Low Cycle Design

[StressGuy]

I know that allowables in B31.3 are based on a maximum cycle life of 7000 and you apply the reduction factor for designs in excess of this.

I also know that the code doesn't allow you to take credit for a higher allowable for less than 7000 cycles. But, from a practical standpoint, is there much that can be gained in terms of a design stress by considering a lower cycle life?

I'm faced with a unique problem in that adding enough flexibility to accomodate the stresses based on 7000 cycles is creating pressure drop issues with the amount of tubing required to handle the displacement. The client is asking if a design life of 500 cycles, that they could guarantee, would help in reducing the length required. They are aware that this is going to be stepping outside the code.

So, I'm looking for a couple of things. First - is there any point to pursuing this?

And, if so, can someone share some references that I can look into to tackle this issue.

Thanks,


Edward L. Klein
Pipe Stress Engineer
Houston, Texas

All opinions expressed here are my own and not my company's.

 

[davefitz]

As I see it , there are 3 issues:

a)code requirements and related "buraucracy" as opposed to technical requirements

b)Fatigue as treated in section Viii div 2 or section III is based on a classical model of fatigue from unnotched test specimens. Latest fatigue theories and practical experience shows nearly all failures are in weld zones,, and classical model is not the most accurate method of predicting actual failures. Newer codes ( EU PED codes, for example EN-13445-3 part 3 , or BS 5500)directly address fatigue failures in weldments )

b) As Clint Eastwood said it, " how lucky do you feel today?" If you do not meet the specific bureaucratic requirements of the ASME code, you are putting yourself in a vulnerable position if the part does fail. It is very easy to convince yourself that you can predict all likely future events related to the pressure part, but as a practical matter, "everybody plays the fool, sometime". There are significant low cycle fatigue events associated with the intial fabrication and erection processes which are ignored as part of the "shakedown stresses" and assumed to be covered by other factors. Also, equipment operators are notorious for not beina able to read the mind of the designer/engineer and tend to operate the equipment in ways not imagined by the designer.

 

[chungdy]

Use of maximum 7000 stress cycles for secondary stress is a simplified way of checking fatigue failure. I think you can solve the problem by performing detailed fatigue analysis using cumulative damage concept described in ASME Section III, Subsection NB-3222.4, if you can afford economically.

 

[hacksaw]

B31 is based in part on experience. I takes a wealthy customer to design piping an a specific number of cycles. Basically they will be embarking on a design by experience escapade. Good luck!

Several issues with process piping:

1. failure is rarely deterministic. you never know the actual state of the material before stressing.

2. B31 is really an estimate based on one heatup/cool down cylce each day. 7000 cycles- is 20 years more or less. It is a nominal but safe starting minimum.

3. the B31 criterial for fatigue failure is at best an allowance based on testing and experience. It is only able to set the minimum metal thickness amounts. You are free to increase the metal thickness subject to the design stresses.

4. you do not have to comply with B31; however, you are tossing aside a lot of crediable experience when you do.

(Prior to adoption of the boiler code some 5-10 % of the US population was killed or injured as a result of boiler/poping failures ... each year. It is the best design Code we have.)

5. In adjusting the safety factor (i.e. ignoring reasonable derating according to cyclical stresses) what you are doing is moving from one regime of failure (leak before rupture) to catistrophic failure.

6. If you know the streeses exactly, and the material is perfect, you can take all sorts of liberties if you are dealing with structures. With piping, all bets are off simply because of the normal variation in piping quality, field fabrication, pipe fittings, supports, and unknowns about the process.