FLYING R/C Lunar Module Quadcopter project

[georgegassaway]

"M" powered Lunar Module!

Fun fact I realized about the Lunar Module Quad. Its propulsion system can deliver at least 8000 N-sec of total impulse, an M motor! If the hover thrust and flight duration (up to 10 minutes on 3000 mAh) are converted into newton-seconds, it is at least 8000 N-sec, a full “M” is 10200 N-s. While for horizontal flight, it uses more thrust than during hover, so even for a 10 minute flight it had done a lot more than just hover, it moved around horizontally so that would be more than 8,000. Or if I upgraded it to 4000 mAh to increase flight time about 20-25%. So, either or both of those factors might nudge into a “N” class motor. OK, so not the same thing as using rocket power, because of the lack of fire and smoke. But otherwise, incredible efficiency of thrust to weight of the system (hell of a lot lighter than an “M” motor), and cost per flight of "pennies", not hundreds of dollars! ). And I do not need to fill out a flight card, and can fly my “M” N-sec model in the front yard!

M14-P4-AE (M class, 14 Newton average thrust, 4 propellers, Air & Electricity "propellant".

Last time, I posted about the strut mold, side 1 was curing. So, after curing, the clay was peeled away, careful not to dislodge the master strut assembly from the RTV rubber. The negative impression as seen in the peeled away red clay, is exactly what the 2nd layer of the RTV mold will look like.

Note how the alignment holes pressed into the clay created raised pins on the first RTV layer. Bits of left over red clay were easily removed.

I shifted the 5-sided mold fence upwards, to prepare to pour the second half of the mold (no photo of this). After getting it just right, I realized there would be some leaks, so I needed to mix up a bit of RTV and brush into the areas that would leak, and let that cure. A day later, carefully brushed a special Mold Release onto all the exposed RTV, so that the second layer would not bond to the first layer. Wanted be really sure, so after a lot of time to dry, gave it a second coat and a lot more time to dry.

Mixed up more RTV, brushed RTV along all the master part surfaces and edges meeting the first RTV layer, to avoid air voids. Then poured RTV to fill up the rest of the mold.

At left below, the mold is curing. Note that I added a removable piece 1/4” tube (at a diagonal angle)that will last most of a hollow hole into the 3/8” tube portion that a leg will lift thru. After casting a part, the rest of the hole will be drilled out. Also in place are two 7/64” drill bits, I did to have any 7/64” rods to use. These will be used during casting, for 7/64” rods to cast holes in the mounting lugs, for 4-40 screws to pass thru. When a casting is done, while the resin is still "green", the temporary insert rods and tube will be removed, then the mold left alone for a full hard cure before the mold is peeled apart and the cast assembly removed.

On the right, the mold with the wooden 5-sided mold fence removed, and 1/4” tube and rods removed. I started to peel the two parts apart along the edge a bit and the mold release worked well. However, the 2nd layer of RTV felt like it needed a bit more time to cure so I left it intact for another day. So, no cast leg struts yet.

While waiting for curing I’d been doing some related work, partly by flying my 250 Quad for practice and trying things out. My 250 Quad had not been getting a good GPS lock. I swapped to a different GPS module and it got GPS lock very well, so the old GPS module is bad. First time I've had a bad one, so that adds to the experience base.

Alt Hold mode is a fun way to fly at low altitude, but the barometer does vary a bit and that can be trickier close to the ground. So, I looked into using “Sonar” (ultrasonic sensor).
https://ardupilot.org/copter/docs/sonar.html

I ordered one and it arrived Friday. Its realistic altitude detection range will be about 12-16 feet (4-5 meters). When too high up or poor signal, the F.C. will default back to the barometer.

So, here it is after adding a 100 uf capacitor and wiring in a shielded cable.

It was not working with the F.C at first, but I did use a voltmeter to confirm it was working, the analog voltage got higher when the distance increased, and lower when I move a sheet of paper closer to it. I finally found out the problem, it was in the Mission Planner software's parameter settings, found and enabled the correct analog pin value for the controller. Once it was working, then made some tweaks in Mission Planner to calibrate it.

I’ll be testing this out THOROUGHLY on the 250 Quad, before deciding whether to add it to the Lunar Module. I need to modify the 250 Quad, so the Sonar can be mounted in the belly, where the battery is now. So I’ll have to reposition the battery as well as make up a belly mount for the Sonar which will allow it to work yet also protect it in hard landings.

Of course, the real Lunar Module had an active distance-above-surface sensor. In that case it was Radar, sending and receiving radio waves to determine distance for several miles rather than sound waves measuring in feet. It was mounted to the bottom of the Descent Stage.

(above from: https://www.braeunig.us/apollo/LM-descent.htm)

Below, a drawing of the exterior of the Landing Radar (inverted, on left, and facing down on right)

So, if I do add this, it will be somewhat realistic to have an active distance-scanning "something-AR" onboard, sensing the altitude above the surface during the landing.

 

[rharshberger]

I use a similar method for rtv mold boxes only instead of wood strips I use Lego's and seal the base with Elmers white glue which peels off easily after the Legos are removed from the base.

 

[georgegassaway]

I use a similar method for rtv mold boxes only instead of wood strips I use Lego's and seal the base with Elmers white glue which peels off easily after the Legos are removed from the base.

Great idea, I need to remember that for some future project where a square or rectangular box would work well, and often I have done some projects with those shaped boxes.

In this case, due to the shape, large size, and therefore large amount of RTV needed, the 5-sided box/fence was optimum. I barely had enough, as in I literally scraped the bottom of the container. I need to get some more, hopefully Hobby Lobby has it in stock (40% coupon).

I may very well end up casting the landing footpads too (about 2.25" OD). I usually would cut up a round tube to use as the box/fence for a round part like that, but I'll see if a square lego box might work out well enough compared to the available tube diameters (A 2.6" tube may not be large enough diameter to allow for the alignment pins outside of the part, a 3.0" tube would require so much extra volume that a square lego box might beat it, alignment pins/holes on the corners outside of the round part). Hmm, just occurred to me that with Legos, the box would not have to be square, the center could be "bumped out". Sort of like below but not bumped out nearly as wide, more like one or two rows per side, as suitable. Say bumped out about the width of the black line in the drawing.

 

[georgegassaway]

Got a couple of struts cast. Below is a photo (left) showing the bent .039” music wire used for reinforcement, bent to shape. Also a second piece to add more reinforcement to the 3/8” tube the the 1/4” leg dowel will fit thru. To its right is the RTV mold, with 1/4” steel rod (waxed) in place to cast most of the 1/4” hole for the leg. Also visible near the bottom are the ends of two 3/8” 4-40 screws that were used for creating the holes for other 4-40 screws to attach the strut to the Descent Stage. To the right, the .039” wires laid in place before beginning the casting

I mixed up Alumilite resin to cast the strut. Quickly poured the resin into both sides of the mold, and quickly placed the 2nd part of the mold on top of the base portion. I pressed down on it to squeeze out excess resin, sort of like excess batter being squeezed out of a Waffle mold.
Allowed the resin to cure for a bit, to the “green” stage. Used a screwdriver to rotate each screw abut 90 degrees, so they would not get stuck in place if the resin had cured fully. Same thing for the 1/4” rod, used pliers to rotate it a bit to unfreeze it (if this was not done in the green stage, then after a full cure the part might break when trying to rotate it to free it). Then let it cure another half hour so it could cure in the mold without warping.

Below, the rough-cast part (middle) after removal from the mold halves. There is a web of “flashing” all around that has to be trimmed away.

The original master part, and the cast part after some trimming. Holes were drilled through. The resin mix was a bit off, so the appearance was a bit mottled.

The casting still needs some more clean-up, although for test flying purposes that does not matter. Actually the whole assembly will be covered with foils of different colors, or black wrappings, so little to none of the cast part will be visible itself.

First one weighs about 15 grams. A good result, I had estimated 45 grams for each entire leg assembly, and was concerned this cast strut assembly, with wire, might run 20-25 grams.

Second one, I added some .06” graphite rod to be cast inside the “X” braces. The resin began to cure bit early, I could not squeeze out as much excess resin, so the “web” was pretty thick in spots and the cast part heavier and the round parts more oval-ish. The part is stiffer partly due to the extra thickness, and partly from the .06” graphite rods in the "X".

So far I only cast those two. I’m learning how to modify the casting process, I know why it started to cure sooner, resulting in the unwanted extra thickness. I’ll cast 2-3 more struts by Wednesday night. Do not know when I’ll try to fly the LM with the new closer-to-real legs, too windy for the next few days.

I did get in some more flying with my 250 Quad. It was not flying as well as it had a year ago, wobbling and bucking in wind, in ways it had not done before. I had done an “autotune” flight last week to tune it better, but it made it worse, it had a wobble in the roll axis (think aircraft roll axis, not rocket - “Aileron” stick axis). I got the pitch better, but it was bad in roll. Changed the roll PID values and that settled it down. In the video below, it begins with the bad roll settings, you can see it rocking back and forth a lot when in Loiter mode. Then a takeoff after changing the roll PID’s and it flew a lot better. Still not ideal, but better.


I want to note that when I mention “Loiter” mode, it is using GPS to try to hold itself over the same spot on the ground. So the roll axis wobbling seen in Loiter mode was the model overcorrecting to try to stay in the intended location. When I mention “Alt Hold” mold, then no GPS is involved, it will drift freely with the wind, if I’m not actively flying it into the wind at the same speed as the wind.

This low altitude flying is not how I plan to fly the Lunar Module. It’s more useful though for me to evaluate what I’m doing with tweaks to the 250 quad. And simply better overall practice flying for me since it is more challenging to fly, and better for practicing the landings.

[video=youtube;JLiVR7l_yH0]https://www.youtube.com/watch?v=JLiVR7l_yH0[/video]


Ah, what the heck. A “Not Ready for Prime Time” photo below, showing the two cast struts temporarily in place. The fit and positioning turned out really good! And that strut assembly really helps make it look more real. The structural strength/stiffness of the leg assembly seems very good (Much later I will add the inverted “V” struts that run from the base of the Descent Stage, upwards to the top of the legs, but those will be cosmetic, not load-bearing). In-between waiting for full solid stiff curing while casting more struts, I need to do some structural and cosmetic repairs to the Descent stage, have not fixed the damage from that throttle-down crash 2 weeks ago.

 

[burkefj]

George, the legs really add something, that's starting to take shape nicely.

Frank

 

[KenECoyote]

George, the legs really add something, that's starting to take shape nicely.

+1 That's seriously impressive work!

 

[rstaff3]

Quite impressive, thanks for sharing!

 

[georgegassaway]

Thanks for the comments.

Had a VERY GOOD flying session with the Lunar Module Friday. First time it flew in 2.5 weeks, after the “throttle down” crash that broke the crude landing gear and a bit of damage to the Descent Module. I put off doing any more flying until getting the cast struts made so I could try them out, and in the meantime I practiced more with my 250 Quad.

Here’s a close-up bottom view of one of the cast struts attached to the model.

So, the outline sure looks more realistic with those struts.

Also made up a temporary base plate, with a plastic cup to represent the Descent Engine. The model will get better ones later, the base plate will be vac-formed and the nozzle probably will be vac-formed after making up proper balsa pattern to form over.

The image below shows the model in it’s “prep box”. Cut 4 notches into the box, for the motor arms to fit into, to hold the model centered and supported by the arms only. This is for prepping, installing/removing the battery packs, and securing the base plate. The final version of the Ascent Stage would not be able support the weight of the model inverted on the ground, and also a lot of parts on the Ascent Stage would get broken or smushed. At some point for car transport I'm going to need to get or make up a protective box to put the whole model inside of, and the arms will also support the model in that box.

And some other views of the model.

When installing the legs, there was more friction than I wanted, so I sanded a few of the 1/4” dowels. This turned out to be a flight issue, a couple of legs rotated from the vibrations, the one seen on the left (LM’s right leg) would rotate so much the landing pad looked skewed. It even FELL OFF once, which led to an interesting landing. After a few landings, I just applied an external strip of tape to keep it fixed. So I need to add a bit of material back or do something else so the dowels won’t rotate or slide out so easily. Yet I do not want to tack-glue, as I want it to be easy to replace a broken leg with a spare.

The flying went SO WELL, there were no hard landings. Between the use of two battery packs, it made thirteen landings…… all under good control. "ALT HOLD" and "Loiter" make this model fly so much better (Now as a racing Quad , or Acro model it would suck.... but that ain't what it is).

I used a couple of cameras. My Canon Digital camera sitting on a box, just to get fixed views. And GoPro on my bike helmet to get video following the model. But the GoPro had a problem for the first session so I went with just the Canon. First flight was good, but due to a loose leg I landed it soon after (Video was blurry). But I really wanted to get some good close-up airborne video of the model. So after seeing the model was flying so well on that first short test, “loiter” mode working well, I took off, put it into hover, loitering at about 8 feet up. I put the transmitter down, grabbed the Canon, and got some close-up video of the model in the air. One of the legs had rotated, I actually tried to grab that leg to rotate it back straight but when I did try that for a moment it didn’t want to rotate easily enough so I gave up on that (I know it may seem nuts to grab onto a multicopter, but when working properly it wants to keep itself level regardless of any outside force....even my hand. Some people who fly Phantoms in limited spaces, such as on a boat, or with terrible ground conditions like big rocks, grab them by the landing gear and throttle down ). Got hold of the transmitter and resumed flying it myself.

For the second session, the GoPro was working. So I got fixed video and GoPro video following the model. I have made a video, I edited a lot of stuff out but still it’s pretty long. But hey, 2.5 weeks of no flying, great successes, good landings, I wanted to keep in a lot of the flying (I did edit out when the model went out of view of the fixed camera for more than a couple of seconds) . This video will probably be the “definitive” video for a long while, since the weather is not going to be very good. And I need to move on to getting other parts made up…. like landing pads and start on a proper 3D Ascent Stage using scaled-up cardboard model patterns printed onto poster paper stock.

Well, it is going to take a LONG time to upload the video to Youtube. So in the meantime, I’m merging two clips together for a short video for now, will post the good one Saturday (it may take overnight to Upload.)

[video=youtube;sVBJNi0eoGM]https://www.youtube.com/watch?edit=vd&v=sVBJNi0eoGM[/video]

 

[dhbarr]

Matte grey finish for arms / props, to disappear?

 

[georgegassaway]

Matte grey finish for arms / props, to disappear?

Considered it. But the props are gray and appear anyway in video (no good enough quality clear props). So not sure if painting the arms gray would do enough. Motors could not be painted, just too problematic to mask everything, any paint inside the motor could be a big problem.

So, the main video got uploaded, here it is:

[video=youtube;a4lWdjO04wE]https://www.youtube.com/watch?v=a4lWdjO04wE[/video]


Also, this afternoon I did one flight for the purpose of getting some still photographs of the model in the air. Took off, put it into a good hover in Loiter mode, and let it hover in place while I shot photos. Wished it was sunny. But since it was not, and I could see the first couple of pics the base was dark, I used the flash to get some better shots. It'll be great to get some close-up stills like these once it's completed and looking good.

 

[dhbarr]

That is so cool, every time I see it.

 

[georgegassaway]

Some updates....

Have made a horrible miscalculation with the Ascent Stage of my 1/16 R/C Lunar Module Quadcopter!

OK, just kidding. Decided to get in a little practice of how to assemble the cardboard model patterns that are in 1/48 scale. So the partially built 1/48 Ascent stage is sitting on top of the GPS module "tower" that keeps the GPS module and its compass above the other electronics. Did some computer work with the 1/48 cardboard pattern files to copy and paste 300% scaled up patterns for Ascent Stage parts onto 11 x 17 size paper ("poster" stock, though not as thick as poster paper). Got them printed at OfficeMax. A few are visible in the background of this photo.

One of the planned extra features of the model is for the engine nozzle to be lit up like it was actually firing. I have some very bright LED modules about 1" in diameter, with individual LED elements. I mounted four deep inside the mockup for the engine nozzle (Plastic cup), and added a frosted disc. It lit up well indoors.

Tried it Friday in bright sunlight. Nice overall photo of the model, but the engine lighting was pretty wimpy.

So, I removed the frosted disc. You can see the four LED modules, and the rows of LED elements. So, the brighness works but doesn't look very good.

I have some other things I'll try for that. One is to try to find a clear or white reflector that is at least 3 preferably 3.5" in diameter (or if rectangular I could cut into a 3.5" circle). Also, to get a circular LED assembly. It would still have issues with being able to see the individual LED elements, but all equally spaced and a bit easier to work with since those could be mounted near the bottom of the engine nozzle, and use a clear reflector or some kind of dispersion lens (I got a flat Fresnel magnifying sheet, but it just makes the elements look bigger).

I had hoped to get some video and photos with the moon, as today was partly sunny and the next few days will be overcast, then the moon will rise after dark and the opportunity will have to wait another month. When I shot the above photos, the sun was out for awhile, but overcast to the East at about the time the moon was due to rise. Just didn't seem like the sky would clear so I did some other stuff. Then 10 minutes before sunset I realized the moon was out, mostly. So I quickly got the model ready to fly again.

GoPro camera had worked earlier in the day but quit while I was trying this. I used my Canon digital camera in video mode, hand held it early to show was was about to happen. Then while using the neckstrap, let go of the camera, leaving it running but not pointed as I got the transmitter and took off. Once the model was in loiter at about 30 feet up (GPS keeping it from drifting away altimeter holding altitude), I put the Tx down and used the camera to get some video of the model with the moon.

[video=youtube;h98aR0b4ZJE]https://www.youtube.com/watch?v=h98aR0b4ZJE[/video]

Then I stopped the video and took some photographs. Due to how dark it was, I had to use flash, so the sky looks black but it was not quite that dark (As seen in the screenshot from video below)

Could not get a shot that had both the LM and the moon in focus. The model needed to be farther away so I could use infinity focus and zoom in tighter. But it was windier than I was comfortable in letting the model fly that much higher up and much farther away horizontally.

BTW - I had replaced the flashing LED that is wired in series to make a super bright LED flash to simulate the Rendezvous strobe light. But after awhile it stopped flashing and stayed on. I think the voltage eventually did it, made the flashing LED either die "on", or killed the flashing LED but its shorted and giving power to the super bright LED (rated 12V). So I won't fix it again. The flashing LED was a temporary measure anyway as I plan to use an Arduino for some features, like flashing the strobe more accurately than the flashing LED. And, the final version of the engine lighting will also be Arduino controlled, so the nozzle is only lit up when the electric motors are set above idle (could "Y" into the throttle channel to use a Speed control for that, but there's a better reason to do it with the Arduino).

Next build stuff will be the Ascent Stage using mostly the Cardboard patterns that I got printed at 300%. I will probably make a prototype Ascent Stage first to learn from. Learn how to assemble, how strong or weak it is, and so forth so I can find out what works and what needs to be done differently. So I can make a better one after learning from that one.

 

[TangoJuliet]

George, check in the R/C jet world. There are some lighting manufacturers that sell LED ring lights to simulate afterburner thrust.

 

[georgegassaway]

Thanks for the idea. I have seen those before, and they look pretty good for jet models but won’t light up the way I’m looking for on this. Those LED strips are nice but for the effect I’d like to get, for the entire nozzle to seem lit up, I’d have to line the whole nozzle inside surface with them. But they would not fit together well enough due to the shape, not like trying to say apply a bunch of them lengthwise inside of a straight tube.

I finally checked my GoPro video from earlier Friday. Here’s some footage showing the visibility of the engine LED lights when it’s at a proper angle to see them up inside. In the video, at about 40 seconds, I put it into Loiter mode (no drift, GPS keeping it in place) and used my digital camera to take some photos. Did the same sort of thing for the night pics too. You can see the GoPro view change to the ground, as I got ready to, and then did, lay the transmitter onto the ground before using the digital camera for pics.

[video=youtube;KPTwRE7xRq4]https://www.youtube.com/watch?v=KPTwRE7xRq4[/video]


I want to make it more visible from a wide angle, so I may get a circular LED array like the one below to mount near the bottom of the engine.

Still will need to get some sort of lens, clear reflector, or something to scatter the light so the individual LED elements will not be visible. Here is an example of a clear reflector that might do it (has a faint watermark on the image).

For my bicycle turn signal lights, I used yellow or red reflectors, cut to fit, with the same LED modules behind them and those scatter the light pretty well while being pretty bright (do not kill a lot of the light intensity as the frosted translucent disk did). So I think that a clear reflector may work, more a matter of finding the right size for some reasonable cost.

 

[BEC]

George, this is an amazing project. I'm just about speechless.

As you may know, full scale airliners use radar altimeters for precision close to the ground rather than barometric altitude, just as you're planning to use with the ultrasonic range finder.

 

[georgegassaway]

I gave the model a partial Facelift. Temporarily added an incomplete forward cabin section, and also an incomplete aft assembly, to the dummy Ascent stage. Now, the dummy Ascent stage is about the correct thickness for the middle section, so with these parts stuck together they are close to representing the fore-aft locations of the parts.

I learned a lot in assembling the forward cabin. The fit is not quite right, I need to tweak the original cardboard model file and also be sure I get a couple of folds done better. But as I said earlier, I figured the first one would not be the final one since there would be things to learn.

When I do make the central section, I will make some of it using foam board, such as used for the dummy Ascent stage. But the structure will be a bit different, and it will be “skinned” with the cardboard print.

Changing gears a bit.... by 1971, Estes had a promotion called “Follow Apollo”.

Fly a model rocket on the same day as an Apollo launch, fill out a small form and mail it in to get a “Space Certificate”.

I might have flown a model the same day as an Apollo launch, but I do not think ever sent in anything.

I mention that, because today, 44 years later, I followed Apollo 17. Well, Apollo-17 launched on December 7th, 1972. But today was the 44th anniversary of the LANDING, at 2:55 PM EST. I was flying the model today at 2:55 PM EST, first landing a bit before , last landing some time after.

So, here is how the model looked in the air with the partial facelift:

And……

Here is how it looked landing….

kicking up “dust”

Here’s a video. Good thing that the moon’s dust was not as deep as some feared…..

[video=youtube;YqrBkXjkWgY]https://www.youtube.com/watch?v=YqrBkXjkWgY[/video]

Also, a video from some flying on Saturday. I got it higher up than I’ve shown in previous videos. The engine nozzle light really stood out.

[video=youtube;5TvF7jvLycU]https://www.youtube.com/watch?v=5TvF7jvLycU[/video]

 

[burkefj]

The engine light against the hardpacked snow was nice, also the slow blowing as it landed was nice as well. I was thinking it would be fun to be able to remotely detatch the ascent stage and leave the descent stage behind I don't remember where your flight battery is kept....


Frank

 

[georgegassaway]

So, the model needs some more accurate landing pads. The 1/32” plywood discs have been OK but do not look right. I found some DIY clear plastic buttons listed as 2.25” in diameter. Well, the smooth front part was 2.5” across… they meant that the size of a printed button insert can be 2.25”. Fortunately the other side is about 2.25”, and has a decent curve near the edge. But it has a rectangular depression for the pin to go through. So, I filled that in and sanded it.

Used that modified part for creating an RTV mold. Used a piece of 3” tube coupler as the outer fence to contain the RTV when it was poured. After the mold was created, used it to cast a solid copy (right).

Mounted the solid copy on top of a short piece of 2” tubing, and vacuum formed it. After forming, trimmed the excess plastic away to finally get the landing pad shaped part.

When I formed the part, I used .030” plastic, as I did not have the .04” that I wanted to use. I figured that a thickness of .030" under the leg might be a little weak. So I cut a 2” disc out of .015” plastic, and glued that inside, to make the pad .045” thick except for the outer edges.. Applied some Elmer’s glue along the edges of the 2" disc to help fill in the gap a bit. Although when the model is complete, the pad will be covered by foil anyway.

Cut and trimmed some 5/16” and 3/8” plastic tubing to make up the actual socket mount for the landing leg (1/4” dowel) to be glued into. I used one of the the model’s legs to mark off the proper angle to cut the part. Then when it was time to glue the socket, the model’s leg was used as a jig to assure the socket and pad would be at the correct angle.

Forgot to take a pic of embedding the landing pad into some clay and the installation of the mold box and keying holes into the clay.

I use a similar method for rtv mold boxes only instead of wood strips I use Lego's and seal the base with Elmers white glue which peels off easily after the Legos are removed from the base.

I had heard of that long ago but never tried it (did not have any Legos around). I temporarily “borrowed” some Legos and used that great tip to make up a mold box to put in place over the clay. It works out GREAT! After the first pour and letting it cure overnight, I realized I should have more “depth” for the hole that holds the 1/4” rod that creates a 1/4” hole during casting. The Lego mold box allowed me to do something I could not have done any other way (well, not as easily). Using the thin Lego pieces, I built up a sort of “roof” over most of the RTV, with an opening, and added a small Lego wall to create an extension block of RTV for the 1/4” rod to be held more securely during casting. The good thing about RTV is that it bonds to itself, so after curing it is as though it was cast in one piece. As seen at right after removing the “roof” pieces, there was a little bit of seepage underneath the Lego “roof” bricks, but not too bad. I would have had to use a LOT more RTV if that method had not been used for adding the extension.

So, after that was cured, the clay was peeled away. You can see the keying posts in the RTV part that were caused by the holes in the clay.

I added more Lego pieces to make the mold box taller and slid it down. Brushed on mold release twice to all of the RTV surfaces, and poured RTV to cast the second part of the mold. After curing, this is what it looked like before removing the Lego mold box.

And finally, here’s the mold - the two halves and the empty Lego mold box

When casting a landing pad, I first laid in some pieces of 1.4 ounce fiberglass cloth to reinforce the center area of the landing pad. I’ve done that sort of thing before to reinforce some other cast parts, so I think that will make the pads strong enough to not break easily. As it is, ever since the dumb throttle-down crash, none of the landing pads have been damaged at all (The flight mode allowing that mistake no longer is enabled, all the flights modes use a variation of Alt Hold now).

The photo below left shows the mold with the first landing pad cast in it. Excess resin running down the sides, sort of like overfilling a waffle iron (I apply resin to both halves face-up, then quickly flip one on top of the other and press down to squeeze out the excess.). Also visible is the 1/4” rod that casts the hole for a 1/4” leg dowel, which is rotated to loosen it a few minutes after casting, while the resin is still “green”. The 1/4" rod has some blue and red heat-shrink, the blue is a stop ring to assure the rod is set to the correct depth (I used a razor blade to trim the top of the extension block of the RTV mold at an angle 90 degrees to the rod, for the stop ring to work best)

Well, unfortunately, I had some trouble removing the first casting due to a bit of a flaw with the inside part of the mold. In prying the edges loose, the rim cracked in one spot. Aside from the mold issue, the cast part’s outer rim is marginally thin (one reason it cracked easily). If I had vac-formed .040”, maybe it would have been thick enough there. Given the mold issue and the rim being so easy to damage, i think I’m going to have to vac-form a new pad from .06” plastic (I think .05 would be ideal but there’s nothing between .04 and .06). I had even attempted to vac-form a .06” part by stacking two .030” sheets together and forming the sandwich. But when it formed there were some trapped air pockets that caused some bulges, so I just went with the .030” vac-formed part.

Close-up below shows a cast pad right after removing from the mold, before removing the flash and doing some clean-up. And on the right, what the pad looks like on the leg.

I think what I’ll do for now is cast a few more of these marginal parts so I can add them to the LM for any upcoming flying, and maybe find any issues that are not already obvious. Then later vac-form a .06” piece (need to get some) , add another socket assembly, and cast a new RTV mold. I can re-use some of the new RTV mold halves in some ways to avoid using as much RTV as was needed to make the mold. So, it’s disappointing but not a disaster. I’ll put off doing that for awhile, while I move on to some other parts and assemblies. And when I do that I won’t document the revised mold, the process will be pretty much the same as above, just tweaked.

So yeah, all the above steps to be able to make an RTV mold to cast the landing pads.

UPDATE: Cast more landing pads and added them to the model.

I'll end with this one. Not counting my avatar, this is the first real animated GIF I've made (using GIFphy). It's the last landing that I made on Sunday, the 44th anniversary of Apollo 17's landing.

 

[georgegassaway]

Well, been a long time since I posted any update. Because till now there’s been nothing to update. Recovering from breaking my arm: https://www.rocketryforum.com/showthread.php?138074

Then when I was able to build…. I just didn’t have my “mojo” to really get back to building. But i’m getting back to it.

Have not flown it since December either. Missed the anniversary of the Apollo14 landing, wanted to, weather was just too rotten. And was going to fly it at a club rocket launch last Sunday but it was just too windy (12-15 mph) to risk flying it.

My recent efforts have involved working out issues with the R/C gear. I had always planned to go to a better receiver, to use a Spektrum AR9020 receiver. When it finally occurred to me that the foil and aluminized mylar would mess around with RF reception, …ohh….crap! I needed to work out ways to solve that. The 9020 comes with two “satellite” receivers, which attach to the primary receiver via a cable. So if the main receiver and satellite #1 does not get a good signal, but the remaining satellite #2 receiver gets a good signal, then the main receiver will use Satellite #2’s signal for controlling the model.

It turns out the 9020 can use three satellite receivers, so I could add a third.

Now to shift gears a bit, though the story will converge. A lot of the medium and smaller “drones”, are using satellite type of receivers, not conventional ones. This is because more and more Flight Controllers, especially smaller ones, are using S-bus, or serial signals from one data line coming from a receiver. Same kind of data line and signal format as Satellite receivers use.

Reason for this is that for a Flight Controller to be able to use a regular RC receiver, it needs to have at least 6 pins to connect to the control signal wires coming from 6 channels of a receiver. So many of the smaller FC’s are moving away from that, to just use a serial bus instead to save precious space along the edges of the board. Also another reason is that Satellite receivers are smaller and lighter than regular receivers, really important to those who are doing Drone Racing or making very small light sport models. And frankly less wires to screw around with and no chance of getting control channel signal wires mixed up.

There is another Quad project I want to do (almost as neat as this, but a lot simpler), but now it will be delayed to fall. That project will need to use one of the newer FC’s in order to use the most recent version of Arducopter software for a special flight mode. And that newer FC only accepts serial signals, won’t work with regular receivers that have individual control channel outputs.

So for that model project, I got a Spektrum 4648 receiver, Satellite style which they make specifically for use by multicopters. A big plus that it has over the regular satellite receivers is that the two antennas the 4648 uses have long “coax” extensions, so the actual antennas (the tips with 31mm of exposed wire) can be located several inches apart and easier to orient differently to maximize the chances of one getting better reception, as the 4648 automatically chooses whichever antenna has the best signal at the moment (Diversity). I think the regular satellites do that too (unless they are dipole), but their antennas are very short and not easy to bend to different orientations without risking metal fatigue.

https://www.spektrumrc.com/Products/Default.aspx?ProdID=SPM4648

So, back to the Lunar Module model. The 4648 would be better to use as the satellites than the regular satellites, if only due to the longer antennas. So to test, I plugged it in as the 3rd satellite, and it works. So, I plan to replace the two regular Satellites with two more 4648’s. Which will give me SEVEN antennas in separate locations, to work around avoiding foil and aluminized mylar interference as much as possible. Only need a good signal received by any one of those seven at any moment.

But….. one very risky complication. The safest way to prepare to fly R/C, is turn the transmitter on first, then the model. When flying models like this, I usually do that…. but not always. Sometimes I turn the model on first, transmitter later.

In the still-new world of 2.4G R/C systems such as used by Spectrum, there are no frequency “channels” like with old 72 mHz. They use Spread Spectrum. Every transmitter has a unique electronic ID, not unlike a Cell phone (which is why when someone calls you, only YOUR phone gets the call). To get a receiver to work with a specific transmitter, the receiver has to be electronically “bound” to it. That is usually done by plugging in a “bind plug”, into the receiver’s Bind/Data port or bind/battery port. The binding plug is just a regular servo type connector that is shorted from ground to signal. So to bind the receiver, plug in the bind plug, power the receiver, which flashes when in bind mode. Then press the bind button on the transmitter and turn the Transmitter on at the same time. The Transmitter will detect the receiver being in bind mode and the two will bind (the Tx also receives, and the receiver also transmits!). After that, the receiver will work with that transmitter and only that transmitter. Also, this is true for specific model memory as well. So if the transmitter is still left set for flying a Radian model, and I try flying the Lunar Module, nothing will happen! Like “calling the wrong number”. As contrasted with old tech 72 mHz type transmitters where if you fly with the wrong model memory, there’s likely be a disaster (such as reversed elevator, or reversed ailerons, or a mixer that is not mixing, or mixer that is mixing but ought not to, or terribly messed up trims).

Here is the complication with the 4648, due to the unique way I’m using it as a satellite. It has no bind plug. So it’s method of binding is that if on power-up it does not detect a transmitter signal that it is ALREADY bound to, within about 15 seconds then the 4648 goes into “Bind Mode” on its own, flashing an LED that is on solid when it does get a valid signal from the TX it is bound to. Which is very convenient….. when used as the one and only receiver in a drone.

But when used as a satellite, and I plan for all three Satellites to be 4648’s, if I turn the model on first and not the transmitter within about 15 seconds, the 4648’s will go into bind mode and not recognize the transmitter once it does come on. And with the receivers being hidden inside, I’d have NO IDEA that had happened.

So on takeoff, rather than four receivers with seven antennas working, I would only have the one 9020 receiver and its single antenna working.

I puzzled over this. Looked for some advice on RC Groups. At one point I was thinking of just leaving the bind plug plugged into the 9020 all the time, and just re-bind it to the TX every time I powered it up. If I forgot to…..nothing would happen as all receivers would be in bind mode. Then I’d realize why nothing happened, and start over. But I was warned that was not a good idea, to have the model flying around with the bind plug still in. And yes, there was a bit of a risk in that, so I dropped that idea.

So, what I figured would do it would be a “Lost Signal Alarm”. Lots of those sold for R/C use, also help to find lost models if you get close enough to hear it. Thing is, every one of them that I found for sale, had a delay of 30 to 60 seconds before they would warn about losing the transmitter signal. That was not good at all. Because if I turned on the transmitter before the lost signal alarm warned me, but after the 4648’s went into bind mode, I’d have no warning that 3 of 4 receivers and 6 of 7 antennas were not working properly for flight.

I could not find any alarms that acted fast, other than a special homemade design that I would have to burn (transfer programming into) a PIC chip and build from scratch. Heck, I’d be better off doing a small custom Arduino than that. And I do plan to add a small Arduino later for some neat things, but due to this delay I need to put that “nice to have but not needed right now” stuff off on hold for next fall/winter and try to get the LM completed visually.

Then a guy named FrankS suggested using an electronic switch to turn on a piezo beeper. Well, I have a very old (1980’s vintage) Ace R/C Electronic Switch someone gave me long ago. So I wired it up with the proper servo connector and piezo beeper to the output, and plugged it into a channel the Lunar Module does to use for flight control. Now, that alone would not accomplish anything since without a Transmitter signal to turn the switch (and beeper) on, nothing wold happen when the receiver was powered up but the Tx was not. Now, if it was an electronic switch that was Double Pole, I could just wire the beeper to the Normally Closed switch and use the transmitter to open that NC switch so the beeper would be silent when the Tx was on and Rx working correctly. But I didn't have a Double Pole type electronic switch, didn't want to wait to get one, and didn't want to mess with a Double Pole relay to get the Ace switch to do it. Because there was another way.

A lot of receivers, including the 9020, have “Fail Safe” capability. It is possible to set the receiver during binding to default each channel to any given position, *IF* the receiver does not get a signal from the transmitter. So, if the R/C connection FAILS, the receiver can go to a "Safe" mode to try to save the model, or reduce the severity of a crash. Say a powered plane, it not only sets the ailerons, rudder and elevator to neutral, but cuts throttle, if the signal is lost (better to glide or spiral down into somewhere than dive into the ground at full throttle. My first good R/C plane that I learned to fly with suffered that fate, all was destroyed other than the .049 engine).

So, I set the receiver's fail safe to cause the switch and beeper to sound if there is no transmitter signal.

And…. it works exactly like I need to! So if I hear that warning beep, I’ll realize I need to start over, turn model off, transmitter on, then turn the model on. Here’s a video demonstrating:

[video=youtube;7w_1AYIaPM8]https://www.youtube.com/watch?v=7w_1AYIaPM8[/video]

So, this has been a “short and simple” explanation!

In the photo below, Electronic switch near lower left. 4648 at the bottom. Two regular satellites at right. 9020 receiver middle-left near the top, battery at upper left. All four RX’s with solid signal LED’s.

Anyway, all the above is an example of stuff that has to be figured out and tested before doing final R/C installation. I will also be doing some brute force ground range testing using foil and aluminized mylar in different ways, plus tests without anything, to find out how much foil can affect RF reception (I expect a LOT), and adhesive mylar (I expect less effect). If aluminized mylar is significantly less, I’ll try to use less foil and more mylar in it place.

 

[georgegassaway]

“Return to Flight”, of the Lunar Module Quadcopter. I’ve been working on it again, though do not have much to externally show for it yet. New Ascent Stage being built, revised R/C system as documented in previous post, and other nagging details.

I had not flown it since December, and plan to fly it at a local rocket launch Saturday as the first public flights. And the weather was nice today (Tuesday), not too windy. So I put in a flight to get back into practice of flying the LM Quad. Flying went well. I usually document the flying with a GoPro on a helmet I wear, and/or a digital camera sitting on a box. But I didn’t want to deal with that today. But I did get video by using my iPod Touch’s camera face-up to shoot video of the takeoffs and landings. And incentive to land close. I edited out the parts where it flew off-screen.

The four round LED’s inside the engine bell…. I want to replace those with a larger diameter single round LED, or something that will look better than that (BTW - onboard voltage is nearly 12V, three LiPo cells).

[video=youtube;VPBCOWzW37c]https://www.youtube.com/watch?v=VPBCOWzW37c[/video]

BTW - there is a Facebook group named "Space Hipsters". I temporarily added their logo to the front landing pad last December. So, that is why it looks blue.

 

[georgegassaway]

Finally, a lot of progress on the new Ascent Stage, so much that I'm splitting it into two posts. The Ascent stage is based on using 300% scale-ups of 1/48 cardboard model patters printed onto “poster paper” stock (not really as thick as poster paper but it’s OK). I used a spray adhesive to bond the two main pieces, the front and back of the middle structure of the Ascent Stage, to foam board. Then cut them out.

Cut out two foam board pieces to the needed width to glue the front and back pieces together with the correct thickness, and glued them all together for a very sturdy structural box-like main assembly. Also cut out the very complex faceted pieces used for the left and right sides and glued them up for the proper shapes. In the photo below, the main Ascent stage assembly is face-up (front cabin with windows were added there later)

After doing that, later I epoxied in a couple of pieces of 3/16" balsa, 1" wide, running from front to back. Each with a 3/4" 6-32 bolt already glued in, sticking downwards. Those were attached nearly flush to the base of the Ascent stage, along the left and right edges. Two holes were drilled into the top of the Descent Stage, and chamfered, for the bolts to pass though. So, the Ascent Stage is secured by using two 6/32 knurled nuts, attached to those bolts via the inside of the Descent Stage. I forgot to weigh the Ascent stage, need to remember to do some next time I have it removed. But it may only be 2 ounces, I'd be surprised if it was over 3 ounces (UPDATE - I was surprised to find it weighs about 4 ounces, 118 grams. Future work on it probably won't add more than 1/2 ounce if even close to that.

The Flight Controller uses two LED’s to indicate when it has a solid GPS lock, and also to indicate when the motors are armed. Until now I had used a generic LED/Beeper assembly sold for that purpose. But to use it had to mount it externally where it did not look right. i’m not trying to build this to be some judging-worthy scale model (this is 100% for fun), but wanted to do it more realistically. The real LM had navigation lights, used for docking (OK, so those were not turned on for landing, but if i’m going to have any externally visible LED’s at all, these are the most suitable places). Green on the right side (astronaut’s right), and red on the left. So, I mounted a green led into the scale location on the right side, for GPS lock. And a red LED on the left side, for Motor Arm status. And have a piezo beeper elsewhere. So, below is an image of the Descent Stage, wth those cardboard pieces. At left of photo is the right fairing with green LED (green color is a bit washed out) and at right of photo its the left fairing with red LED.

The brighter green and bit of blue, seen inside the Flight Controller at the base of the GPS receiver tower, are status light indicators (blue one for GPS), which I can't see with the Ascent Stage on. There is also a red LED (motor status) in the controller which was not visible in that pic.

Previously, I had already made up the forward cabin and aft assembly, which had been temporarily attached to the crude Ascent stage for a few flights in December. I removed them and put them on the new Ascent Stage.

Friday, I wanted to do a test flight again, trying some new transmitter settings. But the new Ascent Stage was not ready. So I flew it without any Ascent Stage, and also left off the base with engine bell. Sure looks funky that way. The tower on top (made of nylon) which holds the GPS receiver/compass, raises the receiver so it is less affected by the model’s electronics (self generated EMI ) than if it was say mounted very close to the Flight Controller and Speed Controllers/motor wiring. GPS lock has always been rock-steady with this model.

 

[georgegassaway]

Saturday was the first “public demonstration” of the model. I took it to a local rocket launch and took some photos before flight. Had the new Ascent Stage on it, though a few pieces still missing.

Made 5 flights. Worked out pretty well. The moon was out, not a great phase but it worked pretty well, I wasn’t even expecting that. Once the model is done and all prettied up, then I want to get a photo like that with at least a 3/4 moon.

One bad thing. I had recently changed the low voltage alarm setting to a lower voltage because far too often it goes off way too early. So, I would fly awhile longer since I knew there was still some flying time left. But than I didn't know how much. So, I wanted to adjust it down to a lower setting so when I would hear it, I’d know it’s time to land, but still with a decent time cushion. Well, on the last planned flight of the day, it was 50 feet up when the low voltage alarm went off. I figured I had a minute, and started to land, moving it closer and starting descent. But at 30 feet it started to descend faster, and while I went to full throttle it kept descending faster. It hit hard on all 4 legs, causing totally different damage to three of them. One snapped the 5/16” Plastic tube (white) that the 1/4” dowel leg slides inside of. No big deal to fix. One had the cast inner “X” brace break. So, I need to cast a new one. And one had its cast “footpad” broken, the part that the leg fits inside of punched thru the rest of the cast footpad. So, I need to cast a new one. So, in theory a 1 hour fix, probably morel like 2 hours. But simple easy stuff.

And also a good sign for how well the model can survive a “hard landing” . That’s only the second hard landing i’ve had with it (first was way harder), and the first time for it with the final version of the landing legs. And I'd much rather be fixing things like that than the primary structural support of 1/16" music wire mounted inside the Descent Stage (if any of those ripped out, it'll be a PITA to fix)

So, I’ll be adjusting the warning beeper’s voltage threshold to a higher setting, and whenever it goes off I’ll go ahead and land.

Below is a short video from the launch. In addition to the LM at the launch, I also flew a Blade 350 Quad with a camera, to try to get some video of rockets being launched. This was my first time trying that (used FPV and a video TX onboard so I could see exactly what the camera saw, to aim the camera and get into the right position). So, the video shows the rocket launches first. I only got a little bit of video of the LM flying at the launch, takeoff and near the moon. I cheated a bit by re-using a landing scene from earlier in the week.

[video]https://youtu.be/Mm5pVfXbxKo[/video]

[video=youtube;Mm5pVfXbxKo]https://www.youtube.com/watch?v=Mm5pVfXbxKo[/video]

 

[TangoJuliet]

I really love this project George, and hope to see it in person perhaps this summer.

I have a question in regards to the design, and maybe you answered this already and I missed it, but is there a reason why you didn't align the 4 arms of the motor supports with the LM landing gear? It just seems to me that it would visually help to disguise the "quad-copter".

 

[georgegassaway]

I do not think that going from an "X" layout to a "+" layout would really disguise it.

I mean, if it was "+", then a head-on view would always have the front motor, prop, and beam in front of the Ascent stage. Yes, the rear motor/blade/arm would be blocked but that's still 3 of 4 and one literally "in the face" of the model.

Disadvantages of being in "+", or in line with the legs, would be:

1 - Some thrust lost due to the 2.25" diameter landing pads and landing gear struts.

2 - Front beam would line up with the ladder and "porch" (To Be Added this month). Top of Ladder would have to be cut a bit short. And really hard for the astronauts to crawl down past it (Arduino programmed 1/16 robotic Astronauts to climb out of the model and down the ladder to be completed and added 1 April 2018). Really, more seriously, having the beam aligned with the porch/ladder would be a lot more intrusive to the realistic appearance.

3 - Experience flying "X" style quads. To fly a "+" style would mean that it would be a unique model to pilot, as far as visual cues. While this certainly is a unique model as far as appearance, it is "just another quadcopter" when it comes to basic piloting skills (in some ways, less demanding). So when I fly some other quads, especially my 250 sized home-built that has exactly the same flight controller configuration, that flying experience is directly useable as practice flying for the Lunar Module.

Now, with an extremely precision-throttleable (plus or minus 1/2 ounce of thrust) hybrid motor that could burn for at least 2 minutes, with a separate compact gimbaled nozzle, and some guidance programming guru, OK, could make it look a LOT more realistic. I look forward to seeing someone doing that.

Me, I'll stick with the quadcopter that costs about 20 cents per 8-10 minute flight (mostly battery depreciation costs to replace after 100-200 charge cycles, the ultimate "reloadable motor energy source").

OK, below is a pic after the battery got too low Saturday and it had the "hard landing", as referred to at the end of my previous post.

Lower right corner of the photo, the front leg, the upper 1" of plastic tubing snapped and flew away, leaving the upper part of the leg disconnected so it slid up on the cast "X" brace assembly. Looks the worst but easiest to fix (cut new plastic tube).

Lower left, inner part of the cast "X" brace broke, piece laying to the left. Need to cast a new one.

Opposite side, the cast landing pad's cast collar that the 1/4" leg dowel fits into, punched out thru the bottom of the pad. Need to cast a new one.

But again, not a big deal to fix. I actually flew it again later in the day after re-attaching the dislodged front leg.

 

[TangoJuliet]

I hadn't really considered the "front" of the craft in my observation and also wasn't considering the orientation of drones flying in an "X" rather than a "+". You answered my question quite clearly, thank you.

 

[georgegassaway]

So, took off all the landing gear stuff in order prepare the Descent Stage so I could apply balsa sealer and sand and fill and sand and fill to get a great smooth finish before painting....

Uh, no. The Descent Stage finish is adhesive film of one kind or another, on bare balsa. Mostly "gold" mylar. Also some orange mylar (helium balloon) where the original Kapton film looked darker. Used Microscale's "Micro Liquitape", brushed onto the mylar, to make the mylar adhere to the balsa.

Used black vinyl (CriCut from Michael's using a 40% off coupon) for the black areas. Printed and cut out the UNITED STATES lettering and US Flag, using double-sided scotch tape.

And.....

Also added them to the upper surface, except for where the base of the Ascent stage covers over them. The "nickel" color in two places near the middle is from the opposite side of the orange mylar balloon, it was not as shiny as silver mylar usually is.

I still need to add some orange mylar to the base of the Descent Stage structure. And will be making up a better removable lower hatch assembly that also has the nozzle.

Now that the mylar is applied, I'll be removing the existing 6 channel receiver and replace it with the 9020 receiver and three satellite receivers as described earlier. I wanted to wait till now i order to work out just where to place all the receivers and antennas. Especially relative to the Quad 4 area (US Flag) that has a limited area of black vinyl that the signal can pass through.

BTW - I did do a range test a few weeks ago with a receiver and a servo with a piece of 1 x 12" balsa to "wave" back and forth to see when it lost signal. Did three tests, with mylar in front of the receiver antenna, with thin foil in front of the antenna, and nothing. I was surprised to find that the mylar was only slightly better than the foil. And that the range reduction was only about 1/3 (worst-case range distance 2/3 of normal).

Lots of other things to do, but with this it's no longer "naked balsa" and looks a lot more realistic.

Will soon be moving on to the landing gear. Have a mold curing for casting a part to slide up to the top of the main leg attachment, to secure two the top ends of inverted "V" struts that have not been on this model before. Then move on to finishing the legs (some paint, mostly mylar, foil, and black vinyl). And I already have the ladder, got one 3D printed from Shapeways, as well as RCS thruster nozzles.

 

[georgegassaway]

OK…. lots of stuff.

I’ve finally added the missing inverted “V” struts. I made up an interesting master part to represent the top part of the main leg, were it is tapered and the inverted V struts attached. It called for quite a bit of casting mold trickery. A 2-piece mold, with two 3/32” steel rods and a 3/16” rod to cast the desired holes in the cast part. The rods were coated with mold release before casting. In the photo, the original master part at left, used for creating the 2-piece RTV mold. To the right of the mold, a finished cast part. And to the right of that, another cast part showing how it would be used with the main leg and inverted V struts (upside down in this view).

So, in this photo, at right, five sets (one spare) of main legs (5/16” Plastic tubing) with the cast V strut attachment at the top (The alumimum Main legs had gold-ish Kapton added facing inwards, I used some 1/2" wide gold foil tape ). And to the left, an example of the main horizontal “X” strut assembly after painting and upgrade.

The “Heat Shield” and main engine nozzle on the model finally got the attention they needed. Used a lot of the cardboard Lunar module patterns for building the structure. Also made up a balsa assembly as a “base” to mount it to, and which fits into the open hole of the Descent Stage much like an 8-sided manhole cover (yes, I know why manhole covers are round). The nozzle is the cardboard pattern, more scale than the one I had. Later I’ll change it to a curved nozzle (have to get a balsa pattern made up, then vac-form the nozzle)

I added some orange mylar to match the base of the Descent Stage, and some dull mylar to represent the nickel-colored foil that was over the heat shield assembly. A couple of things sticking out are a couple of antennas in different orientations. These not only help in general but specifically in the case of flying very high vertically, as none of the other antennas inside would have a “clean view” thru the model, the mylar on the base would be between them and the transmitter.

Also, the above photo documents that yes, the original single receiver (very reliable Spektrum AR6255 with two antennas) was replaced by a 9020 receiver and three 4648 "Quad" satellite receivers, for a total of 7 antennas.

So here’s a view of the base of the model with the new heat shield assembly, nozzle, and also with leg struts attached.

One of the motors acted up. Never a flight problem, but sometimes the motor would start to spin the wrong way, or twitch, before spinning up properly. The Flight Computer software detected that as a motor problem and disarmed the motors. Often before even throttling ip to try to take off. And a few times, did it right after takeoff, going into an automatic slow descent. And sometimes arming fine and flying fine (again the “bad” motor never affected the actual flying, it was the software disarming most of the time that was the main issue. And I didn’t want to fly with a motor that might go bad in the air). I finally ended up replacing the motor.

Here’s a photo taken after replacing the bad motor, testing in on April 13th. Because nothing ever went badly related to the number 13 for Apollo, or being launched on the 13th...... or April..... (oh, right!). I did not want to risk anything happening to the Ascent Stage, so flew it this way. The GPS receiver is at top of the assembly. If the Ascent Stage was on it, this photo would be looking at the “Front” of it, where the ladder would be. Also notable and seen well here are the inverted “V” struts (gold mylar) that run from the new cast pieces at the top of the main legs down towards the bottom of the Descent Stage.

And..... a short video during the test flying. For part of it, my camera was hanging by the neckstrap and randomly getting video, until I set the transmitter down and used the camera by hand (the model had a good GPS/Compass lock, sometimes it wanders around in a horizontal spiral when left to loiter). I also used this flight and another one later to "dial in" the correct setting for the low Voltage alarm. If I get a steady alarm, then it has at least 60 seconds of flying time left. Time enough to land safely. It was a bit windy, about 8-10 mph..

[video=youtube;dcGDagkrrJ8]https://www.youtube.com/watch?edit=vd&v=dcGDagkrrJ8[/video]

More soon…..

 

[georgegassaway]

Got out an old Dremel Moto-Lathe to turn a special part out of a dowel. It’s for the black fairings that covered over the upper “V” struts. Which are 1/16” Music Wire on this model.

Pic below shows the wooden part at bottom and 2-peice RTV mold at left. Cast several copies. Used a 1/16” Dremel wheel to cut a 1/16” slot into the castings for the 1/16” music wire to fit centered inside. Was not hard to cut the slots, Alumilite Casting Resin machines easily. Painted them Flat Black.

At upper left, the Egress Platform, or “Porch”. It was not hard to make, a rectangle of corrugated plastic, and bending some 1/16” Aluminum rod for the handrails. The Ladder and RCS thrusters are 3D printed parts. Vincent Meens is an incredible scale builder (of static space vehicles), and researched the Lunar Module exhaustively. He has created 3D files for every part of the LM, available via Shapeways. The original parts were frost-clear, this pic is after painting.

And so….. finally…… this photo showing the Ladder and Porch added, as well as the upper V strut black covers. Also, the lower portions of the landing legs were covered with gold mylar, and the landing pads covered with gold foil.

Ascent Stage forward RCS arms added, and the RCS thrusters. Also added an aluminum strut from the forward cabin’s lower left to the fuel tank structure on the left (LM’s left, to the right in the photo)

The only scale things left to add, at some point, are the antennas. And maybe one day the RCS plume deflectors that were added after Apollo-10. Though I do plan to make a better Descent engine nozzle, and do some other tweaks later (like using an Arduino to flash the forward strobe light at the correct rate, and brighter. Right now it’s just a flashing LED in series that is driving it)

And…. flying!

One of the legs rotated a bit so the landing pad is a bit crooked. i’m trying to work a balance between being a bit loose like that vs being so secure that it would be difficult to replace leg parts if any break from a hard landing/crash. Also the cardstock nozzle is a bit off, as it is taped on so I can replace it easily when I replace it with a vac-formed nozzle.

View of the old crude Ascent Stage used mostly for visual orientation, and the new Ascent Stage.

“Ready to step off the LEM now….”

I shot some video late, but the compass/GPS were acting up, “toilet bowl” effect where when it should stay at a set GPS position, it wanders around horizontally in an ever-widening spiral. So, I could not leave it in Loiter mode to take some photos and video as I would have liked to. And didn’t have the helmet GoPro available. So, I wore my Canon digital camera above my belly using the neckstrap, and blindly aimed it at the model, trying to keep the model low enough to stay in view for a lot of the time. But also at times I set the transmitter down to use the camera by hand (model stays level but drifts with the wind). And even a time or two when I had the camera in one hand and was able to tweak the throttle with the other. Not ideal conditions, but there is some pretty neat footage that I was able to get. It was in Alt Hold mode for that, so I was hand-flying it at some times, and other times I let it drift with the wind before bringing it back.

[video=youtube;TeeLxePIckU]https://www.youtube.com/watch?v=TeeLxePIckU[/video]

 

[les]

Cool!

 

[georgegassaway]

(moved to next page)