ASME. Design by analysis. Stress linearization 2

[-RuDi-]

Hello everyone!

I need a help from someone, who is advanced in ASME codes or others. The problem is pipeline system with internal pressure and temperature distribution (hot fluid inside the pipeline). I calculated temperature field and applied it in my finite-element structural model (3D solid elements, linear elastic material models). The model includes flange connections with applied bolt preloads. The analysis converged, so I got the resulting stress distributions. I am familiar with the aspects of stress linearization procedure and understand the operations with stress tensors. But I don't fully understand, what criteria to apply in this case when I have primary (pressure) and secondary (temperature) loads acting simultaneously. How to separate Pm, Pb and Q stresses from the resulting fields? How to assess the strength of bolts in this case? Uneven temperature distribution has a great influence on the bolts, because they remain cold while pipeline heats up and expands. I'm not sure that this stresses in bolts are secondary in this case.
Will it be correct to calculate Pm and Pb by additional FE analysis where all coefficients of thermal expansion are equal to zero?
Any ideas?

 

[saplanti]

I am not expert on FEA, but if the bolts are critical why don't you change bolting system to compansate the thermal growth on the flanges. You may use either spring washers or use special bolts can expand under load or other applications available in the industry.

Sometimes you need to solve things using fundamentals. Additionally you need to consider some safety margin, numbers do not reflect the reality all the time.

 

[IdanPV]

I think you will find answers for your questions in ASME VIII-2, 5.2.1.2 which state that:
"For components with a complex geometry and/or complex loading, the categorization of stresses requires significant knowledge and judgment. This is especially true for three-dimensional stress fields. Application of the limit load or elastic–plastic analysis methods in 5.2.3 and 5.2.4, respectively, is recommended for cases where the categorization process may produce ambiguous results."

Consider the use of Elastic-Plastic analysis instead.

 

[-RuDi-]

Saplanti, IdanPV, thanks for your replies!

I prefer to avoid elastic-plastic analysis when there are bolt connections in the model. I might be wrong, but I think that in this case forces and moments are underestimated. This can lead to non-conservative assessment of bolts. Unfortunately, this is the only way as I see it.

Saplanti, what do you mean by bolts that can expand under load?

 

[saplanti]

I mean that you can design and/or select the bolt length which can strech under the tensile load within an allowable tension. The extension side of the bolts can be filled by adequate tubes.

 

[TGS4]

If you are following the procedure in ASME Section VIII, Division 2, Part 5, then you will have noticed 5.2.2.4, Step 1, which states:

Step 1. Determine the types of loads acting on the component. In general, separate load cases are analyzed to evaluate “load-controlled” loads such as pressure and externally applied reactions due to weight effects and “strain-controlled” loads resulting from thermal gradients and imposed displacements. The loads to be considered in the design shall include, but not be limited to, those given in Table 5.1. The load combinations that shall be considered for each loading condition shall include, but not be limited to those given in Table 5.3.

Effectively, you need separate load cases for the load-controlled loads and the strain-controlled loads. You cannot separate them in-situ, as it were.

 

[saplanti]

RuDı,

You additionally should be in contact with your process engineer, they migh be considering the temperature action and correcting the system accordingly.

Most of the cases, they might prefer the welded connection instead. If the connected equipment is a mobile equipment the process engineer might come up with a cheaper and effective solution by keeping the flanged connection without going deeper calculations and complex solutions.

 

[KevinNZ]

TGS4
If you regard loads from thermal expansion from attached piping in the same case as "load-controlled” loads, does this give a conservative design?

I ask this because vessels are over specified with a single allowable load and do not differentiate between load types.

 

[TGS4]

KevinNZ - yes. Unless otherwise specified, loads from attached piping, even if from restrained free thermal expansion, should be considered as load-controlled loads, and the resulting membrane stresses in the shell should be classified as primary.

The reality is that they are neither fully primary nor secondary, but classifying the stresses as fully secondary may be unconseravtive, and therefore given the binary nature of the choice, the only reasonable classification is thus: primary.