PREDIMRC

Manual

Update of 08.04.07, valid for PredimRC_2.1

Index

1. INTRODUCTION

2. BLOCK-DIAGRAM PREDIMRC

3. SOME DEFINITIONS

3.1. GEOMETRICAL MAGNITUDES

3.2. AERODYNAMIC SIZES
3.3. CONSTANT PROFILE

4
4
5

4. KEY STAGES OF THE DESIGN

4.1. TO DEFINE ITS NEED

4.2. TO ESTIMATE THE RE OF REFERS FLIGHT (RE_REF)
4.3. TO CHOOSE THE PROFILES
4.4. TO CALCULATE POLAR
4.5. TO INTEGRATE THE POLAR ONES IN PREDIMRC
4.6. TO OPTIMIZE LENGTHENING
4.7. TO DIMENSION AND REGULATE THE MODEL
4.8. TO OPTIMIZE the GEOMETRY Of WING
4.9. TO EVALUATE THE PERFORMANCES OF ITS MODEL
4.10. TO DIMENSION THE PROPULSION
4.11. TO DIMENSION THE CONTROL SURFACES
4.12. TO DIMENSION THE SERVOS

6
6
6
7
11
14
15
17
18
20
21
21

5. CONCLUSION

22

1. Introduction
PredimRC is a software of design of flying models functioning on Microsoft Excel. Conceived initially for
predimensioning, it became with the wire of the evolutions a tool of aerodynamic design complete and
powerful. The majority of the flying models can thus be conceived with a maximum of effectiveness and
in a minimum of time.
PredimRC is based on the following outstanding points:
-

polar of profiles numerical exits of blowers

original method of the calculation of optimal lengthening for the machines of performances
(60pouces, F3J, F3F, F3B, F3D,)
calculations with variable Re, on the globality of the model (wing, stab, fuselage)
complete equations of the taking into account of the wake of the wings for the problems of
stability, of centering and chock of stab (Mr. Sherrer/T. Plato)
original method of calculation of the point of operation of an electric model

PredimRC makes it possible to obtain the following results:

-

optimization of lengthening
optimization of the geometry by VLM
curves of performances of the model, for 3 profiles X 2 masses
performances with the engine
adjustments (centerings, chocks) and stability
dimensioning of the servos

In practice, the results given by PredimRC were checked on many models - micro model with the
sailplane of competition F3F while passing by electric racers - and showed the great robustness and
reliability of this software.

REMARKS:

PredimRC can be used in two radically opposite ways:

direct design, starting from a white sheet.
retro-design, starting from a model existing more or less advanced.

PredimRC is compartmentalized, which makes it possible to separate two aspects from the design:
the performance part, which uses the polar profiles exits of numerical blower.
the adjustments part, which uses the data profiles Cm0 and Alpha0.

The adjustments parts and performances are of course complementary, but one can thus perfectly
use one without the other according to what one seeks. It is a very important aspect: useless to
seize polar profiles if one does not seek to calculate the performances, but simply to regulate the
model for a healthy flight.

Do not hesitate to use the integrated assistance of PredimRC: the majority of the cells are
commented on, with target values or explanations. If the comments are not posted, go in small
Outils/Option, Affichage miter, then to only choose indicating.

PredimRC uses macros VisualBasic. By defect, Excel asks for the confirmation of activation of these
macros, which thus should be accepted. It is also possible to configure Excel to accept these
macros automatically, in small Outils/Macros/Scurit. On certain nonFrench versions of Excel, if the
macros do not function, it is necessary to replace the Graphic word by Chart in the macros.

PredimRC is reserved for a private use, and can thus be neither copied, neither distributed, nor used
within the framework of a commercial application without the opinion of the author. In the same way,
this software is delivered in the state, the responsibility for the author could not be committed in
some manner that it is in the case of an accident implying a model designed thanks to this software.

1. PredimRC block-diagram
2. Some definitions

3.1. Geometrical magnitudes

Lengthening (
scale
defines mathematically by:
A=

): characterize the importance of the

scale in front of the cords of wings or
stab.

Neutral line: line characteristic of a fuselage, which corresponds to its incidence of minimal drag. It is
comparable with the line of vol.
Chock: angle enters the wing or the stabilizer compared to the neutral line of the fuselage. That it is for
the wing or the stab, this angle is positive when the leading edge is higher than the trailing edge.
Volume of stab: a dimensional value which reflects the capacity of the stab to hold the wing in the
various configurations of vol. the volume of stab must be all the more large as the apparatus is likely
evolutions with the great angles, which often goes against the research of the minimal drag. It is also the
case if Cm0 of the profile is important.
Hearth of the complete model: not neutral stability of a model, where balance the variations of the
forces of bearing pressure (wing, stab and fuselage) during a variation of incidence (wanted, after an
action with the depth, or undergone because of a turbulence). In practice, centering must always be in
front of this point to ensure the stability of vol.

3.2. Aerodynamic sizes

Alpha (

): angle of incidence of the profile

compared to the air.

Cx: model or wing, profile, coefficient of drag. The Cx characterize resistance to advance.
Cz: coefficient of bearing pressure, a profile or a wing. Cz evolves/moves about linearly according to the
angle of incidence of the profile.
Cm: coefficient of moment, a profile or a wing. Cm reflects the couple of swivelling around the hearth of
the profile (25% of the cord) generated by the air flow. It is positive for a hang-glider (stable), and
negative for a standard profile (unstable). In the last case, it is the stabilizer which must counter this
couple, all the more important as the absolute Cm value is high.
Trail (Fx): force parallel with the advance of an object. The trail is expressed in NR (newton, 10N
1

1kg), and is calculated by the relation: Fx = S.Cx.V , with. = 1.292kg/m (standard density of the air), S =
2 surface (in m ), and V rate of advance (in m/s).
Bearing pressure (Fz): force perpendicular to the advance of an object. The bearing pressure is
1
expressed in NR and is calculated by the relation:
Fx = S.Cz.V
induced Trail: trail related to the lengthening of the wing. Null for Cz = 0, it increase with Cz, but of as
much less than lengthening is important. Physically, the induced trail is generated on the level of salmon
by a circulation of air of the under-surface (overpressure) towards the suction face (depression). The
more important lengthening is, the less this circulation affects the remainder of the wing. Is calculated by
Cz relation: Cxi = z.A
Reynolds number (Re): a dimensional coefficient which includes the speed of evolution and the
dimension (cord) of a profile. Postulate: a profile of cord X evolving/moving at the speed Y behaves

manner identical to this same profile of X/2 cord evolving/moving at the Y*2 speed, because they
evolve/move with the same Reynolds number. Re = 70.Corde (m). V (m/s)
Polar (S): curve (S) characteristic (S) of the performances or the behavior of a profile, a wing or a
model. Polar the most traditional is:
-

Polar profiles: Cz according to Cx, Cz according to Alpha, Cm according to Cz. One generally
traces these polar for different Re.

Polar model: rate of fall according to the speed of flight, smoothness according to the speed of
vol. One generally traces these polar for various masses.

Twist average aerodynamics (CMA): virtual cord, equivalent from an aerodynamic point of view to the
whole of the cords of a wing. It is at the base of the calculation of the centering and the arms of lever of
an aircraft. Mathematically, this cord is the average of the cords balanced by elementary surfaces. More
the math students will recognize the definition of an integral, in fact:
env /2
2 with S = surface wing, C (there) = cord and C (there) .dy = elementary surface.
CMA =. C (there). Dy
S

Wake of wing: it is the flow of downward air generated by a wing. According to its arm of lever and its
height, a stabilizer undergoes this flow more or less, and its chock must be corrected consequently to
obtain the good line of vol.
Interaction: qualify the trail generated by the intersection of two surfaces. Typically, one finds the trail of
interaction to the junction of the fuselage with the wings or the stab. By defect, one estimates that it
accounts for 10% of the total trail of a model.
Stroke static: percentage which indicates the degree of stability of a model, defined by the report/ratio
of distance CG/hearth in the average cord. This value is also valid for the hang-gliders. What it is
necessary to retain:
-

If the static margin is negative: the model is divergent, the least disturbance of trajectory (action
with the depth or movement of air) is amplified.
If the static margin is null: the model is neutral.
If the static margin is positive: the model all the more takes again its natural trajectory quickly
that the static margin is high. Concretely, this value can go from 0 for an apparatus speed or
stunt-flying to 10% for a calm apparatus where stability is privileged. Attention, for a model with
stabilizer, these values are valid only for one aft C G limit taking account of the contribution of
the fuselage.

3.3. Constant profile

Cm0: coefficient of moment of the profile with bearing pressure (Cz) null.
Alpha0: angle of incidence of the profile with bearing pressure (Cz) null.

4. Key stages of the design

Before starting, most important to understand is that the design of a model is an iterative step. Each
stage can call into question the initial assumptions, and it is thus necessary to rehabilitate them then Reto unroll the process of design. It is then the experiment of the originator who will allow to arrive as fast
as possible at the best compromise.
In order to facilitate comprehension of it, all the method is unrolled through an example, in fact the
design of a sailplane of relaxation 3 axes with stab in T.
4.1. To define its need
That amounts raising the question: which is the model which I wish to design, and for which flight
envelope? If the question is commonplace, the answer is less because it poses the bases of all that will
follow. Here the elements minimum to be defined, roughly for the moment:
- which scale?
- which wing surface?
- which beach of mass?
- which nominal speed (or average) of flight?
This stage of the design, it simply acts of very broad orders of magnitude, that the design will make it
possible to refine, to even call into question if they are not adapted.
For the example which interests us, here starting assumptions:
- scale: approximately 1,80 m
- surface: approximately 30 DM
- masses: from 700 to 1000 G
- mean velocity: 50 km/h

4.2. To estimate the Re of reference of flight (Re_ref)

This Re will use to calculate polar the profile (S) with nearest to the flight envelope of your model.
Working PredimRC with variable Re, the polar ones calculated for Re_ref will be extrapolated with all the
Re of vol.
The important thing here is not to find a value precise, but simply an about logical base of work
compared to the use of the model. For example, a racer F3D flying to more than 300 km/h will
have great chances to have Re_ref definitely more raised than that of a launch-hand optimized
less quickly to evolve/move almost 10 times!
A priori definite data in stage 3.1 make it possible to calculate this Reynolds number of reference.
Nothing good wizard: Re_ref = 70 X Speed (m/s) X average chord (mm) With: Speed (m/s) = Speed
(km/h)/3.6 and Average chord (mm) 10 * Surface (DM )/Scale (m) If this part appears too indigestible
to you, it is possible to retain Re_ref of 200000, it is enough pass key for many models.
For the example: Re_ref = 162000, rounded to 200000.

4.3. To choose the profiles

This stage is vital for the model and its performances, even if it is necessary to be conscious that the
profile does not do all: the geometry of the wings and the stab, the fuselage, the chocks, the trail of the
orders, play a part quite as important in the final performances.

How to choose a profile? This question is not so innocent that, because the number of profiles available
has what to drown more informed of the model maker. In fact, the choice of a profile is directly
dependant on the use of its model, and a comparison with other profiles. And of course, one compares
the profiles for a Re representative of the flight envelope chosen, which corresponds to Re_ref defined
into 4.2.
Here some broad outlines: them profiles of beginning: the trail should not be too weak, and Cz_max
must be high to improve the behavior at low speed. them profiles of scrapes: the principal criterion is to
have a weak trail in rather important Cz, about 0,3 to 0,5, as well as high Cz_max.
them profiles speed: the trail must be minimal in weak Cz. For categories F3F, F3B, F3I, F3D, which
requires in addition to turning tight, it is necessary to have raised Cz_max, which is often contradictory
with the mini criterion of trail. To circumvent the problem, more and more profiles are optimized for the
use of shutters which shift Cz_max and Cz de Cx_mini.
them profiles of stunt-flying: with or without shutters, it must present broad a enough negative beach of
Cz for evolving/moving well on the back. Then, if it is about stunt-flying sailplane, it is also necessary to
seek the smoothness and scrapes it, which requires a compromise often difficult to find.
them profiles of hang-gliders: their main feature is the car-stability, defined by positive Cm0.
Other considerations that the Cx and Cz can also return in account: It Cm: if it is weak, that makes it
possible to reduce the size of the stab, and thus the associated trail. Without counting the more
important neutrality of the model whatever the speed. it answer to the control surfaces: certain profiles
have more or less effective answers, which can be important for stunt-flying. It critical Re: it is minimal
Re in lower part of which a profile functions badly. For our profiles of reduced models, that can go from
20000 to more than 100000 following the profiles.
For the example: three profiles are selected a priori: - FAD05S-9%, a
rather general-purpose profile perso and teasler - Clark there, famous
for its beach of high bearing pressure - Eppler 195, which had its hour
of glory
FOOT-NOTE: one can also select only one profile, and decline his operation with various values of flap
deflection. It is typically the case for a sailplane of competition.
4.4. To calculate the polar ones
After having thus preselected some profiles, they should be passed out of numerical blower to evaluate
their performances and to retain most interesting (PredimRC manages 3 of them simultaneously). In
fact, the reference is XFOIL, used on line (but it is rather not very ergonomic) or by the means of
interface like XFLR5. The method is below detailed for this software, but the majority of the numerical
blowers function in a similar way. Of course, such a software is not limited to some functions presented
here, with you to discover all the power of it!
First of all, throw XFLR5, then to choose the workshop of analysis of profile: To use the button then to
open to go to seek a file of profile to the format .dat. The profile is then drawn by the software. One can
of course open profiles as many as necessary.

To go then in the menu of polar to define the type of polar to calculate, which will apply to the profile in progress:

To then put the Reynolds number adequate, here Re_ref calculated into 4.2: The bar of menu then shows on the left
the name of the profile
and on the right the
polar in progress one:

And to launch the calculation of polar, it is enough to fill the beach with incidence for the analysis (typically -6 with

13), then to click on analysis:

Polar that one can export in textual file for then recovering it in PredimRC: Before passing to another profile, it
remains to measure Cm0 and Alpha0 of the profile in progress.

For Alpha0, a right simple click on the bottom of screen makes it possible to rock on the posting of the polar one

One can then measure the sought value graphically, while zoomant with the scroll of the mouse on the zone of Cz=
0. Here, Alpha0 = - 2.
For Cm0, it is complicated a little more, it is necessary to create a graph personalized having for Cz ordinate and X-

coordinate Cm, always with the click on the bottom of screen (finely variable). While zoomant, one reads in
Easy way: to facilitate the search for Alpha0 and Cm0, to use small the Settings to post a more precise squaring.
FOOT-NOTE: if one seeks only Alpha0 and Cm0, without being concerned with polar profiles, there exists
other simpler manners to obtain these values: it can be simply given in the file of certain profiles. - grace a small
software like ProfilKonverter (http://members.aon.at/p-51/download.html).
The results relatively right are compared with Xfoil: the approximation seldom exceeds 0,2 for Alpha0 and 0.01 for
Cm0, which is largely sufficient for trimming a project well.

4.5. To integrate the polar ones in PredimRC

For having launched PredimRC, it is necessary to position on miter 1 (Profiles). For more convenience, it is to better
remove the existing data, but it is not obligatory. Initially, to fill the names of the profiles chosen, then Re_ref
calculated into 4.2. Then, in small the file of Excel, to click on opening. In the window which opens, to choose it

To click directly on finishing, the separation of the data being well made by Excel for the files resulting from XFLR5.
All the data are thus imported in an Excel sheet. It is then enough to copy the Cz data (= CL)/Cx (= CD), for the
zone of positive Cz, within the limit of 33 lines (maximum envisaged in the zone of importation of PredimRC):

In PredimRC, to stick the data with a special joining:

To choose the option values, then OK. The imported points appear in the graph. It may be that does not function in
the case or the decimal separator is the comma. In this case,

Then remain to regulate the parameters Cxmin and Czopt (with being read in the imported values, adapting to the
need), then the coefficient K and Cz of unhooking. The goal is to make as well as possible stick the curve calculated
on the imported points. Not to forget to defer Cm0 and Alpha0 of the profile. The advantage of this method of

interpolation is that PredimRC is freed thus from the small errors or

To reproduce the same thing for the other profiles:

Important FOOT-NOTE: if a profile is not used (or if one uses only Alpha0/Cm0 and not his polar), it is
necessary to put a nondecimal character (by ex: -) in Cxmin corresponding so that PredimRC does not take

it into account in the curves of performances and search for lengthening.

4.6. To optimize lengthening
This stage is optional if one does not seek the performance in term of smoothness. This zone of PredimRC is thus
voluntarily independent of the remainder of the software (except the polar profiles). It should be noted that the
smoothness is not the prerogative of the sailplanes: on a plane, a good smoothness results in a less power
necessary to fly at a given speed.
The research of optimal lengthening rests on two antagonistic phenomena: - To reduce lengthening increases the
cord profile, therefore its Re, and the trail profile decreases. - To increase lengthening decreases the induced trail.
This calculation depends only on the wing, the contributions of the stab or the spindle in the polar ones creating only

one offset without influence on optimal lengthening.

Then, one seeks Cz of flight (curve in top on the left) corresponding at the speed estimated into 4.1, and optimal
lengthening corresponding (curve in bottom in the medium). By deferring these values in the adapted cells, the
cursor (in blue) makes it possible to graphically check the selected value. The value of lengthening thus found will
give the best performances of smoothness for desired speed.

To help with the choice of Cz of flight (typically, 0.1 to 0.2 for a model speed, and 0.3 to 0.4 for a calmer model) and
ideal lengthening corresponding, the two curves of smoothness on the right give following information:

Potential smoothness max (in top): each point of this curve gives the maximum smoothness
possible for the optimal lengthening of this point. This curve is thus a curve with variable
lengthening, a kind of ideal theoretical.
Smoothness wing for selected lengthening (in bottom): it is the curve of smoothness for the
lengthening fixed by the user (cell: Cz of flight to be optimized). Compared with the curve
maximum smoothness, it makes it possible to visualize the loss of smoothness due to selected
lengthening for all other Cz that optimized.

Then, to facilitate the future drawing of the model (miter 3), this lengthening is represented into scale and cord of
root of an elliptic wing (optimal output for selected lengthening).
The use of the originators of certain models of competition (60pouce, F3F, F3B, F3D), which must at the same time
go quickly in straight line and turn to tighten without lose speed, PredimRC makes it possible to quantify the
smoothness in turn, as well as the distances covered during this phase (to be compared with the distance covered
in straight line). The graph smoothness wing in tight turn thus gives the evolution of the smoothness according to
lengthening, for the conditions of turn seized just audessus (speed and Cz wing). The blue cursor indicates the
smoothness selected graphically previously. These indications make it possible to the originator to balance optimal
lengthening, in order to find the best compromise between smoothness on the level and smoothness in turn.
Caution: the optimization of lengthening can result in wanting to use important lengthenings. It will then be
necessary to make sure of the mechanical resistance of the wings with the technique of suitable
construction. For this purpose, PredimRC indicates the effort which the wings in tight turn must support (or
in resource, with the weight of the model near).
In our example, Cz of flight of 0.3 is selected for the optimization of lengthening. According to the mass (700 or 1000
G), the smoothness will be thus optimized for speeds of flight from approximately 40 to 50 km/h, for the profile
FADS05-9% which has a better smoothness for this Cz than the other profiles. Optimal lengthening correspondent
will be 12, with a cord of root of approximately 200 mm for a scale of approximately 1,90 Mr.

4.7. To dimension and regulate the model

The creativity can be expressed here freely, on the condition of complying with some simple rules:

If the smoothness is a criterion of dimensioning, it is necessary that lengthening approaches to

elliptic lengthening (calculated starting from the scale and of the cord of root) and the lengthening
optimized into 4.6. See also the coefficient of Oswald into 4.8 which gives a finer indication of the
output of the wing. the dimensions and the shape of the fuselage have a considerable impact on the
performances and centering.
For the models with stabilizer, the volume of stab is a very important data, it conditions the capacity
of the model to evolve/move with the great angles.
Attention with the geometries too alambiqus. Often, a simple and elegant design is an good
indicator of correct design

Foot-note: starting from version 1.9, PredimRC integrates the formulation of T. Plato for the taking into
account of the contribution of the fuselage in the calculation of the aft C G limit. This limit is more
representative of reality, and one can note that it is often more before without the fuselage, and this more
especially as the fuselage is bulky. This formulation has also the enormous advantage of making
homogeneous the concept of margin static for the hang-gliders and the apparatuses to stabilizers.

Here the interface of seizure of the geometry:

For the adjustments, here the logic used in PredimRC: it stability is function of the choice of the static
margin, which must be selected between 0 (perfectly neutral model) and 10% (very stable model), and
which positions the center of gravity compared to the aft C G limit (= hearth of the complete model).
Generally, whatever the model, a value of 5% is a good approach for the first vol. it Cz wing for neutral trim
of stab: this value is given as an indication, it acts of Cz for which the center of bearing pressure coincides
with the CG chosen, which gives a neutral stab for this point of vol. It is used as value of control of design,
particularly while underlining (when it too high, i.e. >1) a stab too slightly dimensioned or a too large
spindle. It Cz of flight: it conditions the chocks of wing and stab, which will be thus adapted to a type of
flight in particular. Generally, one selected Cz of flight of about 0.1 to 0.2 for a machine speed, up to 0.4
even 0.5 for an apparatus of beginning or of scrapes. PredimRC alerts the user in the event of incoherent
choice (if Cz<0 or >0.5). Attention, it is an obviousness, it is necessary also that the profile is adapted to Cz
of selected flight. It Cz of stab: so that the stab of course functions all the beach speed of flight, it should
not exceed 0.3 in absolute value, under penalty of being likely to take down before the wing.
One can thus note that centering and the chocks (by the means of Cz of flight) are uncoupled. That can shock, but

this approach is however very close to the true life. Indeed, a final tuning of a model during in-flight tests shows
that one can modify a centering without having a significant impact on the choice of the chocks, and vice versa.
What is confirmed by the theory, which shows that centering is conditioned by purely geometrical considerations
(hearth of the complete model), whereas the chocks are it by purely aerodynamic problems.
Thus: one centers for the model is stable, and one fixes it so that it is pleasant (and powerful) at the mean
velocities of flight (thus Cz) to which it is intended.

The graph hereafter represents the model seen of top, which makes it possible to control coherence (and

esthetics) of its work:

4.8. To optimize the geometry of wing
PredimRC integrates a calculation of distribution of Cz and bearing pressure, using method VLM (Vortex Lattice
Method, developed on Excel by John Hazel). The principle consists in segmenting the wing in several small pieces
(40 here), and to analyze the interactions between each piece. It is rather close to the principle of calculation by
finite elements used by the industrial computation software (RDM, thermics, flow,). For reading the continuation, it
will be necessary well to make the distinction between Cz and bearing pressure: Cz is a coefficient, the bearing
pressure is a force (see paragraph 3.2).
One uses these graphs in the following way, while varying Cz_aile by the means of the angle of attack of the wing
(button with arrows):
- Distribution of Cz: by putting a value of Cz_aile close to unhooking, the ideal is to have a decreasing distribution
of the root towards salmon, so that unhooking occurs initially with the root. It is a pledge of healthy behavior at
low speed or at the time of one started.
- Distribution of bearing pressure: ideally, it must follow an elliptic distribution of bearing pressure, which gives
the weakest trail induced for a given lengthening. Of course, it is desirable to optimize the distribution of bearing
pressure for Cz of flight for which one sought best lengthening.
- Factor of Owald: indicate the output of the wing compared to an elliptic wing of the same surface and
lengthening. Its maximum value (to be sought) is thus of 1.

Two lines of action are possible to optimize the distributions:

modifying the cords: the only precaution to be taken is to prevent choosing a too small salmon cord if the
selected profile supports weak Re badly.
modifying the twists: to handle with many precautions, particularly in the case of machines being able to go
very quickly. Indeed, an evolution of twist can be perfect for Cz of flight, but induce a nonhomogeneous
evolution of bearing pressure in other Cz, which can go until inducing too great efforts for the wing. A
traditional case consists in putting too much twist at salmon of a large sailplane with strong lengthening, as it
is the case on certain models commercial: at high speed, one can see the end of the wings taking an anhedral
because of the negative efforts of bearing pressures in end of wing, which sometimes can go until a rupture of
the wings.

Foot-note: if the worksheet LiftRoll of John Hazel (sheet who was thus used as a basis for these
calculations) is limited to strictly 4 panels, method VLM used by PredimRC functions whatever the number
of panels per half-wing (1 to 5), which is nevertheless definitely more flexible.

4.9. To evaluate the performances of its model

PredimRC makes it possible to post the curves of performance in two different ways: - curves of reference with an
elliptic wing ideal of the same surface and lengthening that in the course of design. - performances by the method
VLM, which takes into account the complete geometry of the wings. Foot-note: to forget to click only the button
Calculations if the geometry changes.
The aerological conditions (altitude, temperature) of the site of flight can also be taken into account, which shows
the impact of these parameters in the performances of a model.

These performances are given in the following graphs:

The first graph gives polar speeds of the complete model, One reads there the rate of fall according to the speed of
practical vol. Of manner, in the case of a sailplane, one can interpret these curves like giving best speeds
(stabilized) possible for a given bearing pressure.
Always according to the speed of flight, the curve of smoothness is revealing performances of an aircraft. This curve
translates the distance covered report/ratio/starting height. Here, it is seen that the maximum smoothness is given
by Clark_Y, but that the FAD05S-9% are shown more powerful apart from the maximum smoothness. One notes
there also the important effect of the wing load: the more one charges, the more the smoothness is obtained at a
raised speed, and better is the maximum smoothness. On the other hand, more one charges and more mini rate of
fall is high
The last curve makes it possible to find Cz of flight according to speed and the mass, as well as Cz of stab. One
finds there also the evolution of the factor of Oswald according to Cz_aile.

Foot-note: thanks to small the stage coaches, one can decontaminate the taking into account the
contributions of the stab (also including/understanding the drift), of the fuselage, the interaction between
the elements (10% of the total trail), or of calculation VLM. That makes it possible to better
include/understand the contribution of each element. As for the buttons +, they act on the scale factor to
improve legibility.

In our example, if one considers the speed of flight of 50 km/h, one can read the following things on the curves:
- According to the mass and the profile, the rate of fall goes from 0.9 to 1,5 m/s. To fly in a stable way, the
model must then steal in a mass of ascending air of the same rate, or a possible engine must provide an
equivalent power (=masse * revolved * rate of fall).
-

Always according to the mass and the profile, the smoothness goes from 9,8 to 14,8. That wants to say
that, in planed, the model will advance of 9,8 with 14,8m for 1 m of descent in the mass of air.

According to the mass, Cz of corresponding flight goes from 0,2 to almost 0,3

Another manner of approaching the things: in the example, Cz of flight chosen for the adjustments is of 0,3. While
referring to the Cz/V graph, one from of deduced a speed of flight of approximately 41 km/h to the minimal mass
(700 G), and of 49 km/h to the maximum mass (1000 G). These speeds are those towards which the sailplane will
naturally once tighten the released handles, one can thus compare them to speeds of better comfort of piloting. That
is very interesting for a sailplane of performance that one ballaste according to the conditions weather: while
ballastant, the smoothness increases as well as the speed of flight, without there needing to change the
adjustments.

4.10. To dimension the propulsion

This dimensioning requires only traditional values of entries, to seize in the interface cidessous:
On the other hand, the research the speed of flight to the engine (on the stabilized level) is more original: PredimRC
puts in glance (see graphic above) the power absorptive by the model with the power transmitted to the mass of air
by the motorization, the whole according to speed. The latter is maximum on the ground (null speed) and is null
when the speed of the model corresponds to

mode with vacuum of the engine. The point of intersection of the curves gives this speed of level flight. This method
is particularly relevant, in the condition of knowing Kv of its engine well. To note, that, as for the other curves of
performances, one can post the powers in the case of an elliptic wing of reference or with calculation VLM (just).
In the example, the speed of level flight is of approximately 60 km/h. One can notice that, on this level of rather
modest power, the profile has relatively little importance in the result.

4.11. To dimension the control surfaces

PredimRC does not manage this aspect of the design of a model, but it is good to point out some basic principles:
- twist of a control surface depends on the profile to obtain the best trailed effectiveness ratio/. In general, it is
about 25% of the airfoil chord, but that can vary from 20 to 30%. This choice is particularly sensitive on the
sailplanes of performance whose shutters and ailerons are used out of camber flaps, with often a coupling to
the depth to improve the performances in tight turn.
- For the ailerons not full-span (very traditional case): their length must be close to half of that of the wing, and
the aileron must be located at nearest to salmon.

4.12. To dimension the servos

The interface was desired ultra-simple, but complete. So seemingly the levers of swing bar or puppet seem to miss,
it of it is nothing because these arms of levers are included in servo clearances and controls.
It should be noted that one can as calculate the couple necessary to control a pendular stab, with the standard
assumption as the axis of rotation is 5% front the hearth of the stab (located at 25% of the average cord).

The case of a system of control per integral incidence was also taken into account, on the basis of formalization
developed by the author. There too, few values to be seized, with a great reliability of result. The arms of lever are to

be respected as well as possible

For both type of control surface, a graph gives the curve of evolution of the couple according to the race of servo, of
0 until maximum clearance (100%). In the case of the control surfaces traditional, a whole of buttons to notching
makes it possible to choose for which control surface the graph must be posted. It should be noted that, for the
system with integral incidence, the neutral position of the control surface does not correspond inevitably to a null
couple on the servos.
Foot-note: the speed seized for calculation of the couple requires some tweezers: it is a question of
considering a speed easily reached by the model, noted here standard speed, and not of taking a too high
value which will oversize the servos unnecessarily. For this reason, some usual values are given in the
comment of assistance of PredimRC. In the same way, the values of couple are calculated with the most
important clearance, which gives a certain safety margin. Indeed, particularly at high-speed, the control
surfaces are very seldom used with their full clearance. Reason moreover not to overestimate the standard
speed of its model.
For the model of the example, standard speed selected is of 80km/h. the calculated couples are traditional enough
for this type of model: they correspond to microphone-servos traditional of 5g with 9g for the ailerons and depth, and
to servo of 15g for the drift.

5. Conclusion
Here, it is finished! For those which will have traversed this document length into broad, you will have
included/understood it: beyond the tool, it is a true design method of model which I propose to you here, complete
and powerful. With you to play

F.A.