Aug 21, 2020, 06:59 AM
launch low, fly high
Discussion

# F5J parametric aircraft performance evaluation

I have seen some rudimentary analyses on RCG recently that purports to show performance calculations for various F5J aircraft designs. I have taken issue with some aspects of the performance analysis that have been done, and have spent some time recently to do what I think are more appropriate calculations so as to provide a better comparative evaluation.

An aside, this more accurate evaluation does not favor designs that I have been associated with. Quite the contrary. The rudimentary design evaluations favor higher aspect ratio designs such as the Plus X, and penalize designs such as the Explorer. Why am I spending the time to do this evaluation when the flawed analyses shows designs that I am associated with in a good light? Good question… I have rather strong opinions in terms of fairly representing factual data as well as the underlying physics that support the data. Bad analyses should always be called out. We have certainly seen more than enough bad science being put forward in recent months…

This will be a somewhat long discussion; it will take a few posts to get everything sorted out appropriately. First of all, the rudimentary analysis that I had objections to is found on this link. They calculated the minimum sink rate as based solely on the span loading, ignoring the hobgoblin of Reynolds number, which haunts all model aircraft designers. Below is a plot of a typical F5J airfoil lift versus section drag. Note carefully that the drag increases quickly with reducing Re, and also increases quickly with higher lift coefficient. The rudimentary analysis assumes that there is ZERO change in the airfoil drag for either Reynolds number changes or lift coefficient (as well as other essential simplifying assumptions). In rough terms, changing the Re by 10% changes the airfoil drag by about 5% when at light weight, all else being equal. This is a big change! The increase in drag at higher lift coefficient is also substantial, as you can see on the plot below.

### Images

Aug 21, 2020, 07:01 AM
launch low, fly high
I have done two different analyses. One uses airfoil data at varying Re and trailing edge deflection so as to get a hopefully reasonable assessment of the straight line aircraft performance for various aspect ratios and flight weights. Another analysis fixes the airfoil associated profile drag to be a constant regardless of Re or lift coefficient, which is a basic assumption that is used in the rudimentary analysis previously referenced. There is a primary variance, in that I fixed the fuselage drag (not drag coefficient, but drag as expressed in equivalent flat plate frontal area) as the fuselage length is largely defined by the wing span, not aspect ratio. The fuselage cross-section is largely defined by the contents placed within. The fixed length and fixed internal packaging defines the fuselage drag. As the length is fixed, and the diameter is fixed, then the total surface area is fully defined. This changes the assumed span loading analysis as fixing the airfoil drag to be invariant with Re changes or lift changes is insufficient to meet the constraints required to assess the sink rate solely by evaluating the span loading as the fuselage drag coefficient changes with changes in the wing area.

I have done the plots with the primary independent variables being Aspect Ratio and surface loading. The vertical axis is the sink rate, shown in feet/second (sorry, my spreadsheet has some residual Imperial units… I promise I will not use it to fly on Mars!) The 12 g/dm^2 curve is the minimum surface loading allowed for F5J events. It appears that there is at least one person with an aircraft that is below this limit as per the data in the build evaluation link. I think every manufacturer can achieve this minimum weight limit at present. I have seen some planes that need about 25g of ballast to be installed so as to get up to the required minimum weight, in particular the lowest aspect ratio and the highest aspect ratio designs, so we are not limited by manufacturing capabilities to achieve the minimum weight.

With the airfoil data incorporated, one sees that the lowest weight minimum sink performance is rather flat, and not appreciably affected by Aspect Ratio for a rather broad range. This has been seen in multiple competition experience, as a reasonably diverse set of designs have similar performance in minimum sink conditions. Minimum sink performance isn’t where the various designs show differences, despite the span loading evaluation process evaluation. The differences arise when the wind shows up, as some designs have better speed range, and others have an increased need for ballast in the wind.

### Images

Last edited by Joe W; Aug 21, 2020 at 07:11 AM.
 Aug 21, 2020, 08:02 AM launch low, fly high Thread OP The next topic will be analysis for turns. This has significantly increased computation. I will do only a subset of the analyses due to the amount of processing time required.
 Aug 21, 2020, 09:31 AM Dark Side of the Red Merle Joe, Thanks for this, if I read this correctly does this mean that for certain weights a slightly lower aspect ratio is better? Thanks Curtis
 Aug 21, 2020, 12:47 PM Registered User Joe Thanks for the good work. I assume the AR ratio performance is affected by wing design and airfoil sets and you selected a good over all wing designs and airfoil sets. When you do a turn analysis the wing design is even more critical. Art
 Aug 21, 2020, 01:40 PM Registered User Thanks for this, Joe!
Aug 21, 2020, 05:12 PM
launch low, fly high
Quote:
 Originally Posted by Curtis Suter Joe, Thanks for this, if I read this correctly does this mean that for certain weights a slightly lower aspect ratio is better? Thanks Curtis
There is a distinct possibility for this to be true.

There are so many different aspects to aircraft performance. Minimum sink is just one small aspect that contributes to overall soaring performance. There are many other aspects, such as how a plane turns, the handling qualities, possible aeroelastic issues, high speed capabilities for dashing to a thermal and more... It depends on what weighting one puts on each metric as to where the optimum design may end up.
Aug 21, 2020, 05:17 PM
launch low, fly high
Quote:
 Originally Posted by old1104 Joe Thanks for the good work. I assume the AR ratio performance is affected by wing design and airfoil sets and you selected a good over all wing designs and airfoil sets. When you do a turn analysis the wing design is even more critical. Art
Yes, this analysis is subject to the airfoils chosen. I went with a single set for the entire range, which decidedly simplified the analysis. The airfoil set was originally designed for an Re that was closer to the slightly lower AR than the Plus X (assuming minimum weight aircraft is the driving design point).

There may be a small shift in the optimum depending on the airfoils selected. There is also a dependence on the various compromises that are done during the airfoil design stage. Some of the F5J aircraft are more optimized for minimum sink flight, others optimized with a bit more capability to penetrate into he wind.
 Aug 22, 2020, 03:47 AM Hugh Blackburn Hi Joe That is very interesting. The "rudimentary analysis" in which one might just use a simple 2-term drag polar (say assuming fixed values of CD0 and span efficiency, but letting aspect ratio and wing loading vary) suggests that the best sink rate should be proportional to the square root of wing loading and also aspect ratio to the -0.75 power, right? (If I read this right) Your first (presumably, more representative) carpet plot suggests that one approaches that kind of dependence on wing loading but the aspect ratio effect isn't nearly as strong (in fact, it is very weak). The second carpet plot (with fixed airfoil profile drag coeff) goes rather closer towards what the rudimentary analysis suggests in both wing loading and aspect ratio dependencies.
 Aug 22, 2020, 09:21 AM Genoma² Paris, Fr Dear all Looking at aero dynamic parameters only is like "looking at your finger that mask the forest". If you compare a good and optimized wing to a not optimized one, there is more or less a maximum of 2 to 3 cm/s sink difference. There are several other parameters that are far more important at the end. as an example, what is the sinking rate provided by the propeller? Can it be optimised? What about fuselage drag, interaction drag, and what about any parasite drag? In my opinion, you'd better try to improve them than to improve a wing. Further more, you also have to take into account factors such as resistance / deformation / G load to be supported that may change the optimization point. This is directly linked to wing load where we have to minimize it accurately. As an example, a buzzard or an eagle have an aspect ratio close to 6. This is their optimum point for sure. Far away from the 20 we are talking here (is it our optimum? I don't think so). In F5J, you better have to circle tight at very low altitude than to go at high speed. The plane is passing 60% of the time circling. The plane then never fly during such phase in a straight flight nor in a stabilised flight. This means that dynamic behavior with associated stability factors (on all axis) have to be included into the general equation with the consequence on fuselage length, vertical and horizontal area and of course with associated weight consequences. And any consequence may be weighted accurately thanks to the construction Technic retained. Other esotheric factors may also be integrated. As an example, in order to reduce costs or to increase earning, a high aspect ratio may be chosen and someone may demonstrate that this has a good aerodynamic reason. In reality it does not have such so high importance. Then, finally the pure aero aspect is not so important. Dynamic aspects and Constructions aspects is in my opinion far more important. And of course, on the top of the main factors is the Pilot. That represent more than 70% of the result. So you'd better try to optimize him/her than to optimize the wing... A Plus (not the Plus X), is for me the worst plane developed for F5J (Sorry Joe). But you can have very good results with it. 2 main reasons : This plane is really well made and Joe (as also others) is one of the very top pilot. He can win with "any kind of plane"... Construction and pilot change drastically the final result. Caution then to take the good and correct factors and not to omit some Marc
 Aug 22, 2020, 05:03 PM Registered User How would one quantify the spinner-prop drag on a F5J model? Could the drag be significant? 3-5%? The Plus was a F5J design that addressed the prop drag issue and thus IMHO was an innovative approach and as such an important iteration in F5J design. Last edited by Tom Gressman; Aug 22, 2020 at 07:04 PM.
 Aug 22, 2020, 07:59 PM Registered User ah very interesting points. i recently deduced, why birds of prey, classic land soaring birds, have a sinkrate of about 1 m/s; i reason it's because; a "standard" thermal rises fastest in the the core, more than the rise speed of the complete thermal. the bird reaches the top of the thermal in a fairly short time, where the rising speed of the air is 1 m/s or more than the complete thermal rises. therefore the bird does not "need" to have a better sinkrate than that (to stay in the top and climb along with the thermal at the thermal climb speed). ofcourse they do have other optimisations to keep the weight and biological energy use low. they do not fly with a bad airframe. edit: what they gain by not investing in a better sink rate, is low drag at high speed, energy retention. maneuverability near obstacles or prey. i am not sure how good, but the "overland speed" in good conditions may be very good in total when they combine those "modes" this also explains "pooling", "bunching" of aircraft and birds in thermals. also explains how someone might fly below someone and gain altitude faster than someone in the top of a thermal, but then not get past but rather stay with the other bird or plane. i had this several times in my flying history. looking back. edit 2: a statement which may flow out of this, is; that someone who can find thermals better, can use a heavier plane, with no disadvantage in the (normal, standard, usable) thermal. the pilot may then, be able to find a thermal sooner outside. while not loosing too much altitude, and covering much ground. a less able thermal finding pilot, needs a better sink rate, to find (!) the thermals. when there are weak or unusable thermals, this model\pilot wins. Last edited by m4rc3l; Aug 23, 2020 at 10:38 AM.
Aug 23, 2020, 02:09 PM
Genoma² Paris, Fr
Quote:
 Originally Posted by Tom Gressman How would one quantify the spinner-prop drag on a F5J model? Could the drag be significant? 3-5%? The Plus was a F5J design that addressed the prop drag issue and thus IMHO was an innovative approach and as such an important iteration in F5J design.
On a real glider, experience made with or without propeller in the nose (like us) shows a quiet interesting difference in L/D rtion : 43 vs 40 points.

On F3B, experience also shows a difference of 1s in speed task.

Make your computation and you will see that this is more than 3 to 5 % of total drag... and more than 1 or 2 cm/s sink

Marc
Aug 23, 2020, 11:12 PM
launch low, fly high
Here is a bit more detail in regards to calculating sink rate, this time for a fixed design.

The first plot is for an Aspect Ratio of 20, with the trailing edge camber set at +2 degrees. The sink rate is plotted with respect to the aircraft lift coefficient, and uses the airfoil profile drag as calculated by XFOIL. Note that the optimum lift coefficient is in the range of 0.8, and the optimum lift coefficient changes with aircraft surface loading.

The second plot uses the same performance for the 12 gm/dm2 surface loading, and then scales the higher surface loading performance by the square root of the loading change. This is equivalent to evaluating the sink rate with a fixed drag coefficient. Evaluating the sink rate by using the span loading assumes a fixed drag coefficient, although it actually fixes the profile drag coeffient for all lift coefficients. Note that the change in the sink rate when going from 12 gm/dm2 to 20 gm/dm2 increases by 44% when using a fixed drag coefficient! Evaluating via span loading results in an excessively penalized performance for higher weights.

The previous evaluation showed how poorly the result is for comparing different aspect ratios via only span loading. This evaluation shows the magnitude of the error in the result when comparing different surface loadings for a fixed design.

In short, evaluating the minimum sink via span loading, without consideration for the effects of Reynolds number, generates incorrect results and should not be used for aircraft performance comparison.

### Images

Aug 23, 2020, 11:13 PM
launch low, fly high