js3

My Feedback: (3)

Hello Everyone,

I'm interested in what the performance differences are between the following three airfoils and why someone might choose one over another:

S8052
S8055
S8064

I know the S8064 is the newest one and is used on the Great Planes Viper. The S8052 was the airfoil on my favorite racer, the Bird of Prey. I don't know of any plane or anyone who has tried the S8055. Also, how can you tell if a particular section is turbulent or laminar? Can this be determined just by looking at the section or must the data be analyzed with XFoil or a wind tunnel simulation? I'm not an aerodynamicist so if these and other questions are sophomoric, please forgive me.

If Dr. Selig himself would respond here that would be GREAT!

js3

My Feedback: (3)

Do any of you aero guys have any comments? Bruce, Ben, Paul, or Dick? Anyone?

BMatthews

Sorry, I'd looked through a couple of those but got distracted and never answered.

Just looking at the cambers we see that the first two are of a similar envelope but with differing cambers and peak camber values. That's all I can say from that. I don't have info on the 8064.

What you need is to find the lift/drag polar charts for them and look at them closely. What you want to see is for the higher reynolds numbers a curve that holds the range of the "left side" low drag region all the way down to and even a touch below the Cl=0 line so you are sure you have minimum drag even when the elevator overshoots a trifle and the G load goes slightly negative for a moment. But in the turns you need quite a bit of lift ability to ensure a minmal drag in the turns. So you're looking for as tall and vertical a "left side" as you can get where the lower corner is at or slightly below the 0 line. So the one with the deepest and most consistent "left side" is the one you want.

Now you just need to find the lift/drag polars. If you don't know where they are then download Profili2 and pay the cheap upgrade price to go from the free version to the pro which unlocks the Xfoil utility and allows you to generate calculated data for such curves.

js3

My Feedback: (3)

Thanks very much for the reply, Bruce. I have Profili 2 loaded and registered so I'll start there. It's just so much information. Kind of overwhelming for a hack like myself. When I looked at all of those graphs and charts I didn't know where to begin and what it all means.

Thanks again!

Tall Paul

I have a GP Viper, but no idea as to what airfoil it has.
Haven't flown it in years.

BMatthews

If you can merge the curves on one chart and post 'em up and we'll help you translate. Otherwise just save or screen capture each chart to some sort of program that'll let you save the image and reduce it to about 4x6 or 7 inches at something like 72 dpi.

In the meantime if you want to get a bit of a handle on what is happening with your Quickie in a race trot on over to...

http://www.grc.nasa.gov/WWW/K-12/airplane/foil2.html and play with the Foilsim applet there. You may need to go and install Java to run it but it's well worth the trouble.

It first opens with a full sized wing but you can alter that. Select the Size Shape tools and run the chord down to 0.82 ft and 4.1 feet to reflect the size of your Quickie. Play with the angle and camber entries to get a 1.5 to 1.6% camber. The speed defaults to 100 mph which is close enough for this. Note that the lift can be shown as a Lift Coeficient or pure lbs of lift. Select the pure lbs. Now alter the angle of attack until the lift is down to 4 lbs. I believe this is the minimum weight requirement for Quickie 500. If I'm wrong then alter the AoA until it matches the mandated weight. Now click over the output so it reads in Cl. First off you'll find that it ends up with a minus 1.46 degree AoA. That's fine. And then you'll see that the Cl only needs to be a paltry .046. This is why I'm saying that you need to ensure the low drag lift range extends down to Cl=0 or even a bit lower.

Now up the angle until the Cl = around 0.65 or so. For many low camber airfoils I've seen this is about where the coefficient of drag starts to really slope over to the "bad" right side of the chart. Once you have the angle of attack set to pull a Cl of 0.65 switch back to lbs of lift. You'll find that it's 55 lbs of lift. So this means you can pull about a 14G turn with a 4 lb model or 11G turn with a 5 lb model and not really pull up to the higher drag area of the flight envelope. If you do some more calculations to determine the radius of the turn to achieve these G loads and don't exceed them you'll find that you can pull through the turns with little speed loss.

So print out the charts and study them. Then play with Foilsim a bit to learn just what your model is doing. Granted the airfoil in foilsim is rather generic but it does the job. But the Cl's and lift are still there. Use the lift/drag charts to see what Cl's your airfoil is happy at and then Foilsim to determine what the angles and lift range to the happly Cl spreads are.

js3

My Feedback: (3)

Thanks a lot for that info, Bruce. I'll play with both Profili and Foilsim.

But what about my question regarding laminar vs turbulent airfoils? Is there an easy way to tell the difference?

BMatthews

No, there isn't. A 'foil isn't laminar just because of it's shape. And in any event at typcial model reynolds numbers you can't keep the airflow laminar much beyound the high point anyway. There just isn't enough energy in the air or something. The full sized airfoils that retain the laminar flow for longer are generally quite obvious by their very far back high points. The Mustang used one for example. But at model sizes they generally do not behave well and have a very vicious stall... or so I read a long time ago.

You see, very few airfoils have a laminar flow over the whole chord. At some point they all break down to a turbulent boundry layer and upper flow form. With care that turbulent area stays thin. Without care the area gets thick for a while and you have a separation bubble. If it reattaches it's just a performance robbing nuisance. If it doesn't reattach it's a stall.

From reading about airflow and pressure suction peaks and other stuff Selig and Drela and a few others have written they shaped the airfoils to ease the air around their shapes rather than try to force it. At our speeds and sizes this is the best way to do it. The shapes that result are carefully designed to coddle the airflow into staying in contact with the curvature and delay separation until the last possible instant. So while laminar flow shapes may be best for larger aircraft the pressure peak avoiding shapes seen in the higher performance Selig, and Hepperle and other options are the way to high performance for our needs.

ptxman

My Feedback: (2)

If you go to the following ezone link, post #6 (by ptxman) I replied to a somewhat related question but also but also expanded to include some issues regarding airfoil selection specifically as it applies to pylon racing (F5D in that particular case).

http://www.rcgroups.com/forums/showthread.php?t=469869

In some respects Q500 is a bit easier because of the constant chord planform (same Re along span) & max airfoil thickness (although I think the rules still allow you to blend if so inclined). But IMHO, in order to really get to the heart of quantifying that airfoil A is better that airfoil B, I think you really need to pull a bit more information together along the lines of what I'm describing. Also check out Martin Hepperle's website for great related info.

MSelig

I've attached some XFOIL predictions that include the S8064, S8052 and some other Q500 airfoils from some airplanes that were/are popular. This comparison is by no means exhaustive. For a lift range from 0 to ~.6, the S8064 is looking pretty good to me.

js3

My Feedback: (3)

Thanks very much, Michael! I've been playing around with Profili trying to get it to display the polars that Bruce indicated and I just can't seem to figure out how to do it.

Please correct me if I'm wrong but isn't the 8064 a touch thicker than the 8052? So if I used the 8064, I could use a slightly shorter chord and still have the appropriate thickness as required by the rules?

Thanks again for your reply to this thread! [sm=thumbup.gif]

ptxman

My Feedback: (2)

Yes, the 8064 does look good! Ill have to check that one some more for interest sake. To refine this a bit more, it would be beneficial to see:

1) all foils normalized to the same MT%. Reason is, among the competitive foils, the differences start clustering quite close together & often people overlook that airfoil A (@ 11.8%) looks better than airfoil B (which happens to be MT=12.0%) by a small margin, but 'B' can be thinned/normalized to 11.8 & starts to look closer to the A as an example.

2) also a plot/comparison of Cm (pitching moment) of same foils over the Cl range of interest, particularly at low AOA or cruise mode: Reason: in order to comparatively evaluate the 'total' drag meaning the contribution the stab in addition to the wing, a minimal Cm profile is a desirable characteristic. It would not be uncommon for the drag to be 20% of total in this sort of application. If a particular wing foil has great polars but comes at an asociated high Cm, it may loose enough ground against another foil which has minimal Cm. Basically Ive resorted to spreadsheet comparisons to handle this & weight the various modes. Its not always obvious just looking at the polars.

3) It would be nice to see some names to the Q500 &/or some of the current crop of MH foils for comparison.

Good stuff!

ptxman

My Feedback: (2)

Are you aware there is a Yahoo group dedicated to Profili & moderated by the software author? Lots of Q&A's there.
http://groups.yahoo.com/group/software_profili/

Similarly there is a Yahoo Xfoil user group with lots of aerodynamics expertise on board.

js3

My Feedback: (3)

Oh yes! I'd forgotten about the Profili user's group. Perhaps I'll post a question to the list.

BMatthews

Yes, thanks very much Michel!

Js3, you can now clearly see what I meant by the need to pass the Cl=0 line in a clean way. All these do so.

Now comes the hair tearing out decision. The 8064 and 8052 are the clear winners here. Both show good low drag coefficients through the generally used Cl=0 to about 0.6 range with the 8064 being a clear winner.

But look what happens when the Cl rises above the 0.6 value. The 8064 flattens out and begins to gain a lot more drag for gains in lift compared to the 8052 which angles but does not lay anwhere near as flat in the upper range.

What does this mean out on the race course? To me it means that the 8064 airplane will have a little less drag on both the straights and in the turns. But if you are in the habit of flying close and tight the hard turns may well force the Cl to a higher than 0.6 value and suddenly you generate a lot of drag that will slow the model and force it to need to accelerate out of the turns. So the 8064 has the POTENTIAL to be a better choice but you need to limit your up elevator travel to the point where it's still responsive but is not capable of pitching hard enough to push the Cl over 0.6 when at speed. This may mean you need to fly a slightly more open turn radius to avoid this.

The 8052 isn't a miracle airfoil by any means but while it has a touch higher drag in the main area slight transgressions into higher Cl's or a tightly flown course will not push this option as deeply into the high drag region. Being able to fly a tighter course may make up for the speed gain of the 8064 airfoil. Will it? That depends on what sort of turn radius can be supported by the two models. It may well be that even a snappy turn 1 "ricochet" won't push the model above the Cl=0.6 in which case the 8064 is the hands down choice. But if the a tight course requires a 0.7 to 0.8 Cl vs a more open course that is limited by a 0.6 then the numbers need to be crunched to see just how open a course you need and if the extra speed will make up for this added distance. A 10% speed advantage isn't much good if you need to fly an additional 15% distance to maintain it.

I know I'm adding questions instead of answering them but I'm not a racer and doing the calculations to figure out the G loads at speed and thus what the required Cl's are is far too much like work. But if you can do the turn radii and add up the distances then it's at least possible to determine how much extra wing drag you'll have from the lift and since the rest of the drag is going to be constant you can sort of figure out which airfoil is the way to go.

js3

My Feedback: (3)

All,

Thanks very much for your replies! I think I've succeeded in getting Profili to output the information Bruce indicated. The airfoils I've chosen for this comparison are the Selig S8052, S8055, S8064 and another "in vogue" airfoil, the NACA 66-012 which is used on the Vortex--winner of the last several years' AMA Nationals.

Images 1, 2, and 3 are at Reynolds #1090000 which at my altitude is about 160 mph.
Images 4, 5, and 6 are at Reynolds #1020000 which is about 150 mph.

Please respond with your comments regarding this comparison.

Thanks again!

MSelig

What do things look like when zoomed in on the low drag region? A very good pilot can feel a 3% difference in airfoil drag. In my experience, ~3% is the threshold. We can add a lot of caveats to that ~3% ... but I won't.

Note in my polar plot above the Cd scale does not go to zero. In my plot I can see about a 5-10% difference between the top two at a Cl around 0.5.

js3

My Feedback: (3)

Michael,

When you say "zoomed in on the low drag region", is this what you mean (see attached)?

Thanks!

MSelig

I mean make the drag axis go from, say, zero to 0.010 instead of 0.100.
That will stretch things out and make it easier to see differences.

BMatthews

You're certainly on the right track. But not being a racer myself I'm not sure what speeds the models fly at on the straights and in the turns. Even if they fly at 150 in the straights you may want to redo the charts for a lower speed and corresponding Re that reflects the speed drop in the turns. Say something like 70% of the straight line speeds.

But even from what you have already it's VERY obvious that the Selig options are far better than the NACA airfoil Not so much on the straights but the Selig options will have far less drag for the same lift in the turns. So that would let you turn as tight with less speed loss or tighter for the same speed loss.

ptxman

My Feedback: (2)

Im actually surpised to hear that, but it proves another trueism in performance modelling. Sometimes what is 'winning' doesnt align with what the data suggests, or worded another way, better (thumbs, reflexes, consistancy, engine, prop setup) can override & overcompensate for the aerodynamic differences. The real answer would be what would the Sxxx airfoils do in the hands of the same winning pilot, same power system & all other things being equal. I'd also be interested to see the Martin Hepperle Q500 series of airfoils presented alongside the ones you've shown. (I think I know teh answer but you may find it interesting). For some reason, again maybe a N-Am thing, the MH Q500 airfoils have not been as popular despite good looking polars, yet in other areas like F3D & Q40 you can definately see the influence.

BMatthews

Actually a lot of time what wins in racing is the best engine and pilot that flys the tighest and most controlled course. But all else being equal I like the performance of the Selig airfoils I'm seeing here so far. I look forward to seeing the expanded charts to be able to see what is happening on the straights where the performance around the Cl=0 is more important.

n8rloves2fly

My Feedback: (5)

I need help getting an airfoil design. can anyone call me and show me how to download one reall quick.
402-660-6708

n8rloves2fly

My Feedback: (5)

I need help getting an airfoil design. can anyone call me and show me how to download one reall quick.
402-660-6708