Home page‎ > ‎

Tuft testing & vortex generators

September 2008
updates July 2016
Peter Chapman

Photos and video were taken of in-flight tuft tests in 2007 and 2008, on my & my father's CH-601 HDS. I hope they are useful in helping understand a little about airflow over an aircraft's wing and the HDS in particular.

Below are photos, then videos. Photos are shown side by side at the same indicated airspeeds, before and after vortex generators (VG's) were added.

The photos can be a little boring, for it takes time to go over them in detail and compare exactly what various tufts are doing at different places. The lowest speed photos get a little more interesting. The videos bring things to life, and that's where it is fun to see how airflow separates over large areas of the wing as the stall approaches.

The vortex generators are "Feathers" VG's from http://www.stolspeed.com.

More recently, in 2014, we took off the Feathers VG's and changed to Hall Brothers metal VG's available from Aircraft Spruce.
The Feathers ones worked great but became brittle in winter and tended to break off, particularly if accidentally touched. (E.g., brushing off snow on an aircraft housed outside). We kept on replacing some each year. We had a fairly early batch of Feathers VGs so it is possible that newer ones are improved, although even at that time the plastic had good anti-UV ingredients. They may work well on an aircraft in a less severe environment.

The Hall Brothers ones are "U"-shaped aluminum. A large proportion of those in the kit have a slightly curved base, which makes them fit well on the curved wing surface. (The kit also includes some flat bottomed ones in case that works better for the buyer.)

Since they have two 'fins' per unit, we spaced them to give a similar 'fin' density across the wing, or actually slightly less. What we did just happened to look reasonable: one piece about every 9" on most of the wing, and every 6" out by the tips for extra effect.

Back to the tuft testing with the Feathers VGs:

These VG's are made of  UV resistant polycarbonate, have rounded tips, and a slightly curved bottom to better fit on the wing. They attach using a 3M adhesive that applies like an adhesive tape, that allows for removal of the VG's without damaging the paint.. I haven't closely examined all the VG's on the market, but I'm impressed by the design details on these ones.

At time of [original] writing, they haven't yet gone through a Canadian winter, but we'll see how that works out. At least one other HDS owner in Canada has Feathers VGs and keeps his plane tied down outside.

Opinions vary on where to locate VG's. One theory is to keep them well forward, such as at 10% of chord. On the thick HDS airfoil (approx 18% t/c).  I chose to locate the VG's at approximately 15% of chord. As a reference, the spar rivet line is at about 27% according to my rough measurements.

Some messy details on picking the 15% chord:
The 15% is to the front of the VG's (as is Stolspeed's convention), and may not be the true 15% point because the required distance infront of the spar rivet line was measured on the airfoil's curved surface, rather than being perfectly parallel to the airfoil's chord line. Also, in the inner wing section the distance from the spar was kept constant, as looks right for the untapered center section. (A 15% line would technically be angled because the chord length changes due to the tapered trailing edge / wing root fairing.)

Spanwise VG spacing was pretty much as suggested by Stolspeed: Every 90 mm for most of the wing, but 60 mm near the tips to ensure that the tips wouldn't stall first. Given how well the tips stayed flying until the last moment before a full stall break, both with and without VG's, I think one can probably do away with the 60 mm and stay with the 90 mm everywhere. But I haven't tested that. (The 601 HDS still gets a wing rock just at the edge of the stall, but tufting suggests it may be due to quickly varying amounts of detached flow from inboard to the center span, while the tips stay flying better.)

A total of 37 were used per wing. (HDS span is only 23 ft.) Plenty more VG's were left in the box.

Both from theory and practice, the airfoil and wing are not prone to stalling through leading edge separation. Instead the stalled area grows in the classic manner from the inner wing root and moves outboard.

The aircraft is a Zenair Zodiac CH-601 HDS, Rotax 912 powered, flown with two aboard and fairly light on fuel.


About the photos:

Airspeeds are indicated airspeeds, in statute mph. Calibration is unknown. Therefore airspeeds are to be taken for relative comparison purposes only.

Position error of course is likely high when the indicated speed drops to 40 or below!

Initial indications are that 30 mph indicated is more like 50 mph actual, and 50 mph indicated is more like 60 mph actual. That sounds somewhat realistic. These are initial results from GPS based testing on a too-windy day in May 2009. At cruise speeds, indicated airspeeds have always seemed reasonably correct.

Typically there's a top photo looking at most of the wing, and a lower photo that focuses on as much of the wing root as I could see. For the slowest speed cases, I included a photo of the panel showing the airspeed indicator.

Photos are kept smaller to allow side by side comparison on larger monitors.

Unless noted, photos were taken with power pulled back (not necessarily all the way), either maintaining level flight or slowly descending.


  • With vortex generators, the airflow is often at least as clean as the airflow had been 10 mph faster without VG's. This is a significant improvement. (Compare photos in the right column, with those in the left column but one row higher up.)
  • With the VG's we could slow down significantly more than without the VG's. We could get down to flying steadily under 30 mph indicated before wing rock, buffeting, and a stall break took place. Previously, only a little below 40 mph indicated was possible before all that would happen. At such speeds, the  IAS is typically inaccurate on light planes, so one can't quantify the reduction in stall speed and increase in range of controllability. 
    • No tests were done to determine whether top speed changed. None was apparent. To test it properly needs careful attention to rpm, manifold pressure (with a variable pitch Ivoprop), density altitude, and so on, as well as very still air.

  • Without VG's, there is a surprising amount of inward flow over the ailerons. This is present even when not down near the stall, such as at 70 mph.
  • Note that the airflow on this particular aircraft is poor along the inboard edge of the wing locker, worse than it is further inboard. It could be due to the poor profile of the wing locker edge, creating a wavy surface and perhaps an air leak from inside the wing. Yet even the forward most tuft is usually more disturbed than others further outboard, so I am starting to think that it could also be related to the discontinuity at the very nose, where the gap cover design crudely joins the untapered center section and the tapered outer wing.
  • Even at the lowest speeds achieved before getting into buffeting, the outer wings still had largely attached airflow. Only the inner wings and sometimes the ailerons, were seeing a lot of separated airflow. When we did stalls (mostly only on video), that was when the airflow over the larger areas of the outer wing would actually detach.
  • Even at 110 mph (with VG's), the inner aft tufts show a little turbulence and don't stay flat along the airfoil surface. While the area isn't stalled, I am guessing that the boundary layer is fairly thick and the air is of lower energy. Airflow isn't perfectly clean everywhere, even when well away from the stall.

  • Stall behaviour felt about the same before and after the VG's. With VG's, while the ailerons might have more authority before the stall, the actual stall seemed about the same, with the same degree of moderate roll off to one side when it did break. When a plane's aerodynamics are changed, delaying the stall is one thing, but it is also important that when the stall does arrive, handling qualities are still OK and stall behaviour isn't worse.

While normally one doesn't fly around at under 40 mph IAS, perhaps the VG's will improve handling at low speeds such as in turbulence on final approach, and provide an increased margin before the stall point, if one flies takeoffs and landings similar to before. I don't see VG's as being magical, just useful: We don't use the VG's to "fly slower" very much at all,  but use them to have better margins, better control, and less drag at normal approach, landing, and takeoff speeds.

Installing the VG's:

Stolspeed suggests using a chalk line, but masking tape was sufficient. It does bend, however, so one needs to have a number of reference marks to ensure that it stays straight. The bottle on the wing is alcohol to clean the wing and bottom of the VG's before applying the 3M adhesive and sticking them on. The distance between the VG's and the spar varies across the span, but the perspective in the photo makes it look almost as if it is a constant distance, which it is not. (The nose of the VG's varies from 11 cm ahead of the spar rivet line, to 18 cm ahead, along the wing surface.) 

VG's installed, at ~ 15 % chord:

Tuft test photos:
Each successive row shows the wing at a lower speed, as shown in the left hand column.

(Nov. 2008: Table reformatted to work better in Internet Explorer.)





No VG's:


No photos at this speed.


With VG's:




70 mph IAS
(in a climb)

No VG's:

With VG's:





No VG's:


With VG's:















No VG's:

With VG's:


(Actually slightly under 40 IAS, perhaps 38, just above where wing rock starts)


No VG's:

The first photo below shows some classic airfoil stall behaviour at the first full row visible inboard  -- tufts go from pointing aft, to pointing up and towards each other, to pointing straight forward. Some of the tufts look like they want to leave the wing, like rats from a sinking ship!



With VG's:

30 mph

No VG's:

Not applicable without VG's!
With VG's:
(At a speed just above where wing rock starts.)

(Although there is parallax, from my right seat position,
 the IAS is 30 or even slightly below)
(Ignore the incorrectly set altimeter.
 It was set 1000' low so we're really 2000 ASL, or 1300 AGL -- still low for flight testing!)


One more photo, with vortex generators, during a stall:

The wing has dropped, a large area of it had detached flow (unlike the smooth flight pics above), but the airflow at the tip has almost instantly reattached.


Before the VG's were added:        https://youtu.be/WMSddgL78VM
Shows some slow flight at the edge of the stall, with some wing rock, followed by a segment showing slowing from 60 mph IAS down to 40 before adding power and speed.
~ 2 minutes

Another before-VG video:           https://youtu.be/xBtXS7sNCcs
This is a more organized and detailed video. For each segment the speed called out by the pilot is shown on screen.
The video includes takeoff, climb, three sets of speed reductions to near the stall, and the landing. One segment includes a full stall break. Has sound, unlike the previous video.
~ 2 1/2 minutes

After the VG's were added:    
Shows a reduction in speed to 30 mph IAS, some flight at that speed or below, at the edge of a stall, wing rocking, before adding power and speed.
It shows how well the wing is flying at 30 indicated, but lacks an actual stall break which would be interesting to see.
Only had my digital camera with me, but it was sufficient for a video.
~ 30 seconds

    (Videos were on blip.tv long ago, then disappeared, and are now on youtube)

If you have comments or found this useful, let me know: