Low Temp. Brittle Failure of Carbon P.V. steels 5

[nzoldun]

Can anyone point me to reference works, papers, et.al, on the exact mechanism of failure when a relatively thick vessel wall is below the transistion temperature and then subject to a (large) thermal shock. How does the fracture start, (existing micro-crack?) and where does the energy come from to propogate it? (internal pressure?, extreme thermal through-wall gradients?).

 

[davefitz]

The calculation procedure would use fracture mechanics.
see <

http://www.kuleuven.ac.de/bwk/materials/Education/master/wg12/t1000.htm>

Other lectures in this series describe the new European approach to fracture and fatigue damage, and they now focus on the failure of welds as being the principal failure location. All welds contain microcracks, plus the weld zone has dissimilar material properties ( fracture toughness, modulus of elasticity, etc)compared to the parent material which tends to cause strain or stress concentrations even when there are no geometrical stress raisers.

The source of energy for opening the crack in the scenario you listed include the following:
a) thermal stress from the shock- considered to be self limiting in ductile material , but brittle material may disqualify this assumption
b)residual stress from the fabrication process that was not relieved during post weld heat treatment ( roughly equal in magnitude to yield stress at the heat treatment temperature)
c)structural stress from gravity loads and piping loads at nozzles.

ASME requires the pressure vessel to be hydrotested at a min temp of 70F to avoid catastrophic brittle failure. One notable case that had occurred in 1967 in Scotland ( see the photo of the failed steam drum from a 1967 hydro test in Scotland , as in Harvey's &quot;pressure vessel design&quot;.

 

[TVP]

The link provided by davefitz had a mistake (KU Leuven is in Belgium not Germany). The correct link is provided below:

http://www.kuleuven.ac.be/bwk/materials/Education/master/wg12/t1000.htm

 

[JOM]

Here's a case study of a cold brittle failure from the Belgian Welding Institute.

http://mahbsrv.jrc.it/downloads-pdf/forum.pdf

I'm sure this would be of interest to you. There is another recent case of failure at the Longford gas plant, Australia. That failure was intensely studied. Cheers,
John.

 

[TVP]

Excellent reference, John!

 

[JOM]

TVP,

That's a nice investigation report. It was published very soon after the incident. I was surprised that the brittle fracture occurred at 6 degrees C. That doesn't seem very cold to me.

The Longford incident involved a large reboiler (shell & tube heat exchanger) that dropped in temp to minus 40 or lower. Then hot stream was introduced and the thermal shock caused a crack to propogate from a small weld defect.

As far as i know there has not been any published report from the metallurgists who investigated, but the accident investigation report does cover the fracture mechanics in detail. That is available from Info Victoria at Au$77 :-

http://www.bookshop.vic.gov.au/infovic/anonymous/products.asp?ProductID=24419

Cheers,
John.

 

[Metalguy]

Residual welding stresses can fracture a vessel even when it's unpressurized and at constant temp.

I had the experience of being within about 50' of a large thick-walled vessel when a 6&quot; nozzle weld cracked about 1/2 way around. The crack didn't open up more than ~1/32&quot;, but it made quite a bang and a big rust cloud formed above the area.