Speed and small sexy bodies may be cute for supercars – but they can be a lethal cocktail for inexperienced pilots. Beech had this problem with the V-tail Bonanza, Piper with its Malibu and Cirrus with the SR-22. The good es is that in time, all these problems were solved.
Non type certified aircraft are tested to far lower standards than type certified aircraft and if a plane is just too fast for its own good – it is generally the unsuspecting pilot who will be the first to discover its limits. One of the most spectacular such events happened in South Africa in 2010 when a VL-3 broke up in flight – with fatal consequences.
It is a hopeful sign that the latest incarnation of the VL-3 is called the Evolution. Hopefully this glamorous little sportplane has evolved from its troubled progenitor.
The original Aveko VL-3 has a scary history in South Africa – and indeed around the world. A look at its accident stats shows 28 significant accidents (including such minor items as a nose gear failure in South Africa) yet with 30 fatalities. With more than one fatality per crash, some may consider it a lethal little beast.

I have some personal experience ¬¬– in 2010 I was flying the Presidents’ Trophy Air Race when competitors ahead reported a plane having gone down after breaking-up in flight.
The crew of a VL-3 had put the nose down to descend to a check point. The aircraft behind reported large bits falling off. The CAA accident report found that the published Vne had been exceeded, causing flutter. I suspect it began with the ailerons, followed by the elevator. With the elevator gone, the nose pitched hard down, the main spar broke and the wings came off. I cannot unsee the gut-wrenching pictures the first helicopter on the scene brought back of the two dead crew in their wingless tub.
This crash caused much controversy about the real nature of Vne – the Velocity to never exceed. Citing this tragedy, Peter Garrison wrote a seminal piece on Vne, which should be required reading for anyone who wants to fly a fast little plane – especially a non-type certified one. The title is telling: “Are you feeling lucky?” https://www.flyingmag.com/technique-proficiency-technicalities-are-you-feeling-lucky/
Peter Garrison wrote; “An airplane was descending toward a turn point in a valley when the pilot of a following airplane saw what appeared to be paper scattering behind it. An instant later, a shattered wing separated from the fuselage, falling to earth a hundred yards from the main wreckage. The airplane was a Czech-built Aveko VL-3 Sprint, an NTCA (non-type-certificated aircraft) that bills itself as the “world’s fastest ultralight.”
After the race crash, the Vne for the VL-3s in South Africa was reduced from 165 KIAS to 145 KTAS – just about its straight and level maximum speed. And most importantly, note the shift from Indicated to True Airspeed.

There was clearly a problem – so Aveko went back to the drawing board. But the doubts about the safety of little sports plane had set in, and Aveko went bust. Then two intrepid Belgians brothers, who were distributors of the VL-3, bought the factory. The VL-3 was redesigned with more carbon fibre to make it stronger and stiffer. Essential aerodynamic changes were made – the most obvious of which is a balance horn for the elevators.
The manufacturer claims that the design has been thoroughly re-validated by EASA and the Vne is now back at 183 Knots Indicated. Remarkably, the JMB website claims that the VL-3 has been tested to be flutter free at 450 km/h =240 KTAS. Sounds good, but the rub is in old the old conundrum of Vne as a True or Indicated air speed. My whizzwheel confuser says that at a typical South African ISA plus 20 at FL200, 183 KIAS is 260 KTAS. So hold and below! – the factory is saying that they have not in fact tested it to the real VNE – which is after all a True airspeed. Thanks to the dramatic power increase this is still a non type-certified aircraft that canal too easily bust through its Vne.
For additional insights former EAA Chairman David Toma writes; “So the never exceed speed (VNE) is a limit indicated airspeed that the pilot is to not exceed, but it isn’t always for the reasons that we think. Here are the constraints that a designer should be considering when specifying the aircraft’s VNE.
“The first would be inertial loads that could be imposed on the airframe during gust loads as specified by the chosen certification specification. Such loads would require a higher dynamic pressure if we are to look at the worst-case scenarios and this can possibly lead to the assumption of flying closer to sea level and hence one could use indicated airspeed directly for this constraint.
“The other constraint would be flutter, which as we all know is dependent on true airspeed as compared to indicated. Lastly one must not forget about acceptable handling qualities. If beyond a specific airspeed the aircraft has unacceptable handling qualities, then a new applicable limiting velocity is specified. Having considered all the variables, a decision is to be made as to how to communicate this to the pilot. The easiest choice is through an airspeed indicator and an indicated speed is chosen for the VNE. However in cases where an aircraft has a high service ceiling, one indicated airspeed becomes rather limiting and a table of VNE speeds is given for varying altitudes. That also saves the pilot from continuously whipping out the whizz wheel as they climb in order to calculate a true airspeed.”
This makes JMBs decision to turbocharge the VL-3 even more dubious. The original 912 produced around 75 hp at a typical cruise altitude of FL075, whereas the Rotax 915iS is good for 135 hp to 15,000 ft. Since Vne is actually a true airspeed, it is high altitude flight that is the most dangerous. I fear the highly turbocharged 915 will test the 183 KIAS Vne, once again with fatal consequences.
So let’s see how far the original Aveko VL-3 has come in its latest incarnation – the JMB VL-3 Evolution.

On the ground
The subject of this report is ZU-TSC, a 2022 VL-3 Evolution which has Rotax’s 915iS 141-horsepower powerplant.
On the apron in front of the large hangars at Jo’burg’s Eagles Creek airfield, the overriding impression is that the VL-3 is small, short and sexy – it looks like a shrunk Lancair Legacy. Apart from its small stature, what is also evident is that the quality of the fit and finish are to supercar standards.
The composite surfaces are mirror smooth and JMB has moved away from the earlier version’s boring white paint schemes to elaborate colour schemes glistening under a clearcoat varnish. The 50 shades of grey with a colour that can only be described as Lamborghini orange, sets this plane apart.
The large bubble canopy is front-hinged. Swinging it open reveals a beautifully detailed and finished interior.
Giving it a ramp presence, like a sport’s cars fat takkies, is no less than a four blade, constant-speed, Duc propeller with, unusually for Rotax, a traditional hydraulic hub. The prop’s governor is controlled via a blue lever on the throttle quadrant. (Electric constant-speed props are an option).
As befits a sexy little hot ship, the wings are small. In American units, just over 105 sq ft, with a 14 lb/sq ft loading. Wingspan is just 27 ft 7 in. (or 8.44m for South African hangars). Each wing has a 70-litre fuel tank, fitted with flush lockable lids. For an engine that can swallow more than 30 litres of fuel and hour, 70 litre tanks are too small. The local agent, Warren Massey, says that South African models will have a gross weight increase from the EASA LSA 600 kg limit to 750kg. This will enable 160 litre tanks and a decent load carrying ability in the cockpit, but for any useful range, 200 litres would be better still.
For low-speed handling, the wings have split flaps over two-thirds of the trailing edge. Many manufacturers have gone with slotted, or even the complexity of Fowler flaps for greater aerodynamic efficiency. JMB uses a simple manually operated split flap to achieve the same purpose—and it gets results by making the flaps gigantic. The split flap has the advantage of doing without worm gears, draggy external hinges, or other complex electro-hydraulic mechanisms. On the VL-3 the flaps extend to a barn door 55 degrees. Great for drag – but they require careful handling for a go-around. Fortunately the flaps are operated by a single large handbrake ‘Armstrong’ lever which makes them quick to retract.
Climbing into the cockpit requires a step up onto the low wing, which has a walkway with thin non-slip strips, presumably to keep the wing surface as smooth as possible. You can then step over the cockpit sill and with a bit of dexterity, by leaning on the front edge of the turtle deck, avoid having to step on the beautiful grey stitched leather seat squabs.
The sculptured bucket seats are reclined, with firm but good lumbar support. Once seated, the 45-inch cockpit seems surprisingly wide, so you don’t have to overlap shoulders. The seat doesn’t adjust fore or aft, so setting the recline angle, then adding or subtracting cushions, is the only way to alter the seating position. This is a plane you sit in, rather than on. Thanks to the reclined seating position the top of the panel is roughly at the pilot’s eye height – so the view forward is not great. The nosewheel retracts rearwards into a large tunnel that takes up space in the footwell.
Two NACA scoops supply fresh air to the cockpit via eyeball vents, and glider type pop-out vents on the sides of the canopy are great for an air blast on hot days.
A whole airframe parachute is located ahead of the canopy, and a prominent red handle, bolted, in what seems something of an afterthought, to side of the centre pedestal, fires it.
All the controls and instruments are in easy reach, despite the reclined seatback. Two large Garmin G3X displays fill the panel. An optional extra is a neat feature to have: a wide-angle video camera on the belly showing all three undercarriage legs so you see the actual gear legs and wheels. Its one-upmanship on a Cessna 210’s wing-mount mirrors. Next to the surprisingly small gear toggle switch, oddly located at the top of the panel, there are also the usual ‘veggies’ – three green undercarriage up or down lights.
The undercarriage is electrically controlled and hydraulically actuated. Looking from the front, the size of the landing gear doors is apparent – and they do not align with the airflow but rather form a large airbrake. Again a reason to treat go-arounds very cautiously and to make sure you clean the gear up promptly – especially considering its slow transition.
Something to note is that the aircraft has such a high power to weight ratio that is also coupled with low pitch forces that I don’t foresee a problem for a go around in the dirty configuration. The 182/206/210 Long body Mooney that require a lot of trim changes with flaps and have higher control forces, end up throwing your nose skywards if you simply slam the throttle forward for the go around without anticipating the required trim change.

Flying the VL-3 Evolution
For our flight test, David Toma flew with the factory demonstration pilot Joren Vermeulen (he’s Belgian, not a Saffer). Starting the 915 takes some practice as the design philosophy is completely different compared to the more traditional Lycosaurus engines. Here there are lights that are used to communicate with the pilot as to how the engine is running and if there is any exceedances or failures. Flip on the two ignition lane switches and the two fuel pumps and turn the ignition key. The Rotax springs readily into life, and, compared to earlier engines, has less clatter from the prop gearbox, thanks to its larger impulse damper.
Despite the weight of the complex engine, turbo and intercooler in the nose, and the forward CG as we were two-up with lots of fuel and no baggage, the VL-3 is still light and easy to taxi with its tricycle gear and nose wheel steering so there’s seldom need for differential braking. The turn radius is tight, but the small wheels may be harsh on a rough taxiway.
Runup is standard Rotax 915iS: the computer performs a diagnostic test of both ignition systems by toggling the lane switches.
Whilst it’s a turbo, so you shouldn’t just bang the throttle open even though the engine management system limits boost automatically. As you get to full throttle, the turbocharger limits manifold pressure to usually around 47 inches, and this occurs midway during the take-off roll. Using a brisk but progressive application of power, take-off acceleration is impressive considering that we had 37° of the massive split flaps deflected. You can definitely feel that it wasn’t just Eagle Creek’s strongly downhill Runway 08 that got us airborne in under 200 meters. Above 82% power the 915 goes to auto rich and has a huge thirst for such a small engine, sucking up to 46 litres per hour. This is something to keep in mind when flight planning for a long leg as a reduced power setting, and thus speed, is required to achieve the required range.
David Toma reports: “The low seating position heightened the sense of speed down the runway. With a bit of back pressure on the stick, we were airborne. It was immediately apparent that stick pitch forces are noticeably light – a common feature of non-type certified aircraft. Those used to clunky Cessnas will find themselves having to guard against pilot induced oscillations.
We were almost instantly at Vy (best rate) climb speed of 79 knots, indicating a healthy 1,500 fpm over the runway end. It was natural to lower the nose to 100 KIAS and still be climbing skyward at over 1000 fpm. Gear retraction takes about four seconds.
The graphics on the large Garmin G3X EFIS make situational awareness a treat. The two large 10.6-inch screens dominate the instrument panel. There is even an option for Traffic Awareness System (TAS) – like TCAS but without the automatic instruction to both aircraft and, importantly for the soon to be mandatory ADSB.
A huge benefit is the intelligent and smooth Garmin GFC500 integrated auto-pilot that can be programmed to maintain altitude, direction and attitude and even fly a coupled approach. A particularly useful feature on the Garmin GFC autopilot is to simply push the blue “level” button which means the plane flies hands-off straight and level, leaving the pilot free to snap that photo for the memory bank. If you plan on eating lunch rather, then set the autopilot in nav mode including vertical navigation and watch how the aircraft not only corrects for the prevailing winds but also sets up a perfect descent to have you exactly at the joining altitude as you reach your destination!
The VL-3 is true to its looks in terms of sporty handling. Pitch is significantly lighter than roll, particularly at high speeds. David Toma notes, “The ailerons are ‘RV10 stiff’, which to me serves as a reminder of the speeds that are being achieved and that you must be mindful of large control deflections.”
The ailerons use control cables rather than the more frictionless push rods, yet it was immediately apparent that the breakout forces are low, and the roll response is still crisp, yet without twitchiness. Overall control harmony is fair, and the bubble canopy makes for reasonable visibility while manoeuvring.
Even though we were below the Johannesburg TMA, which is far below the aircraft’s best operating altitude, we explored the VL-3’s speed capability at 7000 ft. The EFIS displays power as a percentage of total rated power, and at around 82%, it changes the engine management from producing best power to best efficiency. Thus, at 82% power, the fuel burn is 9.2 gph, while at 80% the fuel burn drops to 7.4 gph. The G3X shows true airspeed and at 78% power, burning 34 litres/hour, we saw 170 KTAS. Throttling back to 55% power we saw 150 KTAS whilst sipping 18 litres! This possibly explains why 140 litres was selected as an acceptable fuel quantity.
Having been lied to many times before by NTCA manufacturers that seem to ignore large static errors in their selected systems; I took it upon myself to fly a standard position error calibration at 139 KIAS which showed 160 KTAS at 7000ft. The result of this test was a static error of only 2 knots in the VL3! This correlates to the figures shown in the POH for this speed which meant that you can get rid of the normally required pinch of salt when it comes to some of the published performance figures.
I tested the control harmony and roll response by cranking it over into about a 75-degree bank. It effortlessly maintained height, even though the stall warning beeped occasionally. With a full throw of the stick, a 150-degree roll reversal was smooth, with little tail wag.
Slow speed handling was found to be textbook thanks to the stall strips and wing fence. I pulled the power back, thankful that liquid cooling vastly reduces the chance of shock cooling the engine. With the flaps and gear up and the nose surprisingly high, the stall break came at 55 knots. I held the wing in a deep stall and tried to keep it straight with rudder until it would take no more abuse and gently dropped the nose. The recovery just required relaxing the back-pressure on the stick.
With gear down, full flap and power off, the airframe shook until the stall break came at 40 knots—with a very mild and, thanks to inbuilt stall strips, a predictable wing drop. We recovered with less than 500 feet of altitude loss. Even deep in the stall, the elevator authority is excellent. Both recoveries are standard with the caveat that it’s easy to unintentionally exceed flap speeds, from experience one quickly gets accustomed to the required pitch attitude that would keep you away from such an exceedance.
Returning to Eagle’s Creek, we joined a Left Downwind for 26. Despite the cleanness of the airframe, slowing down for the circuit turned out to be surprisingly easy. The VL-3 does however have surprisingly slow gear extension and flap deployment speeds, and I anticipated we’d have to descend early and cut the power to slow enough to meet the limitations of 81 knots for gear and 67 knots for flaps. But at a full 55 degrees, the wide chord flaps, large and draggy gear doors and four blade, constant-speed prop were effective aerodynamic braking. Joren impressed me by arriving almost overhead the runway numbers at nearly 1000 ft, and with everything hanging out we fell out of the sky like a proverbial grand piano, and still made it onto the first third of the runway with just a small squirt of power to stop the precipitous descent. Even more impressive was how stable the VL-3 was with the airspeed happily sitting between 65 and 70 KIAS. I got to try one of those approaches at Aviators Paradise and still remain very impressed with both the performance and how low the pilot workload was throughout the manoeuvre.
As the semi-liquid-cooled Rotax isn’t subject to shock cooling, pulling the power back to idle on downwind is fine and we were able to extend the gear and full flaps in plenty of time. Trim changes with 15 degrees of flap were small. Pitch changes due to further flap deflections all the way to 55° were easily compensated for initially thanks to the light control forces but I am glad to report that there is sufficient pitch trim authority in order to allow for hands off flying in the full landing configuration and at the desired reference landing speed. This is another one of those things that seems to be ignored by several other NTCA manufacturers.
With its light weight and sub-40 knot stall speed the VL-3 can easily be stopped within 100 metres if you work the big Beringer disc brakes hard enough to almost slide the tyres.

This is a plane built for speed – and to go fast economically. It is amazing that it has a cruise speed four times faster than its stall speed,. JMB Aircraft claims the Rotax 915-powered VL-3 has a top cruise of 199 KTAS at 18,000 feet, but 170 KTAS at FL085 will be a more typical set of cruise numbers.
We look forward to new aircraft having an increased max all up weight of 750 kg. if this enables 200 litre fuel tanks, the range will be commensurate with its cross country performance.
New aircraft prices range from $205,000 for a Rotax 912 powered VL-3 with basic analogue avionics to a wallet busting $350,000 for a 915-powered version with dual Garmin G3X primary flight display/multifunction displays, an IFR-capable GPS navigator and many nice to have gadgets. I guess when one looks at the pricing on a new C172/182, a Cirrus or a Diamond, the price appears rather reasonable but R6.3m for a two place NTCA will take some getting used to.

VL-3 Models – Specifications and performance compared