Shreyas Ananthan

Transient helicopter rotor wakes in response to time-dependent blade pitch inputs

A time-accurate, free-vortex method was used to predict the evolution of a helicopter rotor wake and the corresponding unsteady rotor airloads in response to time-varying changes in blade pitch. Both steady and maneuvering flight conditions were examined. The modeling was validated using measured rotor responses to transient increases in collective pitch and also for oscillatory collective and cyclic blade pitch inputs. In each case the results showed that there was a temporal lag in the growth and convection of vorticity into the rotor wake, causing significant unsteady effects at the rotor. For transient blade pitch inputs the calculated results showed the bundling of individual vortex filaments below the rotor into vortex rings, a result also verified experimentally. These vortex rings, however, subsequently break down through the development of Kelvin waves. A simulated piloted pull-up maneuver from descending flight was studied, producing evidence that maneuvers can also cause wake vorticity to bundle below the rotor. Large unsteady rotor airloads were produced as the blades encountered this accumulated wake vorticity

Extension of CHIMPS for Unstructured Overset Simulation and Higher-Order Interpolation

Rapid progress of computational science and engineering has allowed our technical knowledge to mature in individual areas of science over the past few decades. Nevertheless, many of the complex interactions between two or more elementary phenomena, such as multi-phase flows, fluid–structure interactions, or reactive flows, remain poorly understood. Analysis of multi-disciplinary large-scale systems has emerged as an important task of modern computational science and engineering. To address this kind of problem, it is more advantageous to build a flexible integration infrastructure in which several independent solvers can be easily coupled with one another, rather than to newly implement all the necessary functionalities and physical/numerical models onto a single code. The reason is twofold: From a physical and numerical point of view, different physical phenomena often involve drastically different ranges of eigenvalues, in which case it is prohibitively difficult to develop a universally efficient numerical method to account for all relevant physics. Therefore, coupling individual solvers best suited for each purpose would be the most effective way to obtain the solution. On the other hand, from a technical point of view, the ability to easily integrate individual solvers ensures that new features or updated models can be included with reasonable effort without disrupting the entire structure. An additional benefit of this strategy is that various existing component modules or solvers can be rapidly combined to solve new problems without a tremendous overhead to create a new environment from scratch. A majority of multi-physics-coupled applications concerns a situation where different physical phenomena occur locally in space. Two typical examples among the recent studies conducted at Stanford University are the simulations for an entire jet engine (Medic et al. 2007) and for rotor blades (Hahn et al. 2006). In the former example, transonic flows in the compressor and turbine are solved by a structured multi-block compressible URANS code, whereas the mixing of low-speed air with liquid fuel from the intricate injector passages and the subsequent chemical reactions are solved by an unstructured reactive large-eddy simulation (LES) solver. In the latter example, by contrast, near-blade transonic flows are solved using a compressible URANS solver, whereas the highly vortical wake region is solved by an energy-conserving incompressible finite-volume URANS solver to better preserve tip vortices. In these cases, code-to-code coupling is realized by each solver receiving data at the domain boundary from a domain of another solver and imposing these data as a boundary condition. Since the grids of each participating solver do not necessarily match in general, search and interpolation are the essential elements for integrated simulations. In modern computing environments, participating codes are usually domain-decomposed onto a very large number of processors, and the data to

Prediction, Analysis, and Validation of Main Rotor Blade Loads in a Prescribed Pull-Up Maneuver

This paper predicts, analyzes, and validates main rotor airloads, structural loads, and swashplate servo loads in a prescribed high-g pull-up maneuver. A multibody finite element structural model is coupled to a transient lifting-line aerodynamic model. The structural model includes a swashplate model to calculate servo loads. The lifting-line model combines airfoil tables, Weissinger-L near wake, time marching free wake, and a semi-empirical dynamic stall model. The maneuver is the Army/NASA UH60A Airloads Program Flight Counter 11029. The primary objective of this paper is to isolate the eects of structural dynamics, free wake, dynamic stall, and pitch control angles, to determine the key loads mechanisms in this flight. The structural loads are first calculated using airloads measured in flight. The measured airloads are then replaced with a lifting-line coupled analysis ‐ that is ideally suited to isolate the eects of free wake and dynamic stall. It is concluded that the maneuver is almost entirely dominated by stall with little or no wake induced eect on blade loads ‐ even though the wake cuts through the disk twice during the maneuver. At the peak of the maneuver, almost 75% of the operating envelope of a typical airfoil lies beyond stall. The mechanism of dynamic stall, in the analysis, consists of multiple cycles within a wide disk area. The peak-to-peak structural loads prediction from the lifting-line analysis show an under-prediction of 10%‐20% in flap and chord bending moments and 50% in torsion loads. The errors stem from the prediction of 4 and 5/rev stall loads. Swashplate dynamics appears to have a significant impact on the servo loads - unlike in level flight ‐ with more than 50% variation in peak loads.