DHC-4 Caribou | 73" Span

[Pieliker96]

Some progress. Fuselage is halfway done, empennage is nearly sealed off, and the tailfeathers are complete. The bay doors are operational as well.
This is the largest fuselage I've done with semimonocoque methods. Joining large pieces together without a jig proves to be doable, if quite unwieldy. In any case I'm finding errors in the plans, correcting them, and making note, generally, of how to put it all together. One thing is for sure: This certainly isn't a build you could do in a single afternoon!

 

[Ligbaer]

The full-scale DHC-4 had variable-pitch props and could put them into beta (negative angle of attack, reverse thrust) to reduce landing distance and backtaxi - something similar can be achieved by having the ESCs reverse the directions of the motor and drive the props backwards. It'll be interesting to see how it works at the model scale

what about using a helicopter assembly instead of reversing escs

 

[Pieliker96]

what about using a helicopter assembly instead of reversing escs

Variable pitch would be neat to do and would open up some opportunities like a constant-speed controller. However it's a lot more complexity, weight, and cost than just having a traditional prop (of which the selection is far greater, especially for scale stuff meant for fixed-wing) and a reversing ESC.

 

[PsyBorg]

@willsonman did a build or two here a while back that used variable pitch. maybe could pull something from what and how he did it. He is one of the better known builders here even though he has not been active in a while.

Here is one of his threads that used a variable pitch power system.
https://forum.flitetest.com/index.php?threads/winter-build-2019-2020-top-flite-p-51-0-60-size.60646/

 

[Pieliker96]

The fuselage is more or less done at this point, which means I can start to focus on the wings.



Gear doors (as always) proved their ridiculous tediousness, so I ended up omitting the front nose flap and driving the main nose gear doors with one micro servo each. The carrythrough spar is in as well, along with the wing connectors: one XT60 and one MPX per side. Wiring was considerable but has cleaned up well so far - though it'll all be hidden behind a cover plate anyways.



And here's a function check of the entire fuselage: nose retract, gear doors and steering, the cargo bay doors, and the tail feathers:

 

[willsonman]

A heli assembly is an overly complex solution to a simple problem. My solution involved nearly $300 worth of parts, including a custom 3D printed aluminum part. For anyone considering it, also consider alternatives.

 

[PsyBorg]

The fuselage is more or less done at this point, which means I can start to focus on the wings.

View attachment 214375

Gear doors (as always) proved their ridiculous tediousness, so I ended up omitting the front nose flap and driving the main nose gear doors with one micro servo each. The carrythrough spar is in as well, along with the wing connectors: one XT60 and one MPX per side. Wiring was considerable but has cleaned up well so far - though it'll all be hidden behind a cover plate anyways.

View attachment 214378 View attachment 214376 View attachment 214377

And here's a function check of the entire fuselage: nose retract, gear doors and steering, the cargo bay doors, and the tail feathers:

Well done. Looks like you spent some time and effort on the door timing and speeds as they simple didn't pop out and in as spastically as a default setting. Its hard to see but how much support did you do for the retracts. That seems to be one of the fail points with foam bard planes. Not much in the way of impact distribution for the normal wear and tear of landing with foam.

 

[FlyerInStyle]

The fuselage is more or less done at this point, which means I can start to focus on the wings.

View attachment 214375

Gear doors (as always) proved their ridiculous tediousness, so I ended up omitting the front nose flap and driving the main nose gear doors with one micro servo each. The carrythrough spar is in as well, along with the wing connectors: one XT60 and one MPX per side. Wiring was considerable but has cleaned up well so far - though it'll all be hidden behind a cover plate anyways.

View attachment 214378 View attachment 214376 View attachment 214377

And here's a function check of the entire fuselage: nose retract, gear doors and steering, the cargo bay doors, and the tail feathers:

That is incredible! cant wait to see it fully assembled. that is some fast progress too! Amazing Job, @Pieliker96

 

[L Edge]

For the reversible prop design, is there much of a power loss in forward flight?

Could you explain how you are going to attempt the "Wheelbarrow"? The video says the flaps are at 40% and the speed is 160 knots. You chose a superb project.

 

[Pieliker96]

Well done. Looks like you spent some time and effort on the door timing and speeds as they simple didn't pop out and in as spastically as a default setting.

Definitely. An advantage of using an arduino / FC is that complex sequencing and servo speed control is possible with even the most basic of transmitters.

Its hard to see but how much support did you do for the retracts. That seems to be one of the fail points with foam bard planes. Not much in the way of impact distribution for the normal wear and tear of landing with foam.

I've got two ply mounting plates embedded into three formers (one double-thick) in the nose. The nose itself is braced to the rest of the fuselage with some gussets to take up the moment around the join between the two. There'll be something similarly robust for the main gear: plates transferring their load vertically into the wing through a box just behind the wing spar.

For the reversible prop design, is there much of a power loss in forward flight?

There's no downside to having a reversible ESC. It'll turn the normal way for normal flight and the opposite way for reverse, no loss of power in either direction.
As far as I know, variable-pitch setups (at least, those for helicopters) typically use symmetrical cross sections to reduce the pitching / twisting moment of the blades (this applies at full scale, not sure how true it is at model scale). Cambered sections are better optimized for one direction but provide relatively poor performance in the other (especially when driven backwards) whereas a symmetrical section performs the same in both directions. So the tradeoff is that a variable-pitch setup would provide similar thrust/efficiency in both directions whereas my setup with the reversible ESC is superior for forward flight and inferior for reverse/braking.

Could you explain how you are going to attempt the "Wheelbarrow"? The video says the flaps are at 40% and the speed is 160 knots. You chose a superb project.

If I can get a sufficiently high lift slope with these massive externally blown flaps, I should be able to get the plane to fly level with the nose pointed towards the ground. In that attitude the nose gear will hit first. I should be able to keep the nose on the ground and the mains off. The main concern will be the positive feedback loop with the nose gear that will tend to try and turn the plane around backwards: as the plane yaws to one direction, the nose gear (if uncorrected) will tend to steer it in that direction, further increasing the yaw. Keeping pressure on the nose wheel light and continuing to fly the airplane will give the best chance of success.

 

[L Edge]

Definitely. An advantage of using an arduino / FC is that complex sequencing and servo speed control is possible with even the most basic of transmitters.

I've got two ply mounting plates embedded into three formers (one double-thick) in the nose. The nose itself is braced to the rest of the fuselage with some gussets to take up the moment around the join between the two. There'll be something similarly robust for the main gear: plates transferring their load vertically into the wing through a box just behind the wing spar.

View attachment 214397 View attachment 214398 View attachment 214399 View attachment 214400




There's no downside to having a reversible ESC. It'll turn the normal way for normal flight and the opposite way for reverse, no loss of power in either direction.
As far as I know, variable-pitch setups (at least, those for helicopters) typically use symmetrical cross sections to reduce the pitching / twisting moment of the blades (this applies at full scale, not sure how true it is at model scale). Cambered sections are better optimized for one direction but provide relatively poor performance in the other (especially when driven backwards) whereas a symmetrical section performs the same in both directions. So the tradeoff is that a variable-pitch setup would provide similar thrust/efficiency in both directions whereas my setup with the reversible ESC is superior for forward flight and inferior for reverse/braking.



If I can get a sufficiently high lift slope with these massive externally blown flaps, I should be able to get the plane to fly level with the nose pointed towards the ground. In that attitude the nose gear will hit first. I should be able to keep the nose on the ground and the mains off. The main concern will be the positive feedback loop with the nose gear that will tend to try and turn the plane around backwards: as the plane yaws to one direction, the nose gear (if uncorrected) will tend to steer it in that direction, further increasing the yaw. Keeping pressure on the nose wheel light and continuing to fly the airplane will give the best chance of success.

Not knowing very little about the ESC's and reversing, I was guessing that they perhaps design a different prop. I explored a STOL transport where you can change the wing AOA from zero up to 33 degrees. One important point I found out is that you needed a specific speed as you changed the angle to keep it stable. Your going from a positive 2 to something lower and if a crappy prop is involved while using the blown flaps, adds another factor.

Putting a dual nose wheel is really going to help.

 

[JClaude]

WAOUW! Nice project!

 

[Marzipan]

that is bloody le gorgeous Pieliker96! I am so excited to see your plans!!!

 

[Baron VonHelton]

How heavy is the landing gear??

 

[Pieliker96]

How heavy is the landing gear??

250 grams for all three retract units, oleo struts, wheels, axles, and wheel collars. Add in supporting structure and wiring and you'll likely be close to 300 grams for the entire retract system.

I chose a target weight of 1900g and (as usual) I'm finding it was overly optimistic. The fuselage itself is already 1 kilo, and the battery is 400g. A realistic weight (with all of the wood and electronics) is 500g per wing - I'm now hoping to come in under 2.5kg.

Speaking of wings, here's some progress on the left one:

 

[Battery800]

Nice! I'll be doing a fully-fledged set of plans then


I designed it for the scale of Power Pack C on 9" three-blade props - much larger and thrust-to-weight starts to suffer, smaller and wing loading becomes a concern. In any case you're welcome to scale down the plans to your liking once they're released - I personally won't be doing so since the design work is already done at the current scale

I might build this for buildruary since I already did the p-40, but I would probably need to have it scaled down by half. Still don’t have foam though

 

[Battery800]

Also, have you thought about building a twin otter? I tried but it was hard since I have no cad experience

 

[Photo001]

Nice job, looking forward to your plans release!

Lane

 

[Ligbaer]

im go

I might build this for buildruary since I already did the p-40, but I would probably need to have it scaled down by half. Still don’t have foam though

dhc-4 group build?

 

[Pieliker96]

I might build this for buildruary since I already did the p-40, but I would probably need to have it scaled down by half. Still don’t have foam though

The current plans are at 13 sheets of foam and two 1'x2' sheets of 1/8" ply - if you're short on foam, scaling it down would help a lot.

Also, have you thought about building a twin otter? I tried but it was hard since I have no cad experience

Making a set of plans from CAD is something that takes a good amount of time and practice. If you're looking for a program to start off with, I recommend Google Sketchup (guide/tutorial here). Find a plane you'd like to practice with and find a set of threeviews for it, with cross-sections if possible. Import them into Sketchup, scale appropriately, then trace the threeviews and make a simple profile model. Once that's done try doing the same but with a box fuselage, then with the cross-sections if you so desire. Here's my traced threeviews, which all of the plane's geometry were derived from.

I finally got the wing retention system sorted and the plugs lined up - it's going to be very convenient to just lock the wings in and fly, no fiddling with connectors. The downside is that the wing is already at 500g without any of the control surfaces or nacelle. It'll be one strong but hefty bird. I'm hoping the flaps - especially when externally blown - do wonders to the lift coefficient and let me get real slow.