Software Tools for the Paraglider Computer-Aided Design Guide

AIAA-2001-2001

Software tools for the paraglider computer-aided design guide

Yuri Mosseev *

Scientific-Technical Center OZON

Russia, Moscow, Pomerantsev pereulok, 9 — 8

Fax/Tel. 7-(095)-2457054

E-mail:

http://www.mtu-net.ru/mosseev/rd.htm

7

Abstract

The CAD/CAE system consisting of the design guide and on-line software was developed for ram-air parachutes. The main features of the design guide, the software capabilities and examples of application are discussed. The system components were compiled in CD in wide spread HTML format supported by common internet browsers and for this reason are easy to handle. Each program can be run directly from browser using home PC. All the codes are supplied with the interfaces between each other, data bases, and are compatible with powerful CAD systems like AUTOCAD, 3D-MAX Studio, etc. As a result, this integrated software enables to consider a ram-air parachute as an aeroelastic object and predict its shape in flight, tensions of textile elements, aerodynamic performances.

Nomenclature

AOA – angle of attack [degrees],

Aopt – optimal AOA for max(VX/VY) or max (CL/CD),

AR – paraglider aspect ratio,

Cp – pressure coefficient,

DCp – pressure difference coefficient,

CL – airfoil/wing lift coefficient,

CD – airfoil/wing drag coefficient,

CM – airfoil/wing moment coefficient,

S, X – spanwise and chordwise coordinate,

XM – airfoil/wing pressure center position,

VX – horizontal velocity,

VY – vertical velocity..

Introduction

An outstanding compilation of ram-air parachute design information was presented by Pointer1 in “The Parachute Manual”, by Knacke2, Lindgard3, Butler4.

The state of art in numerical modeling enables to predict important ram-air parachute performances. Gonzales5 modified Prandtl’ lifting line theory to fit ram-air canopy shape features and predict paraglider lift coefficient depending on spanwise curvature, AR —————————————————————

Copyright © 2001 by Y.Mosseev.

* Ph.D., Head of R&D Department

and small AOA. Ross6 used incompressible Navier-Stokes code to compute CL, CD, Cp for an airfoil with inlet and 3D panel code for ram-air rigid wing. The vortices frames and Lagrange-Eulerian mesh-based methods were applied for rigid and flexible gliding parachutes in boook7 and paper8. An aeroelastic approach was applied by Chatzikonstantinou9 for paraglider in steady flight (FEM + vortex lattice method), by Garrard at all. in works10,11,12 for paraglider inflation and flare maneuvers analysis (FEM + mesh-based CFD methods), by Ibos13 at all. (FEM + Navier-Stockes code) and by Mosseev6,14,15 (modified FEM + ALE, discrete vortices and vortex lattice methods) to predict parafoil aeroelastic performances in steady flight.

The author has tried to compile information and recommendations on paraglider design, theoretical background on airfoils and finite span wings, results of serial structural / aerodynamic / aeroelastic analysis, and appropriate software tools into one CD to organize easy-to-use integrated interactive CAD guide15.

CD Design Guide Frame

The design guide was fully compiled in one CD in wide spread HTML format compatible with the most of internet browsers (ex., MS Internet Explorer, Netscape Navigator). The Java and JavaScript applets were specially developed to make guide interactive and simple to handle, use animation, sound effects, perform some simple computing, as well as to get access to external WWW resources.

More complex programs enable to run structural, aerodynamic and flight dynamics analysis to predict and optimize paraglider performances. These programs have interfaces to import/export data between each other, external powerful CAD and 3D design systems, data bases. When coupling the aerodynamic and structural codes it is possible to consider complete fluid-structure interaction and predict the parachute performances in flight.

The guide contains theoretical background on traditional complete airfoils and finite aspect ratio wings, as well as basic graphs on ram-air parachute performances predicted in serial computing. This information helps designer to analyze specific requirements on parachute system and define preliminary the most suitable geometric parameters of paraglider.

Among multiple guide chapters let’s consider in short only those, where the software capabilities could be demonstrated. Note, usually all these guide stages are applied in iterative manner, so that after parachute structure has been modified to optimize certain performance, one should check whether other performances still good enough.

Airfoil / Rib design

The shape of ram-air parachute in flight, and hence its aerodynamic performance, are mostly depends on the ribs shapes.

To begin with, a complete airfoil geometry can be selected from the built-in data base (thousands of foils known in aerospace industry, most of them are in use or tested in wind tunnels), imported from other sources, or input manually. Code DVM-2 predicts in seconds the main aerodynamic performances in wide range of AOA=[0; 20] degrees including critical angles for any airfoil thickness. On this stage the airfoil is considered as a rigid one. Often an aim of optimizations is to increase CL (large critical AOA) and CL/CD ratio (best gliding effect).

The same code evaluates preliminary optimal AOA for wing and effective AOA(S) for parachutes longitudinal sections as a function in spanwise direction S. After possible AOA range for airfoil is defined with some margin, the position of inlet (location of stagnation points) is selected. Then aerodynamic analysis is repeated for airfoil with inlet. The aim is the same as mentioned above, but actually it means to minimise inlet size and guarantee there is no stagnation point on the external side of parachute surface, as otherwise dents on the surface arise and CL/CD ratio drops dramatically. On the other hand, the inlet must be large enough to provide fast wing inflation and fast reinflation, ex. in case of the end cells closure.

Now the problem is to develop suspension and tape carcass system to keep the rib streamlined shape in flight. Since the suspension lines are attached to the rib only in several points, it tends to waviness shaping declining aerodynamic performances. Moreover, for low-porous fabric some cross-ports should be created in rib to balance air pressure in adjacent cells and help to avoid possible end-cell closure. User can run structural analysis in MONSTR-2.2 code and vary the number and location of attachment points, the flares geometry (if flares needed), strengthening carcass on the rib, size and position of cross-ports, textile warp/weft direction in the rib pattern, etc. Due to an interface between MONSTR and DVM-2 one can define aerodynamic load and integral coefficients for airfoil deformed. Special analyser can predict possible wrinkles on the upper wing surface where local negative deformation in textile rib detected. The wrinkles and bulges on the upper surface must be eliminated, especially those located close to airfoil maximal thickness region, as here they can cause flow separation and CL diminishing. Since code DVM-2 can be applied for sufficiently smooth streamlined surfaces, in case the rib boundary became too wavy the more time consuming aerodynamic code PARAD-2.2 applicable to any shapes should de used.

To improve rib shape in flight and to withstand high aerodynamic loads like inflation shock the set of flares can be added to the rib lower side. Despite flares increase significantly total drag and diminish lift-to-drag ratio, sometimes it is more preferable to use them and provide high parachute strength. The code MONSTR-2.2 enables to optimize flares geometry, ex. generate catenary structure.

Since suspension lines gives is as much as 20%… 35% of parasitic drag, essential attention is paid to number, length, diameter and drag of lines, cascades to minimize parachute drag coefficient. This analysis is performed with the aid of coupled DVM-2 and MONSTR-2.2 codes.

After the study for free flight is finished, an aeroelastic analysis will be performed for the control line deflection to evaluate the paraglider aerodynamic performance during maneuvers like turn and brake.

Cell design

Despite airfoil can be designed and optimized in previous stage, it presents real paraglider shape only in several longitudinal sections where the load-bearing rib is located. The parachute cell tends to become “cylindrical” in shape between the ribs and “averaged” longitudinal geometry usually essentially (up to 30%…50%) thicker then the rib. This “averaged” geometry is used to evaluate how aerodynamic performances change in comparison with those obtained for rib.

To improve the canopy shape several intermediate ribs can be installed between load-bearing ones. On this design stage this number is selected (usually 1…3), as well as the way to connect upper and lower surface via ribs positioned in common I I I, VIV, or any other manner.

The stabilizer geometry is also chosen on this stage during simulation for end cell.

Wing design

This design stage includes first of all the choice of canopy planform, wing span, canopy spanwise twisting and curvature, optimal AOA for free flight regime, steering capabilities (brake regime and turn regime), materials, possible permeability, etc. The couple of codes DVM-2 / 3 + MONSTR-2.2 gives preliminary aeroelastic analysis to define the shape and aerodynamic performances including payload drag influence. Essentially 3D and critical aerodynamic effects like spanwise curvature and twisting effects, analysis for AOA close to global flow stall, are considered with the aid of PARAD-2.2 program (3D version).

To improve the canopy shape, the built-in optimizer is applied. It utilizes the soap film model which simulates fabric as an isotropic film with constant tension in any direction, or combination of soap film and orthotropic film model. These models enable to correct the canopy shape to avoid defects like wrinkles (very harmful from aerodynamic point of view as a reason of too early flow separation) and tension concentrators, smooth surfaces. After such an optimization patterns for upper and lower surfaces are generated for each cell separately.

For the AOA selected as an optimal the direction of plumb line passing through the XM point of central chord, where CM(XM)=0, is defined as a possible line to locate the confluence point of suspension lines. Thus, all the lines length will be defined to provide the best gliding performance.

Then aerodynamic perfomances are predicted for brake and turn regimes considered as a quasy-static process.

Loads and strength analysis

An aerodynamic loads acting on ram-air parachute in free flight or during maneuvers predicted in previous stages usually much lower, then during inflation (parachute opening shock) or shape recovery after partial collapse (slope soaring paragliders). A variety of designs and reefing methods of the ram-air parachute complicate the development of simple universal theory to predict an opening shock. In turn, a complete 3D structure-fluid interaction analysis is too complex and time consuming to be recommended as a CAD/CAE component of the design guide.

For this reason a simple code realizing theory for ring parachute inflation with some corrected dependencies is applied. The maximal load and appropriate phase (canopy current size) are predicted and then a quasi-static strength analysis is performed by code MONSTR-2.2. The quasi-static approach seems to be quite reasonable, as it gives excellent results for ring, conic and cross parachutes. As an aerodynamic load acts mostly on the lower surface, one may limit strength analysis by considering only this surface with suspension lines and flares.

Software Tools

An integrated software MONSTR has been embedded into the design guide CD. It consists of several programs with appropriate interfaces to exchange data between each other and CAD systems. All the codes were certified and registered by ROSAPO (Russian Government Legal Agency) and are good for modern PC like Pentium-II/III with at least 512Mb RAM and CPU frequency 500MHz and above. Typically computing takes several minutes except 3D aerodynamic analysis which takes up to several hours. To speed up computing some simplified analytical/experimental dependencies were built into codes.

Structural program MONSTR-2.2

The program is based on modified finite element method and can perform static/dynamic structural analysis for any kind of textile constructions including ram-air parachutes. It can predict the shape and structure stressing state for known external loads. An aerodynamic load can be assigned from built-in predictors or imported from any of the aerodynamic program discussed below. In turn, the structure shape can be exported to aerodynamic code. The program can work with huge textile deformation and structure displacements, wrinkles and folds on the surface, optimize the structure to avoid the surface defects and tension concentrators. An internal editor enables to generate the mesh for ram-air parachutes in minutes.

Aerodynamic programs DVM-2 and DVM-3

Both codes are based on discrete vortices method for 2D and 3D problems respectively. The DVM-2 is very fast and simple 2D code, but actually has analytical 3D extension subroutine to predict even 3D performances for finite span parachute + payload system. It is used to define pressure distribution on the parachute surfaces, main aerodynamic coefficients for an airfoil and ram-air canopy in wide range of AOA including flow stall, system velocity vector, position of confluence point to provide desirable AOA in flight. The code takes into account the canopy shape in flight (imported from MONSTR-2.2 code if unknown), aerodynamic performances of suspension lines, flares, stabilizers and payload. Actually, in many cases information obtained from DVM-2 code is sufficient for preliminary design.

Code DVM-3 may be used to predict some 3D effects not embraced by DVM-2, like the lift and drag dependency on spanwise curvature, wing twisting, wing nonuniform thickness in spanwise direction, etc. To simplify the design process these dependencies are presented as ready-to-use graphics. Also the pressure on the surface in spanwise direction is predicted more correctly than by DVM-2 code.

Both DVM codes are very fast because the vortices-based methods applied resolve N-dimensional problem as a (N-1)-dimensional one. On the other hand, as a result of this approach they consider paraglider as a streamlined body in flow under small sub-critical AOA. In case of complex body shape, flow separation, large AOA, the more powerful code PARAD can be used.

Aerodynamic program PARAD-2.2

This is a 2D/3D CFD code based on Lagrangian-Eulerian mesh method for compressible fluid for Mach number =0.1 ... 6.0. The software enables to run dynamic problems, but in this paper only static results are presented computed as a time limit due to small dampening forces applied in structural solution. In case the pressure function is not constant in time it is averaged in time. Since the code is CAD compatible, the parachute geometry can be imported right from the drawings or MONSTR-2.2 code and then mesh is generated semi-automatically with minimal operator intervention. In PC version 3D mesh size is limited by RAM value and usually not exceeds 200x100x100. The mesh generation procedure takes only minutes. Since 3D computing is very slow, in many cases 2D approach can be recommended (ex., to predict airfoil performances).