1、 STP-PT-043 ASME FLAWED CYLINDER TESTING M“E STANDARDS TECl-INOLOGY, LLC Date oflssuance: December 17, 20 I 0 This report was prepared as an account of work sponsored by ASME Pressure Teclmologies Codes and Standards and the ASME Standards Technology, LLC (ASME ST-LLC). Neither ASME, ASME ST-LLC, th
2、e authors, nor others involved in the preparation or review of this report, nor any of their respective employees, members or persons acting on their behalf, makes any warranty, express or imp! ied, or assumes any legal liability or responsibility for the accuracy, completeness or usefulness of any
3、information, apparatus, product or process disclosed, or represents that its use would not infringe upon privately owned rights. Reference herein to any specific commercial product, process or service by trade name, trademark, manufacturer or otherwise does not necessarily constitute or imply its en
4、dorsement, recommendation or favoring by ASME ST -LLC or others involved in the preparation or review of this report, or any agency thereof The views and opinions of the authors, contributors and reviewers of the report expressed herein do not necessarily reflect those of ASME ST -LLC or others invo
5、lved in the preparation or review of this report, or any agency thereof. ASME ST-LLC does not take any position with respect to the validity of any patent rights asserted in connection with any items mentioned in tllis document, and does not undertake to insure anyone utilizing a publication against
6、 liability for infringement of any applicable Letters Patent, nor assumes any such liability. Users of a publication are expressly advised that determination of the validity of any such patent rights, and the risk of infringement of such rights, is entirely their own responsibility. Participation by
7、 federal agency representative(s) or person(s) affiliated with industry is not to be interpreted as government or industry endorsement of this publication. ASME is the registered trademark ofthe American Society of Mechanical Engineers. No part of this document may be reproduced in any form, in an e
8、lectronic retrieval system or otherwise, without the prior written pernlission of the publisher. ASME Standards Teclmology, LLC Three Park Avenue, New York, NY 10016-5990 ISBN No. 978-0-7918-3353-7 Copyright 20 I 0 by ASME Standards Technology, LLC All Rights Reserved ASME Flawed Cylinder Testing ST
9、P-PT-043 TABLE OF CONTENTS Fo:reword . iv Abstract v INTRODUCTION . 7 2 PRESSURE VESSEL CONFIGURATION 8 3 TEST PLAN 9 4 FLAWS 10 5 FATIGUE TEST l3 6 BURST TESTING . 15 7 RESULTS AND DISCUSSION 16 8 SUMMARY AND CONCLUSIONS 19 Appendix A 20 Appendix B . 33 Appendix C . 46 Acknowledgments 72 LIST OF TA
10、Bl ES Table 1 -Randomly Assigned Tank Numbers, Flaw Depths and Number of Cycles 9 Table 2 -Test Tank Flaw Depths Using the Nominal Structural Composite Thickness of 11.4mrn (0.449 inches) . lO Table 3 -Pressure Cycling Tests Results . 13 Table 4- Summary of Burst Data . 15 Table 5 - Comparison of Bu
11、rst Data . 17 LIST OF FIGURES Figure I -Composite Pressure Vessel, PIN 240075-023 . 8 Figure 2 -Location of Flaws on the Tanks . 11 Figure 3-Locations of Longitudinal and Transverse Flaws 11 Figure 4- Cross-Sectional View of Flaw . 12 Figure 5-Record of Flaw Depths, in Inches, after Machinjng 12 Fig
12、ure 6 -Longitudinal Flaw in Tank 3B Before Cycling . I 4 Figure 7-Longitudinal Flaw in Tank 3B after 20,000 Cycles . 14 Figure 8-Burst Pressure Versus Depth of Flaw and Cycling 18 Figure 9-Burst Pressure Versus Cycling and Depth offlaw 18 111 STP-PT-043 ASME Flawed Cylinder Testing FOREWORD The ASME
13、 BPV Project Team on Hydrogen Tanks, in conjunction with other ASME Codes and Standards groups, is developing Code Cases and revisions to the Boiler therefore, this tank was tagged as one of the two “spares“ for testing. Flaws that range from 1.27 rnm (0.050 in.) (approximately 10%) up to 40% throug
14、h the composite were machined into 12 tanks. One additional tank without the machined flaws was burst as a control. Tanks were to be burst after 0, 5000 or 10,000 cycles. However, since cycling the tanks with the flaws to 10,000 cycles did not significantly decrease the burst pressure, it was decide
15、d to cycle the remaining set to 20,000 cycles rather than the originally planned 5000 cycles. The use of a non-load sharing HOPE liner in these pressure vessels allows higher number of pressure cycles without the risk of leaking due to liner fatigue. SINs were randomly assigned to each test and are
16、listed in Table 1. Table 1 -Randomly Assgned Tank Numbers, Flaw Depths and Number of Cycles Flaw Depth Test Tank SN (I) (% of structural thickness) Cycles 5 1007-014 none 0 4A 1007-005 1.27 mm (0.050“) 10% 0 lA 1007-002 20% 0 2A 1007-009 30% 0 3A 1007-011 40% 0 4C 1007-010 1.27 mm (0.050“) 10% IOk I
17、C 1007-001 20% IOk 2C 1007-015 30% IOk 3C 1007-013 40% IOk 48 1007-006 1.27 mm (0.050“) 10% 20k IS 1007-003 20% 20k 28 1007-004 30% 20k 38 1007-012 40% 20k (I) Chosen at random I 007-008 Damaged during flaw procedure I 007-007 Slightly heavy from manufacturing process, used as flaw practice tank 9 S
18、TP-PT-043 ASME Flawed Cylinder Testing 4 FLAWS For the tanks with flaws, two flaws were machined into the composite laminate. Since these composite pressure vessels contained a sacrificial glass overwrap, the first step was to remove a portion of the glass to expose the structural composite layer. T
19、he removal of this layer was performed on all of t!he tanks regardless of whether a flaw was to be machined into the structural layer. The depths of the flaws are listed in Table 1. The flaw requirements are as follows. Flaws were made using a I mm diameter ball-end cutter to a depth listed in Table
20、 2. The length of each flaw was 57.0 +/- 0.13 mm (2.245 +/- 0.005 in.) as measured from the center of the cutter. This flaw length was five times the structural laminate thickness of 11.4 mm (0.449 in.) based on the flaw test requirement ofiSO 11119-3. The longitudinal flaw was cut at the mid-length
21、 of the pressure vessel. The transverse flaw was cut at the mid-length of the pressure vessel approximately 120 from the longitudinal flaw. Schematic drawings of these locations are shown in Figure 2 through Figure 4. Dimensions were measured and recorded for each flaw and are shown in Figure 5. Eac
22、h flaw was photo-documented and is shown in Appendix A. The structural laminate was a combination of approximately 43% helical reinforcement and 57% circumferential reinforcement patterns. The reduction of these reinforcements due to the flaws are also listed in Table 2. Table 2-Test Tank Flaw Depth
23、s Using the Nominal Structural Composite Thickness of 11.4 mm (0.449 in.) Longitudinal !Flaw Depth Transverse Flaw Depth Reduction in Red uction in +f. 0.13 mm ( 0.005“) +f. 0.13 mm ( 0.005“) Helical Circumferential Test Tank SN (I) mm (inches) mm (inches) Reinforcement Reinforcement lA 1007-002 2.3
24、 (0.090) 2.3 (0.090) IB 1007.003 2.3 (0.090) 2.3 (0.090) IC 1007-001 2.3 (0.090) 2.3 (0.090) 17% 22% 2A 1007.009 3.43 (0.135) 3.43 (0.135) 2B. 1007-004 3.43 (0.135) 3.43 (0.135) 2C 1007-015 3.43 (0.135) 3.43 (0.135) 33% 28% 3A 1007-011 4.57 (0.180) 4.57 (0.180) 3B 1007.012 4.57 (0.180) 4.57 (0.180)
25、3C 1007-013 4.57 (0.180) 4.57 (0.180) 45% 38% 4A 1007.005 1.27 (0.050) 1.27 (0.050) 4B. 1007-006 1.27 (0.050) 1.27 (0.050) 4C 1007-010 1.27 (0.050) 1.27 (0.050) 8% 13% 5 1007-0 14 glass removal only glass removal only 0 0 10 ASME Flawed Cylinder Testing Longitudinal flaw / at mid-cylinder Figure 2-
26、Location of Flaws on the Tanks Longitudinal flaw at mid-cylinder / STP-PT-043 Transverse flaw 120 from longitudinal flaw at mid-cylinder / Transverse flaw 120 from longitudinal flaw at mid-cylinder Figure 3 - Locations of Longitudinal and Transverse Flaws II STP-PT-043 - 1 57 mm (2.245 in) Depth - A
27、SME Flawed Cylinder Testing Surface of structural layer 0.5 mm R Figure 4 - Cross-Sectional View of Flaw T bl 2 R d ffl dim . ASM:E Fla ed Cylind T . a t : ecor o aw ensoons “ er eslll1g IAogitlofilal lAHleitoodlnl ln.,. n . , I toocmoduw TraonttH l:bu , “ gtan Loaci1dlal Lodiut Tr-a-.tnc Traanuse l
28、la:wDtptb Fbwll .o“ l 2.“:;1-11 I .oBCf . .:) . 2A 1007-0.135 0.135 . (v .ll 009 .560 .;t.lq O .rl. ;J. “I IJ(p -“II 28 1007-0.135 I 0.135 .900 004 -116 ?CO ;(.fl . I “!“I -f5 l!lS“ ;.fO 2C 1007-0.135 0.135 .“! a-ffe 015 - -Ill .l -10 .r:;S ;.t.ll -1007-;,- .Sb: ;!.11 ;.lfl8 3A 0.180 0.180 .SPO S-;t
29、 :iO ,lSI a.O s- .0 Zt-.a51 2 Z4/ 1007- glass removal tlA. thickne ss) Cycles cycles min recorded recorded s I 007-014 none 0 - - -4A I 007-005 10% 0 - - -lA I 007-002 20% 0 - - -2A 1007-009 30% 0 - - -lA 1007-011 40% 0 - - -4C I 007-010 10% IOk 10,006 1.44 (209) 25.5 (3697) IC 1007-001 20% IOk 10.0
30、06 1.44 (209) 25.5 (3697) 2C 1007-015 30% IOk 10,006 1.44 (209) 25.5 (3697) l C 1007-013 40% IOk 10,000 1.94 (282) 25.2 (3660) 4B 1007-006 10% 20k 20,007 1.49 (2 16) 25.3 (3672) IB I 007-003 20% 20k 20,007 1.49(216) 25.3 (3672) 2B I 007-004 30% 20k 20,007 1.49 (2 16) 25.3 (3672) l B I 007-012 40% 20
31、k 20,006 1.68 (243) 25.2 (3658) Note: Tanks lC, 2C and 4C were cycled together and lB, 2B and 4B were cycled together. 13 STP-PT-043 . CD o_ Oc: . 0 E ASME Flawed Cylinder Testing R Figure 6 -Longitudinal Flaw in Tank 36 Before Cycling _S . PM - ,. - “-:;o l l ll l l II I i I U llll 111 Jl 1111 i 11
32、111111 flflltiiJ Figure 7 - Longitudinal Flaw in Tank 36 after 20,000 Cycles 14 ASME Flawed Cylinder Testing STP-PT-043 6 BURST TESTING Each cylinder was tested hydraulically with water to rupture at ambient conditions. The pressurization rate was less than 0.48 MPa/second (70 psi/second). The maxim
33、um pressure obtained and the burst initiation location were recorded and are shown in Table 4. The pressure versus time was recorded and is shown in Appendix B. Photographs of each cylinder after burst are shown in Appendix C. There are four photographs of each burst. The first two are overall views
34、 of the carcass and the last two are closer views of the flaw locations. Table 4 -Summary of Burst Data Orde r Test Tank SN Flaw Depth Cycles Burst, MPa (psi) Locat ion I 5 1007-014 none 0 74.30 ( I 0,776) mid-cyl 2 4A 1007-005 10% 0 76.12 (1 1,041) mid-cyl 3 lA 1007-002 20% 0 69.63 (I 0,099) mid-cy
35、l 4 2A 1007-009 30% 0 59.27 (8,597) mid-cyl 5 3A 1007-011 40% 0 52.98 (7,685) mid-cyl 6 4C 1007-010 10% IOk 7 1.46 (I 0,364) mid-cyl 7 IC 1007-001 20% IOk 65.57 (9,5 10) mid-cyl 8 2C 1007-015 30% IOk 63.54 (9,216) mid-cyl 9 3C 1007-013 40% IOk 53.59 (7,773) mid-cyl 10 4B 1007-006 10% 20k 65.37 (9,48
36、 1) mid-cyl II IB 1007-003 20% 20k 67.80 (9,834) mid-cyl 12 2B 1007-004 30% 20k 60.29 (8,745) mid-cyl 13 3B 1007-012 40% 20k 56.23 (8.1 56) mid-cyl (1 Chosen at random 15 STP-PT-043 ASME Flawed Cylinder Testing 7 RESULTS AND DISCUSSION Burst data are listed in Table 5 and these pressures are graphed
37、 in Figure 8 and Figure 9. All bursts originated at the longitudinal flaw location mid-cylinder. This was expected since this flaw cut the hoop direction fibers preferentially, which resulted in a larger decrease in the burst performance than the transverse (circumferential) flaws. Burst-to-operatin
38、g pressure ratio was calculated as the ratio of the burst pressure to that of the service pressure, which in this case was 24.8 MPa (3600 psi). . . Bu st Pressure (MP a) burst- to - operattng pressw-e ratw = 24_8 To compare the effect of flaw size and cycling with the burst pressures, the ratio betw
39、een the burst pressure and the burst pressure of the tank without flaws and cycles was used. Burst Pressure (MP a) Burst Retention= x 100% Burst Pressure of Tank 5 (no flaws) To compare the effect of cycling only, the ratio between the lburst pressure and the burst pressure with the same flaw size b
40、ut no cycles was used. Burst Pressure (M Pa) Cy cle Ret ention= . x 100% Burst Pressure of Tank wtth same f law but no cycles The burst history on this configuration was reviewed. The average of five bursts was 77.2 MPa (1 I ,200 psi) and ranged from 71.7 MPa (10,400 psi) to 81.4 MPa (11,800 psi). T
41、ank 5 (no flaws or cycles baseline) burst at 74.3 MPa (l 0,776 psi) which is a burst-to-operating pressure ratio of 2.99. This burst was lower than the historical average but it was within the range of burst pressures, and even though there was no machined flaw in the structural layer, the glass ove
42、rwrap had been removed in the location of the flaws. Tank 4A, which had flaws cut 10% through the structure, actually burst slightly higher at 76.1 MPa (11,041 psi). This higher burst pressure can be explained due to normal tank-to-tank variation in the manufacturing process and it should not be int
43、erpreted to mean that adding a small flaw increases burst performance. Other than this slight increase, generally a deeper flaw resulted in a lower burst pressure, which was expected. In addition, this trend of decreasing burst pressure with increasing flaw depth is observed when tanks that have bee
44、n cycled 10,000 and 20,000 times are compared. This trend is clearly seen in the Burst Loss column of Table 5 and also in Figure 9. An exception to this was observed with a 2.23 MPa (324 psi) increase in burst pressure at 20,000 cycles for the 20% flaw over the I 0% flaw. The effect of cycling on a
45、given flaw depth is shown in Figure 8. For a flaw 10% through the structural layer, the burst pressure decreases with increasing cycles. This indicates that, for a small flaw, cyclic fatigue loading does have an effect. However, the burst-to-operating pressure ratio was still 2.63 after 20,000 cycle
46、s. The indicated effect of cyclic loading on the deeper flaws was different. For the 40% flaw, cyclic fatigue loading actually increased the burst pressure when compared to the 16 ASME Flawed Cylinder Testing STP-PT-043 zero cycle tank with the same flaws. This trend was also observed with the 30% f
47、law and somewhat with the 20% flaw. The explanation for this behavior with cycling is as follows. Flaws machined into the laminate with the l mm diameter carbide ball tool, even though there was a radius, resulted in a geometric stress riser. Upon loading, the stress at this point was concentrated u
48、ntil sudden fracture occurred which nucleated the burst of the vessel. For the two tanks that were cycled with the 40% flaws, a re distribution of the stress concentration may have taken place. First, it is possible that the severity of the stress riser was mitigated by fatigue cracks originating at
49、 the locations of highest stresses. Secondly, these cracks formed by the cyclic loading may reduce the tendency of a sudden, intense fracture at the flaw that would nucleate the burst. In essence, it has been shown here that with severe flaws, cyclic loading to service pressure increases the pressure at burst. However, the differences in burst pressures between cyclic fatigue levels are small enough that some
