Modern aerodynamics models do not capture important unsteady forces and delays--the failure of the quasi-steady approximation

arXiv:2608.13424 · physics.flu-dyn · Submitted 2026-08-13 · Read on arXiv

Listen

Radio episode about this paper

Transcript

Introduction to the show: ident: Physics. The day in general physics: instruments, methods and results across the field.

Kai: We're looking at the paper: "Modern aerodynamics models do not capture important unsteady forces and delays--the failure of the quasi-steady approximation".

Paper summary: Kai: So, to recap, this paper is testing how well our current aerodynamic models handle unsteadiness in wind turbine blades, specifically where we use the quasi-steady approximation to simplify things. The central claim is that this approximation isn't valid for a wide range of frequencies.

Mira: They found that when you look at dynamic stall—that vortex formation and flow separation—the standard models significantly underestimate the loading on these large rotors, especially near stalling conditions which are common in modern turbines.

Lev: From my perspective building error correction systems, this means if we design controllers based on those steady models, the actual response might be way off because of these unmodeled dynamic events.

Kai: They used a high pressure wind tunnel at Princeton to probe these conditions using the NACA0021 airfoil and tested reduced frequencies up to zero point one one.

Mira: The key finding they highlighted is that they measured low pressure signatures corresponding to both dynamic stall vortex formation and downstream convection all the way down at k equals zero point zero zero five.

Lev: That measurement of the static pressure distribution shows a clear physical reality that current quasi steady frameworks simply can't capture in their assumptions about unsteadiness.

Kai: The paper is essentially saying that if you are designing for modern, massive wind turbines, you need to move beyond those simple quasi steady approaches because they miss important physics.

Mira: This finding matters because the industry has been increasing turbine size so much, and those larger machines amplify the unsteady effects inherent in operating them.

Lev: So for anyone trying to model this on a real system, it suggests that the standard assumptions about how fast inflow changes are too simple for these large-scale machines.

Conclusion: Kai: So, looking at the paper "Modern aerodynamics models do not capture important unsteady forces and delays--the failure of the quasi-steady approximation", it boils down to this: standard aerodynamic models fail when things get dynamic.

Mira: The authors show that because of dynamic stall vortex formation, which temporarily increases lift and delays flow separation, you can't ignore these effects when designing for large blades.

Lev: It means the loading on those massive wind turbine blades will likely be higher than what conventional models predict if they don't account for this unsteady behavior properly.

Kai: The implication is that even slow, low frequency motions can produce lift forces that complicate modeling and design efforts in a way that current tools overlook.

Mira: It presents an opportunity, though, because it shows us exactly what physics is causing those unmodeled forces, and if we can predict and account for them now during the design phase.

Lev: The main thing is this sensitivity to k value means that any approach simplifying unsteady load development for slow motions needs to be re-evaluated against these findings from the study.

Department of Mechanical and Aerospace Engineering, Princeton University

physics.flu-dyn

Submitted: 2026-08-13

Updated: 2026-11-17

Comments: Supplemental materials are attached to the main text after the bibliography starting on page 18

License: http://creativecommons.org/licenses/by/4.0/

Importance score: 75/100

The gist: Modern aerodynamics models do not capture important unsteady forces–the failure of the quasi-steady approximation.

Key concepts

Quasi-Steady Approximation
This is a common modeling technique used to simplify how an airfoil reacts to slow changes in airflow. It assumes the flow adjusts instantly to changes in conditions. The study found this assumption breaks down across all tested frequencies, meaning it cannot accurately capture the true aerodynamic response.
Dynamic Stall
This is a complex unsteady phenomenon where the boundary layer separates from a wing's surface and rolls up into a vortex. This process temporarily increases lift and delays flow separation. If not modeled correctly, large dynamic stall events can cause lasting vibrations and structural damage to wind turbine blades.
Reduced Frequency (k)
This is a non-dimensional parameter ($k = rac{ ext{pi} f_c}{U_0}$) that describes the degree of unsteadiness in the flow. The study showed that for all tested motions, the quasi-steady model fails, indicating that even slow motions (low k) produce significant unsteady effects.
Unsteady Aerodynamic Forces
These are forces and torques generated by rapidly changing airflow around a rotating blade. Conventional models often ignore these forces. The research suggests that neglecting them can lead to significant underestimation of the actual loading, power output, and structural stresses on modern wind turbines.

Terminology

Summary

Modern aerodynamics models do not capture important unsteady forces–the failure of the quasi-steady approximation. p. 0

Where the paper says it

Modern aerodynamics models do not capture important unsteady forces–the failure of the quasi-steady approximation Victoria M. Malarczyk1 , Marcus Hultmark1∗ 1Department of Mechanical and Aerospace Engineering, Princeton University, Princeton, NJ, 08540, USA.

This finding suggests that the loading on wind turbine blades will be significantly underestimated when using conventional models, especially near stalling conditions, which modern wind turbines typically encounter. p. 0
Where the paper says it

We find that the quasi-steady approximation is not valid for a large range of frequencies where it is normally applied. This finding suggests that the loading on wind turbine blades will be significantly underestimated when using conventional models, especially near stalling conditions, which modern wind turbines typically encounter.

The Problem with Quasi-Steady Approximations

Unsteady aerodynamic forces are inherent to many engineering applications, particularly those involving large-scale rotating blades like modern wind turbines. The quasi-steady approximation is often used to simplify the modeling of the unsteady aerodynamic response of an airfoil to slow changes in inflow conditions. p. 0

Where the paper says it

Over the past few decades, wind turbine rotors have grown rapidly and are now exceeding 200 meters in diameter, causing them to operate in conditions where limited empirical data are available, and where models have not been validated. Here, we use a highly pressurized wind tunnel to probe these conditions and evaluate the quasi-steady approximation, often used to simplify the modeling of the unsteady aerodynamic response of an airfoil to slow changes in inflow conditions.

However, this approximation is found not to be valid for a large range of frequencies where it is normally applied. p. 0
Where the paper says it

Here, we use a highly pressurized wind tunnel to probe these conditions and evaluate the quasi-steady approximation, often used to simplify the modeling of the unsteady aerodynamic response of an airfoil to slow changes in inflow conditions. We find that the quasi-steady approximation is not valid for a large range of frequencies where it is normally applied.

This suggests that the loading on wind turbine blades will be significantly underestimated when using conventional models, especially near stalling conditions, which modern wind turbines typically encounter. p. 0
Where the paper says it

We find that the quasi-steady approximation is not valid for a large range of frequencies where it is normally applied. This finding suggests that the loading on wind turbine blades will be significantly underestimated when using conventional models, especially near stalling conditions, which modern wind turbines typically encounter.

Unsteady Flow and Dynamic Stall

Incoming winds are unsteady and turbulent, and as blades rotate through the atmospheric boundary layer, they experience unsteady inflow conditions. The aerodynamic response of the turbine blade to these unsteady conditions is a likely cause to the rise in fatigue and structural failure of turbine blades. p. 1

Where the paper says it

Furthermore, any misalignment of the rotor plane with the inflow wind gives rise to effective velocity variations on the blade that affect how forces vary in time (8). The aerodynamic response of the turbine blade to these unsteady conditions have been identified as a likely cause to the rise in the fatigue and structural failure of turbine blades, disproportionately affecting large-scale blades (9, 10).

Dynamic stall most notably involves the separation and roll-up of the boundary layer into a vortex that temporarily increases lift and delays flow separation. p. 2
Where the paper says it

Even if the blade has fixed pitch, Figure 1D demonstrates how the change in effective angle of attack can occur from unsteadiness in the incoming windspeed 𝑈0. Dynamic stall most notably involves the separation and roll-up of the boundary layer into a vortex that temporarily increases lift and delays flow separation (13,14).

When improperly accounted for, a large dynamic stall event is capable of setting off lasting vibrations that excite aerodynamic flutter of the blade. p. 2
Where the paper says it

Given the cyclic nature of blades, as they rotate through the atmospheric wind shear, the airfoil at a given spanwise location may experience near-stalling conditions repeatedly during normal operation (15) and may trigger some or all of the stages of dynamic stall. When improperly accounted for, a large dynamic stall event is capable of setting off lasting vibrations that excite aerodynamic flutter of the blade (16, 17).

Experimental Methodology

The study utilized a highly pressurized wind tunnel to probe these conditions and evaluate the quasi-steady approximation. The unique test conditions were made possible through use of the High Reynolds number Test Facility (HRTF) located at Princeton University. p. 9

Where the paper says it

Figure 2: Schematic of the experimental setup and wind tunnel facility. (A) Schematic of the High Reynolds number Test Facility (HRTF) at Princeton University.

This facility supports static pressures of up to 240 bar and freestream velocities up to 8. p. 4
Where the paper says it

001 ≤ 𝑘 ≤ 0. 2 Experimental setup The data presented herein were collected in the High Reynolds Number Test Facility, which is a wind tunnel that can operate with air that is pressurized up to 240 bar static pressure (for more information about the facility, see Ref.

5 m/s resulting in Mach numbers less than 0. p. 3
Where the paper says it

Large wind tunnels are costly to run, while high freestream velocities require correspondingly high pitching frequencies to match 𝑘, making tests infeasible in practice and limiting the parameter space that can be evaluated. The current study makes use of a pressurized wind tunnel, using the increased density of the working fluid to match the high 𝑅𝑒𝑐 relevant to modern wind turbines, while doing so at freestream speeds less than 10 m/s.

03. The NACA0021 airfoil model was used, with a chord length of c = 0. p. 4
Where the paper says it

A detailed schematic of the wind tunnel, the NACA0021 airfoil model, and the static pressure measurement system (including the PCB and surface tap locations) is shown in Figure 2. The NACA0021 airfoil profile used in the model is symmetric and can be considered moderately 5

17 m and an aspect ratio AR = 1. p. 2
Where the paper says it

Here, 𝑓 is the frequency of the unsteady event, 𝑐 is the airfoil chord length, and 𝑈0 is the incoming wind speed. The reduced frequency represents the ratio of the convective time scale to the time scale of the effective angle of attack perturbation of the airfoil.

5.

Key Findings on Reduced Frequency

The degree of unsteadiness in the flowfield is categorized by the reduced frequency k = pi fc/U0, which is a non-dimensional parameter. Typically, for perturbations of k > 0. p. 2

Where the paper says it

The reduced frequency represents the ratio of the convective time scale to the time scale of the effective angle of attack perturbation of the airfoil. Typically, for perturbations of 𝑘 > 0.

05 the flowfield response is assumed to be dynamic, such that during the progression from attached to separated flow it encounters flowfield features that cannot be captured using a quasi-steady framework. p. 2
Where the paper says it

Typically, for perturbations of 𝑘 > 0. 05 the flowfield response is assumed to be dynamic, such that during the progression from attached to separated flow it encounters flowfield features that cannot be captured using a quasi-steady framework (18).

The study found that the quasi-steady approximation fails for all tested motions. p. 7
Where the paper says it

For the approximation to be valid at a given 𝑘, the curves shown in Figure 4B-C should all collapse atop the 𝑘 = 0 case. It is evident that it is not the case for any of the 𝑘 of the tested motions, and the quasi-steady approximation fails for all.

Measurements of the static pressure distribution show definitive low-pressure signatures of both dynamic stall vortex formation and downstream convection down to k = 0. p. 7
Where the paper says it

It is evident that it is not the case for any of the 𝑘 of the tested motions, and the quasi-steady approximation fails for all. Measurements of the static pressure distribution show definitive low-pressure signatures of both dynamic stall vortex formation and downstream convection in response to ramp-up motions down to 𝑘 = 0.

005.

Implications for Design

The typical unsteady aerodynamic modeling approach assumes that any unsteadiness in the response can reasonably be neglected if the dynamic stall vortex is not fully formed. The findings suggest that for wind turbines of growing size, even slow, low frequency motions can produce elevated lift forces that complicate modeling and design efforts. p. 7

Where the paper says it

Nonetheless, this finding still represents a non-negligible unsteadiness in the form of both a stall delay and an elevated leading edge suction signature that must be accounted for in order to accurately evaluate the unsteady loading on airfoil sections for a modeled wind turbine blade at very low 𝑘 maneuvers. This suggests that for wind turbines of growing size, even slow, low frequency motions can produce elevated lift forces that complicate modeling and design efforts.

This indicates that conventional models overlook significant physics causing unmodeled forces, torques, and power. p. 7
Where the paper says it

A wind turbine blade effectively never operates in perfectly steady wind conditions. Our finding presents a conundrum for any approach that attempts to simplify the modeling of unsteady load development about an airfoil for slow motions, as this common approximation appears to overlook significant physics that can cause forces, torques and power to go unmodeled.

Conclusion on Modeling

The paper concludes that the quasi-steady approximation fails for all tested k values. The aerodynamic response possesses a high degree of sensitivity to k, and any approach simplifying unsteady load development for slow motions overlooks significant physics. p. 7

Where the paper says it

A wind turbine blade effectively never operates in perfectly steady wind conditions. Our finding presents a conundrum for any approach that attempts to simplify the modeling of unsteady load development about an airfoil for slow motions, as this common approximation appears to overlook significant physics that can cause forces, torques and power to go unmodeled.

This finding presents an opportunity if the resulting increased forces and power output can be predicted and accounted for in the design. p. 7
Where the paper says it

This will evidently contribute to unanticipated failures and higher maintenance costs for wind turbine blades specifically as they are being designed for virtually non-existent, simplified hypothetical operating conditions. However, it should also be pointed out that this finding presents an opportunity as, after all, increased forces and power output from a system that benefits from just that can be highly beneficial, but only if it can be predicted and accounted for in its design.

The study demonstrates that the approximation appears to overlook significant physics causing forces, torques and power to go unmodeled. p. 7
Where the paper says it

A wind turbine blade effectively never operates in perfectly steady wind conditions. Our finding presents a conundrum for any approach that attempts to simplify the modeling of unsteady load development about an airfoil for slow motions, as this common approximation appears to overlook significant physics that can cause forces, torques and power to go unmodeled.

Data Availability

Both the dynamic (unsteady) and static baseline data are publicly available at https://doi.org/10.34770/h9sj-fk79. The results are based on entirely uncorrected data, with no additional considerations of blockage or three-dimensional effects applied. Table S3 and Table S4 provide the tunnel conditions for static and dynamic tests, respectively. p. 4

Where the paper says it

5 m/s and 153 bar static gauge pressure (Table S3). For the dynamic tests, the wind tunnel was operating at approximately 6.

The data processing involved a discrete midpoint integration scheme with the panel method. p. 1
Where the paper says it

This, in combination with the push for offshore wind energy, has resulted in enormous machines where all aspects of them are massive, including the tower height, blade length, and rotor diameter. The large lengthscales involved increase the individual power output of the turbine both by accessing both a larger area of the incoming wind and faster wind speeds higher up in the atmosphere.

The lift coefficient was found through rotating the normal and axial forces through the angle of attack. p. 2
Where the paper says it

Dynamic stall most notably involves the separation and roll-up of the boundary layer into a vortex that temporarily increases lift and delays flow separation (13,14). Given the cyclic nature of blades, as they rotate through the atmospheric wind shear, the airfoil at a given spanwise location may experience near-stalling conditions repeatedly during normal operation (15) and may trigger some or all of the stages of dynamic stall.

The drag component is found by using c d = cn sin(alpha) + ca cos(alpha). p. 0
Where the paper says it

We find that the quasi-steady approximation is not valid for a large range of frequencies where it is normally applied. This finding suggests that the loading on wind turbine blades will be significantly underestimated when using conventional models, especially near stalling conditions, which modern wind turbines typically encounter.

The quarter-chord moment coefficient was found from shifting the point about which the moment is calculated from the leading edge to the quarter chord point. p. 6
Where the paper says it

coefficient reading shown in Figure 4C. The maximum suction value, across all taps, is achieved near the leading edge (typically tap 2) with elevated suction observed in the chordwise extent of the airfoil stretching to the mid chord (taps 1-10).

The study provides a more generalizable result than a specific production airfoil by choosing a more canonical airfoil profile with a relevant thickness ratio. p. 5
Where the paper says it

It is important to note that modern wind turbines use quite complex, proprietary airfoil geometries that vary along the span of the blade, yet the thickness of the airfoil correlates most directly to its stall and unsteady behavior. Thus, despite the detailed geometry of the airfoil not being exactly representative of what is implemented in a modern wind turbine, by choosing a more canonical airfoil profile with a relevant thickness ratio, the present study provides a more generalizable result than a specific production airfoil.

The authors thank Dan Hoffman for help with the airfoil setup and operation of the pressurized wind tunnel facility. p. 4
Where the paper says it

001 ≤ 𝑘 ≤ 0. 2 Experimental setup The data presented herein were collected in the High Reynolds Number Test Facility, which is a wind tunnel that can operate with air that is pressurized up to 240 bar static pressure (for more information about the facility, see Ref.

The authors acknowledge funding from the Army Research Office through Grant W911NF-22-1-0187 and the New Jersey Wind Institute Fellowship Program. The data and materials are available in supplementary material. p. 0
Where the paper says it

edu Aerodynamic unsteadiness is inherent to the operation of many engineering applications, especially those that involve large-scale rotating blades, such as modern wind turbines. Over the past few decades, wind turbine rotors have grown rapidly and are now exceeding 200 meters in diameter, causing them to operate in conditions where limited empirical data are available, and where models have not been validated.

The paper includes Materials and Methods, Figures S1 to S4, Tables S1 to S4, and Caption for Data S1. p. 0
Where the paper says it

edu Aerodynamic unsteadiness is inherent to the operation of many engineering applications, especially those that involve large-scale rotating blades, such as modern wind turbines. Over the past few decades, wind turbine rotors have grown rapidly and are now exceeding 200 meters in diameter, causing them to operate in conditions where limited empirical data are available, and where models have not been validated.

The data set contains the static pressure coefficient measurements from the 32 surface taps and the corresponding integrated lift, drag, and moment coefficient values experimentally obtained by the NACA0021 airfoil model under both static and dynamic (unsteady) conditions. p. 9
Where the paper says it

(I) Representative static pressure measurements obtained from the surface taps on the airfoil model. (J) Chordwise tap locations on the NACA0021 airfoil model.

The data was acquired in the High Reynolds number Test Facility at Princeton University. p. 9
Where the paper says it

Figure 2: Schematic of the experimental setup and wind tunnel facility. (A) Schematic of the High Reynolds number Test Facility (HRTF) at Princeton University.

The measurements were taken sequentially from tap 1 to 32 which repeated until the sampling window was terminated. p. 9
Where the paper says it

(I) Representative static pressure measurements obtained from the surface taps on the airfoil model. (J) Chordwise tap locations on the NACA0021 airfoil model.

The multiplexer operated at 2 kHz resulting in individual pressure sensors being sampled at 62. p. 9
Where the paper says it

’ (F) Photo of the hand-polished NACA0021 airfoil model with pneumatic connectors to the PCB exposed. (G) Image showing the printed circuit board (PCB) with 32 differential pressure sensors and a reference static pressure manifold.

5 Hz. The tunnel diagnostic capability involved measuring the freestream velocity using a Pitot-static tube and static pressure using a transducer. p. 3
Where the paper says it

Large wind tunnels are costly to run, while high freestream velocities require correspondingly high pitching frequencies to match 𝑘, making tests infeasible in practice and limiting the parameter space that can be evaluated. The current study makes use of a pressurized wind tunnel, using the increased density of the working fluid to match the high 𝑅𝑒𝑐 relevant to modern wind turbines, while doing so at freestream speeds less than 10 m/s.

The working fluid density could be computed from the static tunnel pressure and temperature measurements using the ideal gas law. p. 3
Where the paper says it

Large wind tunnels are costly to run, while high freestream velocities require correspondingly high pitching frequencies to match 𝑘, making tests infeasible in practice and limiting the parameter space that can be evaluated. The current study makes use of a pressurized wind tunnel, using the increased density of the working fluid to match the high 𝑅𝑒𝑐 relevant to modern wind turbines, while doing so at freestream speeds less than 10 m/s.

The study investigates the onset of unsteady behavior and evaluates the validity of the quasi-steady approximation at wind turbine relevant Reynolds numbers, over two whole decades of reduced frequencies (0. p. 3
Where the paper says it

As such, compressed gas facilities are uniquely positioned to enable detailed investigations into unsteady phenomena at high 𝑅𝑒𝑐. The focus of the present study is to use this unique testing environment to investigate the onset of unsteady behavior and evaluate the validity of the quasi-steady approximation at wind turbine relevant Reynolds numbers, over two whole 4

001 ≤ k ≤ 0. p. 4
Where the paper says it

decades of reduced frequencies (0. 001 ≤ 𝑘 ≤ 0.

1).

Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.

Improvements for AI systems

  1. Improved Aerodynamic Modeling for Large Rotors: The system can accurately predict loading on wind turbine blades by moving beyond conventional quasi-steady approximations, especially near stalling conditions where the loading on wind turbine blades will be significantly underestimated when using conventional models.

  2. Reduced Frequency Dependent Simulation: The improved AI system can simulate the aerodynamic response across a wide range of reduced frequencies (0.001 to 0.1) by incorporating the observed high degree of sensitivity to k and demonstrating that the quasi-steady approximation fails for all tested motions.

  3. Dynamic Stall Prediction: The system can accurately model dynamic stall events, including vortex formation and convection, by recognizing the low-pressure signature associated with the dynamic stall vortex, which is critical for understanding unsteady effects even at low reduced frequencies.

  4. Design Optimization Under Unsteady Conditions: The improved AI system can optimize wind turbine blade designs by accounting for elevated lift forces that complicate modeling and design efforts caused by slow, low frequency motions, leading to better prediction of unanticipated failures and higher maintenance costs.

Abstract

Aerodynamic unsteadiness is inherent to the operation of many engineering applications, especially those that involve large-scale rotating blades, such as modern wind turbines. Over the past few decades, wind turbine rotors have grown rapidly and are now exceeding 200 meters in diameter, causing them to operate in conditions where limited empirical data are available, and where models have not been validated. Here, we use a highly pressurized wind tunnel to probe these conditions and evaluate the quasi-steady approximation, often used to simplify the modeling of the unsteady aerodynamic response of an airfoil to slow changes in inflow conditions. We find that the quasi-steady approximation is not valid for a large range of frequencies where it is normally applied. This finding suggests that the loading on wind turbine blades will be significantly underestimated when using conventional models, especially near stalling conditions, which modern wind turbines typically encounter.

Related papers