Study on Testing Method of Mechanical Load of Blade Root
XueTianwei XingZuoxia
(School of New Energy Engineering Shenyang University of Technology, Shenyang 110023, China)
Abstract:The method to measure loads of blade of wind turbine is introduced. The measurement of load signals
which not accurate will directly affect the results of the software system is difficult. This paper principally
discusses how to measure the loads with foil strain gauges. The methods of choosing gauges and determining
measuring points in wind energy application are explained. The principle of strain gauges is described and
Wheatstone bridge method is used in the measurement in this paper. The influencing factors are summarized and
the methods to reduce interference are developed in the results of blade measurements so as to ensure the
accuracy and the reliability of the results.
Key words:wind turbine, blade root load, foil strain gauge, uncertainty analysis
1 Introduction
In recent years, with the prices gradually rising caused by increasing depletion of traditional energy, the new
energy industry development has paid extensive attention. Wind energy has the characteristics of clean and
renewable and has formed a rapid development trend at home and abroad. The wind turbine blades which are a
core part of the power generation play an important role in the conversion of electromechanical energy. Along
with the increasing demand of renewable energy, the requirements of size and capacity of wind turbines are more
stringent. Especially when the blade size increases, the cost of manufacturing and transportation could be
increasing. The irreparable damage of blades will bring huge economic loss. The blade design is optimized to
improve the performance and security life of wind turbines. The load test of blades is an assessment processof
performance analysis and optimization in operating wind turbine status by measuring blade loads. However, the
measurement results of blade loads will affect the reliability of analysis directly.
Wind turbine blade load measurement is an analytical methodof using a special strain gauge to determine
strain or stress of the surface of measuring points. The Sensitive gates of strain gauges will change with the
deformation of measuring points by external load, which may result in resistance changes. The resistance change
and the strain of blades should become some kind of proportional relationship within certain range. The change
could be measured by the resistance strain indicator and strain values of measuring points could be calculated
with the sensitivity coefficient which is calibrated by manufacturers. From the stress-strain relationship,
calculated stress values could provide to the process of blade design for reliable stress analysis. Because the
strain gauge used for model test and industrial measurement has advantages of small size, light quality, portable,
easy to use, high accuracy and easy showing and recording, it has been widely used in electricity, transportation,
construction, energy and other industries[1].
2 Measuring Object
Wind turbine blade load measurement is a process of statistics and analysis in time series of different
operating conditions in accordance with relevant standards and technical requirements. The blade root through
which the entire blade load passed to the hub is the most important part of the blade and very representative. As
the specification requirements, new design blades must be tested to ensure that the structure of performance to
meet the requirements. The measuring blade material is steel from a manufacturer. Air force and gravity mainly
affect the blades. The measurement quantities are including flatwise bending and edgewise bending[2].
3 Test Method
3.1 The selection of strain gauges and measuring points
In this paper, a metal foil strain gauge has chosen as a measuring element, particularly for the special
circumstance of wind power industry. The metal foil strain gauge has the following advantages in engineering
application[3]:
a. The foil strain gauge has the features of accurate size and uniform lines, so the dispersion of sensitivity
coefficient is small;
b. The surface areas of sensitive gates is large, so the heat dissipation is good and the gates could support
larger current which presents strong signal to improve the sensitivity. In addition, large surface areas could
enhance the adhesion to transfer deformation, thus the accuracy is assured;
c. The transverse effect of strain gauge is small;
d. The foil gauge has an excellent insulation with a small creep and mechanical hysteresis, it also has a long
fatigue life;
e. It can be mass-produced and the production efficiency is high.
The type of foil strain gauges needs to choose by the stress state of blade root. Selection method is to meet the
test condition in order to ensure the quality of measurement results. Firstly, it is needed to choose the type of
strain gauge. As the direction of main strain is known and unidirectional, uniaxial strain gauge is chosen for the
tested material(steel). According to the requirements of the resistance strain indicator and the sensitivity value,
the resistance value of strain gauge is determined. In order to improve the stability and sensitivity of output, a
high resistance value(eg, 350Ω) is often chosen. A smaller gate length of strain gauge ensures measurement
accuracy, but could affect the quality of installation. In this paper, 6mm gate length is used for the installation
environment. In the measurement of edge and flat bending of blade root, a high and uniform strain per load level
is determined. The strain of cylindrical region of blade root is relatively high, yet the influence of local stress
caused by blade bolts need to avoid. So a certain distance(1m-1.5m) from the blade flange is necessary[4,5].
Fig.1. Determining the measuring point and the installation orientation of strain gauges
3.2 Working principle
During the measurement, the resistance of gauges changes by the geometry of test material. For the resistivity
of the material is constant, the resistance increases with the length and decreases with the cross-sectional area.
The resistance rate and the cross-sectional area of sensitive gates change by Poisson’s ratio, resulting in changing
the resistance of strain gauge.
If strain resistivity is defined as ρ0, the strain resistivity in strain is defined as ρL, so ρ0 − ρL = ∆ρ, the
theoretical formula of resistance strain effect is as follows:
∆R
R≈
∆L
L+
∆ρ
ρ0
−∆S
S= Gk ∙ ε (1)
Where, R is the resistance of original length(L) of the gate.
R =𝜌0L
S (2)
Where, S is the cross-sectional area of the gate; ∆R, ∆L and ∆S are the variations of the resistance, the length
and the cross-sectional area relatively; Gk is the sensitivity coefficient of the strain gauge; ε is the axial strain
value of measuring points, ε = ∆L/L.
3.3 Test method
The change of ambient temperature could affect the material size of blades which may change the resistance.
Therefore, mechanical deformation and temperature will both affect the resistance. In order to measure the strain
only caused by deformation accurately, it is necessary to eliminate the influence of temperature. This paper uses
Wheatstone DC bridge to measure blade loads.Wheatstone bridge has advantages of sensitivity, accuracy and
stability[6]. Both strain gauges are installed in the position of blade root. As a temperature compensator, one of
the gauges is not in the direction of mechanical deformation.
The bridge circuit can be regarded as a circuit of voltage divider. As shown in the following figure, each
branch of the circuit undertakes same excited voltage(Eex). Two sets of circuits can measure the signals of
flatwise and edgewise.
Fig.2. Wheatstone bridge circuit
Where, ∆U is the output signal, R1(R2) is the gauge of edgewise(flatwise), R3 and R4 are the temperature
compensators. When the deformation occurs under blade loads, the change of R1 and R2 will affect output
signals. According to the calibration, a correspondence between signals and loads could be established and the
collection is finished.
4 Test results and uncertainty analysis
The load case is reported by measuring signals directly. By measuring the strains of three blade roots of a
wind turbine in normal power generation state, the results are as follows:
Fig.3. The measuring results of edgewise moments of three blades
Fig.4. The measuring results of flatwise moments of three blades
As the relationship between the measuring signals and the strains is linear, the characteristic will keep by the
amplifier. Comparingwith the results of amplifying flatwise or edgewise signals of three blades in this paper, it is
found that three groups of signal waveforms approximate the sine wave cycles within a certain amplitude range.
A difference among the amplitudes exists as the calibration, however the results are basically satisfactory.
Fig.5. The results between the predicting and the measuring flatwise signals
In the figure5, a direct comparison between the actual signals and the ideal waveforms in a period of normal
power generation is shown. It can be clearly seen that the fluctuations caused by alternating loads in the peak and
valley is severe. If the directions of the installed gauges and the loads are not completely coincided, it will lead
to a vertical offset in the curve. The angle between the pre-baseline of the measuring point and the main strain is
φ, and the angle between the installation direction and the main strain is φ,, the strain variation ∆εφ is:
∆εφ = (1 + μ) ∙ ε ∙ sin2∆φ (3)
Where, μ is Poisson’s ratio.
Relative error(eφ):
eφ =∆εφ
ε1
= (1 + μ) ∙ sin2∆φ (4)
If μ = 0.28and ∆φ ≤ 5∘, eφ ≤ 1%. The relative error increases with the value φ[7].In order to ensure that
the measuring signals are accurate, above influencing factors are all needed to analyzed, and appropriate ways to
avoid errors are proposed promptly.
5 Conclusion
A strain measurement method is given by a load test of blade roots of a wind turbine. Against the shape of
blade, the load condition and the environment, the solutions to avoid influences are appropriate. The example of
the measurement results of three blades of a wind turbine is shown above. The repeatability of the results is good
and the prediction error is small to meet anticipation. For the load spectrum analysis and the performance
optimization of wind turbine, a reliable data is needed.
REFERENCES:
[1] Li Tang, Jun Wang, Ying Wang.“The strain gauge load transducer of the wind power vane static load test[J]”, China
Measurement Technology, 2008,34,1.
[2] GB/Z 25426-2010 “Wind Turbine Mechanical Load Measurement”.
[3] Liangcheng Ma. “Strain Measurement and Sensor Technology[M]”, Beijing: China Metrology Press, 1993.
[4] Kurt S. Hansen, Knud Ole Helgesen Pedersen &Uwe Schmidt Paulsen. Online wind turbine measurement laboratory[C].
Proceedings of the EWEC conference, 2006, 2-3.
[5] Wenjun Zhang, Boping Lv. “Research of testing method of Rotor shaft loading of one type of helicopter[J]”, Testing and
Calibration, 2006,26,1-2.
[6] Brent L. Ellis & L. Montgomery Smith. Modeling and Experimental Testing of Strain Gauges in Operational and Failure
Modes[J]. IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT,VOL.58,NO.7,2009(7),1-2.
[7] Yung-Li Lee, Jwo Pan, Richard Hathaway & Mark Barkey. Fatigue Testing and Analysis[M]. Amsterdam: Elsevier Inc,
2005.