Saturday, April 23, 2011

Yak vs. Yak - Round 2

Scores from Round 1 will be posted shortly. Next up is the Great Planes Yak-55M Sport/Scale Aerobatic EP ARF 50". Another mouthful. Another Yak.  Notably absent from the official Great Planes supercalifragilistic naming convention, is "3D."  However, looking at the barn door control surface with 45 degrees of throw available, and the awesome T:W ratio of nearly 700W, half-weight twin engine, I'm cautiously optimistic.

The "Great Planes Supercalifragilistic Yak-55M" would be a much sweeter name for this starship (next time they'll probably ask me first).  It is, by any non-RC-indoctrinated standard, huge (super), drop-dead gorgeous (cali), and like all airplanes somewhat delicate (fragilistic).  That said, GP's interlocking, laser-cut balsa and ply construction covered in Monokote is one of the strongest builds around--more often than not flying away miraculously unscathed from some of the most shocking, high speed dirt pancakes.
The larger version of the red star on the tail is actually part of the
decal page so you can center it as shown in the upper photo set
This GPY55M is perhaps my most unbelievable deal yet. ARF quality is unreal. Price was even unreal-er. Towerhobby's promotional price of $139 stacked nicely on the usual $20 off coupon on top of free fast shipping. Total cost? An eye watering $109.  It's a crime that the flimsy little UMX Beast, with rather limited real aerobatic capability, cost me 60% more than this stunning 50" GP ARF:
The extremely expensive UMX Beast (shown with aftermarket 3mm prop adapter)
looks diminutive next to the GPY55M's huge, airfoil x-section tailplane
On the Ground:

This ePerformance series Yak goes together faster than the GP 41" class Yak-54.  Build time is advertised as "just a few hours" and for this kit it is true.  The GP Yak-55M has a higher degree of assembly than the smaller GP ARFs, and the wing attach system is quick and perfect. The sum total of wing assembly is sliding in the carbon fiber tube, gluing four small carbon fiber guide pegs (shown below poking through the fuse sidewall at the leading and trialing edge), and then a simple aluminum thumb screw does all the work:
Thumbscrews attach the wings in seconds.  Battery access is
through the top magnetic canopy hood, which clicks positively in
place with side fences to prevent a side-slip canopy departure in the air
The ARF model is perfectly jig built and beautifully finished in Monokote.  The hardest part (and it's not hard) of the "Almost" in "Ready to Fly" is inserting the metal U-bar that connects the two elevator-halves, then sliding-in the horizontal tailplane and aligning it perfectly while tacking with a bit of reasonable working-time glue. 

Aligning the horizontal stab gives a good idea of jig quality at the GP assembly house.  I like to compare multiple measurements to ensure a straight tailplane, since a true tail is so important to square aerobatics.  Whether you measure from wing tips to tail tips, wing roots to tail tips, or from various fuselage center points to tail tips, all the measurements align within a fraction of a 1/16th of an inch.  Horizontal level is on the money too, without adjustment, and the vertical tail is already installed dead perfect.  Spectacular job GP!

The only thing consistently lacking in these otherwise awsome GP ARFs is the wheel pants, which are typically very thin plastic.  The 55M's pants (Yak-style trailing fairings) are a little better, with basswood interior hubs for nice solid attach points.  But the hollow, molded plastic wheel fairings would only serve as two mini parachutes, so I opted for thin 2.5" wheels instead of the the included 2" wide tires, for aerodynamics sake:
Shown with 12 x 3.8 propeller
2.5" white-hub thin wheels
(the kit includes 2" wheels along with fitted 2" wheel spats)
Weight thus far is very encouraging, with a 56-58 oz GP total weight estimate.  The model as shown above, including twin tandem Super Tigre .10 engines swinging a 12x3.8 and spinner, is a dainty 36.8 oz.  My 55M still needs four x 0.7 oz metal gear micro servos to move in, plus a receiver, and motor batteries.  I'm hopeful to keep fly weight under 15 oz/sqft. CG is spot on, already, which sounds about right after two MG servos go into the tail and the batteries reside just forward of the main spar.
As shipped from the factory except for the piano wire
landing gear connection. Even the cowl's interior  firewall comes
factory assembled, magnets and all.  Tons of interior space.
Initial test runs look awsome, with fully redundant ST .10 x 2 motors running on fully independent power systems shoveling-out mega-power with little effort and feather weight (4.8 oz total motor weight, uninstalled).  Initial prop testing is finished:
  • APC 11 x 5.5 = 273W per motor, 23.8A per motor @ 10,880 RPM 
  • Sum = 546W,  47.6A
My installed thrust scale can't read high enough, but a rough uninstalled calculation is about 79 oz, or 1.6:1 at around 57 mph flat out.  Installation error from the flower pot size, scale radial cowl might lower that to a more realistic 1.4:1.  The nice thing is the low power draw, which should be good for about 13 minutes of mixed-throttle flying time with two 4.3 oz 1300mAH 3-cells and should squeak in under 50 oz RTF weight.  With double independent delivery, you get 40C for free from cheap 20C batteries, or one could up performance even more with two 40Cs for an 80C power system.
Flight tests show this configuration is as strong or stronger than a
Power 32 motor on 3S, but 3S is not enough for 3D performance.
Second prop test is now complete:
  • APC 11 x 3.8 SF = 327W and 28.7A per motor @ 10,340 RPM 
  • Sum = 654W,  57.4A
Thrust feels substantially stronger, desperately trying to rip the plane out of my tight white knuckles.  During the this trial I flew the throttle to exactly a 1:1 thrust to weight ratio, using two 6.7 oz 2200mAh 40C Lectra LiPos, or 80C in total, with an RTF weight of 52.1 oz.  That it occurred at almost exactly at the 50% throttle mark at just a smidge under 11A per motor as measured by my Wattsup. 

With the second power system option (11x3.8, 4400mAh @ 80C), one could hover for about 12 minutes, uninterrupted.  Wing loading is respectable for a 50" class model at just 15.7 oz/sq ft, up from 13.4 oz/sq ft using the 2600mAh solution.  For only a ~30% flight time reduction (a lesser hit than the straight mathematical 40% since the weight reduction also lowers thrust required over the entire flight envelop), the 2600mAh 40C battery system seems like a better way to 3D this horse.

A Z83D prop would be over-the-top efficient for the competition-minded.  the rough figures indicate my ST .10-twin with a Z83D prop would roughly double the thrust to weight ratio of any other commercially available complete power system and with higher efficiency.
What is perhaps most advantageous about the super light .10-twin is how under-worked the motors are, especially at cruise power, making it amazingly efficient compared to a single, heavier brushless outrunner.  The low operating temp makes either motor much less likely to fail than one over-worked engine, and completely redundant if one motor does fail.  In sum total, it is perhaps 16 times more reliable:
  • 4x as reliable for any twin engine design over any identical single design (factor = .5 x .5 before losing all power)
  • Double (probably more like 4x) the motor life per motor from half the Amp draw (factor = 2 x 2)
  • = about 16x less likely to lose all power in the air from motor failure
Receiver power is also redundant and independent too, eliminating the primary cause of cheap radio-induced crashes from brownout, which very common with Spektrum radios.

Interestingly, it seems that the practical kV limit kicked in, above, well before the motors were tapped out.  1250 kV * 0.80 (typical real world constant) * 10.9V (under load) = maximum of ~10,900 RPM.  I can fit a third motor to the shaft, but it is clear that this .10-twin delivers more than enough power for 3D.

In the Air:


Initial flight testing is complete, with some very interesting results.  First things first, this Yak is an excellent flier.  Even with only 9" additional span, it is exactly twice the size of the GP Yak-54 at around 55 oz current flying weight.  Wing loading is a respectable 15 oz for its scale, but its certainly not the AT Yak-54 Forrest Gump feather, not by  long shot.  Still, the overall experience is intense fun, hell, I can hardly stand it.

Second things second, the twin ST .10, while providing amazing 3S thrust, was simply too high a kV at 1250 to ride this horse to glory once the cowling rode shotgun.  The after installation losses, the big ass Yak could barely hover at battery mid-life.  CG was controllable with the 3S 2600 mAh shoved into the motor mount bay; the 55M's internals are very well designed, basically a huge, flexible empty space.  Maneuverability was simply awesome.  Power delivery was less than ideal for a 3D beast, the 11" prop had too little acceleration bite, especially with the APC prop's high efficiency driving thrust output into a higher RPM regime. 

Thrust comes from mass flow, you can move a little air a lot or a lot of air a little.  This Russian thug has a flower pot for a radial cowl, almost 7" round.  A high kV, 3S solution, even my twin, simply could not turn a wide enough prop to get the thrust column out and around the cowling.  That physical constraint essentially drives the approach of pushing a lot of air a little.  Best thrust comes from spinning an 11 incher at almost 11K RPM, a little air a lot--but right into the cowling.   At 1250 RPM per volt, the twin .10 is perpetually stuck in "5th gear."  Gear reduction driving a larger prop diameter could work marvelously (essentially lowering the kV), but the solution was too involved for this test.  I entertained using a more effective slow fly prop, as lift generation is pushed outward from the hub, making it less efficient, but more effective for a big cowl.  But even so, 11" seemed a bit too fundamentally small to run through this 3D bird's veins.

The cowling diameter vs kV issue is a real shame, as the ST .10 twin generates the exact same, great motor thrust to weight ratio as the .10 single, about 16x, and it does so with a relatively light weight total power system (4.8 oz motor + 8 oz battery + 2 oz for two 30A ESCs = 14.8oz).  Doesn't matter, cause you can't use an 11" prop effectively on this big Yak.  Installation error is a cruel witch.
 
The recommended motor is a low kV Rimfire 32 and 2200mAh 4S (8 oz motor + 9 oz battery,  + 2 oz 60A ESC = 18.8 oz power system).  Since this is a Yak v Yak showdown, I decided a conventional second try was the best way to go.  Unfortunately, my LHS didn't stock the Rimfire motor, so I went a similar spec, 700W to 1000,  770 kV, e-flight Power 32. With a good stock of 3S and 4S batteries and  2-6S 60A pro ESC, this should be a lot of fun.

The 11" twin motor drove a 95 sq inch air column, of that 95", abut 40 sq inches was consumed by cowling, leaving about 52% of the thrust column to do solid work .  The recommended prop is a 12".  While only 9" wider, it shoves a 113 sq in air column, leaving 64% of its higher thrust output unscathed.  A 13" prop drives 70% of its even higher thrust around the cowl boundary.

Is this a trick question? e-flight allows 11x7 to 14x10, so 13x6 seemed about right, with the ability to swap between 3S and 4S without over-stressing the motor with some kind of 14" 4S hatchet.  And so, off we go to our knife fight, armed with a Howitzer.
The big GPY55M has it's work cut out for it, the ATY54 and GPY54
are incredibly fun and much more economical to outfit and fly.
The motor is a monster at 8 oz, even with a lower motor T:W of around 13, pure thrust is over the top for this airplane at around 110 oz, a cool 2:1 aircraft T:W.  After I saw what this plane can do at 2:1 T:W on 4S, I immediately added a few 0.25" square carbon fiber beams in support of the thin plywood motor mount box.  Holy smokes, this huge radial cowl drives some amazing overkill.

I decided to control myself and accomplished the first 32 motor flight on 3S.   On 3S, aircraft T:W seemed a little lacking, even for sport flight.  Over the tops were strained; no sustained hover, the 55m slipped lower at full bore.  Hmmm.  Time for 4S.

Boom.  Like a cannon shot, this thing rocketed off the runway and tore upward to Heaven.  What a difference a cell makes!  This airplane had finally come alive, and oh man, how alive.  With one minor problem - impending death.  The 32 motor on 13x6 swallowed 4S batteries like Shaq downing a mini box of Pringles.  5:00 minutes on 3500 mAh, if I strrrrrrrrrretched battery life.   Yikes.  What next?

Alright, its time to see if I can find some middle ground.  Props, Watts, and Amps follow, on 4S:
Prop
Full Throttle Amps
Full Throttle Watts
Min Hover Throttle
Min Hover
Amps
Min Hover Watts
13 x 6 MAS K series
54
838
50%
25
385
12 x 6 APC Thin E
39
614



12 x 6 x 3-blade MAS
45
632



12 x 4 MAS K series
34
535
80%
28
430
12 x 3.8 APC Slo Fly
61
985
55%
26
395
What a range. 12 x 4 K tomorrow.
 ...
On 4S, the 12x4 K was too lethargic in both airspeed speed and thrust.  Amazingly, it didn't seem much better on battery life, at around 6 mins T.O. to land.   The higher throttle requirement throughout the flight envelope seems to negate most or all of the full throttle Amp draw advantage.  12x6 APC and MAS K on deck.

The 12x6 APC had plenty of top speed but was uninspiring on the low end of the airspeed range.  APC designs continue to disappoint.  Efficient blade design, yes, but that just makes the prop need to spin faster before it generates the lift you expect.  Perfect example of engineers obsessed with optimizing the most intuitive, but wrong set of metrics.

12x6x3-blade looks like it could deliver a great blend of performance and battery life.  It should generate 50% more thrust at the same RPM and give a good unloaded top speed with a 6 pitch.  It is on deck now waiting for the mini hurricane to end....