OpenVSP/examples/scripts/python_scripts/report.md

OpenVSP Verification & Validation Studies

OpenVSP Home | OpenVSP Wiki

Introduction

This document outlines a series of verification and validation (V&V) studies that have been conducted for the VSPAERO aerodynamic solver. This HTML document pulls its graphs from the results of a single python script file named "Master_VSP_VV_Script.py" located in the examples/scripts/python_scripts directory distributed with the OpenVSP software. In each of the V&V test cases listed below, there exists published theoretical or experimental results that can be verified for a particular geometric model. The results from VSPAERO are then able to be validated against the published results to determine the accuracy of the solver. Each test begins by modeling a wing or series of wing geometries through the OpenVSP API. The geometric parameters are listed in a table for each study. If a parameter is not listed in the table the default value in OpenVSP should be assumed. Each generated model is saved, and therefore can be loaded in OpenVSP to verify the model accurately represents the geometry of the V&V study. Depending on the particular test case, the results needed for validation must then be identified or calculated. Once results for validation are established, VSPAERO is setup and executed through the Analysis Manager. The VSPAERO analysis inputs are listed in a table for each study, with default values used for items not identified in the tables. Following VSPAERO execution, results are obtained through the Results Manager, compared to the published results, and displayed.

Notation

AR $\equiv$ Aspect Ratio
$\alpha\equiv$ Angle of Attack
$\alpha_0\equiv$ Zero-Lift Angle of Attack
B $\equiv$ Mach Parameter
b $\equiv$ Wing Span
$\beta\equiv$ Sideslip Angle
c $\equiv$ Wing Chord
$C_{Di}\equiv$ Coefficient of Induced Drag
$C_L\equiv$ Coefficient of Lift
$C_{L\alpha}\equiv$ Lift Curve Slope
$C_M\equiv$ Pitching Moment Coefficient
$C_{M\alpha}\equiv$ Pitching Moment Coefficient with Respect to $\alpha$
$\Gamma\equiv$ Circulation
K $\equiv$ Ratio of C_{L\alpha} to $2\pi$
$\Lambda\equiv$ Sweep
$\Lambda_{c/2}\equiv$ Sweep of Half Chord Line
M $\equiv$ Mach Number m $\equiv$ Subsonic (m < 1) or Supersonic (m > 1) Character of Wing Leading Edge
S $\equiv$ Planform Armake_tableea
s $\equiv$ Half Span
$V_{\infty}\equiv$ Freestream Velocity
w $\equiv$ Downwash Velocity
y $\equiv$ Location Along Wing Span

Equations

Lifting Line Theory (LLT) provides the following equations outlined in the NACA TN 3911 A Method for Predicting Lift Increments Due to Flap Deflection at Low Angles of Attack in Incompressible Flow by Lowry and Polhamus:

C_{L\alpha} = \frac{2\pi AR }{2 + \sqrt{(\frac{AR^2 B^2}{K^2})(1 + \frac{tan^2 \Lambda_{c/2}}{B^2}) + 4}} B = \sqrt{1-M^2} K = C_{L\alpha2D}(\frac{180}{\pi})/(2\pi)

Lowry and Polhamus also present a method of determining the downwash velocity and circulation distribution across a wing span:

w(y)=\int_{-s}^{s} \frac{[-d\Gamma(y_0)/dy]dy_0}{4\pi(y - y_0)} \Gamma(y) = {\pi}{V_\infty}(\alpha - \alpha_0)c + {\pi}w(y)c

In Aerodynamics for Engineers, Bertin and Smith develop the monoplane equation:

\frac{{\pi}c}{4s}(\alpha - \alpha_0)sin(\theta) = \sum_{n=1, odd}^{\infty}A_n sin(n\theta)[\frac{{\pi}cn}{4s}+sin(\theta)]

Glauert's method can be used to solve the monoplane equation for (A_n), which can be used to determine the distribution of lift, drag, circulation, and downwash velocity:

\Gamma(y) = 4{V_\infty}s \sum_{n=1, odd}^{\infty}A_n sin(n\theta) w(y) = \frac{V_\infty}{sin(\theta)} {\sum_{n=1, odd}^{\infty}{{A_n}{sin(n\theta)}}} {C_L}(y) = \frac{2}{{V_\infty}c} \int_{-s}^{s} {\Gamma}(y)dy {C_{Di}}(y) = \frac{2}{{V_\infty}^{2}S} \int_{-s}^{s} w(y){\Gamma}(y)dy

The next group of equations are presented for three-dimensional wings in steady supersonic flow in Aerodynamics of Wings and Bodies by Ashley and Landahl:

m = \frac{B}{tan(\Lambda)} C_{L\alpha}tan(\Lambda) = f(m)

This final group of equations are general equations used throughout this V&V study:

\%_{error} = |\frac{experimental - theoretical}{theoretical}| \times 100\%

The central differencing formula:

f'(x)\approx\frac{f(x+h)-f(x-h)}{2h}

Case 1: Hershey Bar

Various Hershey Bar wings with AR ranging from 10 to 60 are modeled through the OpenVSP API for this V&V case. The Hershey Bar wing has been studied and used for aerodynamic verification and validation studies extensively. The Hershey Bar wing has a unit length chord and no sweep. For the studies presented below, a symmetric NACA0012 airfoil is modeled for each Heshey Bar wing. A series of studies are conducted that investigate qualities such as tesselation, aspect ratio, and advanced settings, and their effect on VSPAERO accuracy.

Aspect Ratio Study

In the first Hershey Bar study, the resulting coefficient of lift C_L from VSPAERO's vortex lattice solver is compared to the approximate C_L determined for each angle of attack from Lifting Line Theory. The first plot listed below displays these results.

Next, the lift curve slope C_{L\alpha} is calculated for each vortex lattice and panel method VSPAERO result, defined by a particular AR, by subtracting C_L at two angle of attack flow condition cases and dividing by the total change in \alpha. This method, known as central differencing, cannot be used at the minimum or maximum angle of attack values. For those cases, C_{L\alpha} is identified as the slope of the C_L vs \alpha curve immediately before or after the angle of attack endpoint. These methods are known as forward and backward differencing. The calculated C_{L\alpha} values are compared to C_{L\alpha} determined by Lifting Line Theory at the associated aspect ratios. A 2D lift curve slope of 0.1096622 per degree is assumed, given a theoretical K ratio of 1.0. This plot can be seen in the second graph below.

Aspect Ratio Study Geometry Setup

Airfoil	AR	Root Chord	Tip Chord	\Lambda (°)	Span Tess (U)	Chord Tess (W)	LE Clustering	TE Clustering	Tip Clustering
NACA0012	5 to 60	1.0	1.0	0.0	41	51	0.2	1.0	1.0

Aspect Ratio Study VSPAERO Setup

Case #	Analysis	Method	\alpha (°)	\beta (°)	M	Wake Iterations
1	Sweep	VLM	-20.0 to 20.0, npts: 8	0.0	0.1	3
2	Single Point	Panel	1.0	0.0	0.1	3

Angle of Attack Study

Using the same OpenVSP Hershey Bar wing geomentry as the previous study, aspect ratio is held constant and a VSPAERO alpha sweep analysis is conducted. The goal of this study is to demonstrate how the error in C_{L\alpha} changes at various angles of attack.

Angle of Attack Study Geometry Setup

Airfoil	AR	Root Chord	Tip Chord	\Lambda (°)	Span Tess (U)	Chord Tess (W)	LE Clustering	TE Clustering	Tip Clustering
NACA0012	10	1.0	1.0	0.0	12	17	0.2	1.0	1.0

Angle of Attack Study VSPAERO Setup

Case #	Analysis	Method	\alpha (°)	\beta (°)	M	Wake Iterations
1	Sweep	VLM	-20.0 to 20.0, npts: 9	0.0	0.1	3

Tesselation Study

The next Hershey Bar study investigates the effects of tesselation on VSPAERO VLM results by varying U and W tesselation for a particular AR case. Two charts are generated and can be seen below. The chart on the left displays the percent error between VSPAERO and theoretical lifting line C_{L\alpha} results as chord tesselation increases at various span tesselations. The charts on the right display the same error, but as span tesselation increases at various chord tesselations.

Tesselation Study Geometry Setup

Airfoil	AR	Root Chord	Tip Chord	\Lambda (°)	Span Tess (U)	Chord Tess (W)	LE Clustering	TE Clustering	Tip Clustering
NACA0012	10	1.0	1.0	0.0	5 to 41	9 to 51	0.2	1.0	1.0

Tesselation Study VSPAERO Setup

Analysis	Method	\alpha (°)	\beta (°)	M	Wake Iterations
Single Point	VLM	1.0	0.0	0.1	3

Next, the VSPAERO execution time is considered at various chord and span tesselation values. In conjunction with the plots above, the plots below allow for a comparison between percent error and VSPAERO execution time as tesselation is varied in both the chordwise and spanwise directions.

Tip Clustering Study

This next study looks at the influence of tip clustering on C_L distribution along the Hershey Bar wing span. Tip clustering is varied while aspect ratio, chord tesselation, and span tesselation are held constant. The C_L distribution across the span is compared to the lifting line approximate C_L distribution found using Glauert's method. In addition, lift distribution results are generated in Athena Vortex Lattice v3.37 (AVL) for the Hershey Bar wing using the same setup conditions as VSPAERO. The AVL input file, "Hershey_AR10.avl", and results file, "Hershey_AR10_AVL.dat", can be found in the same directory as the Master V&V Script. The plots below display how tip clustering effects the error between VSPAERO VLM and LLT.

Tip Clustering Study Geometry Setup

Airfoil	AR	Root Chord	Tip Chord	\Lambda (°)	Span Tess (U)	Chord Tess (W)	LE Clustering	TE Clustering	Tip Clustering
NACA0012	10	1.0	1.0	0.0	12	17	0.2	1.0	1.0, 0.5, 1.0

Tip Clustering Study VSPAERO Setup

Analysis	Method	\alpha (°)	\beta (°)	M	Wake Iterations
Single Point	VLM	1.0	0.0	0.1	3

Tip Clustering Study AVL Setup

Nchord	Cspace	Nspan	Sspan	M
30	1.0	20	-3.0	0.1

Span Tesselation Study

The impact of span tesselation on C_L distribution along the Hershey Bar wing span is investigated first in this study. While holding aspect ratio, chord tesselation, and tip clustering constant, span tesselation is varied. The first plot below displays the C_L distribution across the span compared to the lifting line approximate C_L distribution found using Glauert's method. In addition, the lift distribution results computed by AVL for an identical geometry and flow condition input are plotted.

Span Tesselation Study Geometry Setup

Airfoil	AR	Root Chord	Tip Chord	\Lambda (°)	Span Tess (U)	Chord Tess (W)	LE Clustering	TE Clustering	Tip Clustering
NACA0012	10	1.0	1.0	0.0	5 to 41	17	0.2	1.0	1.0

Span Tesselation Study VSPAERO Setup

Analysis	Method	\alpha (°)	\beta (°)	M	Wake Iterations
Single Point	VLM	1.0	0.0	0.1	3

Span Tesselation Study AVL Setup

Nchord	Cspace	Nspan	Sspan	M
30	1.0	20	-3.0	0.1

The impact of span tesselation on C_{Di} distribution along the Hershey Bar wing span is looked at next. As in the plot above, aspect ratio, chord tesselation, and tip clustering are held constant as span tesselation is varied. The plot below displays the C_{Di} distribution across the span compared to the lifting line approximate C_{Di} distribution, and the drag distribution determined by AVL. The AVL input and results file are located in the same directory as the Master V&V Script ("Hershey_AR10.avl" and "Hershey_AR10_AVL.dat" respectively).

Chord Tesselation Study

This next lift distribution study is similar to the previous one, except here the influence of chord tesselation is presented. Span tesselation, aspect ratio, and tip clustering are all held constant. The plots below compare the VSPAERO C_L distribution for the Hershey Bar wing to the lifting line theory approximate solution. The same geometry and flow condition inputs were run in AVL to generate lift distribution results. The AVL results are read in and plotted as well. The results file, "Hershey_AR10_AVL.dat", may be viewed in the scripts directory.

Chord Tesselation Study Geometry Setup

Airfoil	AR	Root Chord	Tip Chord	\Lambda (°)	Span Tess (U)	Chord Tess (W)	LE Clustering	TE Clustering	Tip Clustering
NACA0012	10	1.0	1.0	0.0	12	9 to 51	0.2	1.0	1.0

Chord Tesselation Study VSPAERO Setup

Analysis	Method	\alpha (°)	\beta (°)	M	Wake Iterations
Single Point	VLM	1.0	0.0	0.1	3

Chord Tesselation Study AVL Setup

Nchord	Cspace	Nspan	Sspan	M
30	1.0	20	-3.0	0.1

Chord tesselation's influence on induced drag coefficient distribution for the Hershey Bar wing is then looked at. Aspect ratio, span tesselation, and tip clustering are all held constant while chord tesselation is varied. The lifting line approximate C_{Di} distribution calculated using Glauert's method is plotted below alongside the results generated from VSPAERO. In addition, AVL's drag distribution for the Hershey Bar wing is displayed.

Wake Iteration Study

This next lift distribution study demonstrates the effect of the number of wake iterations on VSPAERO lift coefficient distribution. Span tesselation, chord tesselation, aspect ratio, and tip clustering are all held constant. In the first plot below, the lift distribution is plotted for each wake iteration case. Below the lift distribution plot, the VSPAERO total computation time is plotted for each wake iteration case. Combined, these plots demonstrate the increase in VSPAERO accuracy as the number of wake iterations is increased, but at a cost to overall computation time.

Wake Iteration Study Geometry Setup

Airfoil	AR	Root Chord	Tip Chord	\Lambda (°)	Span Tess (U)	Chord Tess (W)	LE Clustering	TE Clustering	Tip Clustering
NACA0012	10	1.0	1.0	0.0	12	17	0.2	1.0	1.0

Wake Iteration Study VSPAERO Setup

Analysis	Method	\alpha (°)	\beta (°)	M	Wake Iterations
Single Point	VLM	1.0	0.0	0.1	1 to 5

Advanced Settings Study

In this study, the various advanced VSPAERO settings are investigated and compared for the Hershey Bar wing. The settings of interest are the preconditioner and 2nd order Karman-Tsien Mach Correction. The three types of preconditioners available are matrix, Jacobi, and symmetric successive over-relaxation (SSOR). For each VSPAERO run case, only a single advanced setting is varied at a time while all others are left at their default value. The last column of the VSPAERO setup table displays the VSPAERO computation time for each case. The lift distribution is plotted for each case and compared to the theoretical lift distribution determined by Glauert's method. This process is repeated for one, two, and three wake iterations.

Advanced Settings Study Geometry Setup

Airfoil	AR	Root Chord	Tip Chord	\Lambda (°)	Span Tess (U)	Chord Tess (W)	LE Clustering	TE Clustering	Tip Clustering
NACA0012	15	1.0	1.0	0.0	12	17	0.2	1.0	1.0

Advanced Settings Study VSPAERO Setup: Wake Iterations = 1

Case #	Analysis	Method	\alpha (°)	\beta (°)	M	Wake Iterations	Preconditioner	Mach Correction
Default	Single Point	VLM	1.0	0.0	0.1	1	Matrix	Off
1	Single Point	VLM	1.0	0.0	0.1	1	Jacobi	Off
2	Single Point	VLM	1.0	0.0	0.1	1	SSOR	Off
3	Single Point	VLM	1.0	0.0	0.1	1	Matrix	On

Advanced Settings Study VSPAERO Setup: Wake Iterations = 2

Case #	Analysis	Method	\alpha (°)	\beta (°)	M	Wake Iterations	Preconditioner	Mach Correction
Default	Single Point	VLM	1.0	0.0	0.1	3	Matrix	Off
1	Single Point	VLM	1.0	0.0	0.1	3	Jacobi	Off
2	Single Point	VLM	1.0	0.0	0.1	3	SSOR	Off
3	Single Point	VLM	1.0	0.0	0.1	3	Matrix	On

Advanced Settings Study VSPAERO Setup: Wake Iterations = 3

Case #	Analysis	Method	\alpha (°)	\beta (°)	M	Wake Iterations	Preconditioner	Mach Correction
Default	Single Point	VLM	1.0	0.0	0.1	3	Matrix	Off
1	Single Point	VLM	1.0	0.0	0.1	3	Jacobi	Off
2	Single Point	VLM	1.0	0.0	0.1	3	SSOR	Off
3	Single Point	VLM	1.0	0.0	0.1	3	Matrix	On

In the next plot below, the total VSPAERO computation time is plotted for each wake iteration and advanced settings case to provide a summary of the "Exe Time (sec)" column in each of the tables above.

Case 2: Swept Wing

Swept Hershey Bar wings with AR ranging from 10 to 60 are modeled through the OpenVSP API for this V&V case. The wing sweep is measured from the half-chord line for each generated geometry. A series of studies are conducted, with some similarities to the previous Hershey Bar studies. The studies investigate VSPAERO VLM and panel method results generated from variations in aspect ratio, tesselation, and sweep.

Tesselation Study

In the first swept wing study, the influence of chordwise and spanwise tesselation on VSPAERO results is investigated for a particular aspect ratio. A test matrix of wings is created by varying U and W tesselation. For each sweep case, two charts are presented below. First, the percent error between VSPAERO and theoretical lifting line C_{L\alpha} results is presented as chord tesselation increases at various span tesselations. A similar error is displayed in the following chart, but as span tesselation increases at various chord tesselations.

Tesselation Study Geometry Setup

Airfoil	AR	Root Chord	Tip Chord	\Lambda (°)	\Lambda Location	Span Tess (U)	Chord Tess (W)	LE Clustering	TE Clustering	Tip Clustering
NACA0012	10	1.0	1.0	30	0.5	5 to 41	9 to 51	0.2	1.0	1.0

Tesselation Study VSPAERO Setup

Case #	Analysis	Method	\alpha (°)	\beta (°)	M	Wake Iterations
1	Single Point	VLM	1.0	0.0	0.1	3

Sweep Study

This next swept wing test looks to identify if a sweep value plays a role in the accuracy between VSPAERO and 3D theoretical results. Sweep is varied to create a series of wing geometries. For each unique case, the results from VSPAERO VLM and panel method are used to identify the lift curve slope, along with the associated theoretical value. These results are plotted against aspect ratio in the charts below.

Sweep Study Geometry Setup

Airfoil	AR	Root Chord	Tip Chord	\Lambda (°)	\Lambda Location	Span Tess (U)	Chord Tess (W)	LE Clustering	TE Clustering	Tip Clustering
NACA0012	5 to 60	1.0	1.0	0 to 40	0.5	41	51	0.2	1.0	1.0

Sweep Study VSPAERO Setup

Case #	Analysis	Method	\alpha (°)	\beta (°)	M	Wake Iterations
1	Single Point	VLM	1.0	0.0	0.1	3
2	Single Point	Panel	1.0	0.0	0.1	3

For each sweep value, the average percent error between VSPAERO and theoretical C_{L\alpha} is calculated across all AR cases. The plot below identifies how average error changes as wing sweep increases.

Case 3: Bertin-Smith

A Bertin-Smith Wing is modeled based off of the planform described in Aerodynamics for Engineers, by Bertin and Smith. The book identifies a theoretical approximation for C_{L\alpha} on page 202 to be 3.433/rad. A VSPAERO angle of attack sweep analysis is then run. For each \alpha case, the lift coefficient is identified from VSPAERO and displayed alongside the theoretical lift coefficient in the first plot below. C_{L\alpha} is then calculated for each angle of attack by using central differencing. However, central differencing cannot be used for the angle of attack endpoints, so forward differencing or backward differencing is used. These results are compared to the theoretical value to calculate a percent difference, which can be seen in the second plot and results table below.

Bertin-Smith Geometry Setup

Airfoil	AR	Root Chord	Tip Chord	\Lambda (°)	\Lambda Location	Span Tess (U)	Chord Tess (W)	Tip Clustering
NACA0012	5	0.2	0.2	45	0.0	6	33	1.0

Bertin-Smith VSPAERO Setup

Analysis	Method	\alpha (°)	\beta (°)	M	Wake Iterations
Sweep	VLM	-20.0 to 20.0, npts: 9	0.0	0.1	3

Case 4: Warren-12

A New Non-Linear Vortex Lattice Method: Applications to Wing Aerodynamic Optimizations by Oliviu, Andreea, and Ruxandra provides the dimensions for a Warren-12 Wing, a geometry that has been commonly used to verify VLM solvers. This geometry is modeled through the OpenVSP API. Theoretical approximations for C_{L\alpha} and C_{M\alpha} are published to be 2.743/rad and -3.10/rad respectively. Also like the Bertin-Smith Study, a VSPAERO sweep analysis is then run. In the first plot below, the lift coefficient determined by VSPAERO is plotted with the theoretical value for each angle of attack case. C_{L\alpha} is calculated from these results using central differencing where possible, and forward or backward differencing where it is not possible. The percent difference in C_{L\alpha} is displayed for each \alpha case in the second plot. This process is then repeated for C_M and C_{M\alpha}. Last, the values and percent differences for C_{L\alpha} and C_{M\alpha} are summarized in a results table.

Warren-12 Geometry Setup

Airfoil	AR	Root Chord	Tip Chord	\Lambda (°)	\Lambda Location	Span Tess (U)	Chord Tess (W)	Tip Clustering
NACA0012	2\sqrt{2}	1.5	0.5	53.54	0.0	6	33	1.0

Warren-12 VSPAERO Setup

Analysis	Method	\alpha (°)	\beta (°)	M	Wake Iterations
Sweep	VLM	-20.0 to 20.0, npts: 9	0.0	0.1	3

Warren-12 Results

Case 5: Von Karman-Trefftz

A series of Von Karman-Trefftz (VKT) wings are generated using the Karman-Trefftz airfoil available in OpenVSP. A series of studies are then conducted with the first varying the VKT airfoil shape and the second and third varying U and W tesselation. Two API functions, vsp::GetVKTAirfoilPnts and vsp::GetVKTAirfoilCpDist, are used to determine the theoretical VKT airfoil C_{P} distribution. The VSP Slicer is executed to identify the C_{P} distribution at a specified planar slice in each VSPAERO result.

Epsilon Kappa Tau Study

In this study, the shape of the Von Karman-Trefftz airfoil is manipulated to create a series of unique VKT wings. Theoretical C_{P} distributions are determined for each VKT airfoil and compared to the results obtained by executing the CpSlicer analysis with an XZ slice placed at the root of the VKT wing. The theoretical and VSPAERO C_{P} distributions are plotted for each VKT airfoil variation below. Note that if the VKT airfoil is symmetric, there is no difference in the exact C_{P} solution for the upper and lower surface. When the airfoil is not symmetric, there is a noticeable difference in the C_{P} distribution between the upper and lower surfaces.

VKT ε κ τ Study Geometry Setup

Airfoil	\epsilon	\kappa	\tau (°)	AR	Root Chord	Tip Chord	\Lambda (°)	Span Tess (U)	Chord Tess (W)	Tip Clustering
VKT	0.1, 0.2	0.0, 0.1	0, 10	30	1.0	1.0	0.0	41	51	1.0

VKT ε κ τ Study VSPAERO Setup

Analysis	Method	\alpha (°)	\beta (°)	M	Wake Iterations
Single Point	Panel	0.0	0.0	0.1	3

Chord Tesselation Study

This test looks to determine the influence of chord tesselation on VSPAERO results for the VKT wing by varying chord tesselation while holding all other variables constant. A CpSlicer analysis is then performed on each VSPAERO result with an XZ slice defined at the wing root. The VSPAERO C_{P} distribution is then compared to the theoretical C_{P} distribution for the particular VKT airfoil, as seen in the plots below. In addition, XFoil's C_{P} solution for the VKT airfoil is plotted. This result is obtained by first writing a Selig airfoil file from the same coordinate points used to generate the theoretical VKT C_{P} distribution. This file is loaded in XFoil 6.99 and run with identical flow conditions as VSPAERO. The VKT airfoil coordinate file, "VKT_e0.1_k0.1_t10.dat", and C_{P} distribution results file from XFoil, "XFoil_VKT_CpDist.txt", can be found in the airfoil directory of the OpenVSP software.

VKT Chord Tesselation Study Geometry Setup

Airfoil	\epsilon	\kappa	\tau (°)	AR	Root Chord	Tip Chord	\Lambda (°)	Span Tess (U)	Chord Tess (W)	LE Clustering	TE Clustering
VKT	0.1	0.1	10	15	1.0	1.0	0.0	12 to 41	17 to 51	0.2	1.0

VKT Chord Tesselation Study VSPAERO Setup

Analysis	Method	\alpha (°)	\beta (°)	M	Wake Iterations
Single Point	Panel	0.0	0.0	0.1	3

Span Tesselation Study

This next test is similar to the previous one, but instead looks to determine the influence of span tesselation on VSPAERO results for the VKT wing. Chord tesselation, leading and trailing edge clustering, aspect ratio, and airfoil shape are all held constant. A CpSlicer analysis with a slice located along the wing root is then performed on each VSPAERO result. The VSPAERO C_{P} distribution is then compared to the theoretical C_{P} distribution for the VKT airfoil and XFoil's C_{P} solution. This comparison can be seen in the plots below. The VKT coordinate file input to XFoil, "VKT_e0.1_k0.1_t10.dat", and C_{P} distribution results file, "XFoil_VKT_CpDist.txt", are located in the airfoil directory.

VKT Span Tesselation Study Geometry Setup

Airfoil	\epsilon	\kappa	\tau (°)	AR	Root Chord	Tip Chord	\Lambda (°)	Span Tess (U)	Chord Tess (W)	LE Clustering	TE Clustering
VKT	0.1	0.1	10	15	1.0	1.0	0.0	12 to 41	17 to 51	0.2	1.0

VKT Span Tesselation Study VSPAERO Setup

Analysis	Method	\alpha (°)	\beta (°)	M	Wake Iterations
Single Point	Panel	0.0	0.0	0.1	3

Case 6: Ellipsoid

An ellipsoid geometry is modeled for this V&V test case. Four VSPAERO single point analyses are conducted in which angle of attack and slideslip angle are varied. To validate the VSPAERO results, the surface coordinate points along each major axis of the ellipsoid are first calculated using the API function vsp::GetFeatureLinePnts. The CP is then calculated at each surface point with the API function vsp::GetEllipsoidCpDist with the freestream velocity set to 100 in the X direction, as is default in VSPAERO. The vsp::GetEllipsoidCpDist function is based off the algorithms presented in NACA TN 196 by Max M. Munk. Three plots are generated for each VSPAERO run case, displaying the VSPAERO and theoretical CP distribution along each major axis of the ellipsoid.

Ellipsoid Geometry Setup

A Radius	B Radius	C Radius	Center	Span Tess (U)	Chord Tess (W)
1.0	2.0	3.0	(0.0, 0.0, 0.0)	40	41

Ellipsoid VSPAERO Setup

Case #	Analysis	Method	\alpha (°)	\beta (°)	M	Wake Iterations
1	Single Point	Panel	0.0	0.0	0.0	3
2	Single Point	Panel	0.0	20.0	0.0	3
3	Single Point	Panel	20.0	0.0	0.0	3
4	Single Point	Panel	20.0	20.0	0.0	3

Case 7: Supersonic Delta Wing

This final study is similar to a study presented by Dave Kinney at the 2016 OpenVSP Workshop, which compared VSPAERO results to the theory of three-dimensional wings in steady supersonic flow outlined in Aerodynamics of Wings and Bodies by Ashley and Landahl. Two supersonic delta wings are modeled and analyzed with VSPAERO, and the results are displayed alongside experimental data below. The subsonic or supersonic character of the wing's leading edge (m) is identified as a function of the Mach parameter (B) and sweep:

B = \sqrt{1-M^{2}}

m = \frac{B}{tan(\Lambda)}

Supersonic Delta Wing Geometry Setup

Case	Root Airfoil	Tip Airfoil	Span	Root Chord	\Lambda (°)	\Lambda Location	Span Tess (U)	Chord Tess (W)
1	NACA 0004	NACA 0006	20	11.0	45.0	0.0	30	33
1	NACA 0004	NACA 0006	20	11.0	65.0	0.0	30	33

Supersonic Delta Wing VSPAERO Setup

Analysis	Method	\alpha (°)	\beta (°)	M	Wake Iterations
Single Point	Panel	5.0	0.0	1.135, 1.366, 1.894, 2.386, 2.861, 3.369, 3.884, 4.404	3

References

1. Abbott, I. & Doenhoff, A. (1959). Theory of Wing Sections: Including a Summary of Airfoil Data. NY: Dover Publications, Inc.
2. Ashley, H. & Landahl, M. (1965). Aerodynamics of Wings and Bodies. MA: Addison-Wesley Publishing Company, Inc.
3. Bertin, J. J., & Smith, M. L. (1994). Aerodynamics for Engineers(2nd ed.). Englewood Cliffs, NJ: Prentice Hall.
4. Drela, M. & Youngren, H. (2017). AVL 3.36 User Primer.http://web.mit.edu/drela/Public/web/avl/avl_doc.txt.
5. Drela, M. & Youngren, H. (2001). XFOIL 6.9 User Primer.http://web.mit.edu/drela/Public/web/xfoil/xfoil_doc.txt.
6. Hoerner, S. F., & Borst, H. V. (1985). Fluid-Dynamic Lift: Practical Information on Aerodynamic and Hydrodynamic Lift (2nd ed.). Albuquerque/N.M.: Hoerner.
7. Katz, J. & Plotkin, A. (2001). Low-Speed Aerodynamics (2nd ed.). New York: Cambridge University Press.
8. Kinney, D. (2016, August 24). VSPAERO Verification & Next Steps. Live presentation at NASA Ames Conference Center, Mountain View.
9. Lamb, H. (1932). Hydrodynamics (6th ed.). Cambridge, MA: Cambridge University Press.
10. Lowry, J. & Polhamus, E. (1957). A Method for Predicting Lift Increments Due to Flap Deflection at Low Angles of Attack in Incompressible Flow. Virginia: National Advisory Committee for Aeronautics. https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/19930084818.pdf. doi:19930084818
11. Munk, M. (1924). Remarks on the Pressure Distribution over the Surface of an Ellipsoid, Moving Translationally Through a Perfect Fluid. Washington, DC: National Advisory Committee for Aeronautics. https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/19930080983.pdf. doi:19930080983
12. Oliviu, S. G., Andreea K., & Ruxandra M. B. (2016). A New Non-Linear Vortex Lattice Method: Applications to Wing Aerodynamic Optimizations. Quebec: LARCASE Laboratory of Applied Research in Active Controls, Avionics, and Aeroelasticity. https://ac.els-cdn.com/S1000936116300954/1-s2.0-S1000936116300954-main.pdf?_tid=ecb8e5ea-d93a-11e7-afba-00000aacb362&acdnat=1512423465_5a88062ebd97b0cac725b33f68055f9d.

Verification & Validation Studies Performed Using OpenVSP