinking that PWHT can be avoided because of the inherent ductility associated with the filler metal.

I do not agree with this approach because you have a locally hardened heat affected zone to contend with that can be susceptible to cracking during welding or in service.

I agree with both statements. 309 and 312 filler are often (mis)used for repairs. While strong and ductile, it still leaves traces in the base materials that cannot be discarded.

http://www.fusionpoint.be

http://be.linkedin.com/in/fusionpoint

 

[Tmoose]

Sounds like these are being made just for you.
Are you requesting/requiring test coupons for each pour that will be subjected to mechanical testing?
Regardless, maybe cutting some test coupons from the failed ( and a few un-failed) pieces and testing them is in order.

ASTM A488 / A488M - 16 Standard Practice for Steel Castings, Welding, Qualifications of Procedures and Personnel

ASTM A27 / A27M - 13(2016) Standard Specification for Steel Castings, Carbon, for General Application

ASTM A781 / A781M - 16 Standard Specification for Castings, Steel and Alloy, Common Requirements, for General Industrial Use

 

[WKTaylor]

emonje ... thoughts...

Based on the tool-bag in the background, this looks like a very large casting [several hundred #s], with very deep/thick sections.

Based on the chemical analysis this material looks mostly like carbon steel with modifiers for casting and heat-treatment [strength].

I am astonished that this thick-section material could be safely/consistently heat treated to yield strength of minimum 700 MPa [Fty 102-KSI min].

I suggest Your lab do a full mag particle inspection of each piece. I would NOT be surprised to see surface crazing/cracks all-around the land around the saddle where the fracture faces are located.

Regards, Wil Taylor

o Trust - But Verify!
o We believe to be true what we prefer to be true. [Unknown]
o For those who believe, no proof is required; for those who cannot believe, no proof is possible. [variation,Stuart Chase]
o Unfortunately, in science what You 'believe' is irrelevant. ["Orion", Homebuiltairplanes.com forum]

 

[emonje]

"Sounds like these are being made just for you.
Are you requesting/requiring test coupons for each pour that will be subjected to mechanical testing?
Regardless, maybe cutting some test coupons from the failed ( and a few un-failed) pieces and testing them is in order."

These castings are our own design, material is also of our in-house specs.
We do require the foundry to do mechanical tests on test coupons from each pour & their test results are showing things within specs..
We have also asked for thermal history & waiting for them.

Third party will do tests (including mechanical) on failed part starting from Monday.

Thanks.

 

[emonje]

"Based on the tool-bag in the background, this looks like a very large casting [several hundred #s], with very deep/thick sections. "

Yes it's a quite thick casting. I should have included some dimensions, my apologies.

The section that broke is 100 mm thick.

 

[WKTaylor]

emonje ...

Was the casting statically loaded, or unloaded, during this time?

IF this casting and alloy are proprietary to Your company, can we assume that they were fully/jointly 'characterized' [several castings statistically sampled] before use, for mechanical and physical properties, such as: X-Y-Z orientations for Ftu, Fty, Fcu, Fcy, Fsu Fbru, Fbry, e, E, KIc, Charpy impact, as cast/HT'ed, as-welded, within service temperature ranges [example for MIL acft: -65F to +160F]. Also, were the [dimensional] expansion/contraction rates characterized [X-Y-Z orientations] within the designated service temp range, etc?

Any idea RE lowest temperature that the failed casting reached 'during the holidays'?? Carbon steel fracture/impact toughness typically decreases rapidly with decreasing temperatures... especially with certain impurities and/or miss-processing and/or geometries and/or constraints.


Regards, Wil Taylor

o Trust - But Verify!
o We believe to be true what we prefer to be true. [Unknown]
o For those who believe, no proof is required; for those who cannot believe, no proof is possible. [variation,Stuart Chase]
o Unfortunately, in science what You 'believe' is irrelevant. ["Orion", Homebuiltairplanes.com forum]

 

[MJCronin]

Where was the casting made ??? Which foundry is responsible ?

MJCronin
Sr. Process Engineer

 

[gtaw]

Was the pin that passed through the two aligned holes bent after the failure?

The location of the failure would seemed to indicate the casting yielded, possibly bending the pin. Just curious.

It is difficult to tell for sure, but is it possible there were two welded repairs, the tack weld to the right in the photo and a weld to the left? The left edge of the lug looks similar to the tack weld on the right side.

The weld interface between the type 309 weld deposit and the casting would constitute what is called the mushy zone where there is partial melting and alloying between the filler metal and the original base metal. The chrome and nickel from the 309 filler metal would increase the hardness of the mushy zone due to the increase in the carbon equivalency. Even without an increase in Ce, the HAZ would form a metallurgical notch. Such a notch without an increase in hardness could still be a cause of crack initiation.

It would seem that it would be better if they had made the repair with a low hydrogen high strength low alloy electrode (8018-B2L) or a low hydrogen welding process using a high strength low alloy steel filler metal. That would mitigate the increase in Ce in the mushy zone and the associated higher hardness assuming the contractor used appropriate preheat before welding. Small repairs, without the benefit of preheat on a thick casting, could still experience high hardness if the weld cooled quickly. Did they repair the casting using 309 filler metal with the mistaken belief that preheat would not be required?

Best regards - Al

 

[emonje]

WKTaylor (Aeronautics)
6 Jan 17 17:51
"... Any idea RE lowest temperature that the failed casting reached 'during the holidays'?? Carbon steel fracture/impact toughness typically decreases rapidly with decreasing temperatures... especially with certain impurities and/or miss-processing and/or geometries and/or constraints. "

The part was in use when it failed, we don't know yet exactly which part of the cycle it failed. It gets loaded gradually then picked up & then the load dumped.

Operating temperature rarely goes below 20 degrees Centigrade, & always well above 0.

 

[emonje]

"MJCronin (Mechanical)
8 Jan 17 14:29
Where was the casting made ??? Which foundry is responsible"

This is the first time we got castings made by this foundry.

We never had any such issues in castings from our regular foundry. They were bit too busy so we decided to try this new one. On paper they looked good but now it seems we should stick to our proven supplier.

Unfortunately some legal mumbo jumbo stops me from disclosing names & locations of the foundries.

 

[emonje]

"gtaw (Structural)
8 Jan 17 16:26
Was the pin that passed through the two aligned holes bent after the failure?
.................
Small repairs, without the benefit of preheat on a thick casting, could still experience high hardness if the weld cooled quickly. Did they repair the casting using 309 filler metal with the mistaken belief that preheat would not be required?"

The pin was not bent. We haven't measured it but it looked as straight as when it went in.

The foundry that did the repairs are being very sketchy about what they did & how & why.
We require the suppliers to let us know of any required repairs & we usually use combination of pre-heat, low Hydrogen wire & GMAW to do such weld repairs. This new foundry apparently went rogue with procedures.

 

[RolMec]

From a similar event,
the outcome of the failure analysis was
Brittle fracture, due to
a) too coarse a microstructure (grain size being not to spec) --> manuf. error
b) de facto non-killed --> manuf. error
c) (hidden) welded spots (construction enterprise welded in some alignment bars and cut those off & ground away the stubs, however the metallurgical notches remained) were the starting points of the fracture.
The complex structure of the part will bring a multi-plane stress state, so adding to a metallurgical analysis a thorough check on load and load direction and perhaps a load / stress analysis might be adviseable.
Regards




Roland Heilmann
Lpz FRG

 

[RolMec]

and here the first pic (just reluctant to be uploaded)
:-( inserting into the post didn't work at all...

Roland Heilmann
Lpz FRG