Special print from “3R international 40 (2001) Special Plastics Pipes”, pages 4-12

Minimum service-life of buried polyethylene pipes

without sand-embedding
Dr.-Ing. Joachim Hessel Hessel Ingenieurtechnik GmbH, Roetgen

VULKAN-VERLAG · HUYSSENALLEE 52-56 · 45128 ESSEN

2
Minimum service-life of buried polyethylene pipes
without sand-embedding
In the past 20 years polyethylene raw ma- Introduction
terials have been developed with the aim Dr.-Ing. Joachim Hessel
In the past the service-life of buried pipes Hessel Ingenieurtechnik GmbH,
of improving the creep rupture strength. has been shown to be reduced due to Roetgen (D)
This has raised the question whether it is stress concentrations caused by external
possible to eliminate sand-embedding of point loads. The service-life of the pipes
buried pipes. is closely related to the material‘s resis-
Investigations were performed in order to tance against slow crack growth. Poly-
determine the minimum quality of poly- ethylene resins are continuously being Tel. +49-2471/92022-0
ethylene pipes for this situation. The developed with the aim of increasing E-Mail: joachim.hessel@
testing conditions cover the loads which creep rupture strength. This increase can hessel-ingtech.de
are to be expected in service. Typical fai- be seen in the creep rupture curves of
lure histories, the test design and the ma- various standards between 1976 and
terial related test conditions in order to 2000 (e.g. DIN 8075). However, in reality
generate a conservative result are descri- many of the current resins have creep 1046 requires that the trench base must
bed. The test results of pipes under inter- rupture values significantly higher than be free from stones and rocks.
nal pressure and additional external point the minimum requirements.
load are reported over a range of test pa- Typical failures of pressure pipes caused
This raises the question, whether the by external, locally concentrated loads
rameters. creep rupture strength of current resins are shown in fig. 1 and 2. The additional
A strong correlation between the long has reached a state of development load shown in figure 1 was generated by
term performance of point loaded pres- where a failure of pressure pipes with ad- a piece of rock that continuously pressed
sure pipes and the resistance of the ma- ditional external point loads should not against the outside of the pipe. The point
terials against slow crack growth as de- be expected during their service-life. of indentation is visible near the crack in
termined in Full Notch Creep tests The following companies took part in the external pipe wall. The cause for the
(FNCT) is confirmed. The results from the these investigations: crack shown in figure 2 was probably a
FNCT are used to predict the minimum
 ATOFINA, Feluy [1], [2] small load on a limited area. The fracture
service-life of polyethylene pipes without
 BASELL, Frankfurt [3]
surface of this crack is shown in fig 3.
sand-embedding. The calculation is ba-
sed on the conservative results of each  BOREALIS, Site Austria [4] The starting point of the crack is located
test series. at the inside of the pipe wall (bottom) and
 EGEPLAST, Emsdetten [5]
travels to the outside (top). The fracture
Four raw material suppliers and three  SOLVAY, Brüssel [6] surface shows very low deformations
pipe producers were involved in these in-
 WAVIN, Twist [7] which is typical for creep rupture failure.
vestigations.
 WIRSBO, Maintal-Bischofheim The creep rupture failure is caused both
[8] by the internal pressure and the additio-
 HESSEL Ingenieurtechnik, nal load generated by an object pressing
Roetgen against the external pipe wall.
Such failures would not occur if the creep
Service Experience rupture strength at the inside of the pipe
wall is above the overall local stress at
Buried polyethylene pipes can be
damaged by point or linear loads this point.
which are acting in addition to
the design loads (e.g. internal Test Materials
pressure, earth and traffic load). Polyethylene pipes 110 x 10 mm (SDR
Linear loads can for example be 11) classified as PE 63, PE 80, PE 100
produced by wood used
as supports below the
Fig. 1: Indentation and associated crack path on the pipe during installation
outside wall of a polyethylene pipe
and not removed before
the trench was backfilled
[9].
Relevant standards do-
cument that a reduction
of service-life can result
from external point or li-
near loads on pipes. For
this reason, DIN 4033 and
DIN 19630 require that
point or linear loads on
the external wall of the
pipe are to be avoided
Fig. 2: Externally loaded area and associated crack during installation. For Fig. 3: Fracture surface in the wall of an externally
path on the outside wall of a polyethylene pipe the same reason, DIN loaded polyethylene pipe

3
Table 1: Pipe samples tested Principle and Limiting Conditions stress of 4 N/mm2 ± 0.03 N/mm2 relating
of the Tests to the remaining unnotched cross-sec-
tions. The nominal test temperature was
FNCT
80 °C ± 0.2 K.
The resistance against slow crack growth
The tensile creep tests were performed
of the samples was tested using the Full
on 3 test specimens per temperature
Notch Creep Test (FNCT) according to
using an aqueous solutions of ®ARKO-
DIN EN 12814-3 Annex A. This test me-
PAL N-100 in demineralised water
thod is also described in supplementary
(2/100, w/w) to accelerate the tests.
sheet 2 of guideline DVS 2203 part 4 and
in ISO-draft: ISO/DIS16770: Plastics – The FNCT failure times for 13 out of the
Determination of environmental stress 19 pipes are shown in fig. 4.
cracking (ESC) of polyethylene (PE) –
Full–notch creep
test (FNCT).
The test speci-
mens were cut
from the pipes in
the axial direction
with parallel sides
and approximately
square cross-sec-
tions. Each speci-
men was notched
perpendicular to
1 s.a. Table 2, 2 ISO/TR 9080, 3 ISO 527 (Tensile load)
4
the parallel length
ISO 178 (Bending load)
in the middle of
the test specimen.
according to DIN 8075 as well as cross- The specimens
linked polyethylene pipes (PE-Xa; PE-Xb) were loaded by a Fig. 4: Creep rupture times of specimens from pipes of different
were used for the investigations (table 1). constant tensile resistances against slow crack growth in FNCT

Table 2: Test program

External Point Load
The maximum stress that the
pipe material will experience
from a point load is the yield
stress. Therefore in this test it
was ensured that the displa-
cement of the point load into
the pipe wall was sufficient to
cause yielding of the material
at the inside of the pipe.
Since the additional stress in
the pipe wall far from the
point of load will be zero all
possible stresses that might
occur in the field due to a
point load are represented in
this test.
There are two scenarios
which are not covered: 1) the
penetration of a sharp object
– e.g. a nail – through the pipe
wall and 2) the complete
crushing of the pipe, e.g. by a
large rock. In the last case the
pipe is no longer functioning,
but the force on the pipe is
comparable to the test load in
the point loading test.
The required surface elon-
gation at the inner pipe wall
(i.e. the above yield elongati-
on) was produced by the dis-
placement of a tool along the
radius of the pipe with a tool

4
 The internal pressure of the pipes was
chosen to produce a hoop stress of 4
N/mm2 during the test (ISO 1167).
To accelerate the tests they were perfor-
med using an aqueous solution of 2 %
Arkopal N-100 and demineralised water.
The test fluid was continuously mixed in
the pipes to prevent separation.
Test materials 1 and 3 were also tested
using water with no detergent.
Fig. 5: Loading by an external point load In order to determine the influence of in-
before starting the internal pressure test ternal pressure 4 pressure levels were
applied for test series no. 3. The magni-
tude of the additional external point load
tip radius of 5 mm. The tool loading was was kept constant.
Fig. 7: Fracture surface of a FNCT specimen
carried out at room temperature with no
taken from a PE 100 pipe
internal pressure in the pipe. The tool Test Temperatures
was loaded until a chord with a length of The tests were performed at 80 °C. In ad-
5 mm was measured at the inside surfa- dition the activation energy was calcula- point is the beginning of thermal degra-
ce of the pipe (fig. 5). This occured at a ted for test series 1 and 3 by performing dation of the pipe [12].
tool displacement of 9 mm from the outer tests at temperatures between 20 and Another method to describe the thermal
surface of the pipe. 95 °C. ageing of polyolefines is the measure-
The elongation at the inside surface can A summary of the test program is given ment of the viscosity number according
be calculated using the following equa- in table 2. to ISO 1628-3. This number is related to
tion: the molecular weight. [13].
Theoretical Basis for Determining With all three methods a correlation is
ε = [0.318 · F · R] / [b‘ · s · E(t)] obtained which can be described by the
the Maximum Service-Life
equation of Arrhenius. This means that
where: Generally the maximum service-life of the thermal degradation takes place after
ε Elongation at the inner surface of the plastics is determined by thermal ageing, a short period of time at high tempera-
pipe wall which causes embrittlement of the mate- tures (e. g. 80 °C) and after a long period
F Radial acting force (single load) rial. The relevant parameters other than of time at low temperatures (e. g. 20 °C)
R Mean pipe radius temperature and time are the environ- according to the activation energy.
s Wall thickness ment and the available oxygen.
It is accepted that if polyethylene with
E(t) E-modulus Water saturated with oxygen has proved conventional stabilisation is tested for
b‘ Supporting width to have the greatest effect on the reduc- one year at 80 °C without thermal ageing
The supporting width of the load (b') can tion of the maximum service-life because this is equivalent at 20 °C to 100 years
be determined experimentally by measu- it causes accelerated thermal ageing. service [14].
ring the force on the tool with the pipe Equivalent thermal ageing times are
under internal pressure. found for either the mixture of a deter- Results of the Investigations
Finite element calculations (fig. 6) con- gent and water or just water without a
Although 6 of the 19 test series are not
firm that with the tool loading a pipe un- detergent. The time to the beginning of
yet completed it is possible to evaluate
der internal pressure of 8 bar the elonga- the effect of thermal ageing is indepen-
the results with regard to the required mi-
tion of the inner surface of the pipe due dent of the elongation of the material.
nimum quality of polyethylene pipes for
to the tool displacement was in excess of The onset of thermal ageing was investi- their installation without sand-embed-
twice the yield elongation of the material, gated by Gaube et al. in the temperature ding.
which is 9 % for all the materials given in range between 80 °C and 40 °C. The cri-
Taking into account the evaluation proce-
table 1. teria for evaluation was that the elongati-
dure [15] the results are conservative,
on to rupture of the aged sample was 50
that means they are on the safe side.
% of the elongation to rupture of the una-
ged sample. Later investigations on
pipes under internal pressure showed Failure Mechanism
that the onset of thermal ageing found by Assuming that the process of slow crack
Gaube testing sheets was identical with growth (stress cracking) is the relevant
the starting point of the sharp decent of long term failure mode for pipes under
the creep rupture curves of pipes. This additional external point load these re-

Fig. 6: Surface elongation in % at the inside

wall of the pipe due to an external point load
and an internal pressure of 8 bar Fig. 8: Fracture surface in a PE-pipe subjected to an external point load and internal pressure

5
 yield point, i.e. before it becomes As can be seen in fig. 14 the relationship
oriented. between the rupture times of the two
A similar phenomenon is found in tests is linear on a double logarithmic
the Cone-test where the crack scale. The correlation coefficient is grea-
initiation is not at the point of ma- ter than 0.98 which is acceptable for
ximum elongation but a certain technical applications. The mathematical
distance away [17]. description of the correlation leads to the
formula:
Influencing Parameters
t1, FNCT = 10 ↑ (m1 · log t1, Pipe + a1) (1)
Pipe Quality
The results of the investigations where
Fig. 9: Fracture surface in a PE-Pipe subjected to
internal pressure only on point loaded pipes and in the
t1, FNCT: Rupture time in FNCT (80 °C,
FNCT, which are complete at the
4 N/mm2, 2 % Arkopal N-100)
time of reporting are shown in
sults should correlate with the resistance fig. 14, where the figures at the data t1, Pipe: Rupture time of point loaded
against stress cracking as tested by the points correspond to the number of the pipe and internal pressure
FNCT. test series in table 2. (80 °C, 4 N/mm2, 2 % Arkopal
N-100)
The fracture surface both in FNCT (fig. 7) This shows that the pipe quality with re-
and in the pipe under external point load gard to the long term performance under m1 = 0.874
(fig. 8) shows low deformations without internal pressure and additional external a1 = -0.151
evidence of macroscopic ductility. This point load correlates with the rupture ti- The shaded area in fig. 14 shows the re-
has been confirmed by investigations of mes of notched specimens in FNCT. This gions where thermal ageing will occur for
Laurent using a scanning electron relationship is supported by the correlati-
conventionally stabilised polyethylene
microscope [16]. on between pipes under internal pressu-
pipe resins.
The same appearance is also visible at re and FNCT published by Fleißner [18].
From these results it can be concluded
fracture surfaces of pipes under internal As expected the samples made from PE
that pipes with rupture times in the FNCT
pressure only (fig. 9). 63 show the shortest rupture times both
above 2000 hours (80 °C, 2 % Arkopal N-
For pipes under external point load the for point loaded pipes and in the FNCT. 100, 4 N/mm2) could be installed without
crack always starts at the inside of the The rupture times in both tests increase sand- embedding for service pressure 4
pipe wall but not at the point of the lar- from PE 80 to PE 100 to PE-X. bar (safety factor 2). This is because the
gest material strain, directly below the Test series no. 4 (PE 80) is exceptional pipes under external point load exceed
loading tool (fig. 10 to 13). From fig. 6 it because the rupture times are in the the limit of thermal ageing at 80 °C,
can be assumed that the crack starts in region of some PE 100 materials (no. 6 which relates to a service-life of at least
the region where the material is at the to 10). 100 years at 20 °C [19].
The results of test series no. 12 (pipe un-
der external point load) and no. 15 (FNCT
and pipe under external point load) can
be found in the area of thermal ageing.
Those test series which are not yet com-
pleted at the time of this report are sum-
marised in table 3.

Internal Pressure of Pipe

All point loaded pipes were tested with
an internal pressure of approximately 8
bar following the relevant test standards
for polyolefin pipes (hoop stress 4
Fig. 10: Cracks in the inside wall of a PE 80 Fig. 12: Cracks at the inside wall of a PE 100 N/mm2). For test series no. 3 additional
pipe subjected to an external point load and pipe subjected to an external point load and
internal pressure (Test series 4) internal pressure (Test series 6) tests were carried out at lower internal
pressures with a notch at the inside pipe
surface in line with the point load
(fig. 15).
A linear relation in a double logarithmic
scale results from these tests in a similar
manner to conventional creep rupture
curves without additional loads. This
shows that an important influence on the
failure times is the internal pressure in the
pipe.
The mathematical description of this re-
lation is given in formula 2:

Fig. 11: Cracks at the inside wall of a PE 100 Fig. 13: Cracks at the inside wall of a PE 100 t2, Pipe = 10 ↑ (m2 · log pi Pipe + a2) (2)
pipe subjected to an external point load and subjected to an external point load and inter-
internal pressure (Test series 8) nal pressure (Test series 7) where:

6
 ween 95 °C and 60 °C. The following for-
mula is derived from the test results of
test series 1:

t3, Pipe = 10 ↑ ((1/T – a3) : m3) (3)

where
t3, Pipe: Rupture time of the point loaded
pipe (4 N/mm2, 2 % Arkopal
N-100)
T: Absolute temperature [K]
m3: 2.87 · 10-4
a3: 2.43 · 10-3
To date there has been no indication that
the activation energy is changing for
lower temperatures. Thus the extrapolati-
on of the values measured at higher tem-
peratures in a short period of time to
Fig. 14: Correlation between test results for external point loaded pipes and from the FNCT lower temperatures and long period of
on specimens taken from polyethylene pipes (Number at data points refer to the test series times is confirmed.
number)
Investigations in the past have shown
that the activation energy determined
from tests in an aqueous solutions with
Table 3: Pipe samples still under test
detergents provides a conservative esti-
mate of the activation energy determined
from tests in water without detergent
[22].

Medium
In order to get results in reasonable times
for all pipe qualities, the investigations
usually were performed under the influ-
1 80 °C, Tensile Stress 4 N/mm2, 2 % Arkopal N-100
2
ence of an aqueous solution with Arkopal
80 °C, Hoop Stress 4 N /mm2, 2 % Arkopal N-100
N-100 (2 % w/w).
Both in the FNCT and for point loading
t2, Pipe: Rupture time of point loaded Fig. 16 shows the linear relation between tests it was important to ensure a homo-
pipe with notch at the inside the inverse of absolute temperature and geneous mixture of the test fluid, becau-
surface and internal pressure the logarithm of the creep rupture times se incomplete mixing produces undefi-
(80 °C, 4 N/mm2, 2 % Arkopal for the test series 1 and 3. The calculated ned concentration conditions and longer,
N-100) correlation coefficients are above 0.98. not comparable rupture times.
pi Pipe: Internal pressure of pipe in bar Thus a constant activation energy is con- In order to transform the test conditions
firmed both for test series 1 between to the service conditions for gas or water
m2 = -0.823 80 °C and 40 °C and for test series 3 bet- pipelines the difference between the in-
a2 = 2.365

Temperature
From experience, creep rupture times
increase according the formula for the
activation energy found by Arrhenius.
Table 4 shows the calculated activation
energies [21] for test series 1 and 3 for
point loaded pipes in comparision with
the values for the activation energies for
the "brittle regions" of the creep curves
of PE 63, PE 80 and PE 100 according to
DIN 8075.
The values for the activation energies for
point loaded pipes in table 3 are consi-
derably lower than those for the creep
curves given in DIN 8075. This means
that the time-temperature correlation gi-
ven in DIN 8075 is not applicable to point
loaded pipes and would result in overop-
timistic values when extrapolated to Fig. 15: Effect of internal pressure on the long term performance of external point loaded
20 °C. pipes (Test series 3)

7
Table 4: Activation energy for creep rupture of pipes
under pressure with and without additional external point
load (brittle failure)

fluences of detergent or water on the

creep rupture behaviour needs to be
quantified.
The test data available at the time of this
report are summarised in table 5. The va- Fig. 16: Effect of temperature on the long term performance of external point
lues agree well with the results from a loaded pipes (Test series 1 and 3)
round robin test on the FNCT [23].

Table 5: Influence of water and detergent on polyethylene surface of the pipe wall t4, Pipe: Rupture time of the point loaded
in the FNCT which exceeded the pipe (80 °C, 4 N/mm2, 2 % Arko-
yield strain of the materi- pal N-100)
als. t4, Notch: Rupture time of the point loaded
The displacement of the pipe with an inside notch (80 °C,
loading tool was varied 4 N/mm2, 2 % Arkopal N-100)
between 9 mm and 0 m = 0.818
4
mm (contact of the loa-
a4 = -0.05
ding tool at the outer
1 80 °C, Tensile Stres 4 N/mm2, Geometric Mean Value from 3 Single Specimens Each
pipe surface) in test se- These results show that the internal
ries 3. notches reduce the rupture times of a
Table 6: Influence of Loading Tool Tip Radius for test series 3 pipe which reaches the thermal ageing
In this connection it is
under the influence of water by a factor
important to note that
of 4.1 compared to unnotched pipes.
the pipes were loaded at
room temperature with- The depth of the notches were varied
out internal pressure but between 0.25 and 0.85 mm using speci-
the tests are carried out mens from test series 3. However, no sig-
at elevated temperature nificant difference was found in the creep
1 80 °C, Hoop Stress 4 N/mm2, 2 % Arkopal N-100 with internal pressure. rupture times.
The results are given in
table 7 and indicate there is no signifi- Scattering of the Test Results
Radius of the Loading Tool cant difference in rupture times for dis- The scattering of the results of test series
The outside of the pipes were usually placements between 0 and 9 mm. 2, 6 and 9 are given in table 8 and were
loaded using a tool with a tip radius of 5 Thus it seems that small displacements used to calculate the cumulative fre-
mm. Additional tests were performed on of an external point load introduce a re- quency at which 2.5 % of the specimens
test series 3 with tool tip radius of 2.5 duction of the creep rupture time. tested would be expected to fail, in a si-
mm and 10 mm. milar manner to that used for creep
rupture tests (ISO/TR 9080).
The test results show that a tool tip Notches at the Inner Pipe Wall
radius between 2.5 mm and 10 mm there Surface
Calculation of Minimum Test
is no difference in the rupture times of the In order to investigate the influence of
pipes (table 6). Requirements
notches with a small depth at the inside
pipe wall surface, tests were carried out In order to calculate the minimum test re-
Displacement of the Loading Tool for test series 1, 2, 3, 6 and 8. The quirements equivalent to a specified ser-
vice-life of point loaded polyethylene pi-
A displacement of the loading tool of 9 notches were 0.85 mm deep and were
located at the inside pipe surface in line pes at 20 °C applied for water or gas
mm caused an elongation at the inner
with the external point load. conveyance the following formula is
used:
Fig. 17 shows the creep rupture times of
Table 7: Influence of Loading Tool the test series with and without a
Displacement for test series 3 notch. A linear relation is found Table 8: Scattering of the test results for point
between the two kinds of speci- loaded pipes
men according to the following for-
mula:
t4, Notch = 10 ↑
(m4 · log t4, Pipe + a4) (4)
1 80 °C, Hoop Stress 4 N/mm2, 2 % Arkopal N-100 where

8
 FTemperature: Time factor, representing the
activation energy between
test- and service tempera-
ture
FMedium: Time factor, representing the
influence of the test fluid
compared to water
FInternal Pressure, Safety:
Time factor representing a hoop stress
other than σV = 4 N/mm2 and the safety
factor
FNCTmin: Minimum rupture time in
FNCT at 80 °C; 4 N/mm2;
2 % Arkopal N-100
A summary of all the data used in the cal-
culation of the minimum service-life is gi-
Fig. 17: Effect of 0.85 mm deep notch at the inside wall on the long term performance of ven in table 9.
external point loaded pipes (Test series 1, 2, 3, 6 and 8)
The calculation of the creep rupture
times of point loaded pipes using the ac-
tivation energy in table 4 (test series 1)
tPipe, Test = (tPipe, Service · FNotch · tPipe, Test: Failure time of the pipe under predicts a service-life of 106 years at 20
FScattering) : (FTemperature · FMedium · test conditions °C when a rupture time of 8760 hours is
FInternal Pressure, Safety) (5) tPipe, Service: Service time of the pipe at exceeded at 80 °C. Thus the material of
20 °C; σV = 4 N/mm2; Water; test series 1 appears to correspond with
The minimum required time in the FNCT Point Load > εs the material for which extrapolation time
at test conditions is obtained when for- limits are described in ISO/TR 9080.
FNotch: Time factor, representing a
mula (5) is inserted in formula (1): The minimum rupture times in the FNCT
notch at the inside wall sur-
face (4) at 80 °C, 4 N/mm2; 2 % Arkopal N-100
FNCTmin = 10 ↑ (m1 · log tPipe, Test + a1) (6) which are shown in tables 10 and 11
FScattering: Time factor, representing a were calculated using formula (5) and (6)
where probability of failure of 2.5 % and the data given in table 9.
The qualifying of polyethylene pipes
Table 9: Data used to calculate the minimum service life against the minimum requirements is
therefore possible by determining the in-
dividual properties and parameters for
the calculation.
Many thanks are given to Dr. Mike
Troughton; TWI (The Welding Institute,
Cambridge) for fruitful technical discussi-
ons and assistance in translation.

Table 11: Minimum rupture times in the FNCT which

Table 10: Minimum rupture times in the FNCT which is equiva- is equivalent to a minimum service-life of 100 years at
lent to a minimum service-life of 100 years at 20 °C for external 20 °C for external point loaded pipes with an internal
point loaded pipes without an internal notch notch

9
Literature [13] Gebler, H.: Langzeitverhalten und Alte- creep rupture strength of polyethylene
rung von PE-HD-Rohren. Kunststoffe 9 pipe). Kunststoffe 77 (1987) 1, p. 45/50
[1] to [8]: Reports of HESSEL Ingenieurtechnik
(1989), S. 823/ 826 [19] Hessel, J., Koch, R., Gaube, E., Gondro,
GmbH, Roetgen C., Heil, H.: Langzeitfestigkeit von Depo-
[14] Wernicke, K.: „100 Jahre Nutzungsdauer
[1] Client ATOFINA, Report R00 04 261-1 für Druckrohre aus PE” Wiesbadener niedichtungsbahnen aus Polyethylen
[2] Client ATOFINA, Report R01 04 508-1 Kunststoffrohrtage 2000 Kunststoffe 2 (1988) S. 155/160
[3] Client BASELL, Report R01 04 447 [15] Hessel, J., Mauer, E., Schleyer A., [20] Hessel, J.: Zur Mindestqualität von Roh-
Schröder-Wrede, V.: Ein neues Konzept ren aus PE für die Erdverlegung ohne
[4] Client BOREALIS, Report R01 04 500
zur Lebensdauerabschätzung von Sandeinbettung. Report No. 4, Wiesba-
[5] Client EGEPLAST, Report R01 04 460-1 dener Kunststoffrohrtage 2000, 04. und
Schweißverbindungen an Kunststoff-
[6] Client SOLVAY, Report R01 04 459-1 05.05.2000
mantelrohren aus Polyethylen. 3R inter-
[7] Client WAVIN, Report R01 04 457-1 national 34 (1995) 3, S. 94/101 [21] Westphal, W. H.: Physikalisches Wörter-
[8] Client WIRSBO, Report R98 04 262-1 buch. Springer-Verlag, Berlin/ Göttingen/
[16] Laurent, E.: Comprehensive Evaluation Heidelberg. 1952
[9] Hessel, J.: Zeitstandverhalten von Poly- of the Long-Term Mechanical Properties
ethylen unter dem Einfluß lokal konzen- [22] Larsen, Chr. T., Hessel, J.: Neue Erkennt-
of PE 100 Resins Meeting the Require- nisse zum Langzeitverhalten von
trierter Spannungen 3R international 34 ments of Modern Installation Techni- Schweißverbindungen an Kunststoff-
(1995) 10/11, S.573/579 ques. Plastics Pipes XI – 4th September Mantelrohren. 3R international 36, Heft 6
[10] Hessel, J., Mauer, E.: Zeitstandzugprü- 2001, Munich Juni 1997, S. 283/ 287
fung in wäßriger Netzmittellösung. Mate- [17] Köstring, V.: Von den Rohreigenschaften [23] Round Robin Test on FNCT 1998. Parti-
rialprüfung 36 (1994) 6, S. 240/243 zum fertigen Produkt: PE 100-Rohre mit cipants: BASF, DSM, FINA, HESSEL,
[11] Nonhoff, G.: Personal Information integrierten Schutzschichten. Vortrag 9, MAINOVA, REINERT-RITZ, SIMONA,
(4/2000) Wiesbadener Kunststoffrohrtage 2001, SOLVAY, TARCO, TGM, TWI, TÜV Süd-
[12] Gaube, E. et al.: Zeitstandfestigkeit und 26. und 27.04.2001 deutschland. Symposium on 0312.1998,
Alterung von Rohren aus HDPE; Erfah- [18] Fleißner, M.: Langsames Rißwachstum Main Office of Deutscher Verbandes für
rungen aus 30 Jahren Rohrprüfung. und Zeitstandfestigkeit von Rohren aus Schweißen und verwandte Verfahren
Kunststoffe 7 (1985), S. 412/ 415 Polyethylen. (Slow crack growth and (DVS), Düsseldorf

10