Final Report Summary - RODTRAC (Robustness of distributed micron-sized roughness-element for transition control)
Executive Summary:
The aim of the RODTRAC project is to understand the impact of acoustic waves and freestream turbulence on stability characteristics of three-dimensional boundary layers in presences of roughness elements. The investigations include both numerical and experimental studies.
Detailed laminar-turbulent transition measurements have been performed in ITAM’s wind tunnel, which has an excellent flow quality suitable for this kind of studies. Experiments have verified the possibility of transition control by distributed micron-size roughness elements. Further, effects of freestream turbulence level and acoustic perturbation on efficiency of this control method have been investigated.
Accurate numerical simulations have been performed to study similar issues. Analysis of interaction of freestream turbulence and surface roughness elements showed new and non-intuitive results, explaining some recent flight tests.
The performed research work has increased our understanding of interaction of surface inhomogeneity and external perturbations sources (e.g. freestream turbulence and acoustic perturbations). The obtained knowledge is of great importance for design of wings with laminar flow.
Project Context and Objectives:
Transition control in three-dimensional flows by means of Distributed Micron-Sized Roughness-elements (DMSR) has been under focus during the last decade. The successful experiments of Prof. Saric and his research group have created lot of attentions and interest. In the reported work by Saric et al. always a significant positive effect of DMSR on delay of transition has been observed. Similar attempts to control transition in three-dimensional flows by means of DMSR have not been successful in the same degree. It has argued that the differences in the outcome of the experiments may be due to small differences in the level of noise (acoustic and free-stream turbulence) in the wind tunnels used in the experiments.
To our knowledge, there are no investigations that really address the sensitivity and robustness of the MSR for transition control in three-dimensional flows. The objective of the proposed activities is to, numerically and experimentally address these issues.
The carefully performed numerical simulations will allow us to characterize the effects of acoustic and vortical perturbations separately or simultaneously. Highly accurate spectral methods used here will make a controlled variation of ‘noise’ level possible and thereby lets us to understand the limitations of the DMSR approach. The direct numerical simulations will be accompanied with stability calculations and receptivity analyses using the Parabolised Stability Equations (PSE). These calculations are much faster and cheaper in terms of computational costs. This will make it possible to do a wide variation of parameters.
Carefully performed experimental investigations in a low-disturbance environment will also generate valuable information and data. Experiments with controlled acoustic perturbations and turbulence level will complete the direct numerical simulations and will be used to validate the numerical results. Variation of flow conditions and distribution of roughness elements will give us the possibility to examine the off-design behaviour of flow in presence of DMSR.
Project Results:
A large set of experimental investigations has been performed to study effects of freestream turbulence and acoustic field on transition in a three-dimensional swep-wing boundary layer. The investigations cases of ‘naturaly’ smooth surface and cases with distributed micron-size roughness (DMSR) elements, with emphasis on robustness of DMSR-elements for transition control. Briefly, the following most important results are obtained in the whole experimental study within the RODTRAC project.
1. The DMSR-elements are able to stabilize the swept-wing boundary layer under study in cases of low turbulence levels (the LTL-cases).
2. No possibility of a significant transition delay by DMSR-elements is found in the studied cases of enhanced turbulence levels (the ETL-cases).
3. In absence of DMSR-elements, the acoustic fields are found do not influence basically the “natural” transition scenarios and location in both LTL- and ETL-cases.
4. In presence of DMSR-elements, the acoustic field either does not influence significantly or influences very weakly destabilizing the transition location in those LTL-cases, in which the DMSR stabilization is absent, as well as in all studied ETL-cases, in which the destabilizing effect can be rather strong, sometimes.
5. The most important results are obtained in the LTL-regimes, in which the presence of the DMSR-elements stabilizes the boundary layer flow and displaces the transition downstream. In these cases harmonic acoustic fields in a certain frequency range are found to be able to provide additional stabilization of the boundary layer and to displace the transition location even farther downstream!
6. Destabilizing effects of acoustic field in presence of stabilizing effect of DMSR elements are found to be either very weak or negligible. In other words, the DMSR-based transition control on a swept-wing is found to be robust with respect to acoustic fields.
Direct numerical simulations have been performed in order to investigate the role of low-intensity freestream turbulence on generation and evaluation of crossflow instability modes in a three-dimensional boundary layer. Independent of the level of the freestream turbulence studied here, if the roughness elements are present the dominating disturbances were the stationary crossflow vortices. The laminar-turbulent transition occurred by breakdown of these stationary vortices due to the secondary instability triggered by unsteady perturbations originating from the freestream turbulence. It was observed that, in the presence of the roughness element, increasing the level of freestream turbulence by one order (i.e. from T u = 0.04% to 0.4%) could shift the transition location downstream up to 10% of the chord. Moreover, the receptivity to freestream turbulence in the presence of roughness elements and the range studied here was found to have a linear behavior. It was observed that the presence of merely freestream turbulence in absence of roughness elements gives rise to the emergence of weak travelling crossflow vortices. These vortices however also broke down near the end of the computational domain. The transition location appeared to fluctuate more as compared to the cases where an array of roughness elements was present. Here, in the presence of the roughness elements, the freestream turbulence seems to act as source of flow unsteadiness which seeds the secondary instability.
In the range of parameters studied here, increasing the roughness height for same turbulence intensity was found to have a stabilising effect. This could be explained by following. The dominant secondary-instability modes which are responsible for breakdown of stationary crossflow vortices here are of the so-called z-mode type. The growth of this type of modes scales with the spanwise shear. It was observed that for the current setup an increase of roughness height from 12μm to 36μm resulted in a weaker spanwise gradient of the flow modified by the presences of stronger crossflow vortices. Consequently, the triggering of the secondary instability was delayed. It must be noted that this behavior points to the presence of a complex nonlinear mechanism as was suggested by Hunt (2011). The observations made in the present study may explain the non-intuitive results of the recent flight experiment by Saric et al. (2015), where transition occurred further downstream on the wing model with painted surface compared to the model with polished surface.
Acoustic receptivity of distributed micron-sized roughness elements is investigated for a swept-wing boundary layer by means of direct numerical simulations (DNS). The interaction of acoustic waves and roughness elements excites unsteady crossflow vortices. It has been shown that the steady crossflow disturbances are likely to dominate over their unsteady counterparts for characteristic acoustic noise levels apparent in the free flight. Unfortunately, a domain dependency study revealed that the previous simulations have been domain dependent which renders the conclusions questionable. The reason for this behaviour is due to existence of two outflow boundaries and their treatment in our numerical solver NEK5000. The open (outflow) boundary condition in NEK5000 arises as a natural boundary condition from the variational formulation of Navier–Stokes, which sets the pressure effectively to zero at both outflow positions. Therefore, the pressure level is set to zero in two spatial locations and obviously by changing the outflow the pressure distribution will change. Just recently we succeed to come up with a modified outflow boundary condition is used to tackle this problem in which the pressure is prescribed at the outflow boundaries. To use this boundary condition the pressure for the basic state should be known a priory. Two-dimensional test cases are designed to validate the modified outflow boundary condition. It has been shown that by employing this boundary condition both baseflow and nonlinear perturbation (acoustic) simulations becomes domain independent. Due to lack of time, we were not able to perform the simulations corresponding to RODTRAC experiments in framework of project. These simulations will be performed and reported in near future.
Potential Impact:
The RODTRAC project has delivered upstream aerodynamics research that has increased our knowledge about the interaction of different source of perturbations on transition in swept-wing boundary layers. The obtained knowledge contributes to improvement of the accuracy of performance prediction for aircraft with laminar wings, allowing design of advanced and innovative aircraft.
The new knowledge can directly be transferred to the European aircraft manufacturers through their involvement in the Joint Technology Initiative – Clean Sky.
List of Websites:
Not applicable.
The aim of the RODTRAC project is to understand the impact of acoustic waves and freestream turbulence on stability characteristics of three-dimensional boundary layers in presences of roughness elements. The investigations include both numerical and experimental studies.
Detailed laminar-turbulent transition measurements have been performed in ITAM’s wind tunnel, which has an excellent flow quality suitable for this kind of studies. Experiments have verified the possibility of transition control by distributed micron-size roughness elements. Further, effects of freestream turbulence level and acoustic perturbation on efficiency of this control method have been investigated.
Accurate numerical simulations have been performed to study similar issues. Analysis of interaction of freestream turbulence and surface roughness elements showed new and non-intuitive results, explaining some recent flight tests.
The performed research work has increased our understanding of interaction of surface inhomogeneity and external perturbations sources (e.g. freestream turbulence and acoustic perturbations). The obtained knowledge is of great importance for design of wings with laminar flow.
Project Context and Objectives:
Transition control in three-dimensional flows by means of Distributed Micron-Sized Roughness-elements (DMSR) has been under focus during the last decade. The successful experiments of Prof. Saric and his research group have created lot of attentions and interest. In the reported work by Saric et al. always a significant positive effect of DMSR on delay of transition has been observed. Similar attempts to control transition in three-dimensional flows by means of DMSR have not been successful in the same degree. It has argued that the differences in the outcome of the experiments may be due to small differences in the level of noise (acoustic and free-stream turbulence) in the wind tunnels used in the experiments.
To our knowledge, there are no investigations that really address the sensitivity and robustness of the MSR for transition control in three-dimensional flows. The objective of the proposed activities is to, numerically and experimentally address these issues.
The carefully performed numerical simulations will allow us to characterize the effects of acoustic and vortical perturbations separately or simultaneously. Highly accurate spectral methods used here will make a controlled variation of ‘noise’ level possible and thereby lets us to understand the limitations of the DMSR approach. The direct numerical simulations will be accompanied with stability calculations and receptivity analyses using the Parabolised Stability Equations (PSE). These calculations are much faster and cheaper in terms of computational costs. This will make it possible to do a wide variation of parameters.
Carefully performed experimental investigations in a low-disturbance environment will also generate valuable information and data. Experiments with controlled acoustic perturbations and turbulence level will complete the direct numerical simulations and will be used to validate the numerical results. Variation of flow conditions and distribution of roughness elements will give us the possibility to examine the off-design behaviour of flow in presence of DMSR.
Project Results:
A large set of experimental investigations has been performed to study effects of freestream turbulence and acoustic field on transition in a three-dimensional swep-wing boundary layer. The investigations cases of ‘naturaly’ smooth surface and cases with distributed micron-size roughness (DMSR) elements, with emphasis on robustness of DMSR-elements for transition control. Briefly, the following most important results are obtained in the whole experimental study within the RODTRAC project.
1. The DMSR-elements are able to stabilize the swept-wing boundary layer under study in cases of low turbulence levels (the LTL-cases).
2. No possibility of a significant transition delay by DMSR-elements is found in the studied cases of enhanced turbulence levels (the ETL-cases).
3. In absence of DMSR-elements, the acoustic fields are found do not influence basically the “natural” transition scenarios and location in both LTL- and ETL-cases.
4. In presence of DMSR-elements, the acoustic field either does not influence significantly or influences very weakly destabilizing the transition location in those LTL-cases, in which the DMSR stabilization is absent, as well as in all studied ETL-cases, in which the destabilizing effect can be rather strong, sometimes.
5. The most important results are obtained in the LTL-regimes, in which the presence of the DMSR-elements stabilizes the boundary layer flow and displaces the transition downstream. In these cases harmonic acoustic fields in a certain frequency range are found to be able to provide additional stabilization of the boundary layer and to displace the transition location even farther downstream!
6. Destabilizing effects of acoustic field in presence of stabilizing effect of DMSR elements are found to be either very weak or negligible. In other words, the DMSR-based transition control on a swept-wing is found to be robust with respect to acoustic fields.
Direct numerical simulations have been performed in order to investigate the role of low-intensity freestream turbulence on generation and evaluation of crossflow instability modes in a three-dimensional boundary layer. Independent of the level of the freestream turbulence studied here, if the roughness elements are present the dominating disturbances were the stationary crossflow vortices. The laminar-turbulent transition occurred by breakdown of these stationary vortices due to the secondary instability triggered by unsteady perturbations originating from the freestream turbulence. It was observed that, in the presence of the roughness element, increasing the level of freestream turbulence by one order (i.e. from T u = 0.04% to 0.4%) could shift the transition location downstream up to 10% of the chord. Moreover, the receptivity to freestream turbulence in the presence of roughness elements and the range studied here was found to have a linear behavior. It was observed that the presence of merely freestream turbulence in absence of roughness elements gives rise to the emergence of weak travelling crossflow vortices. These vortices however also broke down near the end of the computational domain. The transition location appeared to fluctuate more as compared to the cases where an array of roughness elements was present. Here, in the presence of the roughness elements, the freestream turbulence seems to act as source of flow unsteadiness which seeds the secondary instability.
In the range of parameters studied here, increasing the roughness height for same turbulence intensity was found to have a stabilising effect. This could be explained by following. The dominant secondary-instability modes which are responsible for breakdown of stationary crossflow vortices here are of the so-called z-mode type. The growth of this type of modes scales with the spanwise shear. It was observed that for the current setup an increase of roughness height from 12μm to 36μm resulted in a weaker spanwise gradient of the flow modified by the presences of stronger crossflow vortices. Consequently, the triggering of the secondary instability was delayed. It must be noted that this behavior points to the presence of a complex nonlinear mechanism as was suggested by Hunt (2011). The observations made in the present study may explain the non-intuitive results of the recent flight experiment by Saric et al. (2015), where transition occurred further downstream on the wing model with painted surface compared to the model with polished surface.
Acoustic receptivity of distributed micron-sized roughness elements is investigated for a swept-wing boundary layer by means of direct numerical simulations (DNS). The interaction of acoustic waves and roughness elements excites unsteady crossflow vortices. It has been shown that the steady crossflow disturbances are likely to dominate over their unsteady counterparts for characteristic acoustic noise levels apparent in the free flight. Unfortunately, a domain dependency study revealed that the previous simulations have been domain dependent which renders the conclusions questionable. The reason for this behaviour is due to existence of two outflow boundaries and their treatment in our numerical solver NEK5000. The open (outflow) boundary condition in NEK5000 arises as a natural boundary condition from the variational formulation of Navier–Stokes, which sets the pressure effectively to zero at both outflow positions. Therefore, the pressure level is set to zero in two spatial locations and obviously by changing the outflow the pressure distribution will change. Just recently we succeed to come up with a modified outflow boundary condition is used to tackle this problem in which the pressure is prescribed at the outflow boundaries. To use this boundary condition the pressure for the basic state should be known a priory. Two-dimensional test cases are designed to validate the modified outflow boundary condition. It has been shown that by employing this boundary condition both baseflow and nonlinear perturbation (acoustic) simulations becomes domain independent. Due to lack of time, we were not able to perform the simulations corresponding to RODTRAC experiments in framework of project. These simulations will be performed and reported in near future.
Potential Impact:
The RODTRAC project has delivered upstream aerodynamics research that has increased our knowledge about the interaction of different source of perturbations on transition in swept-wing boundary layers. The obtained knowledge contributes to improvement of the accuracy of performance prediction for aircraft with laminar wings, allowing design of advanced and innovative aircraft.
The new knowledge can directly be transferred to the European aircraft manufacturers through their involvement in the Joint Technology Initiative – Clean Sky.
List of Websites:
Not applicable.