An Experimental Study on the Evaluation of Failure Behavior of Pipe With Local Wall Thinning

The pipe failure tests were performed using 102mm-Sch.80 carbon steel pipe with various simulated local wall thinning defects, in the present study, to investigate the failure behavior of pipe thinned by flow accelerated corrosion (FAC). The failure mode, load carrying capacity, and deformation ability were analyzed from the results of experiments conducted under loading conditions of 4-point bending and internal pressure. A failure mode of pipe with a defect depended on the magnitude of internal pressure and axial thinning length as well as stress type and thinning depth and circumferential angle. Also, the results indicated that the load carrying capacity and deformation ability were depended on stress state in the thinning region and dimensions of thinning defect. With increase in axial length of thinning area, for applying tensile stress to the thinning region, the dependence of load carrying capacity was determined by circumferential thinning angle, and the deformation ability was proportionally increased regardless of the circumferential angle. For applying compressive stress to thinning region, however, the load carrying capacity was decreased with increase in axial length of the thinned area. Also, the effect of internal pressure on failure behavior was characterized by failure mode of thinned pipe, and it promoted crack occurrence and mitigated a local buckling of the thinned area.

Design of Ellipsoidal Heads Using Elastic-Plastic Finite Element Analysis

Traditional design techniques result in excess material being required for ellipsoidal heads. The 2001 ASME Boiler and Pressure Vessel Code Section VIII Division 1, UG-32D and Section VIII Division 2, AD-204 limit the minimum design thickness of the heads. ASME Boiler and Pressure Vessel Code Case 2261 provides alternate equations that enable thinner head design thickness. VIII-2 Appendix 3 and 4 methods potentially could be used to further optimize the head thickness. All the equations in the code use one thickness for the entire head. On large diameter thin heads the center or spherical area is often thicker than the knuckle area due to the method of manufacture. Including this extra material in the design calculations results in an increase of the MAWP of large diameter thin heads. VIII-2, AD-200 of the code permits localized thinning in a circumferential band in a cylindrical shell. Applying these same rules to elliptical heads would permit thinning in the knuckle region as well. Engineers have powerful finite element analysis tools that can be used to accurately determine levels of plastic strain and plastic deformed shapes. It is proposed that VIII-2 Appendix 4 and 5 methods be permitted for the design of elliptical heads. Doing so would permit significant decreases in thickness requirements. Different methods of Plastic Finite Element Analysis (PFEA) are investigated. An analysis of a PVRC sponsored burst test is done to develop and verify the PFEA methods. Two designs based on measurements of actual vessels are analyzed to determine the maximum allowable working pressures (MAWP) for thick and thin heads with and without local thin regions. MAWP is determined by limit analysis, per VIII-2 4-136.3 and by two other proposed methods. Using Burst FEA, the calculated burst pressure is multiplied by a safety factor to obtain MAWP. Large deflection large strain elastic perfectly plastic limit analyses (LDLS EPP LL) method includes the beneficial effect of deformations when determining the maximum limit pressure. Elliptical heads become more spherical during deformation. The spherical shape has higher pressure restraining capabilities. An alternate design equation for elliptical heads based on the LDLS EPP LL calculations is also proposed.

Seismic Design of Geothermal Pipeline Supports

Seismic loading is a critical factor in the structural design of piping systems for geothermal power plants in Iceland. It has been shown that the design of piping systems which is based on response spectrum static analysis can lead to overdesigned systems. The loading on the supports will be overestimated and, hence, the supports tend to be too stiff. This increased stiffness both increases the cost and reduces the quality of the seismic design. The systems response is highly dependent on the support stiffness. In this work, the design of a typical expansion loop with respect to seismic loading is discussed, with the goal being to minimize the loading from the ground acceleration. A typical pipe support is modelled, and its stiffness characteristics are evaluated and optimized. Finally, recommendations are made for improvements in pipeline support design.

Dynamic Pipe Stresses During Water Hammer: III — Complex Stress Relationships

Complex three-dimensional dynamic stresses occur in a pipe following a water hammer event. Equations from vibration theory were adapted for use to describe the dynamic stresses at any point along the pipe wall. Hoop, radial, and axial dynamic stress equations are presented to approximate the stresses at a point on the pipe wall. Dynamic stress equations for beams and other simple shapes are also considered. The dynamic pipe stresses are affected principally by the types of water hammer waves or fluid transients, by the wave impacts at elbows or tees, and by the reflections of the waves from these elbows or tees. The three fluid transients considered are a moving step pressure wave, a ramp pressure, and a moving pressure spike. Approximate techniques are presented for evaluating the effects on piping due to the impingement of these transients on an elbow. For an equivalent pressure in a long pipe, application of the step pressure created the largest stress increases of the three transients considered. The vibration equations also prompt a solution to reduce water hammer effects. To this end, slow closing valves are frequently employed. Vibration theory may be applied to quantify the stress reductions afforded by these valves. Pipe stress equations may be manipulated to reduce pipe stresses for a linearly increasing, or ramp, pressure wave traveling along the pipe.

Assessment of Local Decreases in Wall Thickness at the Connection Straight-Pipe to Bend

Welds are ground during manufacturing to free them from offset edges and notches and thus to obtain a more favorable stress distribution. Apart from the above, welds are also ground to prepare them for and improve the conditions of in-service testing and inspection. The grinding of welds may result in a local decrease in wall thickness, so that there may be local deviations from the required minimum wall thickness. In order to fulfill the task of evaluating the strength of such material-loss regions, we have determined appropriate stress concentration factors for typical wall-thickness deviations and various wall-thickness/diameter ratios, which enable us to assess quickly and, if necessary, directly after the on-site measurement of wall thickness, whether a detected deviation from the minimum value is permissible. To be able to evaluate deviations from minimum wall thickness, especially in welds that form a connection to bends, we have determined stress indices for the beginning and the end of bends for common pipe bend dimensions and various bend angles. Compared with the maximum stress indices commonly used in piping calculations for the crown of the bend, the stress indices at the end of the bend are lower than those at the crown and can help to reduce unnecessary conservatism. In the paper, stress indices for various grinding geometries and for the beginning and the end of common bend shapes will be presented, as well as the method used to evaluate strength and the criteria pertaining to the tolerability of decreases in wall thickness.

A Dynamic Investigation of Piping Systems With a Bolted Flange

The dynamic response of piping systems with a bolted flange is analyzed. Experimental and numerical analyses and results are presented and show excellent correlation. An overhanging piping system at various span lengths with a flange at mid-length is used. The testing configuration consists of a standard 2-in. (51 mm) schedule 40 steel pipe and an ANSI B16.5 class 300-pound flange. The presence of a spiral wound wire gasket and high strength flange bolts is also assessed. Included are multiple resonant frequencies and their respective mode shapes for various span lengths and gasket configurations. The experimental procedure utilizes an accelerometer to gather the dynamic response output of the piping system due to an impulse. The resonant frequencies are then determined using a Fast Fourier Transform (FFT) method. The numerical analysis is conducted using the commercial Finite Element (FE) code ANSYS®. Both methods take into account the complex interaction between the flange and gasket and their impact on the entire piping system. The dynamic effects of a bolted flange and gasket on a piping system are critical in their use and a summary of the results for a variety of configurations is presented.

Stress Indices for Circumferential Fillet Welded and Socket Welded Joints

This paper presents the results of an investigation of the Stress Indices for circumferential fillet welded or socket welded joints which are required for ASME Section III Class 1 systems. The history or background of the basis of the various indices is reviewed and summarized. New values for the indices are suggested based on new test data and analysis. These changes will result in a more accurate evaluation of circumferential fillet welded or socket welded joints.

Investigation of Torsional Stress Intensification Factors and Stress Indices for Girth Butt Welds in Straight Pipe

The basis for fatigue evaluation of ASME Section III Class 2, 3 and B31.1 piping is the girth butt weld where the Stress Intensification Factor (SIF) is defined to be 1.0. The SIFs for other components are based on comparison to the butt welds. This SIF of 1.0 for butt welds is based on extensive bending tests on carbon steel straight pipe which are reviewed and summarized in this study. The results of new test data, including torsional test data, are presented. The authors are unaware of any previous torsional tests on carbon steel straight pipe. This new data leads to suggested changes in the codes taking into account the directionality of the loading.

Stiffness Coefficients for Nozzles in API 650 Tanks

The analysis of tank nozzles for API Standard 650 [1] tanks is a complex problem. Appendix P of API 650 provides a method for determining the allowable external loads on tank shell openings. The method in Appendix P is based on two papers, one by Billimoria and Hagstrom [2] and the other by Billimoria and Tam [3]. Although Appendix P is optional, industry has used it for a number of years for large diameter tanks. For tanks less than 120 feet (33.6 m) in diameter, Appendix P is not applicable. In previously published papers [4–10], the authors used finite element analysis (FEA) to verify the experimental results reported by Billimoria and Tam for shell nozzles. The analysis showed the variance between stiffness coefficients and stresses obtained by FEA and API 650 methods for tanks. In this follow-up paper, the authors present stiffness coefficients for tank nozzles located away from a structural discontinuity. Factors to establish spring rates for nozzles varying from 6 to 48 inches and tank diameters from 30 feet to 300 feet and for nozzles at different elevations on the shell are provided. Mathematical equations are provided together with graphs for the stiffness coefficient factors.

Convergence of the Axisymmetric Bessel Function Solution to the Pipe Strap Anchor Problem

One of the paradigmatic problems that frequently arise within a piping flexibility analysis is need to quantify the effect of local stresses created by supports and restraints attachments. Over the past twenty years, concerns have been identified by both regulatory agencies in the nuclear power industry and others in the process and chemicals industries concerning the effect of various stiff clamping arrangements on the expected life of the pipe and piping components. In many of the commonly utilized geometries and arrangements of pipe clamps, it has previously been shown by the author that some pipe clamp anchors can be treated as an axisymmetric elasticity problem. These pipe anchors may be simplified considering the axisymmetric stress and deformation determination within a hollow cylinder (pipe) that is subjected to appropriate boundary and loading conditions, per se. One of the geometries previously considered and addressed is comprised of two pipe clamps that are bolted tightly to the pipe and affixed to a modified shoe-type structure. The shoe is employed for the purpose of providing a theoretically immoveable base that can be easily attached either by bolting or welding within a pipe rack. Because the Bessel function and Fourier solution relies on the summation of a large number of terms (i.e., a series type solution), convergence becomes an issue when time is of the essence. This paper addresses the two variables that need to be controlled to obtain acceptable results with the series type solution.