THE GREMLIN           by GRAHAM WOODS

(Dictionary: a mischievous imaginary being)

It's been a long wait but my brother David finally got his hands on one of these mischievous creatures after waiting over a year ;-) for one to appear. Having had a couple of Mini-Vectors, a pair of Aresti 2Ms and an Aresti 108 he was looking forward to this lightweight VTPR (sic) from the same manufacturer but more of that later. There are, of course, pros and cons to this new model. I found a few niggles but overall the quality of the build is better than expected. The model feels very strong, too strong in fact, it will be a toughie and I reckon it will survive the hardest landings.

I'll begin on a positive note and say that the quality of this all moulded slope soarer is excellent. The place to look is at the join lines where the halves of wings and fuselage are united. This paint design in green, white and black for the top and red, white and black for the bottom is by Sansibear, I believe, and its paint lines were very good with a tiny amount of overspray in just a couple of places. A much better build than other models from this stable with the exception of the Czech built big Aresti 108 IMHO. 

 FUSELAGE, FIN AND RUDDER                           (Click on Thumbnails for larger images)

The overlap joint in the fuselage is excellent and there's even an extra bead of resin and microballoons or thixotrope? at the base of the fin, on the join line, where many fuselages can split on a heavy landing, nice touch. There is Kevlar and carbon around the wing root inside the fuselage but saying this is 2.4 GHz friendly with slabs of carbon exactly where the receiver and its antennas are likely to be placed is doubtful. However, this model is likely to be flown 'close in' to the slope, and therefore at close range, so I'm probably nit-picking here. The fin is reinforced internally with a wood 'spar' and the rudder post. 

The canopy comes with ready fixed spring wire catches and is a very good fit. The actual shape of the fuselage is kind of triangular in cross section, tapering slightly from bottom to top and it has a pleasing curve along its length, i.e. it is not straight and slab sided, almost aerofoil in shape. (See noseweight image below.) The servo tray supplied is of 4mm thick epoxy sheet (overkill) coming in at 37g! This was found to be unsuitable for the Savox 0255 rudder servo used for the rudder so it was replaced with a 6mm plywood facsimile weighing in at just 20g.

The moulded rudder was a bit problematic as the movement on the non-hinged (left) side of the rudder was limited and about 3mm of material had to be removed from the fin here to get adequate movement; 45°+ either direction was achieved. The wiper for the rudder also had to be cut back for the same reason. Working on the model, figuring out how to do the elevator servo mounting (argh!) meant a lot of unintended stress and movement came upon the rudder and the Kevlar hinge of the rudder began to fail/suffer. It needed taping to keep it from falling off. The rudder hinge doesn't use any of the peel ply tape that is sometimes used for surface hinges in moulded models. The rudder hinge, therefore, needed some remedial action to ensure its reliability before flight testing. In fact, Blenderm tape was used on both sides for the rudder hinge.

 TAILPLANE/ELEVATOR

Well I'm afraid to say that I reckon the designer was a bit schizophrenic in his thinking about the Gremlin elevator. I say schizophrenic because his two (or three) ideas of elevator servo mounting and elevator operation are contradictory or incompatible with what one is given in the 'kit'. Trying to make it dual purpose doesn't really work. There is a snake installed for the elevator but no bellcrank was included in the 'bits bag' nor mounted in the fuselage; there is an access hatch cut in the fuselage, presumably to access such an elevator servo bellcrank but this is no use for a fin mounted servo. The supplied epoxy servo tray has elevator (and rudder) servo holes cut ready to take a 9gm servos though.

However, it has been suggested that the Gremlin can use an elevator servo in the fin but in this case said hatch is in the wrong place and any ready installed bellcrank would have been in the way. Someone is trying to have their cake and eat it at the same time! It's a distasteful cake too because neither method is satisfactory. After much musing my brother David decided a servo at the rear was the better option but how exactly to accomplish this was a struggle. In the end, the rudder post was cut away to allow the KST H08 elevator servo in a small, thin ply tray to be glued with 5min epoxy (so it would be easy to remove) through the rear of the fuselage a at the rudder post to one side of the fuselage, part of the rudder post was removed to achieve this access. The hatch here is now pretty much redundant because a servo cannot be mounted in the fin through this hatch. A direct link from the servo to the elevator actuating wire is rigid and slop free, albeit using only a fraction of the servo's movement to get something like the 6-7mm up/down elevator movement David reckoned on using. (The ply servo tray in the fin can come loose on a heavy landing so beware!) I say this because it did!

Getting back to the bellcrank idea, a little thought shows that the geometry on any installed bellcrank would be wrong as the servo movement on a servo mounted on the servo tray would have to be minuscule given the position where the elevator snake exits at the rear so this is wrong too. On the other hand, it has been suggested that the elevator/stab combo could be modified to become an all-moving tailplane (rather than a fixed stab plus elevator) with a bellcrank to facilitate the Gremlin as a VTPR model with a mega tail movement. Getting a +\- 90° stab movement à la Cahour, BPLR etc. or Henderson pulley may be possible, I don't know, that's something we've never done.

A bellcrank would work but the actuating wire slot would have to be enlarged considerably and the moving elevator actually glued to the fixed stabiliser. Making the elevator joiner wobble and slop free would be a hard task I might add. The exit shroud for the rudder would have to removed as well leaving a gaping hole/slot in the fin. Confused? We certainly were! It's your choice. In any case, I have included images here so you can see David's servo-in-fin solution for yourself. This is how it was accomplished on this model, you may opt to do it a completely different way. No instruction is given and saying it is adaptable is simply not on a small, expensive 500 Euro mouldie... you could in fact say this about any model.  

To be honest, I personally see no merit in having a separate elevator and stab and fin mounted servo. On this model the distance between the actuating joiner and hinge line is just 8.5mm. This means close-in hole on the servo arm actually only using a fraction of the servo's movement, effectively not using the full servo resolution. 

NOSEWEIGHT

From the look of the Gremlin at first sight it seems that it has a long nose moment, and it does, but the rather heavy tailplane (actually 5g heavier the the 2 metre Aresti) means adding a lot of noseweight to balance. To this end a piece of lead was cast to exactly fit in the nose. The piece of lead to approximately balance the model on the wing joiner for the initial flight was 218g bringing the all-up-weight to 1700+g and a wing loading of over 20z/ft). In comparison, David's two Aresti 2Ms had approximately 20 ounce loadings and the 'big' Aresti 108, a 17.5 ounce loading. The battery chosen for the Gremlin was a set of four Eneloop Pro cells. Originally the model had been fitted with a 2200mAh 2S Lipo but this was changed when the High Voltage KSTs were removed.

 WINGS

After the drama of the stab/elevator and its fixing, the wings were relatively easy. Four servos with straight push roads  to the flying surfaces and little homemade ply trays to hold the KSTH08 wing servos in place. We thought that the aileron servo well holes were not quite in the correct in orientation and place, btw. Even the wing servo mounting was problematic since lightweight servos were chosen as this was supposed to be a light model. After a heavy landing and a pair of stripped aileron servos all the KST wing servos were replaced with Savox 0255s.

The quality of the wings is mostly good too, apparently with plenty of internal carbon to go with the Sine Wave shear web incorporated in this model which one can espy from the servo wells. The wings appear very strong and torsionally rigid, too much perhaps, therefore very resistant to any bending or twisting force you can apply with your hands. While the wings are maybe over strong for such a small model, the hollow 330mm carbon wing joiner (18mm dia. int. 15mm) tube may appear too weak for its job, we'll see, especially if any ballast (650g/23.2oz) is fitted in said tube as suggested. Because of the tight fit of said tube, I doubt an Pitcheron/Wingeron system would be possible without extensive engineering and ball races to prevent binding in tight turns. 

Like the rudder, we both like lots of control surface movement, on the Gremlin the need was felt necessary to increase the up-aileron movement; being bottom hinged, about 1.5mm of the top wing surfaces (where the ailerons are) was removed to get the necessary throw. More work on a moulded model! The panels have stub aluminium dowels for wing alignment but no thought appears to have been given to holding the wings on the fuselage. The traditional tape method is not really an option as the fuselage hardly has adequate wing fairings on which to stick tape.

Falling back on the rubber band/spring and hook system needs the wing roots being reinforced inside the wing roots with ply to take small hooks as the wing root faces are paper thin. The narrow fuselage makes this awkward though. The aileron and flap horns come ready fixed, they are substantial and pre-drilled. The pushrod exits needed to be opened up to take the clevises used on this model. The wing horns come with exit shrouds and the servo wells have their covers which fit nicely. All in all quite a nice pair of unfinished wings with a straight leading edge.

 VTPR (Voltige Très Près du Relief) and ULTRABATICS

The possibility of the Gremlin being a VTPR or Ultrabatic model arose because of chat and discussions on one of the RC Groups' fora on whether the Gremlin would be RTF, PNF etc. Existing VTPR models like the Fish, Ecalibur, Vagabond typically are 1.5m span and weigh upwards of 400g. This Gremlin is somewhat heavier... early measurements suggest that it may come in at 1500g A.U.W. for a 1.5m span model, and without mega 180° tailplane movement for 'floops, loops and dupes'. It was in fact over 1700g!

David's model does not have the all-moving elevator required for French style VTPR or US Ultrabatics and it is not a lightweight by any means. I'm sure it will fly tight aerobatics on the edge of the slope (i.e. VTPR English Style) but I doubt it will ever fly in the same manner as those Ultra EPP or light, French VTPR models one sees on YouTube. We'll see, the Gremlin hasn't flown yet as I write this build piece during its assembly over Christmas 2016.

FLYING

The first flights took place on a very windy (30-35mph) Christmas Day 2016 at Ivinghoe Beracon. There was too much elevator movement and the CG was wrong. Landed quickly, and, after a couple more proving flights, weight was added to the nose bit by bit bringing the CG to 73mm from the leading edge and the noseweight up to ~225g on the day. Subsequent flights on other slopes (at Cheriton, Firle & BoPeep and Ivinghoe) in lighter then better lift had this noseweight reduced to 204g then down to 180g with a rearward CG at 78mm from the leading edge. This also improved the tightness of the loop.

As a bystander, my impression of the Gremlin from its first early flights is that it rolls axially and very fast, however, the loop was not tight at all, even with snap flap engaged. It seemed quite fast and coped with the 30mph wind with no problem, then again it does have a wing loading of 20oz/sq.ft.

With hindsight, the choice of wing servo was perhaps not the best decision made as both KST H08 aileron servos lost their output gears due to the rotor on the hill causing a heavy landing. They were replaced with Savox 0255 servos as I mentioned earlier. The butterfly needed an increase in elevator compensation (70/75% I think) as well. The elevator movement was set at 7-8mm and the aileron movements reduced a little. Stall turns were fine but large side area and a wide chord means wind can have a larger effect than one may expect.

Personally, I see no reason why one would choose this model over a Mini Vector or Aresti 2M except that it looks a bit quirky looking and it does lay across the parcel shelf of a family car. BITBOC (bung in the back of car) is the acronym coined by its designer and this is what it does.