Where:

• \( t_m \): required minimum thickness
• \( D_o \): Outside diameter
• \( P \): Design pressure
• \( SE \): Maximum allowable stress, defined per material and temperature
• \( A \): Additional compensatory thickness
• \( y \): Factor depending on material and temperature
• \( W \): Weld strength reduction factor based on temperature
• \( S_y \): Material SMYS
• \( E \): Joint factor
• \( F_{B31.4} \): Design factor, max 0.72
• \( F_{B31.8} \): Design factor, max 0.8
• \( T \): Temperature reduction factor
• \( C \): Quality factor based on the material used
• \( M_f \): Material performance factor considering loss of properties due to hydrogen exposure
• \( Y \): Material and temperature-dependent factor

It can be observed that the required thickness calculation in all codes is based on Barlow’s formula, which evaluates the circumferential stress in pipes or cylindrical vessels, solved for wall thickness.

\( t_m = \frac{PD_o}{2S_y} \)

The other factors involved cover, for example, the following aspects:

• Differences in weld strength compared to the base metal, considered in factors like E or W.
• Loss of material properties due to temperature,
• Allowable risk per application, directly considered in factors such as \( F_{B31.4} \) and \( F_{B31.8} \) or in allowable stresses like SE, y, or T
• Material response to specific damage mechanisms, such as those related to hydrogen presence, considered in factors like C or \( M_f \).

This method, based on Barlow’s equation, cannot be applied to the design of fittings that do not have rotational symmetry, as they are subject to more complex stress states. The general criterion for the selection and specification of fittings is that their rupture resistance under internal pressure testing must be at least equal to that of the associated pipe/duct.

The manufacturer must ensure and certify this capability, a process that by code may include analytical calculations and/or hydrostatic tests or FEA modeling.

All these determinations must be based on the mechanical properties of the material used, which must also be chemically and mechanically similar to that of the pipe. The stresses generated in fittings are higher and more complex than those in straight pipe sections, so fitting suitability is not ensured by specifying only the thickness and mechanical properties equal to the associated straight section.

To meet design requirements, it often happens that some parts of the fittings must have greater wall thickness than the associated pipe [B 16.9, Paragraph 2.2.2]. These thickness variations must be considered in strength calculations and must be properly documented and available for purchaser verification, as well as the records of hydrostatic tests, if they were used to determine the fitting rating.

The pressure test

Prior to commissioning, the facilities must mandatorily undergo a pressure test (PT) which is carried out at pressures higher than the design pressure. The excess over the design pressure varies according to the code, generally limited to not exceeding 90% of the Specified Minimum Yield Strength (SMYS) of the installed materials.

Fittings, on the other hand, which exhibit high stress concentrators, may experience secondary stresses higher than the material’s yield strength. This does not imply a failure of the component, as the affected areas represent a small volume of material relative to the whole, and both strain hardening and stress redistribution limit the risk of failure during the PT.

It is clear that localized yielding, no matter how minor, will result in permanent deformations that do not affect functionality and are not noticeable and, in turn, produce a redistribution of stress that enhances the structure for its subsequent use.

Fittings are generally manufactured from hot-formed or welded pipes, and the material undergoes significant deformation to achieve the desired shape. To attain the desired mechanical properties, they are subjected to cold dimensional recalibration and subsequent quenching and tempering heat treatments.

When the necessary thicknesses are not achieved during the manufacturing process, or when the heat treatment does not provide the expected mechanical strength, the risk of failure during the PT arises. In fact, we have recently had the opportunity to witness firsthand cases where the failure rate during the PT increased due to these reasons, as well as due to deficiencies in the design or requisition process.

A failure during a PT does not pose excessive risk given the safety considerations taken into account during the procedure. The main issue lies in the need to replace the failed or affected components, perform new inspections, and repeat the PT. The greatest uncertainty lies in the reliability of the fittings that did not fail.

This type of failure causes economic losses, not only related to the replacement of fittings but also due to delays in acquiring new components, which result in missed deadlines. These events also tend to generate a negative impact on the public image of the companies involved.

Recent Causes of Failures in Pressure Tests

Analyzing the root cause of failures, if they occur during testing, is essential to achieve pre-operational certification. In this regard, in the studies we have conducted over the past year on all failure events related to fittings, mechanical properties lower than those stated in the quality certificates issued by the suppliers themselves were verified. Significant geometrical deficiencies were also evident, mainly in thicknesses.

Likewise, in studies we have conducted on installed fittings that did not show damage during pressure tests, and on fittings available in stock from different batches and suppliers, it was found that the vast majority also did not meet the certified properties.

Additionally, a significant variation was detected in the properties and thicknesses within the same component. The tests were complemented with hardness mappings and metallographic replicas, revealing issues related to heat treatment and forming.

The measurement of the mechanical properties of the fittings, both installed and in stock, was made possible using the ESYS 10 instrumented progressive indentation technique for PMI (GIE), see Figure 1. This technique allows for obtaining the yield and tensile strength of materials without destroying them, which enabled the identification of deviations, leading to quality rejections prior to replacements and hydrostatic tests.

Figure 1: Quality control on fittings acquired for replacement, prior to installation.

The results of all the tests were compared with the values from standardized tensile tests performed on the failed fittings. The results showed excellent correlation between the tensile tests and the ESYS 10 (Figure 2).

Figure 2: Difference between ESYS 10 results compared with the tensile test results on failed components

Global supply and local challenges

The possibility that supplied products do not meet the purchase requisitions is not a new problem. However, a significant increase in the frequency of this issue has been observed in recent years.

Historically, companies worked with local or geographically closer suppliers, which kept commercial relationships contained, and suppliers were well-known actors. Today, the rise of global competition has allowed for cost reductions, but has led to increased risks of poor quality. It is no coincidence that, in all the aforementioned cases, the suppliers were new players from foreign markets with little global track record.

The increased risk of receiving a fitting that does not meet the requirements has grown significantly and forces companies to take measures to mitigate this risk. These measures may include stricter document control, better purchase specifications, and tests that ensure the veracity of the supplied documents.

As shown in Figure 3 (properties of fittings from 4 suppliers compared to the minimum requirements of the standards), the implementation of tests with ESYS 10 for PMI, as a quality control milestone, helps reduce time and avoid installation costs of components that do not meet purchase requirements—something that has become quite common.

Figure 3: Actual properties in relation to the requirements of the corresponding standards. σy: yield strength. UTS: Ultimate Tensile Strength

Conclusions

We can conclude that to avoid the problems mentioned, purchase specifications must be correctly made, supporting documentation for standardized requirements must be requested, and tests must be carried out to verify the properties of the fittings before installation. In this last aspect, the ESYS 10 has proven to be a useful and necessary tool for quality control at the time of acquisition, allowing companies to save costs related to lost profit, repairs, and negative impacts on their public image, thus improving efficiency and reliability.