Sigma Adm. with increasing temperature 1

[starflex]

Hi there,

I have a material selection problem, and I am trying to analyze it under a temperature / creep point of view.

In a valve internals, disc and support are made of carbon steel (Gr 70 and WCB 216), and the bolting to keep them together around the shaft is made of Austenitic SS 316.

The material selection for the bolting looks a bit odd to me (the temperature does not exceen 320 degrees), so I found unusual the selection of SS for the bolting. And, besides that, the bolting failed over the time.

My idea would be also to replace the bolt with something more suitable (perhaps I am wrong), e.g. SA 193 B7 or SA 193 B16, but I would like to support my idea with some data about the material yield at different temperature.

Now, in my very limited knowledge, I am not sure if at 300 Celsius I can use the Yield or I should use a completely different approach, like maximum stress for expected life of 100,000 hours (basically, on creep range). What I am after is some data, for the three materials (B7; 316/b8; B16) that shows the material properties in the range, let's say, 270 / 330 Celsius, so that I can make an evaluation also on that.

Is there something available / which sourse would you recommend ?

Thanks,

Paolo

 

[metengr]

First, how did the stainless steel bolting fail? Overload, fatigue????
Second, using Grade B7 or B16 bolting material seems appropriate and information on service temperature limits and allowable stress values for time dependent properties can be found in ASME B&PV Code, Section II, part D for these materials.

 

[starflex]

Having seen the fracture surface, it looks like fracture more than fatigue. The screw is a socket head hex cap screw, and the fracture is right under the head.
Picture is not 100% clear, I admit. Probably, overload more than fatigue (other reasons in that, maybe too much torque from the actuator).


Thank you for pinpointing to the ASME code!

 

[starflex]

Here again.

Apologies, file was missing from my previous message.

http://files.engineering.com/getfil...2afd7c16&file=Set_Screw_Fracture_Face_1_R.jpg

The Client has reported that another valve (smaller in diameter and with smaller bolts, but same material combination and apparently same fracture surface, although the heads have been rambling for a while in the line and the surfaces are almost polished now...).

He is questioning about stress relaxation, as he remember that 316 is subject to that, although he does not remember the source. We are at 300 Celsius and, in my view, we are well below creep range.

On top of that, I remember using the B8M material (if I am not wrong, it is 316) on valve internals working at much higher temperatures and never had such a failure.

Hints, ideas?

 

[metengr]

The attached picture is blurred and does not provide anything meaningful interpretation. Your hunch is correct that at 300 deg C you are well below creep deformation (time dependent properties). I would guess the bolt failure was caused by either mechanical fatigue from loss of preload (torque) at 300 deg C or possible corrosion-assisted fatigue depending on service environment.

Keep in mind you have a significant difference in the linear coefficient of expansion between carbon steel and 316 ss. This could easily result in the loss of preload in service because the 316 ss would expand at a greater value. Use the B7 or B16 stud material if service conditions dictate.

 

[starflex]

Metengr,
Thank you for tour reply. Glad to see that we come out to the same conclusion.

Totally agree with the different thermal expansion: it's roughly 40% higher and I already pointed put with the client. The media flowing is superheated steam .

I am aware that the picture is really blurry... This is unfortunately what I have been provided , and it doesn't tell too much. On top of that, this recurrent failure (at least four valves of different sizes have had the same issue ) is not discovered immediately, so the heads are around the line, rambling and getting a fracture surface that is somewhat polished anyway. I would guess that a proper close up of the bolt stem would help more, and is already asked for that.

As additional information: the bolts are quite small (14 mm diameter) and short (30mm). Doing a bit of math, and hypothesising a temperature excursion from 300 Celsius when the line is operative, down to ... Probably around 50 / 80 celsius when they shut the line (but the steam generator is still operative ) we are talking about 220 / 250 degrees excursion.

Considering 4 µm/m-K of differential thermal expansion... Would hat be enough to justify loss of preload ?

 

[metengr]

Yes, it is very possible.