Registration and assessment of load cycles in pipelines
An optimized load cycle registration and assessment method was developed on the basis of evaluation of the fatigue test series with arched test pieces. The tests were performed under Wohler-type and transient loading conditions. Test results were evaluated using different load cycle registration a...
Gespeichert in:
| Veröffentlicht in: | Проблемы прочности |
|---|---|
| Datum: | 2009 |
| Hauptverfasser: | , |
| Format: | Artikel |
| Sprache: | English |
| Veröffentlicht: |
Інститут проблем міцності ім. Г.С. Писаренко НАН України
2009
|
| Schlagworte: | |
| Online Zugang: | https://nasplib.isofts.kiev.ua/handle/123456789/48429 |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| Назва журналу: | Digital Library of Periodicals of National Academy of Sciences of Ukraine |
| Zitieren: | Registration and assessment of load cycles in pipelines / O. Torop, V. Schmidt // Проблемы прочности. — 2009. — № 5. — С.109-117. — Бібліогр.: 13 назв. — англ. |
Institution
Digital Library of Periodicals of National Academy of Sciences of Ukraine| id |
nasplib_isofts_kiev_ua-123456789-48429 |
|---|---|
| record_format |
dspace |
| spelling |
Torop, O. Schmidt, V. 2013-08-19T14:18:39Z 2013-08-19T14:18:39Z 2009 Registration and assessment of load cycles in pipelines / O. Torop, V. Schmidt // Проблемы прочности. — 2009. — № 5. — С.109-117. — Бібліогр.: 13 назв. — англ. 0556-171X https://nasplib.isofts.kiev.ua/handle/123456789/48429 539.4 An optimized load cycle registration and assessment method was developed on the basis of evaluation of the fatigue test series with arched test pieces. The tests were performed under Wohler-type and transient loading conditions. Test results were evaluated using different load cycle registration and assessment methods. The finite-element simulations were performed, in order to verify the systematical and random errors caused by the chosen test setup and of the test specimen geometry. Розроблено оптимізованнй метод реєстрації й оцінки циклів навантаження, який базується на аналізі результатів утомних випробувань зразків у вигляді арки. Випробування проводили за постійних (за типом побудови кривих утоми) та змінних амплітуд циклічних напружень. Виконано скінченноелементні розрахунки, що дозволяє оцінити рівень систематичних і випадкових похибок, зумовлених схемами утомних випробувань та геометрією зразків, що використовуються. Разработан оптимизированный метод регистрации и оценки циклов нагружения, основанный на анализе результатов усталостных испытаний образцов в виде арки. Испытания проводили при постоянных (по типу построения усталостных кривых) и переменных амплитудах циклических напряжений. Выполнены конечноэлементные расчеты, позволяющие оценить уровень систематических и случайных погрешностей, что обусловлено используемыми схемами усталостных испытаний и геометрией образцов. en Інститут проблем міцності ім. Г.С. Писаренко НАН України Проблемы прочности Научно-технический раздел Registration and assessment of load cycles in pipelines Регистрация и оценка циклов нагружения в трубопроводах Article published earlier |
| institution |
Digital Library of Periodicals of National Academy of Sciences of Ukraine |
| collection |
DSpace DC |
| title |
Registration and assessment of load cycles in pipelines |
| spellingShingle |
Registration and assessment of load cycles in pipelines Torop, O. Schmidt, V. Научно-технический раздел |
| title_short |
Registration and assessment of load cycles in pipelines |
| title_full |
Registration and assessment of load cycles in pipelines |
| title_fullStr |
Registration and assessment of load cycles in pipelines |
| title_full_unstemmed |
Registration and assessment of load cycles in pipelines |
| title_sort |
registration and assessment of load cycles in pipelines |
| author |
Torop, O. Schmidt, V. |
| author_facet |
Torop, O. Schmidt, V. |
| topic |
Научно-технический раздел |
| topic_facet |
Научно-технический раздел |
| publishDate |
2009 |
| language |
English |
| container_title |
Проблемы прочности |
| publisher |
Інститут проблем міцності ім. Г.С. Писаренко НАН України |
| format |
Article |
| title_alt |
Регистрация и оценка циклов нагружения в трубопроводах |
| description |
An optimized load cycle registration and assessment method was developed on the basis of evaluation of the fatigue test series with arched test pieces. The tests were performed under Wohler-type and transient loading conditions. Test results were evaluated using different load cycle registration and assessment methods. The finite-element simulations were performed, in order to verify the systematical and random errors caused by the chosen test setup and of the test specimen geometry.
Розроблено оптимізованнй метод реєстрації й оцінки циклів навантаження, який базується на аналізі результатів утомних випробувань зразків у вигляді арки. Випробування проводили за постійних (за типом побудови кривих утоми) та змінних амплітуд циклічних напружень. Виконано скінченноелементні розрахунки, що дозволяє оцінити рівень систематичних і випадкових похибок, зумовлених схемами утомних випробувань та геометрією зразків, що використовуються.
Разработан оптимизированный метод регистрации и оценки циклов нагружения, основанный на анализе результатов усталостных испытаний образцов в виде арки. Испытания проводили при постоянных (по типу построения усталостных кривых) и переменных амплитудах циклических напряжений. Выполнены конечноэлементные расчеты, позволяющие оценить уровень систематических и случайных погрешностей, что обусловлено используемыми схемами усталостных испытаний и геометрией образцов.
|
| issn |
0556-171X |
| url |
https://nasplib.isofts.kiev.ua/handle/123456789/48429 |
| citation_txt |
Registration and assessment of load cycles in pipelines / O. Torop, V. Schmidt // Проблемы прочности. — 2009. — № 5. — С.109-117. — Бібліогр.: 13 назв. — англ. |
| work_keys_str_mv |
AT toropo registrationandassessmentofloadcyclesinpipelines AT schmidtv registrationandassessmentofloadcyclesinpipelines AT toropo registraciâiocenkaciklovnagruženiâvtruboprovodah AT schmidtv registraciâiocenkaciklovnagruženiâvtruboprovodah |
| first_indexed |
2025-11-24T21:03:16Z |
| last_indexed |
2025-11-24T21:03:16Z |
| _version_ |
1850497186812395520 |
| fulltext |
UDC 539.4
Registration and Assessment of Load Cycles in Pipelines
O. Toropa and V. Schmidtb
a RWTH Aachen, Aachen, Germany
b TUV Rheinland Industrie Service GmbH, Cologne, Germany
УДК 539.4
Регистрация и оценка циклов нагружения в трубопроводах
О. Торопа, В. Шмидт6
а Университет Земли Рейи-Вестфалия, Аахен, Германия
6 ТЮФ Рейнланд Индaстри Сервис ГмбХ, Кельн, Германия
Р а з р а б о т а н о п т и м и з и р о в а н н ы й м е т о д р е г и с т р а ц и и и о ц е н к и ц и к л о в н а г р у ж е н и я , о с н о в а н н ы й
н а а н а л и з е р е з у л ь т а т о в у с т а л о с т н ы х и с п ы т а н и й о б р а з ц о в в в и д е а р к и . И с п ы т а н и я п р о в о д и л и
п р и п о с т о я н н ы х ( п о т и п у п о с т р о е н и я у с т а л о с т н ы х к р и в ы х ) и п е р е м е н н ы х а м п л и т у д а х
ц и к л и ч е с к и х н а п р я ж е н и й . В ы п о л н е н ы к о н е ч н о э л е м е н т н ы е р а с ч е т ы , п о з в о л я ю щ и е о ц е н и т ь
у р о в е н ь с и с т е м а т и ч е с к и х и с л у ч а й н ы х п о г р е ш н о с т е й , ч т о о б у с л о в л е н о и с п о л ь з у е м ы м и с х е м а
м и у с т а л о с т н ы х и с п ы т а н и й и г е о м е т р и е й о б р а з ц о в .
К л ю ч е в ы е с л о в а : усталостные испытания, цикл нагружения, циклические
напряжения, конечноэлементный расчет.
Introduction . For the fatigue strength calculation o f existing defects the
know ledge o f the dynamic stress o f internal pressure is required. The fault
estimation methods used during calculation are m ostly based on tests in w hich the
failure occurred under W ohler loading conditions [1 -3 ]. Therefore it w as necessary
to convert the occurring quasi-stochastic dynamic internal pressure loading into the
W ohler one with equivalent damage impact. For this sim ple and conservative load
cycle counting methods were used previously [4-7].
In som e cases, the conservative approach results in disproportional high
increase in maintenance costs. The reason is that more and more faults are detected
with increasing operation period and that the integral dynamic damage increases
since startup. Therefore, there is a necessity to replace the more conservative load
cycle registration and evaluation methods w ith advanced methods w hich express
the actual damage due to dynamic internal pressure loading. Since the degree o f
conservativeness o f single load cycle registration and evaluation methods depends
on the component form, the type o f loading and the form o f loading, the planned
realization o f measurements was secured by experimental examination. The test
schem e was planned and executed by T U V Rheinland in collaboration with the
laboratory o f Lightweight D esign o f RWTH Aachen.
© O. TOROP, V. SCHMIDT, 2009
ISSN 0556-171X. Проблемы прочности, 2009, № 5 109
O. Torop and V. Schmidt
E xperim ental Set-U p. The basic material for manufacturing o f specimens
w as taken in the form o f five half-shell segm ents o f longitudinally w elded pipes.
Standard shape geometry for all test specimens (Fig. 1) was defined in collaboration
w ith TUV. Referring to the primary test results, further processing was provided for
the w eld region.
Fig. 1. Specimen geometry.
Comparability o f the test series w ith specim en from different pipes was
achieved by artificially inserting o f notch-like defects, w hich ensured identical
defect geometry.
Finally, five specim en forms were developed:
1. Unprocessed in the w eld seam region specimen.
2. Rem oval o f the w eld reinforcement from both w eld passes, m illed flat
transverse notch at the w eld pass inside, centered grooving on the w eld pass inside.
3. Rem oval o f the w eld reinforcement from both w eld passes, m illed flat
transverse notches at both w eld passes, centered grooving inside and outside.
4. Rem oval o f the w eld reinforcement from both w eld passes, centered recess
(reduction on the outer w eld pass diameter).
5. Rem oval o f the w eld reinforcement from both w eld passes, symmetrical flat
gauges in the w eld region.
On the basis o f the shape geom etries defined for all specim ens, test equipment
(Fig. 2) was designed to induce cyclic circumferential tension stresses initiated by
com pressive forces on the pipe segm ent specimen.
The upper region o f the testing equipment contained the load cell and the
hydraulic cylinder. In the lower part, the specim en carrier w as assembled. The
force transmission was carried out by a pressure rod w hich transfered the force to
the specim en through two flexible pressure stamps as an equivalent to the internal
pressure load in pipelines.
The strain gauges were applied to both sides o f the outer w eld passes for the
collection o f circumferential strains in the region o f tension pulsating stress for all
specimens.
110 ISSN 0556-171X. npo6n.eubi npounocmu, 2009, N 5
Registration and Assessment o f Load Cycles
Lead bands with different thicknesses and forms were introduced in order to
level out the irregularities o f inner surfaces o f the specimens.
The testing plan regarding the loading level, kind o f loading, loading duration
and specim en geom etry was established successively on the basis o f results from
the former tests. The initial situation w as to load the specim ens by uniform tension
cyclic forces or W ohler loading. The maximum and minimum stresses o f the
uniform loading, as w ell as loading duration were different for each specimen form.
Based on the characteristic pipeline pressure curves, T U V generated the
equivalent artificial loading, or transient loading, w hich also was introduced in the
specim ens as strain-controlled loading [9-12].
The experimental data were handed in by the Institute o f Lightweight D esign
RWTH Aachen in form o f data tables, w hich contained data for strain gauges,
measured force and position.
E valuation o f E xperim ental R esults. In the first step, during a load cycle
collection the preliminary data processing was performed in order to collect load
alterations. This was done by means o f different counting methods:
(i) range counting method;
(ii) level crossing method [13];
(iii) Rainflow method.
In the next step, during the load cycle evaluation the registered load alterations
with different oscillation amplitudes were converted into uniform load alterations.
According to the theory on w hich the calculation w as based, the uniform load
alterations cause the same fatigue o f a structural component as the actual load
alterations with different oscillation amplitudes.
The damage effects o f individual load cycles with different amplitudes were
determined by means o f the Palm gren-M iner rule in combination with different
m ean stress correction methods (Gerber, Goodman).
ISSN 0556-171X. npoôëeubi npounocmu, 2009, N 5 111
O. Torop and V. Schmidt
The results o f the W ohler tests with unmachined w eld seam and 2-notch
specim en are shown in Fig. 3. Here the stress range normalized with respect to
tensile strength (loading rate KB) versus the number o f cycles to failure is shown.
Both axes have logarithmic scales. The W ohler lines for SAW- and seam less-pipes
that correspond to DIN 2413 [8] are additionally shown in normalized form. They
were used as a reference during life tim e analysis for pipes w hich m eet the
acceptance specification. The right part o f diagrams contains the experimental
results for the unmachined w eld seam in form o f colored circles. Other points that
are located low er and depicted in form o f colored triangles represent the
experim ental results w hich are corrected with respect to the bending stresses.
Fig. 3. Standardized fatigue strength: Wohler tests on the unmachined weld seam and 2-notch
specimen.
It appeared that all experimental results for unmachined specim ens are located
approximately in the region o f the W ohler line for seam less pipes.
Even considering a very small segm ent on the lower border o f the failure
probability distribution (99.9% probability o f life time extension shown by dashed
line) the calculation basis appears to be improved by a factor o f 4 in comparison to
W ohler lines given in DIN 2413 for SAW -pipes because o f the logarithmic scale.
The knowledge about the optimal load cycle counting and analysis methods
for transient loadings was obtained from the tests on identical 2-notch specim ens
subjected to different loadings. The W ohler tests with 2-notch specim en are
depicted in Fig. 4. The stress range standardized with respect to the tensile strength
(KB) is displayed versus the number o f cycles to failure. Additionally, Fig. 4
contains the results o f the transient tests. The transient test results are plotted with
three different sym bol types.
112 ISSN 0556-171X. npo6n.eubi npounocmu, 2009, N 5
Registration and Assessment o f Load Cycles
Fig. 4. Standardized cyclic strength: Wohler and transient tests on specimen with 2-notch geometry
(hinged stamp adjustment): (O) evaluation with range counting method, (+ ) evaluation with level
crossing method, (A) evaluation with Rainflow method.
The diagram shows that w e get the low est damage by using the analysis with
the range counting method and the highest w ith the level crossing method. The
experimental results evaluated with the Rainflow method show the same number o f
cycles to failure as the range counting method. On the other hand, the analysis with
the level crossing method results in known reduction o f load cycle numbers with a
simultaneous increase in amplitude.
Figure 5 shows the experimental results o f all 2-notch specim ens in a
standardized form separately for different analysis methods applied. The W ohler
tests on 2-notch specim en are shown as reference in the top lines. The result o f the
specim en 3-1 lies a little above the other results although this value cannot be
excluded as an outlier. Together w ith the data points the mean value o f all data are
plotted in form o f crosses.
The results for the range counting depicted in the low est “lin e 1” clearly show
that this approach generally predicts too small damage and therefore this method is
unsafe. The results o f transients evaluated with the level crossing method are
presented in the “lin e 2 .” They are located above the results o f the corresponding
W ohler tests in “lin e 4 ” and show the conservativeness o f the method w hich was
theoretically expected. The mean values o f the W ohler tests without specim en 3-1
yields on average a 15% more conservative damage prognosis for the transient
loads used for the experiment.
The results for all transients evaluated with the Rainflow method are presented
in “lin e 3 .” The m ean value o f the experimental points is congruent with the mean
value for the W ohler tests, especially without the specim en 3-1. Therefore, the
rainflow method com es out as the m ost optimal.
ISSN 0556-171X. npoôëeubi npounocmu, 2009, N 5 113
O. Torop and V. Schmidt
Fig. 5. Standardized Wohler loading (classificated by load acquisition method): Wohler and transient
tests on specimen with 2-notch geometry.
N um erica l Im plem entation . The experiments described above were not
carried out under optimal conditions. After the event it turned out that the ovality
and curvatures of the specim ens were not measured. To evaluate the influence of
these factors and to estimate the errors a finite elements (FE) analysis was
performed by means o f A N SY S 10.0. Additionally the influence o f bending
stresses on the evaluation o f a life time was derived.
In the first step, the sim ulation m odel o f the unmachined w eld seam specimen
for the ideal state was developed. The 3-D v iew o f a simulated geom etry is shown
in Fig. 6.
The follow ing load case was designed:
1. The 80000 kN force w as distributed on two lines parallel to the w eld seam
in the middle o f the upper surface o f each stamp (54 nodes loaded with 1481,5 kN
each).
2. Fixation o f the fixation blocks was applied in joint region on lines in all 3
translational directions. O nly rotation around Z axis is allow ed as depicted in
Fig. 6.
3. Contact was defined between stamps and lead bands.
For the developed m odel, elastic strains in X direction and von M ises
equivalent stresses were evaluated. The evaluation results are shown in Fig. 7. As
expected, the highest stresses occurred close to the middle o f the specimen. In the
center o f the w eld seam region the stresses were smaller because o f the different
w all thicknesses. The change o f both thickness and the Young modulus on the
114 ISSN 0556-171X. npo6n.eubi npounocmu, 2009, N 5
Registration and Assessment o f Load Cycles
border between the w eld region and the specim en caused an offset in strains. For
the geom etry transition w e obtained a high offset in strains at the corner where the
pipe was connected to the fixation block, denoted as M X in Fig. 7.
Fig. 6. Meshed 3-D model with applied loading.
Fig. 7. Distribution of strains in X direction and von Mises equivalent stresses.
Additional simulations showed that the b iggest displacements occurred near
the center o f the seam w eld in the region o f force transmission. For the m odel with
the pipe radius smaller than stamp radius the contact between the stamp and the
lead interlayer occurred in 2 lines exactly on the borders o f the power piston. This
was the reason w hy the load transmission w ent through these regions and caused
high stresses.
ISSN 0556-171X. npoôëeubi npounocmu, 2009, N 5 115
O. Torop and V. Schmidt
For the simulation m odel w ith pipe radius bigger than stamp radius, the stamp
w as in contact w ith the lead interlayer in only one line in the middle causing big
stresses in the region o f force transmission.
Additionally, a specim en with double-notched w eld seam w as simulated. It
appeared that in the notch region a typical notch stress distribution occurred with
the highest offset o f stresses at the tip o f the notch. The notches were added
specifically to allow the breakage in the middle o f the specimen.
In order to investigate the influence o f the pipe thickness on the experiment, a
specim en with different thicknesses on the right and left side o f the w eld was
modeled. A s expected, the stress distribution w as not symmetric.
C onclusions. A n optim ized load cycle registration and assessm ent method
w hich enables an improved fitness-for-service assessm ent and yields a higher
residual service life was developed on the basis o f fatigue test series w ith arched
test pieces. The results o f these tests lead to the follow ing improvements:
as the experimental results for unmachined specim ens lie approximately in the
region o f the W ohler line for seam less pipes the calculation basis was improved by
a factor o f 4 in comparison to W ohler lines given in DIN 2413 for SAW-pipes;
usage o f the 2-notch geom etry in the specim en form reduced the experimental
running time approximately ten times.
The results for range counting showed that this approach generally predicts
too small damage and is to be considered as unsafe. The results o f level crossing
counting were located above the results o f the corresponding W ohler tests and
show ed the conservativeness o f the method. For the R ainflow counting the mean
value o f the experimental points was almost congruent with the mean value for the
W ohler tests. Therefore, the R ainflow method is an optimal analysis method since
it considers the fatigue strength o f defects in m ost accurate way.
Additional FE-simulations enabled the verification o f systematical errors
caused by the chosen test setup and the geom etry o f the test specim en. They also
helped to estimate the random errors caused by irregular geom etries o f the test
specimens, such as misalignment at seams and non circularity. Due to FE simulation
results w e can com e to the follow ing conclusions:
The fact that every specim en has individual curvature affects the experimental
results. This leads to the mismatch between the stamp curvature and specimen
curvature resulting in different bending stress distribution for each specimen. For
exam ple, i f the stamp radius is bigger than the specim en radius then the measured
bending strains in DM S position is smaller than the actual value. This leads to the
unsafe life time estimation.
A s all specim ens had different thicknesses the influence o f thickness w as also
estimated by means o f FE simulations. Results obtained showed that for smaller
thickness value w e get higher values o f stresses and strains in X direction.
Р е з ю м е
Розроблено оптнмізованнй метод реєстрації й оцінки циклів навантаження,
який базується на аналізі результатів утомних випробувань зразків у вигляді
арки. Випробування проводили за постійних (за типом побудови кривих
утоми) та змінних амплітуд циклічних напружень. Виконано скінченноеле-
116 ISSN 0556-171X. Проблемы прочности, 2009, № 5
Registration and Assessment o f Load Cycles
ментні розрахунки, що дозволяє оцінити рівень систематичних і випадкових
похибок, зумовлених схемами утомних випробувань та геометрією зразків,
що використовуються.
1. S. J. Maddox, F atigu e Strength o f W elded S tructures, Second edition, Abington
Publishing (1991).
2. T. P. H. W irsching, R an dom V ibrations, T heory a n d P ra c tic e , A W iley-
Interscience Publication, John W iley and Sons, Inc. (1995).
3. D IN 4 5 6 6 7 (1969).
4. K. T. Endo and H. Nakagawa, “Fatigue o f metals subjected to varying
stress-prediction o f fatigue lives,” in: P relim in ary P ro ceed in g s o f the Chugoku-
Shikoku D is tr ic t M e e tin g [in Japanese], The Japan Society o f M echanical
Engineers, (1967), pp. 4 1 -4 4 .
5. J. B. de Jonge, “The analysis o f load-time-histories by means o f counting
methods,” National Aerospace Laboratory NRL, MP 82039 U , ICAF Document
(1982).
6. N . E. D ow ling, “Fatigue predictions for com plicated stress-strain histories,” J.
M a ter ., 7, 7 1 -8 7 (1972).
7. I. Rychlik, S ta tis tic a l W ave A n a ly s is w ith A p p lica tio n to F a tigu e , Ph.D.
Thesis, Department o f Mathematical Statistics, Lund University (1986).
8. D IN 241 3 , B erech n u n g d e r W an ddicke von S tah lroh ren g eg en Innendruck
(1993).
9. A D -M erk b la tt S2, B erech n u n g a u f W ech selbean spru ch u n g (1998).
10. E N 1 0 2 0 8 -2 , S te e l P ip e s f o r P ip e lin es f o r C o m b u stib le F lu id s - T echn ica l
D e liv e ry C on dition s, Part 2: P ip e s o f R eq u irem en t C la ss B , June 1996,
A N SY S documentation, 2007.
11. E. Haibach, B etrieb sfestig k e it. Verfahren u n d D a ten zu r B au te ilberech n u n g ,
Springer-Verlag (2002).
12. M. Siebel, E. Stieler, U n gleich förm ige S pan n u n gsverte ilu n g b e i sch w in g en d er
B ean spruchu ngen , Volum e 7 (1968).
13. F K M -R ich tlin ie , R ech n erisch er F estig k e itsn a ch w e is f ü r M asch in en bau te ile
au s Stahl, E isen g u ß - u n d A lu m in iu m w erk sto ffen ,V D M A -V erla g (2002).
Received 05. 01. 2009
ISSN G556-Î7ÎX. Проблемы прочности, 2GG9, № 5 l l ?
|