• Welcome to the new Flite Test Forum home! Please note that the migration is still undergoing and certain content, user privileges or features may be missing or limited until the transition is complete. Thank you!

The Other XB-70 Build

#24
As I was planning out the wiring, I realized that I would have to have a way to transport over 5000W of power from the batteries to the ESCs. I plan to have one main wiring run that runs the length of the fuselage, and will be drawing a maximum of 320 amps at 16.8 volts through it. I have a spool of copper single-conductor 12 gauge wire that I intend to use, and I know that I will have to use multiple runs. The total distance from the batteries to the ESCs will be around 6 feet, although depending on final battery placement this number may decrease.

Is there a generally accepted method to find the approximate cross-section or wire gauge that is needed for a specific current and voltage?

Another thing I'm considering is the electrical interference with the lengths of wire that I'm dealing with. I've seen recommendations over on RCGroups to put some beefy capacitors right by the ESCs in order to stablilize the voltage, but I've also seen posts saying that this doesn't really matter.

Thanks in advance, Pieliker96
 
#25
I'm close to the same stage on my build - setting up major wiring - and based on my experience from the C-17, I would HIGHLY recommend extending the SIGNAL wires from the ESCs to the EDFs, NOT the power wires from the batteries to the ESCs. The phase-signal wires are slightly smaller gauge, since there are 3 of them, so you can use 14 or even 16 gauge wire instead of 12 (depending on the length). Also, by extending the phase wires you don't have to mess with the power/capacitor issues on the battery side.

I am still going to have a SHORT extension on the battery side of me ESCs, but I'm talking like 6-inches or less. And for that I will be using 10 or 12 gauge wire.

For the C-17 I was running only four 50mm EDFs, for a total max possible current of 160A. Since that's not a crazy amount of current, I paralleled all the ESCs down to one XT90 battery connection, which plugged into my single 6000mAh 4S Graphene battery.

For my XB-70, however; I plan on using one 5000mAh 4S for EACH 70mm EDF (each run thru an 85A ESC) - if each EDF pulls 70A, that's 420A! The power involved here is too great to bus down to a SINGLE battery connection. So each motor/ESC pair will get their own individual battery connection for power. The Rx controls will still be Y'ed down to one or two throttle inputs, but the power will be separate.

I HIGHLY recommend you consider doing the same, for safety reasons, but I respect your build is your call. :)
 
#26
Much thanks. I still don't know where the battery placement will end up being, so I'm running around 2 feet of battery extension and then 4 feet of motor leads. This will allow me to put the batteries inside the fuselage or the bomb bay for balancing purposes, as I still haven't quite figured out where the CG should end up.
 
#27
Parts have been coming in from my 5 orders to various websites last week. Everything is here now except my order from Hobbyking's China warehouse (my gear struts, nose retract unit, and some other stuff). I've just come to expect slow budget shipping from overseas at this point. I got back from boarding school to open the fan units today.

2018-01-12 21.10.39.jpg

The reverse fan housings and blades are in the back. I'll be replacing the stock housings with these for two of the fans to eliminate any problems caused by rotational torque, as well as helping handling during engine-out scenarios. I know that EDFs generally don't have that much notable p-factor as they take a long time to spool up from a standstill, but based on my experience with the 6-foot model, I'm not taking eny chances.

2018-01-12 21.10.21.jpg

I'll set up a thrust test stand tomorrow morning and get to work finding the optimal intake and exhaust geometry for static thrust. I'm not expecting these units to produce their advertised 1500g at full throttle, but we'll see.
 
Last edited:
#28
Just finished the first thrust test of one of the fans, and holy cow do they have some power! The batteries were at a total of 16.5 volts, meaning that the rotor was spinning at over 45,000 RPM. The tests stand still slid a little despite the twelve pounds of barbells and batteries holding it down. Total thrust produced was 1.6kg, 100g more than the manufacturer stated (although these batteries were close to fully charged). The exhaust jet attempted to rip one of my unfinished airframes off of its storage spot on the wall during a runup to quarter throttle. Assuming that thrust drops 10% with ducting (some of which can be taken back with thrust tube design), I'll have a thrust-to-weight of over 1.5 and should be able to cruise at quarter throttle. I could've easily gotten by with just two EDFs, although being able to pull something like this vertical and have it climb like a fighter jet (or rocket, for that matter) should be well worth the extra cost.

 
Last edited:
#29
Another thing I'm considering is the electrical interference with the lengths of wire that I'm dealing with. I've seen recommendations over on RCGroups to put some beefy capacitors right by the ESCs in order to stablilize the voltage, but I've also seen posts saying that this doesn't really matter.Thanks in advance, Pieliker96
If you're worried about EMI, I can tell you from my days as an MSEE with NASA that we handled interference by twisting the power wires together. Called them "twisted pairs", and they usually did a fair job of keeping EMI to a minimum. If you have a large enough ground plane, attaching the negative lead of larger capacitors to that ground, with the positive lead soldered to the positive power wire will drain power spikes and spurious signals to the ground. Caps should be on the order of 5 microfarads or so, each, with one for each power wire.
 
#30
Thanks, Ray. I'll probably end up using this method instead of caps for cost savings.

I got a little less than I wanted done this weekend, but any progress is always good with a restricted schedule. The final package came in, giving me my XT90s, wheels, oleo struts, and nose retract unit. I soldered the battery connectors and then realized that I'd forgotten to put on the grippy solder joint protection bits. Oh well :rolleyes:

2018-01-21 17.51.06.jpg

2018-01-21 18.34.06.jpg

The main bogies are made out of 4mm brass rod which I wrapped with wire and soldered. The rod itself solders fine with lead-based electrical solder, the wire wrapping is just for extra rigidity. Brass also tends to bend at longer lengths, but at the length-to-thickness ratio I'm working at, it seems as stiff as steel. At one point, the strut got stuck on the retract pin and wouldn't budge vertically, although I could still rotate it without the set screws in place. It took some highly viscous oil and some forceful coercion to finally get it off, and there were no obvious signs for the binding. The sharp edges were cleaned up with a dremel at the same time that the new flats were ground in the shaft, and I haven't had a problem since. I did solder the gear with the struts attached, but I doubt that the heat differential was enough to cause it to stick and not unstick when it cooled. Either way, the problem is solved, and I won't have to deal with it any more as the struts are securely mounted in their final position. Any toe-in or out adjustments will be made by reflowing the solder at the attachment point.

2018-01-21 17.50.33.jpg

The 2.25-inch wheels from HK are incredibly light and cost next to nothing, and seem like they shold last. The cross member that attaches the oleo to the bogie is slightly offset from the center to allow the gear to maintain a positive angle relative they the ground, similar to full-scale.

2018-01-21 17.50.07.jpg

Next steps are to figure out how the mounting of the retracts inside the lower nacelle bit will go, as well as building the nacelle itself.
 
Last edited:
#31
I like your landing gear solution, very slick.

I just watched your latest update video, and heard you mention a target weight of 5kg ... I just want to caution you to double-check your estimates. I know our builds are slightly different in their approach, but we're basically the same size airframe. My basic airframe weight (without flight batteries) is over 17lbs. If I remove two EDFs (since you're running 4), and round-down to account for your lighter-weight retract units, I still don't think I can estimate it any lower than 15-16lbs, or 6.8-7.3kg. Again, that is WITHOUT the main flight batteries.

Another reason I'm concerned about your weight estimate is the fact that you're using a more frame-and-form structure, which uses more foam in its construction than my box-fold "FT-style" frame. More frames means more glue joints, which means more glue - and it is imperative that you don't underestimate the weight of glue! You've got quite a bit of the airframe done now, so I'd encourage you to step on a bathroom scale with it, subtract your weight and then you'll at least know the bare-airframe weight. You can then just add all the other component weights mathematically. This should give you a good ballpark of where your weight is going.

Good luck!
 
#32
Thanks. Due to some special circumstances (the flu claimed ~20% of our student population and the entire language department over the course of a week), school was cancelled. I intend to get close to or finish the Valkyrie in the five days I have, which would include construction of the lower nacelle and installation of the electronics.

Concerning the weight:

All components, as measured by a kitchen scale (don't have a bathroom scale):

Airframe, currently: ~1400g (see post #19, parts measured individually)
+ 6 sheets of foam+glue ~800g (estimated, foam is around 100g/sheet, less with the paper removed)
4 ESCs: 319g
4 Batteries: 1145g
Retracts, legs, bogies: 443g
4 EDFs: 744g
Servos: 112g
---------------------------------------
Total: 4963g + other stuff* (expecting at most 500g)

*wiring, retract door sequencer, receiver, etc.

So it looks like I overshot my goal by at most 500g, or 10%. Not ideal, but certainly not the end of the world, as this was a rather optimistic estimate. If I had to point towards a reason for the large discrepancy in weight between the two craft, I would look towards differences in the power system and functionality of the two craft.

Wall of Text incoming, TL;DR at bottom

The main difference in weight would be caused by Battery selection. I am using the equivalent of two of your batteries (4 x 5000 2s), while you are using six. I don't mind short flight times, especially with the stress and anxiety that flying such a large and expensive plane puts on me, and the lower weight means that I can have less support structure, which means even lower weight. The decision to use only four batteries was also influenced by budget: I already had two of the batteries and used them in a large RC truck, and so buying two more was cheaper than getting a whole new set. This does however mean that I'll be putting much more stress on the batteries, but that risk is relatively tame, as the packs are rated to 50C and my ESCs will blow well before that. This is, however, irrelevant to the point, as your estimate was 6.8+ kg WITHOUT the batteries.

Another factor is the extra features. Your Valkyrie has a functional canard, folding wingtips, and drag chute. These require extra servos, extra wiring, and four 850mah batteries (as shown in the Q&A), as well as multiple receivers. If I had to guess the total weight gain from these features, I would say around 600 to 750g, with the wing fold servos taking up a significant portion.

Your Valkyrie's retracts are also much stronger, and therefore heavier. The retract units, oleos, and tires must weigh more than mine to support the weight of the plane, because they must be made out of higher-quality materials (metal in this case), as well as being beefier in general.

As far as weight regarding construction methods is concerned, The former-skin method is not as heavy as you might think. The formers, as well as one side of the skin, are stripped of paper (where I expect that paper remains on both sides for a box-style fuse section) which significantly reduces their weight. Additionally, the total area of foam required to skin the formers is less than a box-style fuse, as it follows a smaller, curved path. This also applies to the wing and wing ribs, as they don't benefit from the extra structural rigidity in the specific axis provided by the paper. As far as glue mass goes, box fuses will have a roughly constant amount of glue per unit of length, whereas a former-skin approach will not (with the glue being concentrated around the formers). I also don't glue every part of the former to the skin: the skin is pre-formed by bending it until it holds it's shape, at which point it is secured with an amount of glue based on the curvature (parts that curve more will tend to bow back over time, and need to be more secure than relatively flat parts). As a result, the amount of glue used in the construction of both is likely not that far apart. I have yet to do any proper experimentation, but I think that two equally-sized fuselages: one box and one former and sheeting, would have around the same weight. And again, this applies to all other areas of my plane that use this: the wings and center wing section included.

Also, I don't intend to paint anything except the details (USAF Logo, tail number, etc.). This will save considerable weight on such a large airframe, and I'm willing to sacrifice longevity and a glossy shine for weight.

TL;DR: The weight is lower than one might expect because of the lack of extra features, low amount of batteries, surprisingly light construction method, and lack of paint.

Edit:

Although being very similar in size, the two craft differ in just about everything else: weight, construction method, complexity, and power, to name a few. This can be expected from two different people with unique goals and ideas of what each plane should turn out to be, and the differences in the two B-70's are a result of that.

For me, this was a large-scale testbed of the former/skin construction method which would help me learn how to effectively and efficiently use it for future builds. It was also my first venture into ducted fans, proper retract units, and large planes in general.

I'm finalizing CAD of the lower nacelle unit and plan to have it done by the start of next week, unless the cough I got from school ends up being the flu :p
 
Last edited:
#33
Laser cut and assembled the lower nacelle today. Tomorrow is electronics installation, hopefully some taxi tests of the nacelle and finishing the rest of the plane on Saturday and Sunday, if I have the time.

2018-01-25 20.58.09.jpg
 
#34
Very well put. We have very different design goals, resulting in very different end products.

I forgot to consider the removal of paper - that DEFINITELY will help to reduce weight, especially on large builds like this. The paper is effectively 50% of the weight of a sheet of foamboard. Also, I (wrongly) assumed you'd glue every joint of every frame, which as you said is not necessary. (I also forgot to subtract the weight of two ESCs in my estimation).

I'm pleasantly surprised that your component weight buildup actually came in so close to your estimate, nice job!

I think you'll gain more weight in wiring than you guess, but it certainly won't be up to the weight I previously ballparked.

My paint didn't add all that much to the total, actually. Yes I went through 5.5 cans of spray to cover it, but the bulk of that ended up on the EZ-up walls, and in my hair! :p My component weight before spraying added up to about 23-ish lbs, and my all-up weight after paint is 24lbs on the nose. Well, to me that's not a lot, as a percentage of the total anyway. ;)

And yeah, my retract units are BEASTS...but as you said, they kinda need to be.

At any rate, glad you're making progress!

Paint question: Are you replicating airframe #1 or #2?
 
#36
Paint question: Are you replicating airframe #1 or #2?
I hadn't really considered this until recently, but I think I'll stay true to the CAD model and paint it like #1.

I'll be there. I'm not sure if Ben is able, or willing, to transport such a large plane such a long distance.

formation Valkyrie flight?
While the thought of this is highly entertaining and would be awesome, I imagine that a proper formation flight would require both of us to become very familiar with our birds in the air. With this being my largest and most time-consuming build to date, and never having flown formation with any model before, I probably wouldn't be confident enough to do something like this.

Although it would be undeniably awesome.



She stands! Getting the geometry of the retracts and gear doors right was more of a pain than I expected, and I still haven't got the nose gear doors working. Progress was slow today.

20180126_213755.jpg

It sits pretty nose-high, but that'll mean that rotation and liftoff will be easier, especially with the mains so far back of my estimated CG.

20180126_213806.jpg

Retract mounting for the mains, as well as gear door servos. The servos will protrude into the duct, but there's not much I could do about that.

20180126_213827.jpg

And the nose retract/steering setup, which needs a cover over the hole with space for the pushrods to poke through.

I've filled up my SSD with footage for the next build log, and would've put a gif of the mains retracting here if my computer didn't delete the extra files during the cut-paste. :rolleyes: Time for some disk cleaning!
 
#38
This continues to be an awesome build and it's really interesting to have the two XB-70 builds running at the same time to see the different approaches to design / build / problem solving.

What laser cutter are you using?

DamoRC
 
#39
What laser cutter are you using?
I'm "cutting" on this monstrosity that I built from old 3D printer parts. It uses a 500mW blue-violet laser. Due to the low power of the laser and the cellular structure of the foam, it'll only do the equivalent of a score cut. It's still better than tracing out the plans by hand and introducing human error, though, and it can also raster images onto plywood.

2018-01-27 14.17.24.jpg

2018-01-27 14.27.10.jpg

2018-01-27 14.27.28.jpg

Wires, wires, everywhere! This is without the battery or motor leads.

2018-01-27 14.17.40.jpg
 
#40
Progress from last night, as well as today:

I soldered up the ESC wiring. I left the battery leads (not shown here) long and braided the motor leads to make routing the stiff single-core wire less of a hassle.

20180127_162727.jpg

The motor leads added around 350 grams to the weight, and the battery leads around 100g, which accounts for most of the "other stuff" weight budget mentioned in post #32 of this thread. The nacelle without electronics, however, weighs in at 400g, meaning I have another 400g to finish the airframe (4 sheets of foam + glue), which should be more than enough.

I also installed the ducted fans and thrust tubes, which have an exit diameter of 62mm (100% FSA for my fans).

The middle two EDFs are wired to one pair of batteries, the outer two to another. This way, I can lose one set of electronics and not have to worry about asymmetric thrust.

20180128_000220.jpg

I have yet to cut out the decoy nozzles which will hide the thrust tubes as well as make it seem as if the plane has six engines.

20180128_000202.jpg

I went for a taxi test only to discover that the set screws on the mains had loosened, causing the nacelle to tokyo drift in a rather hilarious fashion :p. I'll end up loctiting these during final alignment. The nacelle by itself maintains velocity at less than 10% throttle. Turning radius was a little on the large side for taxiing, but would be perfect for takeoffs (especially with no expo). Before the maiden, I plan to measure the acceleration of the craft to determine the approximate runway length needed for takeoff.

Here's two pictures of everything (approximately) in place, showing the high angle of attack on the ground.

20180128_152833.jpg

20180128_152909.jpg

The finish line is in sight, but still far away. Here's a list of everything I still need to do prior to the maiden:

-Nose retract doors
-Repartition main gear
-Main gear alignment
-Decoy nozzles
-Install and wire servos for elevons and rudders
-Drill holes and tap screws for wing mounting
-Sheet center fuselage
-Attach nacelle to center fuselage, put in Vstabs
-P̶a̶i̶n̶t̶!̶ This bird'll have to prove that it can fly before I invest the time into painting it.
-Stress test electrical system (loading all servos, seeing how the BECs cope)
-Stress test airframe?
-Taxi test, figure out length of runway required for a takeoff

I should have a build log up no earlier than Wednesday, depending on workload.
 
Last edited: