Microsoft Word - 1.docx CHEMICAL ENGINEERING TRANSACTIONS VOL. 77, 2019 A publication of The Italian Association of Chemical Engineering Online at www.cetjournal.it Guest Editors: Genserik Reniers, Bruno Fabiano Copyright © 2019, AIDIC Servizi S.r.l. ISBN 978-88-95608-74-7; ISSN 2283-9216 Experimental Determination of the Static Equivalent Pressures of Detonative Explosions of Cyclohexane/O2/N2- Mixtures in Long and Short Pipes (part 1of 3) Hans-Peter Schildberg*, Julia Eble BASF SE, 67056 Ludwigshafen, Germany hans-peter.schildberg@basf.com In the past 5 years the Safety Engineering Group of BASF had determined the static equivalent pressures (“pstat”) of the eight detonative pressure scenarios which explosive gas mixtures can exhibit in long and in short pipes. More precisely, for different combustibles the pstat-values of the corresponding ternary mixtures combustible/O2/N2 were determined on the stoichiometric line and on the O2-line of the explosion triangle. By doing so, the pstat-values of all other compositions inside the explosion triangle could be predicted by extrapo- lation with an accuracy sufficient for practical applications. Furthermore, a proposal of how to transfer these results to the huge number of other combustibles not investigated so far was provided. A key-parameter in this context was the ratio R between the static equivalent pressure at the point where the deflagration-to- detonation transition occurs in the long pipe and the static equivalent pressure in the region of the stable detonation. In the present work the pstat-values of the new ternary mixture cyclohexane/O2/N2 are reported. Cyclohexane is of special interest because its autoignition temperature (AIT) in air is substantially lower than the AIT-values of all combustibles that had been tested before. According to our hitherto existing understanding the low AIT should noticeably reduce the ratio R. The experiments, however, did not confirm this hypothesis. After presenting the experimental results, which actually confirm the findings for the combustibles investigated so far, an explanation for the unexpected behaviour regarding R will be presented in terms of the differences between the low-temperature and the high temperature oxidation mechanism. As kind of spin-off, this explanation also allows to better understand quantitatively the degree of precompression in the yet unreacted mixture required for the occurrence of the deflagration-to-detonation Transition (DDT). 1. Introduction In chemical process plants explosive gas mixtures with compositions in the potentially detonative regime can occur. Since effective ignition sources can in general not be excluded with certainty and since explosion pressure venting cannot be applied for detonative explosions, the only method to ensure safe operation is detonation pressure resistant design of the affected plant components. Because no guidelines giving recommendations for detonation pressure resistant design had been available, BASF launched a research project 5 years ago to determine the static equivalent pressures (“pstat”) of the 8 detonative pressure scenarios occurring inside long and short pipes. (For quantifying the pstat-values of detonations in vessel-geometry BASF started research in 2019). The results for pipes can be found in Schildberg (2013, 2014, 2015, 2016a and 2018) and in TRGS407 (2016). In Schildberg (2016b) a general overview of the entire detonation issue in pipes is provided (explanation of the 8 different pressure scenarios, experimental technique, definition of pstat, explanation of the key-parameter R, application of pstat for designing pipes detonation pressure resistant, generalization of the results to combustibles other than those tested). The key-parameter R is the ratio between the static equivalent pressure at the point where the DDT occurs in a long pipe and the static equivalent pressure in the region of the stable detonation. This ratio – in combination with the Chapman-Jouguet pressure ratio, which can be calculated in straightforward manner from the thermodynamic properties of the mixtures in question for any value of the initial temperature and pressure – DOI: 10.3303/CET1977175 Paper Received: 25 December 2018; Revised: 19 May 2019; Accepted: 17 June 2019 Please cite this article as: Schildberg H.-P., Eble J., 2019, Experimental Determination of the Static Equivalent Pressures of Detonative Explosions of Cyclohexane/O2/N2-Mixtures in Long and Short Pipes (part 1of 3), Chemical Engineering Transactions, 77, 1045-1050 DOI:10.3303/CET1977175 1045 is the key to quantitatively predict the static equivalent pressures of all design relevant detonative pressure scenarios occurring when explosive gaseous mixtures undergo a DDT inside pipes. A recommendation for the values to be expected for R over the entire detonative regime of the explosion triangle of a common combustible/O2/N2-mixture at ambient temperature is given in chapter 7 of Schildberg (2018). By applying this diagram a pretty good estimate for the static equivalent pressures of any combustible/O2/N2 mixture is possible. Based on the conception of the transition to detonation described in Schildberg (2016b) the hypothesis emerged that combustibles with low autoignition temperatures (AIT) in air (e.g. in the range from 200 °C to 250 °C at 1 bar abs as shown by all alkanes and alkenes with more than 5 C-atoms) might exhibit substantially smaller values for R than found for the combustibles investigated so far, which all had AIT-values between 440 °C and 590 °C (always at 1 bar abs and in air). This hypothesis was based on the assumption that if a mixture M1 autoignites at a lower temperature than a mixture M2, the temperature at which the induction time for autoignition becomes small enough (less than about 10 µs) to allow for autoignition in the precompressed unreacted gas directly ahead of the reaction gases (this is the moment of DDT-occurrence) generated during the initial deflagrative stage of the explosion should also be lower for mixture M1 than for mixture M2. To explain why this hypothesis follows from the abovementioned assumption let us briefly recall the DDT-mechanism: The expanding reaction gases act like an ever faster accelerating piston (final acceleration of the order of 1000000 m/s2 and final speed relative to the pipe wall of the order of 1000 m/s to 1300 m/s) on the yet unreacted mixture ahead of the flame front and, consequently, generate a narrow zone (narrow in axial direction) of ever more compressed unreacted mixture directly ahead of the flame front, because the speed of the piston relative to this mixture is larger than the speed of sound in this mixture, i.e. the mixture gets shocked. Because the temperature in the precompressed mixture rises with the precompression ratio (ratio between initial pressure in the mixture at the moment of ignition and the instantaneous pressure in the precompressed mixture ahead of the flame front), the precompression ratio in mixture M1 at the moment of DDT-occurrence should be lower than in mixture M2, if the assumption from above were valid. Consequently, the ratio between pstat at the location of the DDT and pstat in the region of the stable detonation should be lower for mixture M1 than for mixture M2. To check the validity of this hypothesis the pstat-values of cyclohexane/O2/N2 mixtures were determined in long and short pipes. The AIT of cyclohexane (C6H12) in air at 1 bar abs is 260 °C (CHEMSAFE database). 2. Experimental procedure for determination of pstat-values of cyclohexane/O2/N2 mixtures Figure 1 exemplarily shows the flammable range of the ternary mixture C6H12/O2/N2 at 1 bar abs and 170 °C in the form of an explosion triangle. The yellow dashed-dotted stoichiometric line is based on the reaction equation C6H12 + 9 O2 → 6 CO2 + 6 H2O. The solid yellow line denotes the air line. At the intersection of both lines the concentrations are: C6H12 = 2.28 vol.-%, O2 = 20.52 vol.-%, N2 = 77.2 vol.-% (i.e, air (O2+N2) = 97.72 vol.-%). The explosion diagrams were measured by BASF in a spherical 20 l vessel. The black crosses and open circles in Figure 1 mark the mixtures which were investigated in long pipes. Several of these mixtures were measured both at 80 °C and 130 °C. All 6 mixtures tested in short pipes were stoichiometric and had O2 concentrations between 20.5 vol.-% and 29 vol.-% (these mixtures are not marked by symbols in Figure 1). The red dashed arrow marks the region where the DDT was immediately followed by the stage of stable detonation under omission of the stage of unstable detonation. The detonation tests were carried out in straight pipes made of stainless steel (DIN-code: 1.4541) with lengths between 9.2 m and 9.6 m. The formats were 48.3x2.6 (outer diameter [mm] x wall thickness [mm]) and 114.3x3.6. Piezoelectric pressure sensors (Manufacturer: PCB, Types: M112A05, 0-345 bar and 113B03, 0- 1035 bar) were mounted in the walls at distances of about 320 mm and 640 mm. The details of the experimental procedure, i.e. the mechanical characterization of the employed pipes and the carrying out of the actual detonation tests, was analogous to the tests presented and explained in detail in Schildberg (2013, 2014, 2015, 2016a). The bulging characteristics (i.e. the increase of the pipe diameter in dependence of the inner hydraulic pressure, which was continuously increased until rupture occurred) of the three different pipe melts used in the present tests can be found in Schildberg (2013, 2015). In all tests a displacement body was present inside the welding neck flange at the end of the pipe, i.e. opposite to the side where the ignition source was mounted. By doing so, the reflection of the shock front at the blinded pipe end occurred at a location where the wall still had the normal thickness and not the much larger thickness of the welding neck flange. Because the tests had to be conducted at elevated initial pressures in order to produce residual plastic deformations in the walls of the pipes employed, the tests had to be conducted at elevated initial temperatures. 1046 Figure 1: Explosion triangle of C6H12/O2/N2 at 1 bar abs and 170 °C. The lower and upper flammability limits (LFL and UFL) are: LFL in O2: 0.5 vol.-%, UFL in O2: 52.0 vol.-%, LFL in air: 0.5 vol.-%, UFL in air: 9.0 vol.-%, Limiting oxygen concentration: 8.0 vol.-%. In addition, the lower boarder of the explosive range at pinitial = 10 bar abs and 150 °C is shown. C6H12 concentrations larger than 30 vol.- % could not be investigated at 10 bar abs because autoignition occurred while injecting O2 into the test vessel. Otherwise the vapour pressure of cyclohexane would not have been sufficient to produce explosive C6H12/O2/N2-mixtures (vapor pressures of C6H12 at 20, 70, 80, 130 and 170 °C: 103, 724, 989, 3600, 8027 mbar, respectively). Figure 2 displays how the pipe was heated and insulated. Temperature was controlled by several thermocouples clamped onto the surface of the pipe. Mixtures of C6H12/O2/N2 were produced in a 100 l vessel, which was heated up to 130 °C, according to the partial pressure method. The vessel was first evacuated and then filled with nitrogen up to 1 bar abs. Then liquid cyclohexane was pumped into the vessel by means of a HPLC-pump while controlling the weight loss in the cyclohexane reservoir. The pressure rise inside the vessel due to immediate evaporation of the cyclohexane was in good agreement with the mass injected under the assumption of ideal gas behavior. Thereafter nitrogen was injected up to the required partial pressure. Then an assembly of 4 fans inside the vessel was started, thereby producing a closed loop flow around a baffle plate mounted in axial direction. With the fans running, oxygen was injected as last component up to the required partial pressure. Before the mixture was allowed to flow into the heated and evacuated test pipe, the test pipe was flooded with nitrogen up to a pressure about 2 bar less than the pressure inside the vessel. When mixing was completed, the high pressure valve between the vessel and the test pipe was opened (φi = 6 mm). Thereafter an opening (φi = 4 mm) in the blind flange at the end of the test pipe was opened to the off gas such that the combustible mixture from the vessel pushed out the nitrogen in test pipe by plug flow. When reaching the desired initial pressure for the test (at this moment the nitrogen inside the test pipe had been replaced at least two times by the combustible mixture), all valves were closed and the mixture was ignited. By this filling procedure the fast 0 10 20 30 40 50 60 70 80 90 100 Cyclohexane C6H12 [vol.-%] 0 10 20 30 40 50 60 70 80 90 100 0 10 20 30 40 50 60 70 80 90 100 explosive range at Tinitial = 170 °C and pinitial = 1 bar abs mixtures in region with blue background are not explosive course of the explosions inside the pipes: explosion undergoes a DDT explosion remains deflagrative at pinitial = 10 bar abs and 150 °C the explosive range extends down to this line 1047 expansion of the combustible mixture with concomitant cooling and potential condensation of cyclohexane as well as adiabatic compression effects at elbows or tees in the pipework (risk of autoignition) are avoided. Figure 2: Details of the heating and the insulation. Two electrical heating cables with same power per length were attached on both sides of the pipe over the entire length (white strips in the picture). Then the pipes were enclosed by a 4 cm thick thermal insulation. The green cable in the left photo is from a piezoelectric pressure transducer mounted in the wall. On the right photo the feed gas pipe and the air-operated high-pressure valve can be seen (both components were heated by electrical heating cables and were wrapped in aluminum foil). This valve separates the feed gas section from the test pipe. 3. Experimental results for pstat-values of cyclohexane/O2/N2-mixtures Table 1 summarizes the results of all tests. The following comments a) to h) provide some explanations to this table: a) No transitions to detonation did occur for stoichiometric C6H12/O2/N2 mixtures with O2-concentrations of 19 vol.-% (in the total mixture) and less. b) Correcting the static equivalent pressures for the influence of bulging was done as explained in Schildberg (2013), i.e. it was assumed that when the wall of a pipe starts to bulge outwards there is an immediate feedback on the value of the detonative pressure in the sense that this pressure will be reduced due to the increase in the cross section of the pipe. Hence, the resulting static equivalent pressure derived from such a test will also be less than it would have been if the test had been conducted using a pipe with a thicker wall which would have exhibited a smaller diameter increase. The corrected values were obtained from the measured values by multiplication with the relative increase of the cross section of the pipe. c) In the present tests bulging of the stainless steel pipes (all with DIN code 1.4541) occurred at temperatures between 70 °C and 130 °C. To account for the slight softening of the stainless steel at the elevated temperatures, the static equivalent pressures determined on the basis of the bulging characteristics at 20 °C (Schildberg 2013, 2015) were corrected by multiplying with the ratio between the yield strength (Rp0.2) at the elevated temperature and the yield strength at 20 °C. According to the material data sheets availabe for 1.4541 the minimum Rp0.2-values the manufactures have to guarantee change as follows for the temperatures of 20, 70, 80, 115 and 130 °C: 210, 190.6. 190.6, 180.6, 178 N/mm2. d) To reduce the experimental effort, the same pipes were sometimes used for several tests conducted successively with increasing O2-concentration (see column AM). By doing so, the DDT always occurred in a pipe section having not been exposed to a detonative load before, because when increasing the O2- concentration the predetonation distance drops. Because the focus of the experimental work was on the determination of the static equivalent pressure at the location of DDT occurrence, this goal could still be reached under clean experimental conditions. However, by this reduction of experimental effort the reflection at the blind flange of the pipe sometimes occurred in a section that had already been bulged slightly by the preceding test. 1048 Table 1: Compilation of the experimental conditions and results of all conducted detonation tests with C6H12/O2/N2 at different initial conditions. The lower table is the continuation of the upper table to the right side. To ease reference, the columns B and C are repeated in the lower table. Note that the tables use the German convention for decimal points. Abbreviation “bf.” in column AM stands for “blind flange”. Explanations: *: Calculated using GASEQ (2005) for the respective initial pressure, initial temperature and mixture composition (columns N, P, Q). **: the average decrease of the yield strength with increasing temperature was accounted for. At 70 °C and 80 °C the yield strength is about 90.8 % of the value at 20 °C, at 115 °C it is 86 % and at 130 °C it is 85 % (columns W, AA, AF). (no unst.): There was no region with unstable detonation, since the flame front overtook the initial pressure wave already before the DDT occurred. magenta text in column O for tests 19, 5, 22, 32 and 35: This is the speed of the unstable detonation, because for these tests there was no stable detonation as consequence of the pre-detonation distance being very long and the pipe length being only about twice the pre-detonation distance. blue text in columns T to X for tests 22, 32 and 35: The DDT is scenario 5 and not scenario 1 as for all other tests. The reason is the same as given above in context with the magenta-coloured text. blue text in columns AD to AH for tests 19, 5, 22, 32 and 35: The reflection is scenario 7 and not scenario 4 as for all other tests. The reason is the same as given above in context with the magenta-coloured text. B C D E F G H I J K L M N O P Q R T U V W X 1 2 num- ber of test pipe dimension: outer diameter [mm] x wall thickness [mm] number of the melt of the material use for manufac- turing the pipe inner pipe diame- ter [mm] length of pipe (effec- tive for the en- closed gas mix- ture) [mm] initial pressure of the gas mixture used in the test pinitial [bar abs] initial tempera- ture of the gas mixture used in the test [°C] chosen O2- concen- tration in O2/N2 mixture [vol.-%] stoich- iometric C6H12 concen- tration in total mixture [vol.-%] chosen C6H12- concen- tration in total mixture [vol.-%] resulting O2- concen- tration in total mixture [vol.-%] resulting N2- concen- tration in total mixture [vol.-%] calcula- ted* speed of sound in C6H12/O2/ N2-mixture at initial tempera- ture [m/s] mea- sured speed of stable deto- nation [m/s] calcu- lated* speed of stable deto- nation [m/s] calcu- lated* Chapman- Jouguet pressure ratio pCJ_r [ - ] calcu- lated* Chapman- Jouguet pressure pCJ = pinitial * pCJ_r [bar abs] measured strain caused by DDT [%] static equivalent pressure at DDT pstat_DDT_m (as derived from strain as measured) [bar abs] static equivalent pressure at DDT pstat_DDT_c (corrected for influence of bulging) [bar abs] static equivalent pressure at DDT pstat_DDT_c (corrected for elevated tem- perature**) [bar abs] ratio pstat_DDT_c/ pstat_stable [ - ] 3 4 1 48.3 x 2.6 584549 43,1 9260 20,00 70,00 21,00 2,280 2,280 20,521 77,199 356,00 1955 1855 16,65 333,04 27,5 630,00 1024,14 929,92 3,99 5 19 48.3 x 2.6 836880 43,1 9390 14,10 80,00 25,00 2,703 2,703 24,324 72,973 358,00 2069 1920 17,45 246,00 26,5 500,00 800,11 726,50 4,22 6 20 48.3 x 2.6 836880 43,1 9390 14,00 80,00 30,00 3,226 3,226 29,032 67,742 354,00 2045 1991 19,14 268,00 16,7 506,00 689,12 625,72 3,34 7 21 48.3 x 2.6 836880 43,1 9390 14,00 80,00 40,00 4,255 4,255 38,298 57,447 346,00 2121 2101 22,00 308,00 25,6 504,00 795,08 721,93 3,35 8 9 5 48.3 x 2.6 584549 43,1 9260 20,00 130,00 21,00 2,280 2,280 20,521 77,199 385,00 1994 1853 14,20 284,00 18,12 600,00 837,14 711,57 3,58 10 22 48.3 x 2.6 836880 43,1 9390 14,00 130,00 25,00 2,703 2,703 24,324 72,973 381,00 2032 1917 15,36 215,00 22,5 505,00 757,82 644,14 4,28 11 25 48.3 x 2.6 836880 43,1 9390 12,00 130,00 30,00 3,226 3,226 29,032 67,742 376,00 2010 1984 16,74 200,90 10,14 488,00 591,98 503,19 3,58 12 28 48.3 x 2.6 836880 43,1 9390 12,00 130,00 30,00 3,226 4,600 28,620 66,780 368,00 2095 2032 17,90 214,80 17,81 510,00 707,84 601,66 4,00 13 27 48.3 x 2.6 836880 43,1 9390 12,00 130,00 30,00 3,226 6,700 27,990 65,310 357,00 (no det.) 1921 16,43 197,20 14 23 48.3 x 2.6 836880 43,1 9390 12,00 130,00 40,00 4,255 4,250 38,300 57,450 368,00 2123 2091 19,17 230,00 13,66 500,00 645,93 549,04 3,41 15 29 48.3 x 2.6 836880 43,1 9390 17,00 130,00 50,00 5,263 5,263 47,369 47,369 360,00 2189 2188 21,59 367,00 1,1 380,00 388,41 330,15 1,29 16 26 48.3 x 2.6 836880 43,1 9390 17,00 130,00 60,00 6,250 6,250 56,250 37,500 353,00 2272 2261 23,71 403,00 1,95 420,00 436,54 371,06 1,32 17 24 48.3 x 2.6 836880 43,1 9390 12,00 115,00 60,00 6,250 6,250 56,250 37,500 347,00 2280 2251 24,33 292,00 0 18 19 30 114.3x3.6 878536 107,1 9570 3,17 130,00 21,00 2,280 2,280 20,521 77,199 385,00 (no det.) 1824 13,02 41,27 20 31 114.3x3.6 878536 107,1 9570 3,17 130,00 25,00 2,703 2,703 24,324 72,973 381,00 (no det.) 1888 14,11 44,73 21 32 114.3x3.6 878536 107,1 9570 3,07 130,00 30,00 3,226 3,226 29,032 67,742 376,00 1940 1951 16,19 49,69 3,59 242,00 259,69 220,73 6,35 22 33 114.3x3.6 878536 107,1 9600 2,80 130,00 27,00 2,913 2,913 26,213 70,874 379,00 (no det.) 1912 15,39 43,10 23 34 114.3x3.6 878536 107,1 9600 3,17 130,00 28,50 3,069 3,069 27,625 69,306 378,00 (no det.) 1934 15,83 50,18 24 35 114.3x3.6 878536 107,1 9600 3,17 130,00 30,00 3,226 3,226 29,032 67,742 376,00 2303 1952 16,20 51,35 1,84 221,00 229,21 194,83 5,42 no transition to detonation no transition to detonation, mixture too rich Experiments in short pipes, format 114.3 x 3.6, at 130 °C no transition to detonation no transition to detonation no transition to detonation Experiments in long pipes (exception is test 19, here reflection at pipe end is scenario 7, but DDT is scenario 1), format 48.3 x 2.6, tests at 70 °C and 80 °C Experiments in long pipes (exceptions are tests 5 and 22: reflections at pipe ends are scenario 7; for test 22 the DDT is scenario 5, for test 5 the DDT is scenario 1), format 48.3 x 2.6, most tests at 130 °C Data of pipe Data of C6H12/O2/N2 gas mixture Diameter increase at location of DDT B C Y Z AA AB AC AD AE AF AG AH AI AJ AK AL AM 1 Data of pipe pipe history 2 num- ber of test pipe dimension: outer diameter [mm] x wall thickness [mm] measured strain caused by stable detonation [%] static equivalent pressure of stable detonation pstat_stable_m (as derived from strain as measured) [bar abs] static equivalent pressure of stable detonation pstat_stable_m (corrected for elevated tem- perature**) [bar abs] static equivalent pressure of stable detonation pstat_stable_c (corrected for influence of bulging) [bar abs] ratio α = pstat_stable_c / pCJ [ - ] measured strain caused by reflection at blind flange [%] static equivalent pressure of reflected detonation pstat_ref lection_m (as derived from strain) [bar abs] static equivalent pressure of reflected detonation pstat_ref lection_m (corrected for elevated tem- perature**) [bar abs] static equivalent pressure of reflected detonation pstat_ref lection_c (corrected for influence of bulging) [bar abs] ratio pstat_ref lection_c / pstat_stable [ - ] location of DDT as inferred from pipe deforma- tion [mm] predeto- nation distance expres- sed as multiple of inner pipe diameter end of region of unstable deto- nation [mm] time interval until DDT occurred, counted from the instant of ignition [ms] 3 4 1 48.3 x 2.6 0 8,39 552,00 501,22 588,85 2,53 3740 86,8 8100 23,5 new pipe 5 19 48.3 x 2.6 15,5 505,00 458,54 611,70 3,55 4530 105,1 9390 33,5 new pipe 6 20 48.3 x 2.6 0 6,4 450,00 408,60 462,57 2,47 3060 71,0 6700 16,3 first 6 m reused from test 19, last 3 m new 7 21 48.3 x 2.6 0 11,6 490,00 444,92 489,47 2,27 1710 39,7 2200 26,0 from test 20, at bf. already 6.4% bulging 8 9 5 48.3 x 2.6 4750 110,2 9260 30,6 new pipe 10 22 48.3 x 2.6 6140 142,5 9390 36,0 new pipe 11 25 48.3 x 2.6 0 4,35 440,00 374,00 407,25 2,90 3430 79,6 7100 16,2 new pipe 12 28 48.3 x 2.6 0 2,9 423,00 359,55 380,71 2,53 2980 69,1 4300 12,1 from test 27, i.e. still without bulging 13 27 48.3 x 2.6 0,00 new pipe 14 23 48.3 x 2.6 0 2,69 418,00 355,30 374,67 2,33 1690 39,2 1800 4,7 first 6 m reused from test 22, last 3 m new 15 29 48.3 x 2.6 1,1 380,00 323,00 330,15 0,90 20,3 510,00 433,50 592,50 2,31 <310 <7,2 (no unst) < 1,2 from test 28, at bf. already 2.9% bulging 16 26 48.3 x 2.6 1,95 420,00 357,00 371,06 0,92 20,9 510,00 433,50 581,91 2,06 <310 <7,2 (no unst) <0,7 from test 25, at bf. already 4.35% bulging 17 24 48.3 x 2.6 0 6,4 455,00 391,30 420,08 2,06 <950 <22 (no unst) <1,2 from test 23, at bf. already 2.69% bulging 18 19 30 114.3x3.6 new pipe 20 31 114.3x3.6 from test 30 21 32 114.3x3.6 17,76 320,00 272,00 377,19 10,84 7910 73,9 9570 41,0 from test 31, i.e. still without bulging 22 33 114.3x3.6 new pipe 23 34 114.3x3.6 from test 33 24 35 114.3x3.6 11,9 285,00 242,25 303,34 8,44 7540 70,4 9600 38,5 from test 34, i.e. still without bulging Diameter increase by stable detonation Diameter increase upon reflection at blind flange Characteristic distances and times no stable detonation Experiments in short pipes, format 114.3 x 3.6, at 130 °C no stable detonation Experiments in long pipes (with few exceptions), format 48.3 x 2.6, tests at 70 °C and 80 °C no stable detonation Experiments in long pipes (with few exceptions), format 48.3 x 2.6, most tests at 130 °C no stable detonation rupture due to scenario 7 (reflection of unstable detonation) no stable detonation rupture due to scenario 7 (reflection of unstable detonation) 1049 In these cases, the correction for bulging at the blind flange was done in the same way as for new pipes, i.e. the static equivalent pressure, as taken from the strain/pressure diagrams determined in the hydraulic tests, was multiplied by the cross section of the pipe after the test divided by the cross section the already bulged pipe exhibited prior to the test. (Example: the value in field AG15 (592.50) is calculated by multiplying AF15 (433.50) by (100+AD15)2 / (100+AD12)2 ). e) In column AL, the duration between the moment of ignition and the occurrence of the DDT is specified. The experimental error in these values is in the order of ±1.5 ms, because the duration of the arc discharge used for igniting the mixture was 3 ms. Although values of 0.5 ms could be defined mathematically as difference between t = 0 and the time coordinate of the leading edge of the first pressure peak associated with a detonation, all values less than about 3 ms should be interpreted as indication that the detonative reaction was either directly triggered by the ignition source or that the initial deflagrative stage extended over a few centimeters only. f) In test no. 21 a very long value of 26 ms was found for the time interval until DDT occurrence. In this test a wire with a fivefold larger diameter was erroneously mounted between the two poles of the ignition source. This wire was only glowing but did not melt such that the arc discharge, which is a much more effective ignition source than a thin glowing wire, did not develop. A test with a correctly working ignition source and the same mixture composition and with almost the same initial pressure (test 23) yielded 4.7 ms as time interval, but the DDT occurred at about the same location. This means that the predetonation distance is obviously totally unaffected of these differences in the ignition process. g) All values measured for the propagation speed of the stable detonation agreed extremely well with the values calculated by GASEQ (2005). The propagation speed of the unstable detonation was on the average 7 % faster (140 m/s faster) than the value calculated for the stable detonation. Tests 32 and 35 were disregarded in this context, because in these tests the distance, over which the unstable detonation propagated, was too short for to derive a reliable value. h) Within the time slot available for carrying out the experiments in our concrete bunker we did not succeed in producing scenario 8, i.e. coalescence of DDT and reflection. All 4 tests we conducted with initial conditions chosen such that a slightly longer predetonation distance than found for tests 32 and 35 should result ended up in a purely deflagrative explosion. Figure 3 displays examples for plastic deformations generated by the detonation tests. Figure 4 provides examples for the bulging as function of axial position. References Schildberg H.P., J. Smeulers, G. Pape, 2013, Experimental determination of the static equivalent pressure of gas-phase detonations in pipes and comparison with numerical models, Proc. ASME. 55690; Volume 5: High-Pressure Technology, V005T05A020.July 14, 2013; doi: 10.1115/PVP2013-97677 Schildberg H.P., 2014, Experimental determination of the static equivalent pressure of detonative decomposi- tions of acetylene in long pipes and Chapman-Jouguet pressure ratio, Proc. ASME. 46025; Volume 5: High-Pressure Technology; V005T05A018.July 20, 2014; doi: 10.1115/PVP2014-28197 Schildberg H.P., 2015, Experimental Determination of the Static Equivalent Pressures of detonative Explosions of Stoichiometric H2/O2/N2-Mixtures in Long and Short pipes, Proc. ASME. 56987; Volume 5: High-Pressure Technology; V005T05A015.July 19, 2015; doi: 10.1115/PVP2015-45286 Schildberg H.P., 2016a, Experimental Determination of the Static Equivalent Pressures of Detonative Explo- sions of Stoichiometric CH4/O2/N2-Mixtures and of CH4/O2-Mixtures in Long Pipes, Proc. ASME. 50404; Volume 4: Fluid-Structure Interaction, V004T04A020.July 17,2016; doi: 10.1115/PVP2016-63223 Schildberg H.P., 2016b, Gas phase detonations in pipes: the 8 possible different pressure scenarios and their static equivalent pressures determined by the pipe wall deformation method. (part 1 / part 2), Chemical Engineering Transactions, 48, 241-246 / 247-252; DOI:10.3303/CET1648041 / 10.3303/CET1648042 Schildberg H.P., 2018, Experimental Determination of the Static Equivalent Pressures of Detonative Explosions of Stoichiometric C2H4/O2/N2-Mixtures and of C2H4/O2-Mixtures in Long Pipes, accepted for publication in Proceedings of the ASME 2018 Pressure Vessels and Piping Conference TRGS407, 2016, Technische Regel für Gefahrstoffe 407 (TRGS 407), Tätigkeiten mit Gasen – Gefährdungs- beurteilung, Gemeinsames Ministerialblatt Nr. 12-17 (26.04.2016), p. 328 – 364, ISSN 0939-4729. [Note: a) The TRGS 407 is published by the German Bundesministerium für Arbeit und Soziales (Federal Ministry for Work and Social Affairs; b) attachment A4, page 48 – 56 of TRGS 407, contains the detonation issues] GASEQ, 2005, is a chemical equilibrium program for windows. The code is freeware and can be downloaded from http://www.c.morley.dsl.pipex.com/ or http://www.gaseq.co.uk. 1050