Fatigue evaluation of stacked heat exchanger

[sanshu1111]

We have stacked heat exchanger designed with ASME Sec VIII Div 1. Exchanger experiences Thermal cycle with inlet temperature varying from 41 Deg C - 260 Deg C every hour. Fatigue evaluation for the exchanger was therefore requested as per ASME Sec VIII Div II.

Heat exchanger vendor is carring out FEA analysis to carry out fatigue evaluation only for the interconneting nozzle of the stacked heat exchanger.

When requested to carryout fatigue evaluation for all the elements of the exchanger, vendor is reluctant to do so justifying their stand that interconnection section is most stressed part and also their AI has accepted the same. Technically it may be correct but unless all the elements are checked, it is difficult to say which element will govern for fatigue life.

Request your opinion on the same.

 

[TGS4]

I've gone to these types of discussions with the "show me the math" mentality. Instead of some hand-waving arguments, have them "show me the math" of what they claim.

 

[sanshu1111]

TGS4,

Thanks for your reply. We have been doing the same thing but vendor stand is that code requirements are complied with as AI has already accepted this methodology to evaluate only interconnecting nozzle.
Probably they do not want to run the calculation and submit the same for AI appraisal again.
Can we say that code requirements are compplied in such cases?

 

[TGS4]

In my opinion, no.

Does my opinion count, in this case - no.

Talk to the AI.

 

[jamesl]

In my opinion, if this is a Division 1 vessel, this issue is more a contract issue than a code issue. I understand UG-22 requires evaluation of cyclic loading. However, unlike Div.2 where fatigue analysis is clearly defined in UDS, Div.1 does not clearly spell out when fatigue is required. It is a gray area.

If the buyer wants the fatigue analysis, he needs to clearly define it in the contract and pay for the analysis.
If you paid the vendor to do the analysis of the complete equipment, I don't see why the vendor would refuse to do it.

 

[BPVFEA]

sanshu1111,
It is not always true that nozzle is most stressed part in Heat Exchanger. The reason behind vendor being reluctant is the complexity and time involved in the fatigue analysis. I feel that, for this type of analysis it would take around 15 days for the Analysis+Report. So just find out how much it would cost based on hourly charges at your location.
Vendor has to agree for the analysis if you have paid/paying as per contract.
If Fatigue Analysis is not specified in contract then this is a complex issue. I have seen that such matters are solved with the enginering judgment and negotiations.

If the matter is not getting resolved then

1) Ask the vendor to find out fatigue life of 2-3 more stressed locations (Flange corner, Saddle Support etc.). An expert in Design & Stress Analysis can make out which would the most stressed zones in the exchanger (based on your current configuration).

2) You can suggest him to simplify the analysis (Make in parts). If it us U tube exchanger, then his job is simpler.

Please note that AI should be informed & agreed on this.

 

[sanshu1111]

Requirement of thermal cycle was very clearly specified in the equipment datasheet which was a part of purchase order. Now it looks like that only due to cost vendor is reluctant to carry out fatigue calculation.

Unfortunately AI has already approved the calculation without checking the requirements and giving vendor a chance to argue that AI has already accepted and therefore it meets code requirements.

We are persuing vendor to carryout the complete fatigue evaluation and hopefully we will be able to get what we have ordered.

 

[jamesl]

Specifying thermal cycles and requiring fatigue analysis are not the same thing. A lot of Div.1 vessels in cyclic operation are not designed with fatigue analysis. When we see process data sheet saying the vessel is in cyclic operation, we don't automatically require fatigue analysis in our purchase order. For example, if you have used this type of H.E. many times before, and never had probelms with the cyclic operation, then there is no need to do the analysis.

This issue should be clarified in bid evaluation stage. In my opinion, if the vendor did not include the cost of full fatigue evalution in his bid, and you did not make it clear that full fatigue analysis was required, then it is unfair to ask the vendor to do it for free.

 

[InspectorJim]

The AI's job was the vessel meet the requirements for ASME VIII-1 for which it was being stamped. If the purchaser requests additional design data, that is between the vendor and the purchaser, and not in the realm of the AI. The AI does not typically see purchase order from purchaser to vendor, he will see drawings and supporting calculations for review to meet the Code for which the item is being stamped (ASME VIII Div. 1).

 

[BPVFEA]

The Loadings Clause of UG-22(e) mentions that "cyclic and dynamic reactions due to pressure or thermal
variations, or from equipment mounted on a vessel,
and mechanical loadings" shall be considered while designing the vessel.
Which means (by default) AI needs to check the design of this exchanger under cyclic loads also.
However, if vendors Design Document mentions that "UG-22(e) is not applicable", then it is not in the scope of AI to check this.

 

[InspectorByDay]

The issue is contractual, not a Code issue.

I do believe that the HX meets the requirements of VIII-1 with the analysis on the one component. You specified Section VIII-2 be used for fatigue analysis and my guess is that the Manufacturer is claiming exception based on Paragraphs 5.5.2 of VIII-2.

a) "The provisions of this paragraph can be used to determine if a fatigue analysis is required as part of the
vessel design. The screening options to determine the need for fatigue analysis are described below. If
any one of the screening options is satisfied, then a fatigue analysis is not required as part of the vessel
design.

1) Provisions of paragraph 5.5.2.2, Experience with comparable equipment operating under similar
conditions.

2) Provisions of paragraph 5.5.2.3, Method A based on the materials of construction (limited
applicability), construction details, loading histogram, and smooth bar fatigue curve data.

3) Provisions of paragraph 5.5.2.4, Method B based on the materials of construction (unlimited
applicability), construction details, loading histogram, and smooth bar fatigue curve data.

b) The fatigue exemption in accordance with this paragraph is performed on a component or part basis.

One component (integral) may be exempt, while another component (non-integral) is not exempt. If any
one component is not exempt, then a fatigue evaluation shall be performed for that component."

My best guess is that they are using past knowledge per Paragraph 5.5.2(a)(1) to evaluate a single part and ignore the rest per Paragraph 5.5.2(b).

Obviously, you disagree that past knowledge can prompt an exemption. But, if you requested fatigue analysis per Section VIII-2, I believe Manufacturer met your request. This is where the dispute falls outside of Code and becomes contractual.

This is the reason for the recent addition, though non-mandatory, of the User Design Spec in VIII-1. It sounds like the User was unclear of their required analysis, while the Manufacturer did not ask enough questions.

Remember, the individual providing the quote is a salesperson. From my experience, salespeople know just enough about Code to get the company into trouble. They cannot be relied upon to have taken design issues into full consideration.

This is why the User really needs to be specific when RFQs are issued.