ipe to Pipe Impact Tests
Laurence Brunel
(CEA-CEN Cadarache St. Pau le-Durance, France
Alain Stromboni, Jean Lallement
CEAMCEN, Ponteny-ausoes, France
anstRact
[A study concerning the effects of pipe whip and pipe to pipe sapact in
representative operating ‘conditions of" a graphite gav-cooled reactor heat
‘xchanger as been performed by the French Commissariat. i l'Energie.Atoaiaue.
‘The following tests wore done : pipe whip of 4" ANB grade B schedule XO steel
land impact on 2" or 4 ANG grade B schedule NO target pipe which is aigply or
Figidly supported. During the test pressure, local strain, acceleration,
Sisplacement are recorded (2 high speed amcras operating. at 5000 frames pei
‘ecend)-
Measurements on the whipping phase are compared to thane pradiatad hy Ayame
Glastoplastic caloulations “using the finite element code Tedel and m Rood
Finally, results are discussed in terns of erushing and bending depending on the
‘support’ conditions, in relation to the results from hattelle Pacific Northwest
Taboratory impact teste in PH conditions.
Xt is also underlined that one of the results is in eontradietion with the rule
from the Standard Review Plan ae an inpacted pipe of smaller nominal size than
the swinging pipe (same schedule) presents no oreck and no break after the test
in spite of an important deformation.
‘mernooucrrou
In the framework of safety studies on French graphite ga2-oooled reactor, an
‘experimental program has been performed by the Commissariat i l'Energie Atowique
Torevaiuate the potential consequences of pipe whip and pipe to pipe impact
‘resulting from a pipe rupture. First the results of the tests will be presented,
Then @ comparison drawn with pipe whip calculations performed with finite
clement ‘code. Finally the results obtained will be discussed in relation to
‘ther experimental resulte published frou similar programa as well” aa to” the
Standard Review Plan criteria.
[TEST PROGRAM DESCRIPTION
im of the program
{In the French graphite grs-cooled reactor heat exchanger pipes conveying water,
‘team or. oarton dioxide gaa are numerous and. close togethers Under the
hypothesis of @ guillotine break occuring on one of these pipes, the ruptured
spipe may whip under the force applied on it by the escaping pressurized fluid.
tnadjacent. parallel pipe could be Smpacted and a second break may occur
‘consequently.
‘The foal of this program 1s to perform tests in representative operating
conditions to deteraine what kind of" damgo can bo expected in auch a situation:
‘The pipe impact toot facitity and operating conditions
‘Te experimental program 1s performed on the Aquitaine 11 fectiity. mia
facility hao been “designed for mximm operating pressure and tenperature of
V7.2 Wa and 340 *C respectively and has already boon ued for aiailar programs
(pipe impact ‘on rigid restraints and concrete siab [1] [2]). Ao. show In
Figure 1'the whipping pipe test section is oriented horizontally. It ix titled
with hot pressurized water and connected tothe reservoir by 2 recirculating
Line. "The! target pipe is horizontal too but oriented perpendicularly to. the
\inipping pipe. This situation represents the worst case that can be encountered
between two adjacent pipes. Depending on the test the targst pipe ts Filled with
pressurized cold water or ena.
‘The whipping pipe rupture is initiated by an explosive device which induces
elreurerentiat pipe break in leas than one aillisecond. The forces occuring at
the pipe elbow ‘at this tine mes Uhe pipe whip. the duration of the foot
(nipping and’ inpact) is such that tho jet force oan be considered as constant
during the test (pressure level at saturation conditions) except for the, very
Tirat overpressure gonerated by the explosive device (figure 2):
‘eat conditions
‘wo Kinds of tests were performed depending on target pipe support. conditions.
‘The first two teote were carried out on target pipes clamped at the ends end the
Enind na on a simply mipported pipe, with sapase at aid-span in sli encoes
‘Tho whlpping pipes were in all the tests ™ inchen ANB grade B seheduie 40 carbon
steel pipes: The target pipe wore, for tosts 1 and 3, 2 inches ANS grade, B
Schedule AO carton steel and’ for est 2 the same ne, the whipping pipe:
‘Te test conditions were chosen to be aa close as possible to the operating
conditions, but the asin paramsters were considered to be the Kinetic energy at
impact and a sufficient gap botweon the two pipes (nore thant meter) ‘The
conditions tests are sumarized in table 1. They wore supposed to lead, for all
the Lests, to a Kinetic energy at. tnpact of AO kd with 430 4M thrust force on
tthe whipping pipe ‘supposed to be constant. ‘The target pipes were not. heated
because OF thelr relatively low operating temperature (90 to 100 *C for water
pipes and loss than’ 400 °C for steam pipes) which can be annessed not to
{influence their rupture strength and even to improve their inpact strength.
Instrumentation
‘Te measurements performed during the tests were =
pressure (both pipes)
= tesperature (reservoir fluid, whipping pipe steel)
= Yooatstrain (strain gage on'target. pipe)
= aeecteration (both pipes and the ground for the Inst test)
= ispiacement (one. specific. transducer on target pipe plus two high speed
‘cancras). The Tiret tamera provided a goneral view of the facility during the
est and’ the” second one a ‘close shot of the iapact. Hoth were operating et
about "5000 ‘ranes per second.‘Test RESULTS
‘Te main results of the tests are sumarized in table 2. Ae whipping pipe
Operating conditions remin the same for ail the taste” the pipe wclocity. at
Ifpact Lo ‘constant (39 /omesoured by displacesent transducer). The angle
‘between he whipping pipe and a vertical axis 12 8
In Tact, the real paraneter of the tests appear to be the target pipe stiftness.
‘Te results confirm those of Alsheimer from Battelle Pacific National Labo=
ratory [3]: when the target pipe is rigidly supported even if st is
Significantay crushed doring the vest the whipping pipe ia deforecd much nore
(est 1 compared to, test 4). Indeed, in test 1, the tarest pipe diameter
reduction (defined aa the ratio of the mximm reduction in the outside diamcter
of ‘the ‘pipe to tho initial outside dianeter) is 25 4 against A? 4 forthe
‘winging pipe and in eat 4 these valuea become reapoctively 39 % and 8 $. These
Fesults in diamcter chango mist bo assooiated with thase of bend angles (defined
5 tho angle througn which the pipe was bent by the impact at ‘the point of
inmpact)."tere the target pipe bend angles are respectively for tests | and 3,
430 and 52", showing the Comparative ieportance of the two phases crushing and
ionding. ‘Tia point will be further developed.
Figures 3, % and 5 show respectively the deformation of the throe target pipes
land figures 6, 7, 8 those of the three swinging pipes. Nevertheless the’ main
Fesults of the teats remain that even if" great strain have been reached
{Drobabiy muon more than 15 f, the mime that could be registered) no break
‘and even no erack has boon detected for any test (Figure 9)
YOVERICAL ANALYSIS AND COMPARISON TO THE TEST RESULTS
An analyaia of the whipping ph
performed with the finive 0
identical for all the tests, has teen
font computer code Tedel (omer CEA)
‘mese calculations are based on beam formulation and Tedel enables, with dynanie
Glastoplastic evaluation, the atuiy of non-linear ange deflection of pipes.
‘The code allows calculations with a following load which ia novessary becaioe of
the large swinging pipe rotation. The input paranators for the code were, except
‘the geonetricel asta!
= Tho thrust force, P which was supposed to bo voll known. Soae tests had been
Epeciaily devoted to the measureneats of this forces in n previou experi
‘ental program {2} which alloved us to determine the coefficient K inthe
Fs kP,S.
5°: breach area
the value K = 2.27 was obtained in the ease of suboooled water. Tis leads for
the Grperisentai value to P= 30 KN
= Tho Initial gap betvoon evinging and target pipe + 1.50 m between the two
‘te possible points of comparison between calculations and experimental tests
iMine whipping time between the first explosion and the impact
= the shipping pipe velocity
= the “deformation of the Swinging pipe ata given time and its comparison
‘ith deformations measured on the high speed camera fils
‘Table 3" sumarizos the main pointe of comparison between aleulations and
experimental reauite and a compurivon of wilpping Pipe deforeation at different
Stops can be seen on Figures 10 and 11.
o