Is it correct to use Ambient Temp as Lowest Temp in Sps ?

[BPVFEA]

As per Cl. 5.5.6.1.d of ASME Sec. VIII Div 2,
"The allowable limit on the primary plus secondary stress range, Sps, is computed as the larger of the quantities shown below.

1) Three times the average of the S values for the material at the HIGHEST & LOWEST temperatures during the operational cycle.

2) Two times the average of the Sy values for the material at the HIGHEST & LOWEST temperatures during the operational cycle."

My query is,
In above allowable stress calculation, is it correct to take AMBIENT temperature as LOWEST temperature, if lowest operating temperature is not defined?
Or Can I take Design Temperature as LOWEST as well as HIGHEST temperature in such case.

Thanks in advance

 

[DSB123]

BPVFEA,
Think about it!! The lowest temperature the item will see is not the Design temperature (unless it is sub zero/cryogenic) is it? The lowest temperature , if not specified, is going to be the lowest ambient temperature to which the item is subject taking into account any insulation, if its insulated. I personally would use the average lowest daily temperature as the lowest temperature unless the client has specified otherwise.

 

[TGS4]

It all depends on the CYCLE.

If you have an operational cycle where the temperature varies from 400°F to 500°F while the pressure also varies, then Sps would be calculated as the average of the values at 400°F and 500°F.

If your cycle is from ambient to operating temperature, then follow DSB123's advice on using the average lowest daily temperature.

First activity, though, is to define your cycle. Without that, you're lost.

 

[rneill]

If the cycle is a cold start up I would base my lowest temperature on ambient unless insulated and heat traced.

In Canada I use the January 1% design temperature specified in Table C-2 of Vol 2 of the National Building Code of Canada. In the US I use the Minimum Temperature Isotherms published in Figure 4-2 of API 650 (which are based on 1 day lowest mean temperatures).

Some people would relax these requirements for insulated lines without heat tracing due to residual heat.

For insulated and heat traced lines, I would use the minimum temperature the tracing was designed to maintain.

There is a good discussion of selection of design minimum metal temperatures in the book "Practical Guide to ASME B31.3 Process Piping" published by CASTI.

 

[BPVFEA]

Thanks DSB123 & TGS4
1) I am not able get the definition of "Operational Cycle" from code or any other source.
How Operational cycle can be defined in terms of Pressure and Temperature in the vessel?
2) If definition of "Operational Cycle" is independent of P or T, then the lowest temperature for all vessels would be always the ambient temperature (unless sub-zero). Is it correct? . In that case the Sps should be always calculated based on average of Highest T & Ambient T(unless sub-zero).

 

[TGS4]

BPVFEA - please through 5.5.6.1 of VIII-2 again. Also, take a read through Annex 5.B.

Operational cycles are well-defined in VIII-2. The UDS should clearly define the cycles - as required by 2.2.2.1.d, e, f and h. If the UDS does not, then kick it back to the engineer who wrote it to get them to properly define it.

 

[BPVFEA]

Thanks TGS4,
Actually I should put my query properly, which I am doing now
I have made WRC-107 calculations of one nozzle using one Pressure Vessel Design Software. The software is comparing the calculated Pl+Pb+Q stress with the allowable stress which is based on average at Design temperature & Ambient temperature.
I am of the opinion that the concept of "Average S at Design temperature & Ambient temperature" is applicable only when we carry out Thermo-structural analysis of Nozzle & Shell at two different extreme states of 'Operating Cycle'. Here the stresses due to following loads are considered
(A) External Mechanical loads (Weight + piping thermal expansion loads etc.), &
(B) Differential thermal expansion of nozzle & shell
When we consider only Type-A loads, then we should take allowable based on Design Temperature only (and not the average temperature). As WRC does not include effect of Type-B loads, allowable should be based on Design Temperature. However the above software is calculates S based on average at Design temperature & Ambient temperature.
Since the Cl. 5.5.6.1(d) of Div-2 is applicable to Ratcheting, it has not explained about allowable in Only Type-A situation.
However, when I browsed other codes, (Note-1 Fig. NE-3221-2 of Sec. III, Div-1 Also Note-1- Figure A1,PD-5500) I found following description,
Type-A situation
When part or all of the secondary stress is due to mechanical load, the value of S shall not exceed the value for the highest temperature of metal during the transient.

Type-B & A+B situation
When the secondary stress is due to a temperature transient at the point at which the stresses are being analyzed, the value of S shall be taken as the average of the tabulated S values for the highest and the lowest temperatures of the metal during the transient.

Therefore my understanding is, for stresses evaluated from WRC, the allowable should be evaluated based only on Design Temperature. (And not the average S as per software).
Am I correct? Please comment.

 

[TGS4]

OK - you're starting to focus in on the real issue. Again, I would encourage you to focus on the failure mode first and foremost, and the rules secondarily. You are completely correct that the reason for the Sps limit is to deal with the ratcheting failure mode. So, I propose that you focus on the failure mode, and what causes it, for getting tangled up in applying a mess of rules.

Ratcheting may occur when a component, under load, experiences stresses which, when calculated on an elastic basis, exceed a range of 2 x yield. When the loading occurs over a range of temperatures, a variety of approximations can be used to simulate what really occurs. Imagine a displacement-controlled uni-axial tensile test. In this test, the tension occurs at a hot temperature, but the relaxation or compression occurs at a low temperature. If the stress range were only slightly greater than 2*hot yield, would the stress-strain curve of an elastic-perfectly-plastic material march off to the right (indicating ratcheting)? I would say no, because at the low temperature, the yield stress is much higher, and during cooling, the material action would be only elastic, with no plasticity.

Of course, if your "mechanical" load cycles at the hot temperature, the, well, your cycle would only be at the hot temperature.

As I said before, define your cycle. I don't care whether you're using WRC-107 or the articles in Part itself. You have a cycle (which you need to define), you have a stress range, and the rules are very clear. Of course, if you are going to "fail" a load based on the ratcheting failure mode, I would recommend that your client ask for a second opinion. Ratcheting rarely occurs, and in the vast majority of cases where an analysis has failed 5.5.6.2, it has been subsequently shown to be OK when evaluated to 5.5.7. Of course, you could always try out 5.5.6.2 and/or 5.5.6.3 too.