Help - Sustained stress failed

Could any one help me out with this high temperature stress problem?
Compressed air system with design T=1600F P=131Psig, 800HT material with Dia 12"x1.2"(calculated wall thickness per 31.3 (304.1.2 3b)). Sh=1400 psi and weld joint strength reduction factor is 0.5 so allowable at elbow is 700 psi.
The pipe route is rathar simple e.g from heater towards south with flange and elbow, turning east with or without straight pipe to another elbow, then down 8' and turning west with 7' or more and then to the nozzle of turbine towards north with elbow and flange. I modeled this system, but could not attach to the post. Pipe supports could be anywhere, nozzle displacement can be igored.
How could I make code stress passed in this case.
Thanks a lot,
Flames

Hi Flames

If there is any sustain stress failure in caesar check the following points.

1) Fluid density with proper unit.
2) Insulation density with proper unit.
3) Input pressure with unit
4) support span
5) Input pipe thickness
6) Sh value of material

Flames,

The temperature of 1600F is extremely hot, and it shows in the Sh=1400 psi, and in the weld joint strength reduction factor 0.5 that leaves only Sa = 700 psi at the elbows. There is hardly enough strength for pressure! The pipe could require continuous support along its length to carry weight.

You will need to either select a material with better strength for 1600F, or see if the 1600F could be reduced or limited to occasional excursions, or develope a different allowable strength for the material based on something different than the ASME BPV or B31 Piping Codes (reduced design margins). The temperature of 1600F is at the limits of materials in the B31 piping codes. This is entering a different arena.

Hot ductwork is usually designed with internal refractory insulation to keep the metal temperatures more reasonable and retain metal strength. Is this an aerospace or combustion turbine exhaust application? There are measures taken inside turbine to cool the combustor, to insulate shaft, and to provide flow to cool blade sections in the hot gas stream.

Thanks for all your comments, it is for an application of Energy plant with the code of B31.3. I have a lot of problems with this line - sustain case stress and nozzle loads.

One note on that weld joint strength reduction factor.
This item was added to the 2006 Edition of B31.3. Both longitudinal welds and circumferential welds (and others, e.g. spiral) were explicitly addressed. The 2008 Edition removed the direct reference to circumferential welds. Perhaps you could justify removing the W for circumferential welds.
What does C2 do? When you enter an elbow, we assume there is a circumferential weld at either end of that elbow so we automatically apply W.

If these are pipe bends (with no weld at the C2 "ends") or if you wish to ignore those circumferential welds, you might want to do this:
In Config - allow User's SIF at Bend
In model - check SIFs & Tees for the bend node
- enter the bend node number and set Wc=1

I am using Caesa II 5.00.
Do you mean that 2008 edition has also revised wall thickness formula which would be thinner?
If seamless pipe is used, there is no longitudinal welds thus no weld joint strength reduction factor at all.
Thanks,
Flames

No, the wall thickness for pressure equation still shows that W. Seamless pipe has W=1.
In C2 terms - this is Wl (W - longitudinal).
We also have a Wc (W - circulferential). This is our term that has changed in B31.3 2008. Wc affects the allowable sustained stress limit (as in W*Sh). The 2008 Edition now states that its up to the engineer to decide whhat to do with circumferential welds. We're ading a switch in V5.20 that allows the user to ignore Wc.

Flames,

One other consideration - if you have springs in this system, you can be getting an erroneous sustained stress result. In a typical arrangement with a pipe growing/lifting/moving up means that the cold load of the springs can be much higher than the operating load. Usually, this isn't a big deal. However, in your situation, Caesar wants to check the sustained stress case (against your 700 psi allowable OUCH!) using the higher cold spring loads can show your system being overstressed due to the springs. I've not had to personally deal with these situations, but others in my group have. There are coding techniques (Dave Diehl can probably refer to the article to discussion post about it) that you can use to deal with this situation.

Flames,

I may explain to you the solution properly if you can send me quickly before 5:00 pm today Central time a simple sketch showing your support location. Send it to GeraldGamboa@gmail.com. (what is 800HT material?)

Gerald

Flames,

The spring hanger sizing calculation mentioned by Ed Klein is one item to use for improving the SUS margin that you need. (Spring sized for SUS case)

Another item would be to review the mill tolerance and corrosion allowance used for the wall thickness. Where are you getting the 1.2" wall pipe manufactured? You could specify the wall to be made with a closer minimum tolerance, and not lose so much wall thickness by the usual 12.5% mill tolerance (1/8"+ is lost). If the operational life of the installation is for short time, then a thinner corrosiion allowance might be acceptable.

I am thinking that a larger diameter, say Nps-20, with a 4" thick internal insulating brick and maybe thin lining of 800HT to retain the brick and mortar chips, will let you use a stainless steel shell (or even carbon steel for less expansion problems) and a reasonable wall thickness at the lower temperature of the (externally un-insulated) shell.

Sorry to miss your post these days, I was not able to access the forum recently.
Thanks,
Flames