Testing of Adhesive Joints in The Wind Industry: Abstract
Testing of Adhesive Joints in The Wind Industry: Abstract
Adhesives [t]
Adhesive bonding is used in the wind industry 70000
60000
to join the main parts of a wind turbine blade. 50000
Bond lines can be ether structural or non- 40000
0.25
Flap- wise
0 Flap- & Edge- wise Most studies perform four point beam tests and
-0.25
SC11 SC22 SC12 W11 W33 W13 Edge- wise try to get failure to occur in the constant
moment region between the loading points.
-0.5
However the shear forces on the material
-0.75
present in wind turbine blades are neglected
-1 by this set up. Additionally, it can be very hard
to obtain failures away from the loading points
Figure 4: Strain state in rotor blade
and thus the initiation of damage in the bond
For the loading in flap- and edge- wise line will be influenced by the test set up.
direction, resulting in a service like stress state,
the longitudinal stresses in blade length 3 Adhesive Bonds
direction (SC11) are dominating the other
longitudinal stresses in the spar cap (SC22) and
in the web (W 11 and W 33). The shear stress in 3.1 Certification of Adhesives in
the web (W 13) is about three times higher than Wind Turbine Blades
in the spar cap (SC12). As each turbine blade prototype must pass a
full scale certification test (depending on the
certifier only a static or a static and fatigue test)
2.3 Experimental testing of the blade manufacturer must ensure that the
downscaled models or beams adhesive bonds can pass these tests.
The testing of downscaled models was Furthermore most certifying bodies, e.g.
common before numerical methods became Germanische Lloyd (GL) or Det Norske Veritas
the major tool in the design of structures [7]. (DNV), have introduced some basic
Component or beam testing in the design requirements for the adhesives in their
process of complex technical systems is certification guidelines. For example, the GL
generally accepted [2-8]. The testing of requires:
specially designed beams is performed to • Tg ≥ 65 °C
determine component and material properties • Verified material properties
more realistically than with smaller specimens. • Operation temperature ≥ 60 °C
This approach of testing beams or downscaled • If possible same chemical basis as
blades can be a better method for matrix system
understanding the material behaviour and can
potentially contribute in future certification If the adhesive is used in a structural bond line,
procedures for rotor blades [3]. then GL asks for the use of a GL-certified
material as they need to withstand [1] ambient
Beams have been used to investigate the and service conditions:
properties of different materials. Zhou et al. [9] • Long service life (20 years)
investigated the static behaviour of a glass- 8
• High cyclic loading (2*10 cycles)
fibre-reinforced epoxy sandwich I-beam in a
• High loading (3000 h under full load)
symmetric three point bending test. The beam
was loaded on the middle of the top flanges • Working temperature between -40 °C
not using any tabs, while boundary condition and 50 °C (up to 7000 cycles)
on the sides were close to be fixed (no • Influence of media (salt spray, 100 %
transverse displacement and no rotational air humidity, UV-light)
degree of freedom and only very limited
longitudinal movement). They showed for their No requirements exist with regard to the
beam, that the failure started at the foam core fatigue properties of the adhesives. However,
of the sandwich web. Mandell et al [6] the blade designers have to show that they
designed a four point bending test to deal with this in an appropriate way.
investigate the properties of the web material
of the blade beam structure. Hayes [11], Hayes According to the GL-Guidlines [15, 16] for
and Lesko [12] and Park et al. [13] investigated adhesive joints the mechanical properties are
the structural behaviour of fibre-reinforced to be determined for the adhesive at two
beams for bridges under static and fatigue different bond line thicknesses (0.5 mm and
loading. Park et al. [10] investigated a 3 mm) and the following different material
connection system for such beams. Potter et conditions:
a) dry material: 24 h ± 1 h after curing at A. O`Neil et. al. [18] investigated the fatigue
23°C and 50 % relative humidity behaviour of externally bonded fibre reinforced
b) wet material: 1000 h ± 12 h at 23°C in plastics of three adhesives with lap-shear
distilled water specimens according to ASTM D2983. They
described a “deleterious effect of fatigue
In total four different material configurations loading on the ultimate mechanical properties
have to be tested. Three static tests (6 of epoxy adhesive systems...”.
specimens each), one test on the thermal- I. A. Ashcroft [19] et. al. describe suited test
mechanical properties and one on the creep methods for investigating the fatigue behaviour
behaviour have to be done. of adhesives. They state, that single lap
specimens are useful for comparison and
producing SN-lines of adhesives, however lap-
In the more general GL guidelines for the
strap joints are in favour if crack initiation and
certification of wind turbines a glass transition
crack growth are subject of the investigations.
temperature of more than 65 °C is required.
The fatigue behaviour of paste adhesives was
This has to be tested by a method of thermal
investigated with different specimen
analysis, such as DSC.
geometries to find an adequate test set-up. For
single lap joints, minor changes in specimen
3.2 Testing of Adhesive lay-up and bond line thickness did not change
A variety of investigations on adhesive material the failure mode, however it had an influence
testing are published. The coupon size and on the mechanical performance.
more product-like specimens are addressed as Adhesive tests on perpendicular T-joint
well as adhesive bonded components and specimens under shear loading were
products. Only a short overview of some performed by V. Marcadon et. al. [20]. The
publications on testing of coupon size specimen geometry and loading was similar to
specimens, testing of thick adhesives and the actual application.
more realistic test methods can be given. D. Sambrosky [21] performed static and fatigue
J. R. Vinson gives an extensive review of tests on specimens which are representative
adhesive bonding of polymer composites [17]. for wind turbine blade joints. They investigated
Results of different publications on single, four different specimen geometries, two
double, single double and double-double lap different static testing rates and two fatigue
joints as well as scarf joint until 1989 are loading conditions. The bond-lines were
discussed. However special issues in thick approximately 50 mm wide and had a
bond lines are not addressed. thickness of about 4 mm. The adhesive used is
not reported. As lap shear type specimens
were used, it can be assumed, that the
geometry had an influence on the mechanical
behaviour.
4.4 Design of Henkel-beam The shear web of the beam may be centred or
A major concern during the design process placed slightly asymmetrically. By making the
was the need for a realistic bond-line beam asymmetric, the impact of non-
thickness. Therefore the thickness of the bond- symmetric stress states in the adhesive bond
line is about ten millimetres thick in the gauge due to bend-twist coupling can be assessed.
section and it can be manufactured with wind Figure 11 shows a schematic drawing of the
industry manufacturing technologies. By Henkel-beam.
tapering the spar caps the stresses will be The adhesive chosen for the investigation is a
highest away from the reaction points. new two component Polyurethane (Macroplast
Additionally, increasing the thickness of the UK® 1340) by Henkel. Compared to state–of-
adhesive in the gauge section results in a the-art epoxy systems it is a fast curing (even
reduced bond strength in this region, due to at room temperature), low exothermal and
scaling effects. Therefore similar excellent wetting bonding paste with a high
manufacturing defects as seen in rotor blades flexibility in controlling reaction time. Henkel
are expected. The beam is made from a Macroplast UK® 1340 is certified by
glass-fibre reinforced polymer. The spar caps Germanischer Lloyd for the joining of rotor
have a unidirectional lay-up, while the shear blade components.
web is box made from +/- 45° glass fibre
material and a typical foam core material. 4.5 Testing of the Henkel-beam
Failure Section Load Figure 10 shows the test of the Henkel-beam
Support at the Fraunhofer IWES. Special attention was
paid to the design of the critical beam section
Web and slection of a proper load introduction and
Insert
Support Spar Cap Adhesive support. The test specimen is loaded with a
x
600 mm 200 mm 1200 mm load frame at one side, while the support at the
other side of the beam is realized by two bolts
though the reinforcement in the beam.
„C-Side“mit
Balkenseite ofC-
theForm: „I-Side“ of the
Balkenseite mit I-Henkel-
Form: Four displacement controlled static test of the
Henkel-Beam
C-Seite I-Seite Beam
Henkel-beam were performed on different load
levels (45 mm, 110 mm, 220 mm and 405 mm
displacement). Between the tests the Henkel-
Figure 11: Henkel-beam Design beam was visually inspected.
Figure 12: Testing of the Henkel-beam
7 References
[1] G. Wacker, M. Hauschild;
Figure 16: Failure of shortened Henkel-beam Klebverbindungen in Rotorblättern für
Windenergieanlagen; VDI-Seminar
„Kleben – Innovative
Verbindungstechnik für die Industrie“;
6 Conclusions http://www.gl-
From the survey of the guideline for wind
group.com/pdf/vdikleben03.pdf;
turbine blades it is apparent that only static
Oktober 2008
mechanical properties of adhesive joints are
[2] P. A. Joosse, D. R. V. van Delft, A. M.
usually tested. Although the blades of wind
van Wingerde, Ch. W. Kensche, J.
turbine are highly loaded in fatigue, no cyclic
Korsgaard, et. al.; Using carbon fibres
tests are required by the certification rules. The
for a cost effective large blade; Global
thickness of the experimentally tested bonds
Windpower Conference & Exhibition,
normally ranges between 0.5 mm and 3 mm
France 2-5 April 2002, EWEA, AWEA,
and thus is not totally realistic. The scaling
IWTMA (2002)
factor from the tested bond line thickness to a
[3] H.-G. Busmann, C. Kensche, A. Berg-
real blade can be as high as an order of
Pollack, F. Bürkner, F. Sayer, K.
magnitude.
Wiemann; Testing of Rotor Blades;
The strain state of adhesives in wind turbine
DEWEK 2006
blades is primarily in the axial direction,
[4] K. D. Potter, R. Davies, M. Barett, A.
however significant shear strains are present
Godbehere, L. Bateup, M. Wisnom, A.
as well.
Mills; Heavely loaded bonded
Based on consideration of typical test
composite structure: design,
configurations it was shown that the
manufacture and test of I-beam
asymmetric three point beam configuration can
specimens; Composite Structures 51
provide a stress state in the adhesive bond line
(2001) 389-399
that is similar to the stress state in a wind
[5] http://www.upwind.eu; October 2008
turbine blade.
[6] J. F. Mandell et. al.; Fatigue of
A sub-component was developed for testing
Composite Material Beam Elements
adhesive bond line performance with realistic
Representative of Wind Turbine Blade
static and fatigue loading. It is particularly
Substructure; NREL/SR-500-24379;
important that an asymmetric three point
November 1998
bending test set was chosen, as the behaviour
[7] R. K. Müller; Handbuch der
of adhesives is dependent on the complexity of
Modellstatik; Springer-Verlag Berlin,
the stress state. Additional care was taken in
Heidelberg, New York 1971
the design to ensure that failure occurs away
[8] B. F. Backman; Composite Structures,
from the load introduction points. With this
Design, Safety and Innovation;
design, a realistic loading and bond line
ELSEVIER, Oxford 2005
thickness was possible.
[9] G. Zhou, A. J. Forbes, N. Foster; Static
behaviour of preconditioned glass-fire-
These beam tests will enable the optimisation
reinforced epoxy sandwich I-beams;
of bond-line adhesives and the design of
Composites Science and Technology
adhesive joints in wind turbine blades by better
59 (1999) 963-973
understanding of the fatigue and failure
[10] K.-T. Park; Performance verification of
processes. Additionally, the beam test concept
a new pultruded GFRP bridge deck-to-
offers the opportunity to develop new test and
girder connection system; Composite
certification methodologies for adhesives and
Structures 81 (2007) 114-124
wind turbine blades.
[11] M. H. Hayes; Structural Analysis of a Engineering and Science; Vol. 29;
pultruded composite beam: Shear No.19; October 1989
Stiffness determination and strength [18] A. O`Neil, K. A. Harries, L. Minnaugh;
and fatigue life prediction; Dissertation; Fatigue Behaviour of Adhesive
Virginia Polytechnic Institute and State Systems used for Externally-Bonded
University, 2003 FRP Applications;
[12] M. H. Hayes, J. J. Lesko; Failure http://www.engr.pitt.edu/civil/news/spri
analysis of a hybrid composite ng-07/oneill.pdf; February 2009
structural beam; Composites: Part A [19] I. A. Ashcroft, D. J. Hughes, S. J.
38 (2007) 691-698 Shaw; Adhesive bonding of fibre
[13] K.-T. Park, S. H. Kim, Y.-H. Lee, Y.-K. reinforced polymer composite
Hwang; Pilot test on a developed materials;Assambly Automation;
GFRP bridge deck; Composite Volume 20; No. 20; 2000
Structures 70 (2005) 48-59 [20] V. Marcadon, Y. Nadot, A. Roy, J. L.
[14] T. P. Philippidis; Qualification of Gacougnolle; Fatigue behaviour of T-
Complex Stress State – Effect on joints for marine applications;
Blades Design; Optimant Blades- International Journal of Adhesion &
Report: OB_TG2_R032_rev.000; Adhesives; Volume 26; 2006
[21] D. D. Samborsky; A. T. Sears, J.
www.kc-wmc.nl; November 2008
Mandell, O. Kils; Static and Fatigue
[15] Germanische Lloyd - Richtlinie für Testing of Thick Adhesive Joints for
hochelastische Klebstoffe und Wind Turbine Blades; 2009 ASME
Klebverbindungen – Ausgabe 2002 Wind Energy Symposium
[16] Germanische Lloyd – Klassifikatios- [22] J. S. Tomblin, C. Yang, P. Harter ;
und Bauvorschriften; II Werkstoffe und Investigation of Thick Bondline
Schweißtechnik; 2 Nichtmetallische Adhesive Joints ; Final Report
Werkstoffe – Ausgabe 2006 DOT/FAA/AR-01/33; June 2001
[17] J. R. Vinson; Adhesive Bonding of
Polymer Composites; Polymer