Home Technology On-site contactless visualization of the laminar-turbulent flow transition dynamics on wind turbines
Article Open Access

On-site contactless visualization of the laminar-turbulent flow transition dynamics on wind turbines

  • Daniel Gleichauf

    Daniel Gleichauf has been a research assistant at the University of Bremen at the Bremen Institute for Measurement, Automation and Quality Science (BIMAQ) in the Department of Production Engineering since 2018. Previously, he completed his Master’s degree in Systems Engineering, also at the University of Bremen. His research interest is the thermographic flow visualization on wind turbines focusing on the laminar-turbulent transition.

         , Felix Oehme

    Felix Oehme joined the University of Bremen in 2019 as a research associate at the Bremen Institute for Measurement, Automation and Quality Science (BIMAQ) in the department of Production Engineering. Previously, he completed a diploma degree in process engineering and natural materials engineering at the TU Dresden. His research interests are the flow behavior and the thermographic detection of flow separation on wind turbines.

     EMAIL logo     , Ann-Marie Parrey

    Ann-Marie Parrey has been a research associate at the University of Bremen at the Bremen Institute for Measurement, Automation and Quality Science (BIMAQ) in the department of Production Engineering since 2020. Previously, she completed her physics studies at the University of Bremen. Her research interests are the image processing and feature extraction for thermographic flow visualization on running wind turbines.

         , Michael Sorg

    Michael Sorg studied electrical engineering at the University of Karlsruhe and at the University of Bremen with a focus on communications engineering. After his research period at the Institute for Measurement, Control and Systems Engineering at the University of Bremen, he moved to the Bremen Institute for Measurement, Automation and Quality Science (BIMAQ), where he heads the Energy Systems and Materials Testing research group. His research areas include measurement system technology, especially for the application field of wind energy use and for energy system technology.

         , Nicholas Balaresque

    Nicholas Balaresque has been Managing Director of Deutsche WindGuard Engineering GmbH since 2010, and Managing Director of Deutsche WindGuard Wind Tunnel Services GmbH since 2019. He is responsible for the wind tunnel center of Deutsche WindGuard, which comprises 8 wind tunnels. After studying mechanical engineering at the Technical University Federico Santa María (UTFSM) in Valparaiso, Chile, with the last years at the Technical University Hamburg Harburg (TUHH), he took over the management of the Deutsche WindGuard Aeroacoustic Wind Tunnel (DWAA) in Bremerhaven. His professional focus covers wind tunnel experiments for industry and research, as well as free field measurements on wind turbines.

         and Andreas Fischer

    Andreas Fischer studied electrical engineering, completed his PhD at the Technische Universität Dresden in 2009 and his habilitation in 2014. Since 2016, he is a full professor at the University of Bremen in the department of Production Engineering and is head of the Bremen Institute for Measurement, Automation and Quality Science (BIMAQ). He received the Measurement Technology Prize of the AHMT e. V. in 2010, and an ERC Consolidator Grant in 2021. His research areas cover optical measurement principles for flow and production processes, in-process applications of model-based measurement systems, and the investigation of fundamental limits of measurability.

     ORCID logo
Published/Copyright: February 1, 2023
Download Article (PDF)
tm - Technisches Messen
From the journal tm - Technisches Messen Volume 90 Issue 9

Abstract

Thermographic flow visualization is already an established imaging method to localize the laminar-turbulent flow transition on the rotor blades of operating wind turbines, while a steady flow state is assumed. To understand the potential of thermographic flow visualization for the investigation of unsteady flow phenomena, its capability to detect the change of the flow transition position due to a wind gust is studied. Previously laminar flow regions become turbulent with the gust, which means a sudden increase of heat transfer between surface and fluid and, thus, a decrease of surface temperature. The latter is detected by evaluating the difference of thermographic images before and during the wind gust. The achievable sensitivity and the temporal resolution are limited by the thermodynamic properties of the rotor blade and the fluid flow, as well as by the natural rotor blade heating with the sun’s radiation. As a result of theory and experiments on real wind turbines, the feasibility to detect flow state changes in the order of seconds is proven. This opens upthe analysis of unsteady flow phenomena on wind turbines by means of thermographic flow visualization.

Zusammenfassung

Die thermografische Strömungsvisualisierung ist bereits ein etabliertes bildgebendes Verfahren zur Lokalisierung des Übergangs von laminarer zu turbulenter Strömung an den Rotorblättern von in Betrieb befindlichen Windenergieanlagen. Dabei wird von einem stationären Strömungszustand ausgegangen. Um das Potenzial der thermografischen Strömungsvisualisierung für die Untersuchung instationärer Strömungsphänomene zu erschließen, wird hier geklärt, inwieweit sich eine Änderung der Position des Strömungsübergangs aufgrund einer Windböe auflösen lässt. Zuvor laminare Strömungsbereiche werden durch die Böe turbulent, was zu einem plötzlichen Anstieg des Wärmeübergangskoeffizienten zwischen Oberfläche und Strömung und damit zu einem Rückgang der Oberflächentemperatur führt. Letzteres wird durch die Auswertung der Differenz von Thermografiebildern vor und während der Windböe bewertet. Die erreichbare Empfindlichkeit und die zeitliche Auflösung sind durch die thermodynamischen Eigenschaften des Rotorblatts und der Fluidströmung sowie durch die natürliche Rotorblatterwärmung durch die Sonneneinstrahlung begrenzt. Als Ergebnis von theoretischen Betrachtungen und Experimenten an realen Windenergieanlagen wird nachgewiesen, dass zeitliche Änderungen des Strömungszustandes in der Größenordnung von Sekunden erfasst werden können. Dies ermöglicht die Analyse von instationären Strömungsphänomenen an Windenergieanlagen mittels thermographischer Strömungsvisualisierung.

Schlagwörter: Thermografische Strömungsvisualisierung; instationäre Strömung; laminar-turbulente Transitition; Windenergieanlage
Keywords: laminar-turbulent transition; thermographic flow visualization; unsteady flow; wind turbine

1 Introduction

Due to its direct impact on the wind turbine’s efficiency [1] and its direct link to surface anomalies such as defects or deposits [2], the location of the laminar-turbulent flow transition is of emphasized importance in the evaluation of the boundary layer flow on wind turbine rotor blades. While the desired position of the flow transition is defined in the development process of the rotor blade, the true location on a real wind turbine depends on various factors, such as the speed and angle of attack of the incoming flow. In addition, unsteady inflow conditions, caused by the tower passing, wind shear within the rotor plane or gusts [3], result in an unsteady position of the laminar-turbulent flow transition because of the changing angle of attack. While simulations are used to study the effects of wind gusts on the performance and loads of the wind turbine [4, 5], measurements of the actual changing laminar-turbulent transition on wind turbines in operation are difficult. Therefore a measurement approach is required, that enables the detection of the spatiotemporal changes in the location of the flow transition on wind turbine rotor blades during the occurrence of unsteady inflow conditions.

In order to determine the spatial distribution of the boundary layer on wind turbine rotor blades during operation, thermographic flow visualization has been proven to be a suitable measurement approach since it is non-invasive, works contactless from the ground, and requires no blade modification. The measuring principle is based on evaluating surface temperatures with an infrared camera, which depend on the flow-related local heat transfer as well as on the temperature difference between the surface and the fluid [6]. Especially on wind turbines in operation, this thermographic flow visualization is capable of locating the laminar-turbulent flow transition [7–9], the turbulent separated flow [10–12] as well as evaluating the rotor blade’s surface degradation and contamination [2, 13, 14]. A position change of the laminar-turbulent flow regime due to a changing inflow condition is expected to lead to a change in surface temperature. However, due to the thermal response time of the surface, the thermal contrast between this region and the surrounding steady flow regimes is reduced compared to the contrast between permanently steady flow regimes. Simon et al. analysed the temperature change due to a flow state change from laminar to turbulent for internal and external heating methods and levels [15]. According to their results, an external heating by using a radiant heater is beneficial for a slowly changing flow state, i.e. for frequencies < 1 Hz. However, the absolute maximum heating level expected by the sun’s radiation and optimal measurement conditions under free field conditions equals the minimal heating levels used in the described fundamental study. Nonetheless, these findings motivate the subsequently followed measurement approach of detecting a changed flow state by means of evaluating the thermal temperature difference of a surface region.

Other existing work in the field of measuring unsteady boundary layer flow with thermography is with respect to the rotor blades of a helicopter. Raffel et al. studied an unsteady laminar-turbulent flow transition from an oscillating pitch-motion of an aerodynamic profile and evaluated differential images (differential infrared thermography, DIT), i.e. the temporal gradient of surface temperatures, in order to maximize the contrast of regions with the changing flow regime and the surrounding regions [16]. Gardner et al. carried on this work and supplemented the measurement results with simulations of the laminar-turbulent flow transition position, verifying the localization of the moving transition location [17]. Wolf et al. further studied the effects of different parameters when applying the DIT method in order to optimize the prediction of the transition position [18]. All works use a variety of methods to increase the contrast between the surface region with changing flow condition and the surrounding regions. These include the artificial heating of the surface in order to increase the temperature difference between the fluid and the surface as well as the pitch-averaging over multiple cycles of the pitch-motion. However, on the application on wind turbines in operation, these methods cannot be applied. A thermographic measurement on wind turbines has to cope without these contrast-enhancing methods. Furthermore, the works studied aerodynamic parameters that are typical for helicopters and materials with thermal properties of a helicopter rotor blade. Thus, the capability of thermographic flow visualization for detecting the unsteady laminar-turbulent flow transition position on wind turbines in operation is not yet clear.

Therefore, for the first time, the thermographic visualization of a dynamically changing position of the laminar-turbulent flow transition under unmodified, real aerodynamic and thermal conditions on wind turbines in operation is studied. In particular, a thermographic flow visualization measurement during a wind gust is performed on a real wind turbine under two different natural heating conditions based on the incoming sun radiation and for two different time differences.

First of all, the measurement approach of detecting a changed flow transition position by means of thermography is described in Section 2. This includes the basic measurement principle of thermographic flow visualization as well as the setup and the measurement conditions for the field experiments. In Section 3 the sensitivity limits of the proposed measurement approach for detecting a flow state change are addressed, followed by the presentation of the experimental results in Section 4. The experiments include two natural measurement conditions with a different sensitivity, one with a high and one with a low irradiance on the rotor blade surface. The manuscript finishes with a conclusion and outlook in Section 5.

2 Thermographic visualization of unsteady flow behaviour

2.1 Principle

Locally, different convective heat transfers resulting in different surface temperatures on the rotor blade, which are measurable with an infrared camera, are used to interpret the distribution of the flow states. The convective heat transfer q ̇ conv in each flow regime is directly dependent on the heat transfer coefficient h and the difference between surface and fluid temperature T surface and T fluid:

(1) q ̇ conv = h ⋅ ( T surface − T fluid ) .

An often used approach to estimate the heat transfer coefficient is the Reynold’s analogy, which relates the heat transfer coefficient with the skin friction coefficient c f, and is stated by [19]

(2) Nu Re ⋅ c f = 1 2 .

From the definitions of the Nusselt number Nu = h ⋅ L k f and the Reynolds number R e = U fluid ⋅ L ν with the characteristic length L follows

(3) h = 1 2 ⋅ c f ⋅ U fluid ⋅ k f ν ,

where U fluid is the flow velocity above the surface, k f and ν denote the thermal conductivity coefficient of the fluid and the kinematic viscosity, respectively. According to Eq. (3), an increased friction coefficient c f in the turbulent flow compared to the laminar flow enables the differentiation between both flow regimes by different surface temperatures, due to different heat transfer coefficients. However, the Reynold’s analogy is only approximately valid for the application on rotor blades of wind turbines, since the Prandtl number Pr  ≠ 1 and the pressure gradient in the direction of flow d p d x ≠ 0 . Therefore, Eq. (3) allows a good qualitative explanation for flow-dependent spatio-temporal temperature gradients, but only a rough quantitative measure of the heat transfer coefficient. Note that a precondition of the measurement approach is a difference between the surface and fluid temperature, see Eq. (1). In field measurements this temperature difference exists due to the solar heating of the rotor blade, which increases the surface temperature compared to the surrounding fluid.

In order to detect an unsteady boundary layer flow state, a fast response time of the surface temperature to the change of the boundary layer flow state is necessary. Therefore the heat capacity of the surface should ideally be low and the emissivity high. If the thermal reaction of the surface would be instantaneous, the position of the laminar-turbulent flow transition could be analysed by evaluating each single thermogram during the transient flow change. However, thermal diffusivity of the surface material introduces a time lag, because of which the absolute measured surface temperatures during the transient event show a superposition of both the previous and the changed flow state.

By evaluating only the change of the surface temperatures due to the transient flow event, the thermal flow signatures from before the event are suppressed. In order to visualize the surface temperature variation due to a changed flow state, differential images of the thermographic flow visualizations before and after the flow change event are calculated. As a result, the temporal temperature difference ΔT surface is evaluated. A subtraction of the two heat transfer coefficient profiles h(t 1) and h(t 2) over the chord position before and after the sudden change in inflow speed at t 1 and t 2 yields a difference signal Δh, see Figure 1. This difference Δh is what causes the change of the surface temperature due to the changed flow state. For example, a sudden increase in wind speed effectively results in an increase of the inflow velocity U fluid, which also means an increased angle of attack α at the rotor blade. Consequently, the position of the laminar-turbulent flow transition moves towards the leading edge [20, 21]. Within the region of this position change, the sudden change of the flow from laminar to turbulent results in an increased heat transfer and consequently in a change of the surface temperature T surface.

Figure 1: The difference in the heat transfer Δh between the laminar and turbulent flow state along the normalized chord length causes a temporal temperature gradient in the region of changing flow from turbulent to laminar, if the incoming flow conditions change. The resulting temperature gradient can be made visible by differential thermography.
Figure 1:

The difference in the heat transfer Δh between the laminar and turbulent flow state along the normalized chord length causes a temporal temperature gradient in the region of changing flow from turbulent to laminar, if the incoming flow conditions change. The resulting temperature gradient can be made visible by differential thermography.

2.2 Experimental setup

Thermographic measurements are performed on a 1.5 MW wind turbine type GE1.5sl near Thedinghausen in Germany. The used infrared camera is an IR8800 from the manufacturer InfraTec GmbH, which is sensitive to wavelengths between 7.7 and 10.2 μm and has a noise equivalent temperature difference (NETD) of 25 mK. The camera is positioned in a distance of approx. 100 m from the wind turbine and a 200 mm telephoto-lens is used in order to maximize the geometric resolution and to focus on one section of the rotor blade, see Figure 2. Once per revolution, the camera is triggered, resulting in the acquisition of a thermographic image. The rotational speed of the wind turbine during the measurements is between 11 and 16 rpm. The integration time of the camera is set to 4.45 ms.

Figure 2: Experimental setup of the thermographic flow visualization and wind turbine measurement object. (a) Sketch of the measurement setup with the wind turbine and infrared-camera. The rotor blade to be evaluated is at the time of the image acquisition is horizontal, see (b). (b) Image of the wind turbine analyzed and the field of view (FOV) of the infrared camera in the experimental part of this work.
Figure 2:

Experimental setup of the thermographic flow visualization and wind turbine measurement object. (a) Sketch of the measurement setup with the wind turbine and infrared-camera. The rotor blade to be evaluated is at the time of the image acquisition is horizontal, see (b). (b) Image of the wind turbine analyzed and the field of view (FOV) of the infrared camera in the experimental part of this work.

2.3 Measurement conditions

Two measurements with different thermal conditions concerning the solar irradiance are evaluated. During each measurement a strong gust occurred, which is noticeable by the abrupt increase in the rotational speed shown in Figure 3. The sudden change in wind speed results in a change of the incoming flow velocity U fluid and an increase of the angle of attack α of the incoming flow at the rotor blade. As a result, the transition from laminar to turbulent flow shifts toward the leading edge of the rotor blade, changing the flow state on the rotor blade surface near the leading edge from laminar to turbulent. The calculation of the differential thermographic images is conducted between two flow visualizations shortly before and after the increase in flow speed, i.e. before and at the end of the transient flow change. As a result, the flow change within the time inverval Δt = t 2 − t 1 is evaluated. For the first and the second wind gust, Δt corresponds to 50 s and 20 s, respectively.

Figure 3: Rotational speed of the wind turbine for the two measurements. The sudden increase indicates a wind gust and therefore a transient flow change with a moving laminar-turbulent flow transition towards the leading edge due to an increased angle of attack of the incoming flow. Differential thermographic images of before and after the gust are evaluated. The time stamps of the evaluated images are indicated with vertical dashed lines.
Figure 3:

Rotational speed of the wind turbine for the two measurements. The sudden increase indicates a wind gust and therefore a transient flow change with a moving laminar-turbulent flow transition towards the leading edge due to an increased angle of attack of the incoming flow. Differential thermographic images of before and after the gust are evaluated. The time stamps of the evaluated images are indicated with vertical dashed lines.

3 Sensitivity limits

3.1 Temporal response behaviour of the rotor blade surface

For a thermically ideal rotor blade material with no thermal diffusivity, an instantaneous flow change would result theoretically in an instantaneous change of the surface temperature. However, because of the material capability to store thermal energy and heat transfer limitations, the step response of the surface temperature has a delay and is a superposition of the prior thermal distribution and a change in temperature due to the changed flow state. In order to estimate the delay behavior, a rotor blade sample is heated with a heat radiator. After the surface temperature reaches its steady state, the surface is exposed to a flow and the decreasing surface temperature is measured with an infrared camera. As expected, the course of the surface temperature is steadily decreasing with a large negative gradient, constantly declining as well, until the surface temperature reaches a lower bound asymptotically. A characteristic time constant τ can be calculated by the time until 63.2% of the step height is reached and amounts to 300 s in our experiment. However, caution is needed when trying to use linear system theory and drawing conclusions from this basic experiment to the actual temporal behaviour of the wind turbine rotor blade. First, the thermal response indicates more than a single characteristic time constant. Indeed, comparing the temperature change with the course of an e-function, the first part of the step response has a steeper gradient indicating a faster response. This complex response behaviour is due to the different material layers of the rotor blade surface, which have different thermal properties [15]. Second, the flow condition affects the response time. The actual flow condition for the real rotor blade on a running wind turbine and the flow condition in our basic experiment in the laboratory are indeed different.

Nonetheless, a rough estimate of the achievable response time can be conducted using a lumped capacitance model approach. While a similarity of the different material and geometrical properties can be assumed to a certain extent, an important difference is the different wind speed and Reynolds number. The expected wind speed in a free-field measurement is approx. 30 times higher than the wind speed of the used flow in the laboratory experiment, and the Reynolds number is approx. 60 times higher than in the experiment. According to Simon et al. [15] this reduces the time constant by a factor of 17. Considering further the smaller time constant of the first surface layer, an achievable response time in the order of seconds seems feasible.

According to the studied temporal response behaviour, a detectable change in the surface temperature due to a changed flow state is expected for the analysed time intervals in the field experiments, see. Figure 3. Note, however, that the sensitivity of the thermographic flow visualization does not only depend on the temporal response behaviour but also on the magnitude of the steady surface temperature change, when a laminar flow region became turbulent and vice versa. This temperature difference between steady laminar and turbulent flow regions scales with the temperature difference between the fluid and the surface. Therefore the question remains, what temperature difference between laminar and turbulent flow regions occurs for a solely solar heating of the wind turbine rotor blade, i.e. what sensitivity limit results from the missing artificial heating at wind turbines?

3.2 Solar heating

In addition to the thermodynamic response behavior, the rotor blade heating determines the magnitude of the detectable change of the surface temperature. While artificial heating is commonly used in wind tunnel experiments in order to maximize the contrast, in field measurements on wind turbines in operation the temperature difference depends on the irradiance of the sun, which heats the rotor blade surface. The level of the irradiance E of the sun on the earth’s surface depends mainly on the weather conditions, date and time of day as these factors influence the angle of the incoming sun beams and the distance through the atmosphere reducing the radiation energy. Figure 4(a) and (b) shows the angle of incidence of the incoming sun rays with respect to the ground (horizon) on the longest and shortest day of the year (21.06.22 and 21.12.22), respectively. The angle of incidence is calculated for the location of the examined wind turbine near Thedinghausen, Germany (52°55′59.3″N 9°00′28.1″E) [22]. The orientation of the rotor blade is assumed to be 90° to the ground surface’s normal.

Figure 4: Consideration of the solar heating. (a) Angle of incidence of the solar beams on the rotor blade surface for a summer (21.06.22) and a winter (21.12.22) day. (b) Incoming irradiance E in and absorbed irradiance E absorbed on the rotor blade surface in summer and winter during a single day. The orientation of the rotor blade chord is here 90° to the ground.
Figure 4:

Consideration of the solar heating. (a) Angle of incidence of the solar beams on the rotor blade surface for a summer (21.06.22) and a winter (21.12.22) day. (b) Incoming irradiance E in and absorbed irradiance E absorbed on the rotor blade surface in summer and winter during a single day. The orientation of the rotor blade chord is here 90° to the ground.

The lower the angle of incidence of the sun the higher the heat flux from the sun on the rotor blade surface. For the two days representing summer and winter, the incoming irradiance E in on the rotor blade surface is shown in Figure 4(b) [22]. However, only some of the sun’s energy is actually absorbed into the surface material of the rotor blade. According to Wang et al., the solar absorption coefficient α solar of a rotor blade with the surface material Polyurethane is approx. 0.6 [23]. Applied to the incoming irradiance of the sun at the rotor blade, the absorbed irradiance is

(4) E absorbed = α solar ⋅ E in .

E absorbed of the rotor blade over the course of a day in summer and winter is also shown in Figure 4(b). Accordingly, the maximum absorbed irradiance of a rotor blade with the assumed orientation and location is 375 W/m2 on the 21st of June and 302 W/m2 on the 21st of December.

Note that these heat flux levels are calculated for a clear sky without any cloud coverage. For thermographic flow visualizations on real wind turbines in operation this idealized condition cannot be assumed. On the contrary, the appearance of clouds is often to be expected. Therefore, the reduction of the irradiance on the rotor blade as well as the absorbed portion is evaluated for different cloud coverage levels. The cloud coverage is set to five levels between 0% and 100%. The resulting irradiance analogous to Figure 4(b) is shown in Figure 5. Comparing the maximum daily values, a 25% cloud coverage during summer results in a similar irradiance compared to a clear sky in winter with 283 W/m2 and 302 W/m2, respectively. As a result of this example, not only the seasonal solar radiation but also the cloud coverage is crucial for the heating of the rotor blade surface by the sun’s radiation. Finally, an often accepted cloud coverage of 25%–50% for a field measurement results in at least one order of magnitude lower heating levels of the rotor blade compared to a few thousand Wm−2 what is often applied in laboratory experiments [15, 24, 25].

Figure 5: Incoming irradiance E in and absorbed irradiance E absorbed on the rotor blade surface in (a) summer and (b) winter during one day for different cloud coverage levels between 0% and 100%.
Figure 5:

Incoming irradiance E in and absorbed irradiance E absorbed on the rotor blade surface in (a) summer and (b) winter during one day for different cloud coverage levels between 0% and 100%.

As a result of the lower irradiance in field experiments, a typical thermal contrast between the laminar and turbulent flow regime on the rotor blade surface due to the solar radiation is in the order of 1 K, where the laminar flow regime is warmer compared to the turbulent one. According to the NETD of the camera, this is considered an acceptable contrast since NETD/1 K amounts to 2.5%. If the wind speed increases during a wind gust, the laminar-turbulent flow transition moves towards the leading edge of the blade and previously laminar flow regions become turbulent. Therefore, the expected temperature change in these regions after sufficient time is in the order of 1 K.

4 Experimental results for a wind gust

4.1 High irradiance

Figure 6(a) and (b) show the two thermographic flow visualizations of a rotor blade of a wind turbine in operation. The first image is taken at t 1 = 730 s, see Figure 3(a), shortly before, and the second image at t 2 = 780 s, shortly after the wind gust occurred. Even though many surface contaminations at the leading edge result in many superposed turbulent wedges, regions with laminar flow are visualized by a higher surface temperature T surface, i.e. as brighter regions near the leading edge. The contrast between the laminar and turbulent flow amounts 1.2 K, calculated according to the evaluation areas shown as red and blue rectangle in Figure 6. The sudden change of the incoming flow speed due to the wind gust results in an increased angle of attack. As a consequence, the laminar flow regions become turbulent and the heat flux between the sun-heated surface and the fluid is increased. Therefore the surface temperature decreases, and the previously bright regions at the leading edge get darker in the second image. However, due to internal heat storage, the total surface temperature is, even if cooler compared to the first image, still warmer than the surrounding turbulent flow regions.

Figure 6: Thermographic flow visualization results with the evaluation of the absolute rotor blade surface temperature T surface before (a) and after (b) a sudden increase in the inflow speed, i.e. a change of the laminar-turbulent transition position. Previously laminar regions near the leading edge become turbulent, resulting in a larger cooling of these surface regions compared to the surrounding areas where the flow state does not change.
Figure 6:

Thermographic flow visualization results with the evaluation of the absolute rotor blade surface temperature T surface before (a) and after (b) a sudden increase in the inflow speed, i.e. a change of the laminar-turbulent transition position. Previously laminar regions near the leading edge become turbulent, resulting in a larger cooling of these surface regions compared to the surrounding areas where the flow state does not change.

In order to visualize the regions with a changed flow regime, the differential image between the two thermographic images is calculated and shown in Figure 7. The change from laminar to turbulent flow results in a negative temperature difference ΔT surface. Therefore, those regions that have been identified as previously laminar show a distinct negative temperature difference (black regions at the leading edge) compared to the surrounding regions with a previously already turbulent flow. The measured temperature difference for the two evaluation areas (red and blue rectangle) amounts to 300 mK, which is lower than the previously determined difference between laminar and turbulent flow regions and, thus, indicates an ongoing thermodynamic response of the rotor blade surface. Nevertheless the flow change from laminar to turbulent due to a wind gust, which occurs somewhere within the interval from the first to the second image recording, is already detectable by means of themographic flow visualization.

Figure 7: Differential image of both images shown in Figure 6. The previously laminar regions near the leading edge show a significantly lower temperature difference ΔT surface compared to the surrounding regions that remained turbulent.
Figure 7:

Differential image of both images shown in Figure 6. The previously laminar regions near the leading edge show a significantly lower temperature difference ΔT surface compared to the surrounding regions that remained turbulent.

4.2 Low irradiance

Figure 8(a) and (b) show thermographic images from almost the same radial section of the same rotor blade shortly before (t 1 = 1010 s) and after (t 2 = 1030 s) a different wind gust, i.e. the considered time interval is reduced by a factor of 2.5–20 s. In addition, due to more cloudy weather condition and the winter time, the contrast between the laminar and the turbulent flow before the gust amounts to only 0.4 K, which is a factor of 3 lower compared with the first measurement on a different, sunnier day. While wedge-like image features near the leading edge indicate laminar and turbulent flow regions in the first image, no temperature difference occurs in the second image, i.e. the present flow state is considered being uniform. Note, however, that T surface in the visualization after the wind gust is increased on the entire blade surface, which is not studied here further. The current hypothesis is that the solar radiation changed during the measurement due to the moving clouds.

Figure 8: Thermographic flow visualization results with the evaluation of the absolute rotor blade surface temperature T surface before (a) and after (b) a sudden increase in the inflow speed, i.e. a change of the laminar-turbulent transition position. Previously laminar regions near the leading edge become turbulent, resulting in a lower warming of these surface regions compared to the surrounding areas where the flow state does not change.
Figure 8:

Thermographic flow visualization results with the evaluation of the absolute rotor blade surface temperature T surface before (a) and after (b) a sudden increase in the inflow speed, i.e. a change of the laminar-turbulent transition position. Previously laminar regions near the leading edge become turbulent, resulting in a lower warming of these surface regions compared to the surrounding areas where the flow state does not change.

As a consequence of the low temperature difference between surface and fluid and due to the shorter time interval, the detectable temperature difference due to the change in the flow state from laminar to turbulent is expected to be lower in comparison with the first measurement. The temperature difference ΔT surface between the reference image before and the thermographic image after the wind gust is shown in Figure 9. As expected, the previously laminar (warmer) regions have a lower increase in temperature of 124 mK, while the previously already turbulent (colder) flow regions have a higher increase of 249 mK. The difference in ΔT surface between the surface regions is 125 mK, which is a factor of 3 lower compared to the first measurement, but the changed flow regions near the leading edge are clearly visible as black regions in the visualization. Hence, the measurement method is still capable to visualize the dynamically changing flow state on a wind turbine in operation, despite the low contrast condition and the reduced time interval.

Figure 9: Differential image of both images shown in Figure 8. The previously laminar regions near the leading edge show a significantly lower temperature difference ΔT surface compared to the surrounding regions that remained turbulent.
Figure 9:

Differential image of both images shown in Figure 8. The previously laminar regions near the leading edge show a significantly lower temperature difference ΔT surface compared to the surrounding regions that remained turbulent.

5 Conclusion and outlook

The aim of this work was to prove that thermographic flow visualization is also capable of detecting flow dynamics. According to the thermodynamic behaviour of the rotor blade, the blade surface temperature follows temporal changes of the flow-dependent heat transfer, while the contactless sensing of the surface temperature with an infrared camera provides a high temporal and spatial resolution.

Going beyond the established thermographic measurements on wind turbines with the assumption of a steady state flow, the event of a wind gust and the resulting unsteady boundary layer flow on an operating wind turbine was measured for different solar heating levels. Here, the calculation of a differential thermographic image from the flow visualization before and during or directly after the wind gust was performed. The wind gust results in a sudden movement of the laminar-turbulent transition towards the leading edge, i.e. laminar flow regions that become turbulent flow regions. The regions with a changed flow state could be differentiated from the surrounding regions with unchanged flow regime by a stronger cooling, i.e. a larger reduction of the surface temperature. For the high irradiance measurement condition, a temperature difference of 300 mK between the regions with steady and unsteady flow occurs within the time interval of 50 s. For the low irradiance measurement, a temperature difference of 130 mK is detectable for the time interval of 20 s. Both findings agree with the studied sensitivity limits according to the solar heating and the thermodynamic response behaviour of the rotor blade surface, where a maximal temperature difference in the order of 1 K and a response time in the order of seconds were identified, respectively. Therefore the capability of the introduced measurement approach to visualize flow state changes on a rotor blade surface, for instance due to a change in the incoming flow speed such as occurring during a wind gust, is proven.

Future work should concentrate on the tracking of a blade section of interest during rotation in order to acquire thermographic time series of the surface with an increased imaging rate. This way, sequential differential images can be conducted, analysing the temperature change between consecutive images and with minimal time difference. This is especially important for the evaluation of differences between the highest and lowest point of the rotor, and for studying the dynamic effects of short-time wind gusts and tower passings. While it is expected that a reduced time interval between the two evaluated images can lead to loss of sensitivity due to the thermal inertia of the rotor blade, the respective limits of measurability are not yet clarified quantitatively, i.e. the measurement potentials of thermographic flow visualization on wind turbines in operation are not yet exhausted.

Corresponding author: Felix Oehme, University of Bremen, Bremen Institute for Metrology, Automation and Quality Science, Linzer Str. 13, 28359 Bremen, Germany, E-mail:

Funding source: Bundesministerium für Wirtschaft und Energie (BMWi)

Award Identifier / Grant number: 03EE3013

About the authors

Daniel Gleichauf

Daniel Gleichauf has been a research assistant at the University of Bremen at the Bremen Institute for Measurement, Automation and Quality Science (BIMAQ) in the Department of Production Engineering since 2018. Previously, he completed his Master’s degree in Systems Engineering, also at the University of Bremen. His research interest is the thermographic flow visualization on wind turbines focusing on the laminar-turbulent transition.

Felix Oehme

Felix Oehme joined the University of Bremen in 2019 as a research associate at the Bremen Institute for Measurement, Automation and Quality Science (BIMAQ) in the department of Production Engineering. Previously, he completed a diploma degree in process engineering and natural materials engineering at the TU Dresden. His research interests are the flow behavior and the thermographic detection of flow separation on wind turbines.

Ann-Marie Parrey

Ann-Marie Parrey has been a research associate at the University of Bremen at the Bremen Institute for Measurement, Automation and Quality Science (BIMAQ) in the department of Production Engineering since 2020. Previously, she completed her physics studies at the University of Bremen. Her research interests are the image processing and feature extraction for thermographic flow visualization on running wind turbines.

Michael Sorg

Michael Sorg studied electrical engineering at the University of Karlsruhe and at the University of Bremen with a focus on communications engineering. After his research period at the Institute for Measurement, Control and Systems Engineering at the University of Bremen, he moved to the Bremen Institute for Measurement, Automation and Quality Science (BIMAQ), where he heads the Energy Systems and Materials Testing research group. His research areas include measurement system technology, especially for the application field of wind energy use and for energy system technology.

Nicholas Balaresque

Nicholas Balaresque has been Managing Director of Deutsche WindGuard Engineering GmbH since 2010, and Managing Director of Deutsche WindGuard Wind Tunnel Services GmbH since 2019. He is responsible for the wind tunnel center of Deutsche WindGuard, which comprises 8 wind tunnels. After studying mechanical engineering at the Technical University Federico Santa María (UTFSM) in Valparaiso, Chile, with the last years at the Technical University Hamburg Harburg (TUHH), he took over the management of the Deutsche WindGuard Aeroacoustic Wind Tunnel (DWAA) in Bremerhaven. His professional focus covers wind tunnel experiments for industry and research, as well as free field measurements on wind turbines.

Andreas Fischer

Andreas Fischer studied electrical engineering, completed his PhD at the Technische Universität Dresden in 2009 and his habilitation in 2014. Since 2016, he is a full professor at the University of Bremen in the department of Production Engineering and is head of the Bremen Institute for Measurement, Automation and Quality Science (BIMAQ). He received the Measurement Technology Prize of the AHMT e. V. in 2010, and an ERC Consolidator Grant in 2021. His research areas cover optical measurement principles for flow and production processes, in-process applications of model-based measurement systems, and the investigation of fundamental limits of measurability.

  1. Author contribution: All the authors have accepted responsibility for the entire content of this submitted manuscript and approved submission.

  2. Research funding: This research was funded by Bundesministerium für Wirtschaft und Energie (BMWi) grant number 03EE3013.

  3. Conflict of interest statement: The authors declare no conflicts of interest regarding this article.

References

[1] H. Schlichting and K. Gersten, Grenzschicht-Theorie, Berlin, Heidelberg, New York, Springer, 2006.Search in Google Scholar

[2] C. Dollinger, N. Balaresque, N. Gaudern, D. Gleichauf, M. Sorg, and A. Fischer, “IR thermographic flow visualization for the quantification of boundary layer flow disturbances due to the leading edge condition,” Renew. Energy, vol. 138, pp. 709–721, 2019. https://doi.org/10.1016/j.renene.2019.01.116.Search in Google Scholar

[3] Z. Li, B. Wen, X. Dong, Z. Peng, Y. Qu, and W. Zhang, “Aerodynamic and aeroelastic characteristics of flexible wind turbine blades under periodic unsteady inflows,” J. Wind Eng. Ind. Aerod., vol. 197, p. 104057, 2020. https://doi.org/10.1016/j.jweia.2019.104057.Search in Google Scholar

[4] R. C. Storey, S. E. Norris, and J. E. Cater, “Modelling turbine loads during an extreme coherent gust using large eddy simulation,” J. Phys.: Conf. Ser., vol. 524, p. 012177, 2014. https://doi.org/10.1088/1742-6596/524/1/012177.Search in Google Scholar

[5] L. Menegozzo, A. Dal Monte, E. Benini, and A. Benato, “Small wind turbines: a numerical study for aerodynamic performance assessment under gust conditions,” Renew. Energy, vol. 121, pp. 123–132, 2018. https://doi.org/10.1016/j.renene.2017.12.086.Search in Google Scholar

[6] P. Bæk and P. Fuglsang, “Experimental detection of transition on wind turbine airfoils,” in European Wind Energy Conference & Exhibition 2009, Red Hook, NY, Curran Associates Inc, 2009, pp. 1628–1652.Search in Google Scholar

[7] C. Dollinger, M. Sorg, N. Balaresque, and A. Fischer, “Measurement uncertainty of IR thermographic flow visualization measurements for transition detection on wind turbines in operation,” Exp. Therm. Fluid Sci., vol. 97, pp. 279–289, 2018. https://doi.org/10.1016/j.expthermflusci.2018.04.025.Search in Google Scholar

[8] D. Gleichauf, C. Dollinger, N. Balaresque, A. D. Gardner, M. Sorg, and A. Fischer, “Thermographic flow visualization by means of non-negative matrix factorization,” Int. J. Heat Fluid Flow, vol. 82, p. 108528, 2020. https://doi.org/10.1016/j.ijheatfluidflow.2019.108528.Search in Google Scholar

[9] D. Gleichauf, F. Oehme, M. Sorg, and A. Fischer, “Laminar-turbulent transition localization in thermographic flow visualization by means of principal component analysis,” Appl. Sci., vol. 11, no. 12, p. 5471, 2021. https://doi.org/10.3390/app11125471.Search in Google Scholar

[10] C. Dollinger, N. Balaresque, M. Sorg, and A. Fischer, “IR thermographic visualization of flow separation in applications with low thermal contrast,” Infrared Phys. Technol., vol. 88, no. 2018, pp. 254–264, 2018. https://doi.org/10.1016/j.infrared.2017.12.001.Search in Google Scholar

[11] F. Oehme, J. Suhr, N. Balaresque, D. Gleichauf, M. Sorg, and A. Fischer, “Thermographic stall detection by model-inspired evaluation of the dynamic temperature behaviour,” Appl. Sci., vol. 11, no. 18, p. 8442, 2021. https://doi.org/10.3390/app11188442.Search in Google Scholar

[12] F. Oehme, D. Gleichauf, J. Suhr, N. Balaresque, M. Sorg, and A. Fischer, “Thermographic detection of turbulent flow separation on rotor blades of wind turbines in operation,” J. Wind Eng. Ind. Aerod., vol. 226, p. 105025, 2022. https://doi.org/10.1016/j.jweia.2022.105025.Search in Google Scholar

[13] D. Gleichauf, M. Sorg, and A. Fischer, “Contactless localization of premature laminar–turbulent flow transitions on wind turbine rotor blades in operation,” Appl. Sci., vol. 10, no. 18, p. 6552, 2020. https://doi.org/10.3390/app10186552.Search in Google Scholar

[14] A.-M. Parrey, D. Gleichauf, M. Sorg, and A. Fischer, “Automated detection of premature flow transitions on wind turbine blades using model-based algorithms,” Appl. Sci., vol. 11, no. 18, p. 8700, 2021. https://doi.org/10.3390/app11188700.Search in Google Scholar

[15] B. Simon, A. Filius, C. Tropea, and S. Grundmann, “IR thermography for dynamic detection of laminar-turbulent transition,” Exp. Fluid, vol. 57, no. 5, p. 93, 2016. https://doi.org/10.1007/s00348-016-2178-9.Search in Google Scholar

[16] M. Raffel, C. B. Merz, T. Schwermer, and K. Richter, “Differential infrared thermography for boundary layer transition detection on pitching rotor blade models,” Exp. Fluid, vol. 56, no. 2, p. 30, 2015. https://doi.org/10.1007/s00348-015-1905-y.Search in Google Scholar

[17] A. D. Gardner, C. Eder, C. C. Wolf, and M. Raffel, “Analysis of differential infrared thermography for boundary layer transition detection,” Exp. Fluid, vol. 58, no. 9, p. 122, 2017. https://doi.org/10.1007/s00348-017-2405-z.Search in Google Scholar

[18] C. Wolf, A. D. Gardner, and M. Raffel, “Infrared thermography for boundary layer transition measurements,” Meas. Sci. Technol., vol. 31, no. 11, p. 112002, 2020. https://doi.org/10.1088/1361-6501/aba070.Search in Google Scholar

[19] J. Bons, “A critical assessment of Reynolds analogy for turbine flows,” J. Heat Transfer, vol. 127, no. 5, pp. 472–485, 2005. https://doi.org/10.1115/1.1861919.Search in Google Scholar

[20] E. Gartenberg and A. S. Roberts, “Airfoil transition and separation studies using an infrared imaging system,” J. Aircraft, vol. 28, no. 4, pp. 225–230, 1991. https://doi.org/10.2514/3.46016.Search in Google Scholar

[21] M. S. Genç, İ. Karasu, and H. Hakan Açıkel, “An experimental study on aerodynamics of NACA2415 aerofoil at low Re numbers,” Exp. Therm. Fluid Sci., vol. 39, pp. 252–264, 2012. https://doi.org/10.1016/j.expthermflusci.2012.01.029.Search in Google Scholar

[22] W. F. Holmgren, C. W. Hansen, and M. A. Mikofski, “Pvlib python: a python package for modeling solar energy systems,” J. Open Source Software, vol. 3, no. 29, p. 884, 2018. https://doi.org/10.21105/joss.00884.Search in Google Scholar

[23] C. Wang and Y. Gu, “Research on infrared nondestructive detection of small wind turbine blades,” Results Eng., vol. 15, p. 100570, 2022. https://doi.org/10.1016/j.rineng.2022.100570.Search in Google Scholar

[24] L. A. Joseph, A. Borgoltz, and W. Devenport, “Infrared thermography for detection of laminar-turbulent transition in low-speed wind tunnel testing,” Exp. Fluid, vol. 57, no. 5, p. 77, 2016. https://doi.org/10.1007/s00348-016-2162-4.Search in Google Scholar

[25] W. Lang, A. D. Gardner, S. Mariappan, C. Klein, and M. Raffel, “Boundary-layer transition on a rotor blade measured by temperature-sensitive paint, thermal imaging and image derotation,” Exp. Fluid, vol. 56, no. 6, pp. 1–14, 2015. https://doi.org/10.1007/s00348-015-1988-5.Search in Google Scholar
Received: 2022-12-16
Revised: 2023-01-17
Accepted: 2023-01-17
Published Online: 2023-02-01
Published in Print: 2023-09-28

© 2023 the author(s), published by De Gruyter, Berlin/Boston

This work is licensed under the Creative Commons Attribution 4.0 International License.

Articles in the same Issue

  1. Frontmatter
  2. Editorial
  3. Metrologisches Kompetenzzentrum für Windenergie
  4. Short Communication
  5. Das Kompetenzzentrum für Windenergie (CCW) an der PTB
  6. Research Articles
  7. Rückführung dimensioneller Messungen an Großbauteilen für Windenergiesysteme
  8. Das Virtuelle Koordinatenmessgerät – ein Digitaler Metrologischer Zwilling
  9. Verzahnungsmetrologie für Windenergieanlagen
  10. Parasitic load components for torque and force calibration: a Digital Twin concept
  11. Design der 5 MN·m Drehmoment-Normalmesseinrichtung
  12. Windkanalmesseinrichtung für das bistatische PTB-Wind-Lidar
  13. Ein mobiles System zur vektoriellen Messung der Windgeschwindigkeit
  14. Temperature and pressure effects on the response behavior of anemometers
  15. On-site contactless visualization of the laminar-turbulent flow transition dynamics on wind turbines
https://doi.org/10.1515/teme-2022-0120
Keywords for this article
laminar-turbulent transition; thermographic flow visualization; unsteady flow; wind turbine
Creative Commons
BY 4.0

Articles in the same Issue

  1. Frontmatter
  2. Editorial
  3. Metrologisches Kompetenzzentrum für Windenergie
  4. Short Communication
  5. Das Kompetenzzentrum für Windenergie (CCW) an der PTB
  6. Research Articles
  7. Rückführung dimensioneller Messungen an Großbauteilen für Windenergiesysteme
  8. Das Virtuelle Koordinatenmessgerät – ein Digitaler Metrologischer Zwilling
  9. Verzahnungsmetrologie für Windenergieanlagen
  10. Parasitic load components for torque and force calibration: a Digital Twin concept
  11. Design der 5 MN·m Drehmoment-Normalmesseinrichtung
  12. Windkanalmesseinrichtung für das bistatische PTB-Wind-Lidar
  13. Ein mobiles System zur vektoriellen Messung der Windgeschwindigkeit
  14. Temperature and pressure effects on the response behavior of anemometers
  15. On-site contactless visualization of the laminar-turbulent flow transition dynamics on wind turbines
Sign up now to receive a 20% welcome discount
Institutional Access
How does access work?
Have an idea on how to improve our website?
Please write us.
© 2025 De Gruyter Brill
Downloaded on 8.12.2025 from https://www.degruyterbrill.com/document/doi/10.1515/teme-2022-0120/html
Scroll to top button