GEV-2.1: The Grasshopper v2

Walden

Member
Grasshopper version 2 is in the works. I have almost all of the design details nailed down at this point. Now that I have a handle on using FreeCAD, I can start drawing out the individual pieces in preparation to building the physical plane. That build process should be easier this time (for one, I cut the total number of pieces from 20+ to ~11). I know that the profile view that I included here is mostly unintelligible, nevertheless as I post more pictures it will all become clearer. The general look is very similar to version one. The biggest changes include: smaller overall dimensions (45cm L x 55cm W x 18cm H), increased AOA, larger control surfaces, elevons now extend off of the upper winglets instead of the T-tail and the servos to these elevons are mounted down in the outer aft corners of the main wing, and the power pod is completely rearranged (battery moves up into the wing, the ESC down below, and the receiver gets its own slot behind the spar). Between the narrower fuselage/keel and the increased AOA twice as much of the prop wash is directed under the main wing unobstructed; combine this with the greater maximum thrust (on paper) and this plane should have little difficulty doing takeoffs from water.
 

Attachments

  • GEV 2.1.svg
    367.1 KB · Views: 0
Last edited:

Walden

Member
Update, it took most of my weekend, but now I finally finished the CAD sketches for all 13 unique foam pieces! I will be cutting, test fitting, tweaking the sketches as needed, and teaching myself to make proper digital plans. Next couple weeks I shall be posting pictures as I build to show how these complex parts fit together, so keep a eye out.
 

Walden

Member
As you can see in these pictures, I have been making a little progress on cutting out the parts. So far all part drawings have needed minor revisions (the winglets in particular gave me the most headache at every design stage o_O). My work area is very small, so as always I start out by cutting out rectangles to the overall dimension of each type of part (left and right mirrored parts together). I came down to being one piece away from cutting everything out of just full two sheets. I am relatively confident that if I split up some of those L/R pairs I can have the final plans be a true 2-sheet. Anyway, at least I can say that two of the four most complex parts are out of the way for now.

2.1prt_Tail.jpg


2.1prt_Winglet.jpg


2.1prt_Wing.jpg
 

Walden

Member
Another week, another batch of parts cut and photographed (now I've finished off the other half of the fabulously complex four). These first two photos are examples of how I have consolidated pieces wherever possible to speed up assembly; (2nd photo) ex: the spar, the lower surface of the wing/power pod, and the inboard and lower-aft portions of the floats are all one piece. Also I should point out, unlike most FT plans which will eyeball a specific angle for a bevel, I draw tangent bevels. Ergo, I always calculate and mark the ends of the bevel, score and rip off the paper, and shave off to the line for accuracy. I still have half a dozen minor pieces to cut out. If all goes well I should be posting photos of the assembly process next weekend.

2.1prt_PP.jpg


2.1prt_Spar.jpg


2.1prt_Fuse.jpg


2.1prt_Vert.jpg
 

Walden

Member
Here is a picture of the wooden pieces. First, notice that I trimmed the rear tab off of the two control horns for the elevons, that's to mount them more perpendicular to the incoming angle of their push rods. Like the previous plane, the front leading edges get a bamboo reinforcement for improved impact resistance. However, unlike in the previous plane, I downsized whatever 5 mm skewers there were down to 4 mm to make everything simpler. I weighed all the components (foam/wooden parts, battery, prop, and electronics); combined it added to around 400 g. Compare this with a 546 g final weight on the previous version, and I think those strategic weight savings might be making some difference. We will see.

2.1prt_Wood.jpg
 

Walden

Member
I finally got to gluing this week. First step is to construct the power pod, spar, hatch, and canard, then glue them to each other. It all came together exactly as planned until I realized (due to an error in alignment and measuring out the lower wing) the spar and thus power pod was sitting ~5 mm forward of where it should relative to the canard. Initially I despaired to think I was going to have to redraw and cut most of the adjacent parts, until it became clear that the errors were just isolated within this one part (as well the two littlest foam pieces) and I could likely cut the canard and power pod off without damaging them. Still, it took most of a day to track down and correct the design errors and cut out a new revised spar/lower wing. Dimensions have been double and triple checked in the hopes I don't have I will not encounter similar surprises going forward.

2.1asmb_Hatch.jpg
2.1asmb_Canard.jpg
2.1asmb_Canard2.jpg
2.1asmb_PP_Spar.jpg
2.1prt_Spar2.jpg
 
Last edited:

Walden

Member
Some build progress today. I've got the fuselage, vert, spar/lower wing, and power-pod are assembled. I also sprayed all the electronics with CorrosionXHD this evening, with the partial exception of the rudder servo (remembering to waterproof after it was glued in...:rolleyes: oops, at least the sides are all well sealed). Because, yes, I fully intend to test out those water TOAL after it is finished. You might notice the wire dangling off the vertical stabilizer. That push rod wire will later be connected to the tips of the tail where it joins the winglets; this allowed me to remove the wood reinforcement in the T-tail LE from the previous plane, thus lowering the CG while also reducing stress on the winglet joints. Speaking of CG, I am curious to see if it will be on point because I seem to have a habit of designing the power-pods on my planes a bit snug. Anyway, ready and looking forward to gluing the wings and tail on tomorrow.

2.1asmb_Fuse,Vert.jpg
2.1asmb_Fuse,Vert2.jpg
2.1asmb_ElcrSlots.jpg
2.1asmb_LwrWing,Fuse,Vert.jpg
 

Walden

Member
This last week I glued the wings and tail on. This is when it feels like all of the corrections and measurement re-checking paid off at the moment when I can take a part as complex as these and have it line up as exactly as planned no trim fitting required :cool:. The hubris lasted until I realized that the fully built plane would have ended up too nose-heavy. This prompted me to order a lighter battery (Tattu 4s 850mAH) along with a FT Aura 5 to experiment with, meanwhile I poured over the design to see where else I could possibly shift/cut weight. The battery swap ended up being an unexpected blessing by further lowering total mass/CG height and leaving just enough room in the front of the PP for the FT Aura 5. Final projected weight is down to just under 400 g. However, there was a small complication with this scheme; my elevon servo wires entered through the aft of the PP next to the receiver and thus wouldn't reach the Aura. When I designed the width of the main body, I intentionally sized it to the maximum untapered length of 4 mm bamboo skewers and close to the unextended reach of the 9 g servo wires. In order to re-run the wires I had to reach down into the PP, cut new holes in a precise location through two layers of foam, and feed the servo connectors from the opposite corner of the main wing through these holes blind. It was quite the effort. Many of the most recent weight reducing revisions (not present on this prototype) will probably make it in to the final plans, although final tweaks to the weight distribution will have to be decided after test flights and calibration.

2.1asmb_PlusTail.jpg
2.1asmb_WingTE.jpg
2.1asmb_Centerline.jpg
2.1asmb_PPTopView.jpg
2.1asmb_TailJt.jpg
 
Last edited:

Walden

Member
I love how it feels just like chucking a large paper airplane! (with no need of someone to catch it) :D This is also where that impact resistance design work comes in handy. All forward leading edges are bamboo and tape reinforced, and all winglet joints have been made sturdier. Pretty neutral if maybe slightly nose-heavy without the prop on. Once I have the push/pull rods installed, I will take some side by side pictures with the previous plane, so y'all can really appreciate the differences. 395 g.

2.1comp.jpg
 
Last edited:

Walden

Member
As promised, here are the side by side comparison shots between version 1 and 2, otherwise it is hard to notice all the changes I've mentioned up until now. Also I did a float test ... yep it floats! However, I noticed it clearly needs a good spray coat of water proofing to prevent the paper from getting as soaked. Along with the other changes mentioned, the next revision will probably have those elevon servos be more streamlined/depressed into the wing. Next comes the tedium of of setting up the transmitter and Aura 5. In the Grasshopper V-1 thread I recounted the challenge I had mixing for elevons AND rudder outputs. One design quirk is that on all Lippisch derived ground effect craft is that those flaps on the winglets may look a lot like ailerons but function one at a time more like yaw-controlly-speed-brake-flaps than as roll control. Instead of this control surface setup, I chose to have my flaps mounted on the inward inclined portion of the winglets to function more like an A-tail which lets me have a similar effect while simply utilizing R/L elevon outputs and reduced down elevator travel. When tested on the Mayfly the A-tail let me fly flat turns without much roll. I'm hoping for similar behavior on this craft. It will be interesting to see what effect the Aura 5 has on the final flight characteristics.

2.0-2.1front.jpg
2.0-2.1side.jpg
2.0-2.1top.jpg
2.0-2.1bottom.jpg
2.1floatTest.jpg
 

Walden

Member
This week I have been setting up a new transmitter profile, mounting the FT Aura in the plane, installing the config tool software/drivers, reading the manuals/FAQs, yet it wasn't until I connected everything up and bound the transmitter to the receiver that it became clear: the Spekrum AR620 is not compatible with the FT Aura (well that and after asking for help in the forums, thank you LitterBug!). So, I ordered the Spek RC DSMX 3.3V Serial reciever (the same model Josh paired with the FT Aura in the Nano Goblin video). This sudden change will have a few positive and negative effects:
1. Progress on this plane is frozen until that new receiver arrives.
2. The new receiver can fit bellow the battery in the keel freeing up space to cut through the central spar* and slide the battery back further in the power pod which also moves the FT Aura closer to the CG.
3. By having this extra room in the PP and positioning the battery on top of the CG, it makes the design compatible with almost any size 3s/4s battery.
4. Increasing the PP length would overcome the biggest headache some person might have in creating a 250g, 65% scale version in the future ;).
5. I have a lot more parts to redesign in CAD before the plans are ready for release :cautious:.

*I know, on other planes that would sound like a bad idea, yet sure enough it had no noticeable effect on the stiffness of this airframe.
 
Last edited:

Walden

Member
It has been a while since I posted an update. A couple of weeks ago, while the correct my receiver extension cord was still in the mail, I tested the plane out at the local baseball fields. It had recently rained an one of the fields had flooded, so I had a chance to test how the plane would perform dirt, grass, and water takeoffs. The results left somethings to be desired. Initial tests indoors were promising, however when taking off on grass the added friction was enough to tilt forward on to the skids, in spite of the 8 degrees of inclination of the propeller. On water the plane would taxi great, but then as you would accelerate for takeoff the plane would again tip forward and the prop would catch the water and slow down. Conventional wisdom insists the step on the floats should be inline to just behind the CG, then again very little about this craft is conventional.

On the infield the plane would effortlessly slide around on low throttle (I took the time to tweak the throttle curve to widen this sweet spot). One nagging issue that kept happening as the plane went faster is the left plane would consistently veer left. At first I assumed it was just a trim issue, although after trying to correct for this, I think it was more likely result from the torque of the prop causing the left skid to drag more. :unsure: Then again most ground effect sleds I've seen are single prop don't seem to have that problem. Also, it is probable the elevons are to close to the CG to bring the nose off the ground, nevertheless I did find another way to get fully airborne. :sneaky: Part of the transition from the dirt to the grass made a natural ski jump that would launch the air. The jumps being sudden and unpredictable and my flying skills being what they are I would end up crashing the plane into the ground only seconds later :rolleyes:. I could have practiced a few hand launches, but at this point my batteries were staring to run low, both cross bracing wires had detached, and I need to replace the bent prop.

To quote Millennium 7*, based on initial design criteria the final result could best be described as a "successful failure". A few notable lessons:

1. If I want to make a plane actually fly, I'll need to build a trainer and take the time to practice flying. No shortcuts.
2. She can really take a hit! Besides the bent prop and the cross brace wires detaching from the tail I mentioned, the rest of the plane took half a dozen crashes without any noticeable damage (no new creases, wrinkles, tares, glue seams that cracked open). Why post new pictures because it looks the exact same?
3. Propeller damage is still the Achilles heel. Future designs will be ducted props/BDFs. All made easier by...
4. Twin mid-mounted props are indeed the way to go.
5. It is difficult to design a plane to takeoff on grass AND water. I might avoid the added complication of making every design amphibious until I can make them fly better.
6. Designing a sled that skims along in cord-dominated ground effect is easy; making one fly stable in span-dominated ground effect (0.2-1.0 wingspan) is more of a challenge. (At least I am in good company, ProjectAir released a video of his latest GEV the same week I tested my plane to similar results)

Without sticking to an established design or doing the complex calculations that require would wind tunnel testing, I am kind of throwing darts at a dart board. Going forward I will need to conduct more careful research and glider tests. The next design I have in mind is an origami-style build that will look similar to the Mayfly with twin mid-mounted A-motors either paired to HQ3x5x3 or HQ4x4.3x3 (testing needed).
 
Last edited: