2024-07-23 Fatigue of a Tee

This article is an excerpt of a larger report and gives the current status of the code calculations that have been done to investigate the fatigue life time of a tee. The tee under investigation failed under high cyclic loading.

The conclusion with the code calculations is that

• the tee is strong enough for the maximum (static) internal pressure
• the code calculations give an indication whether fatigue failure due to cyclic loading is to be expected but seem not (yet) to be suitable to predict the life time of the tee under the specified cyclic pressure loading.

Context

Red-Bag has been invited to assess the failure of tees in a high pressure water system with cyclic pressures between 5 MPa and 18 MPa. A tee that fails because of over pressure looks like this:

Figure: result of burst test, image from PVP2014-28265 Proceedings of the ASME 2014 Pressure Vessels & Piping Conference

However the failed tee shows the following first sign of malfunctioning:

Figure: failing tee

The tee is leaking but did not burst. The crack investigation showed signs of fatigue.

Calculations

Static Internal Pressure

The failed tee 7 has been used for the code calculations and the first investigation. There is enough information available and also some material crack investigation has been done.

The following internal pressure calculations have been done with all the same design conditions (internal pressure 17.5 MPa and temperature 42 ^oC), pipe with A106 Gr.B material and elbow/tee with A234 Gr. WPB material:

• ASME B31.3 304.1.2 straight pipe size 10” schedule 160 wall thickness 28.58 mm – PA7001
• ASME B31.3 304.2.1 elbow 6” schedule 160 wall thickness 18.26 mm – EA7001
• ASME B31.3 304.3.3 pipe-pipe (stubon) branch 10”x6” updated wall thickness 39x20 mm – BA7001
• ASME B31.3 304.3.4 extruded outlet 10”x6” updated wall thickness 39x20 mm – TA7001
• EN 13480-3 Clause 6.1 straight pipe size 10” schedule 160 wall thickness 28.58 mm – PE7001
• EN 13480-3 Clause 6.2 elbow 6” schedule 160 wall thickness 18.26 mm – EE7001
• EN 13480-3 Clause 8.4 (CC) pipe-pipe (stubon) branch 10”x6” updated wall thickness 39x20 mm – BE7001
• EN 13480-3 Clause 8.4 (T) extruded outlet 10”x6” updated wall thickness 39x20 mm – TE7001
• EN 13480-3 Clause 8.4 (T) extruded outlet 10”x6” updated wall thickness 39x40 mm – TE7051, to assess the increase of the branch thickness from 20 mm to 40 mm
• EN 13480-3 Clause 8.4 (T) extruded outlet 10”x6” updated wall thickness 39x20 mm with material A860 WPHY – TE7101, to asses the change of material

The internal pressure calculations show the components are suitable for the internal pressure of 17.5 MPa, table with MAWP (maximum allowable working pressure) in MPa:

Table: MAWP as static internal pressure

The TE7051 calculation with the increased branch thickness has a MAWP of 33.94 MPa. The TE7101 calculation with the material A860 WPHY material has a MAWP of 41.46 MPa, due to only the higher allowable stress.

The conclusion for the internal pressure calculation is that the components in the piping system around the tee 7 are, with an adequate margin, strong enough.

Fatigue by Pressure Cycles

The following fatigue calculations are done: 

• ASME VIII div.2 2023 Part3, Part 5 
• EN 13445-3 2021 Clause 17 
• EN 13445-3 2021 Clause 18 

The ASME standard has the following sketches in Annex 5-D Figure 5-D.2 for branch connections: 

Figure: ASME VIII div.2 Figure 5-D.2 branch connection 

The EN 13445-3 Clause 17 contains these sketches which are as close as possible to a tee: 

Figure: EN 13445-3 Clause 17 Table 17-3 branch connection 

The EN 13445-3 Clause 18 contains these sketches, close to a tee: 

Figure: EN 13445-3 Clause 18 Table 18-4 branch connection

The PD 5500 Annex C Table C.2 has similar figures as the EN 13445-3 Clause 18. 

The fatigue calculations are not as straight forward as the internal pressure calculations. This is due to the tee shape, having a complex stress distribution and fatigue is based on statical and historical experimental data. The process and approach for these calculations is somewhat different in each standard. 

The ASME figures show a shape that has similarities with the tee under investigation. However, the radius r1, r2 and r3 are only used to assess against the scope and is not a design parameter as such. The software used for the calculation does not take the radii in to account and assumes the overall design permits the use of the smooth bar. This is a bit strange because the stubon (pipe-pipe branch) does not (always) has smooth transitions 

The tee has been modelled in the EN calculations as a stubon because a tee as a piping component model does not exist in the EN fatigue calculations. 

The stubon branch sketch is shown below: 

Figure: model for fatigue calculations 

The following fatigue calculations have been done: 

ASME VIII Div.2 per 3.F.1.1 to 3.F.1.3 and Annex 5.D (software Xp)

• EN 13445-3 Clause 17 (software Xp) η = 3 and class = 71 (welded)
• EN 13445-3 Clause 18 (software Xp) (160 MPa and 240 MPa for full Δσeq)
• EN 13445-3 Clause 17 (software Xv) η = 3 and class = 71
• EN 13445-3 Clause 18 (software Xv) (160 MPa and 240 MPa for full Δσeq)
• EN 13445-3 Clause 17 (Software Xb) η = 3 and class = 71 

The software Xb is Red-Bag in-house software. It is created for validation of calculations. The software allows a report on per formula basis, refer to the attachments. The software Xp and Xv are commercially available applications. 

The ASME VIII Div.2 and the EN 13445-3 Clause 18 require actual stress input from a FEM calculation. The 160 MPa for reduced range and 240 MPa local stress have been taken from the FEM calculation. 

As an alternative, the ASME can use the stress indices from 3-F for nozzles as a simpler calculation method. The Xp software uses in the calculation only the stress indices for the fatigue. Other type of fittings can not be calculated. 

The FEM calculated stresses have to be used for the EN 13445-3 Clause 18 for both Xp and Xv applications. The stresses have been used in above calculations estimated from the FEM calculations done in Annex D. Further improvement to the actual stresses is still to be done. 

The below table summarizes the expected cycles before failure and the damage factor: 

The damage factor should be below 1, in case the required life time for tee 7 is 5 years. The above cycles and damage factors, according EN, indicate that 5 years will not be met. The damage factor is the actual cycles divided by the maximum allowed cycles. 

The ASME calculation does not seem right, we have looked for input error but could not find it yet. Also, the differences in the results do not give a good confidence about the accuracy of the calculations in general. 

Conclusion

The current fatigue code calculations cannot be used for the alternative design of tee 7. The reasons are amongst others: 

• The dimensions such as the crotch inside and outside radius have limited or no effect in the result of the calculations 
• The detailed fatigue calculations, ASME VIII Div.2 and EN 13445-3 Clause 18 require the interpretation of the FEM stress ranges as input in the calculations. This might lead to inaccurate input and results. 
• Reference calculations, results and validation to actual components is not yet available. 

The effect of corrosion can play a role in the fatigue life time of tee 7. The EN 13445-3 Clause 18.4.5 has a disclaimer on corrosion: 

“Corrosive conditions are detrimental to the fatigue lives of steels. Environmentally-assisted fatigue cracks can occur at lower levels of fluctuating stress than in air and the rate at which they propagate can be higher. The fatigue strengths specified do not include any allowances for corrosive conditions. Therefore, where corrosion fatigue is anticipated and effective protection from the corrosive medium cannot be guaranteed, a factor should be chosen, on the basis of experience or testing, by which the fatigue strengths given in these requirements should be reduced to compensate for the corrosion. If, because of lack of experience, it is not certain that the chosen fatigue strengths are low enough, the frequency of inspection should be increased until there is sufficient experience to justify the factor used.” 

It should be noted that also the water environment is detrimental to fatigue performance of carbon steel, refer also to the graph in annex F. This fact justifies a further disclaimer in addition to the corrosion effect. 

The Red-Bag and Wentzel Dynamics fatigue assessment of the tee continues. The progress of the investigation to an adequate calculation method and assessment of the fatigue life will be posted on this website