Aluminum tube for RDS

[jfrickmann]

I wonder if I can use aluminum tube for an RDS (Rotary Drive System). Can anyone help me figure out what size aluminum tube I would need for a given torque load?

For instance, a Hitec HS125 has a max torque of 3.5kg*cm, and I would like to use alu tubes from Tower Hobbies: http://www2.towerhobbies.com/cgi-bin/wti0093p?&C=RDA

Thanks,
Jesper

[Sparky Paul]

I use an rds on my Nieuport 27 upper wing ailerons.
1/4" aluminum tube, about 18" long from the center drive pickup to the aileron attachment.
The servo arm and horn are about 1-1/4" long each.
It's not a fast airplane and the response is pretty good.
On my HOB T-6, the one with the rds has only a short portion for the torque rod, out in the wing by the aileron.
It's rotated by a pushrod from the servo, as on my Kadet Mulligan.

[BMatthews]

K&S aluminium tubing is very soft and could be easily damaged. In sizes larger than 3/16 diameter it'll easily withstand the servo output but it may be twisted by an accidental blow to the aileron before the servo can move unless you are using a larger size like 1/4 or larger like Sparky did. Even then it wouldn't be my first choice for this.

I'd be inclined to use aluminium arrowshaft tubing. It's far stronger and springy so it won't suffer from accidental damage.

[jfrickmann]

Really? I did not know that! Thanks for that info. I will look around for harder types of aluminium tubing.

[jfrickmann]

I have spent the week researching aluminum alloys and calling around. I have learned that the best alloys are "aircraft grade" 7075 T6 and 2024 T3, which are as strong as medium grade steel. Unfortunately, no one can offer me any of that in the small sizes I need, at realistic quantities and prices. The closest I got was Easton Aluminum, a well know maker of arrow shafts. But their smallest size is 1/4" OD, which is just a little bit too big for me. I also got a quote for 6' 3/16" + 6' 1/4" + 6' 5/16" 2024 T3 alloy - "just" $550 + shipping :eek:
So I went back to K&S' homepage and learned that they actually have two different product lines. The stuff from their "Metal Center" is the soft 3003 alloy, but the stuff from their "Aluminum Center" is actually a descent 6061 T6 alloy (except from the square tubes). You can recognize the good stuff by the wall thickness - 0.035" and 0.049" versus 0.014" for the soft stuff.
I figure that a 6061 T6 alloy, 3/16" OD and 0.035" or 0.049" WT will be plenty strong, so I ordered some and will try it out.

Thanks for everyone's advice and input!

Jesper

[ghoti]

http://www.smallparts.com/products/descriptions/lat.cfm

[Mister UHU]

7075-T6
and
2024-T3,

The closest I got was Easton Aluminum,
a well known maker of arrow shafts.

But their smallest size is 1/4" OD, which is just a little bit too big for me.

I also got a quote for 6' 3/16" + 6' 1/4" + 6' 5/16" 2024 T3 alloy
- "just" $550 + shipping :eek:


K&S

"Metal Center" is the soft 3003 alloy, == 0.014"
"Aluminum Center" is actually a descent 6061 T6 alloy == 0.035" and 0.049"
(except from the square tubes).


I figure that a 6061 T6 alloy, 3/16" OD, and 0.035" or 0.049" WT,
will be plenty strong, so I ordered some and will try it out.


Jesper

6061-T6 is high quality aerospace grade alloy.
Its also can be called by some other names, depending on the maker and country of supply.

The T6 refers to the type of heat treatment, the metal gets after manufacture.

The heat treatment can add quite a lot of strength,
so be wary of heating up the metal, such as for welding, as you might lose strength.

Some of these alloys don't like being bent, so can suddenly fail if over-stressed.


As for Easton arrow shafts, just buy some arrows, and cut them up for the tubing, MUCH CHEAPER !

An archery club may even have some damaged arrows you could have ??

Have you also considered carbon tubing.
It is used for arrows, fishing rods, golf clubs, and many other uses.

I would have thought the bigger problem was making secure attachments to the tubing,
not the strength of the basic tube.

[jfrickmann]

The problem with the smallparts link above and Easton tubes is that they do not go below 1/4". I plan on using the RDS in a Mark Drela inspired electric glider, where the airfoils are very thin. Ideally, I would have liked 5/32", but I can probably sqeeze in the 3/16" I ordered. But 1/4" would not fit.
I did not know that 6061 T6 was also categorized as "aerospace grade". As far as I could see, it is not as strong as 7075 T6 and 2024 T3. Looking at Onlinemetals (http://www.onlinemetals.com/aluminumguide.cfm), I found that 6061-T6 has a yield strength of 40,000 versus 50,000 for 2024-T3 and 73,000 for 7075-T6. So 7075-T6 is definitely the ultimate in strength, and it is also what Easton uses for their arrow shafts. But then I saw that the yield strength of the extremely expensive Titanium tubing sold at smallparts is only 40,000 as well. And since aluminum is lighter, even 6061-T6 is actually stronger relative to its weight! The yield strength of 3003-H14 used for the soft K&S tubes is only 21,000.
I just stumbled on this (http://www.omegaresearchinc.com/Publications/volumeFive.htm) excellent link describing the different alloys and tempering etc. This one (http://combatindex.com/mil_docs/pdf/Hopper/MIL-HDBK/CI-694A-MH-4829-4005.pdf) is also interesting.
I am aware of the fact that tempered metals should not be heated and that the hard alloys do not like being bent. I have attached a drawing of how I plan to make it. As you can see, there is no bending and heating.

[Mister UHU]

You could also try some of the major aluminium firms for basic data,
aluminium is a semi-monopoly industry, with only a few huge companies,
while they won't sell to us small people they will supply very good data,
the smaller re-sellers often just re-use and re-cycle this data,
(and sometimes add unwanted errors of their own !).

Here is a basic 2 page pdf data sheet about 6061 from ALCOA :--

www.alcoa.com/adip/catalog/pdf/Extruded_Alloy_6061.pdf

Right now aluminium alloys containing Scandium are very popular
for applications requiring high strength and low weight.

You should be aware that some alloys that appear to have very good strength in lab tests
(such as alloys containing Lithium)
sometimes can fail catastrophically in real world use.
Generally engineers are very conservative in their use and selection of aluminium alloys.
Choosing only materials with a very well understood and proven track record.

Sometimes aluminium contructions fail due to careless fabrication,
or unforseen aging, weathering, or corrosion;
in the wrong set of circumstances aluminium can be surprisingly reactive,
(such as galvanic reactions, and surface friction).

A lot of the basic work on aluminium alloys was done many years ago,
most of the popular alloys, grades, and processing methods, have been around for years,
the leading ("bleeding") edge of materials
is now composites such as carbon fibre, carbon nano tubes etc,
this is where most of the R&D money is going, and the creative effort is being made.

* * * * * * * * * * * * * * * *

From your drawing I am not clear about your servo attachment arrangement.
Where you have written "EPOXY WITH MICROFIBERS",
couldn't you use the "cap" part from a servo arm ??
(especially if it is made of metal),
cutting off the "arm" part.

You would be better off if the servo screw had a deeper/thicker head,
so it protruded further inside the end of the tube,
you could even try cutting (grinding, filing) a small notch into the head of the screw,
that your "SET SCREW" could screw into, for extra grip against unwanted rotation.

The more space you have available, the easier the junction becomes,
if you want it to be super-slim, then you risk the connection being weaker and more difficult to make.

You should also look at marine boating couplings, as used for propeller/motor drive shafts.
Some of these maybe immediately usable for your application.

You should also look at car drive shaft couplings,
as some of these maybe of use to you ??

If your servo has a square output shaft it could make the fitting much easier,
especially if your tubing had a matching square end.

Another simple joining method is a basic screw-thread,
i.e. an external thread on the servo shaft, and an internal thread on the inside of the tube.
Rotation could be prevented with a set-screw and/or strong glue (eg epoxy).

With some more space,
you could simply drive a bolt (or small rod) into the tube at 90 degrees
(as the "SET SCREW"),
but all the way across, and out the far side of the tube, so it sticks out of both sides of the tube,
you could then fix it to the servo arm.
(Making sure that the tube and the servo are firmly mounted, exactly inline.)

You need to be very sure that your linkage is very frictionless and smooth running,
as there is a very large surface area on your tube that could bind the servo badly.

The traditional method of using a small linkage between the servo and the aileron torsion bar,
has a lot to recommend it, not least is the adjustment of motion (and position) it allows,
your set-up would require gearing if the servo motion didn't match the control surface motion requirement,
which would mostly likely be impractical.

From your drawing dis-assembly looks tricky,
I guess you are going to use the SET SCREW for this ??

I have seen an arrangement a bit like this on RC Groups,
but it was for a much smaller size using a small carbon rod instead of your aluminium tube,
and I think the coupling was made using heat-shrink-tubing over the servo arm ??

You would probably be best off doing some experiments,
with various different designs, before you commit yourself.

[Mister UHU]

If you are able to make good quality small screws/bolts,
then you can solve this problem very neatly :--

smaller screw that screws into the servo//small hex bolt head//larger screw that screws into tube.

All made from one piece of metal (steel, or maybe softer brass ??).

Could be made by turning down steel hex bar on a lathe,
then putting screw threads on either end.

The screw thread for the aluminium tube could also be a self-tapping thread,
but you would need to be sure that you got exactly the correct size.

Provided that both threads were tightened properly this coupling should be loosening-resistant,
especially if glued too.
Grub-screws/set-screws could also be used to lock the system together.


You can do without the hex bolt head,
if you drill a hole across the larger screw instead,
so allowing tightening by inserting a very small rod.
The small rod, or piano wire, that is later cut down to match the width of tube,
stops the larger screw from going further inside the tube, ie acts as bolt head.

Using this design you can turn down a long bolt that fits the tube,
into a small screw thread that fits inside the servo.
So you only need a long bolt that fits the tube
and a set of dies, the smallest of which fits the servo thread requirement.
IE no lathe is required, using this method.

Looks something like :--

.............---------
-------- | |
-------- | |
.............---------


****************************************

It might also be possible to weld (or glue ???????)
2 screws together, head-against-head.
Preferably so the tube screw had a hex head.

****************************************

Or a simple strong steel rod could pass through both servo and tube,
with sleeves to help it fit inside the tube tightly.
Preferably fitting very tightly inside the servo, so that the soft plastic is deformed
around the rod, to make a very tight fit.

As with other options,
glue and/or grub screws could help lock this setup together.

[jfrickmann]

Thank you for the interesting thoughts and suggestions, Mr. UHU! 30 years ago, I had a Graupner plane named Der Kleine UHU ;)
The basic idea is what is described here (http://www.genie.rchomepage.com/RDS%20MODIFICATIONS.pdf) for Tabasco and Genie sized sailplanes. As for the epoxy/fiber cast around the servo head it is adapted from this file's (http://www.genie.rchomepage.com/FILE%2006,%20ROTARY%20DRIVER%20SYSTEM.pdf) section about custom splining.
I will play around with it and see what works, once I receive the tubes. I prefer to keep it simple - no bending (dangerous with aluminium), no machining and welding. Just cutting, drilling and J&B Weld (glue).
I still have plenty of time to play with it, as I need to build the plane also.

[Mister UHU]

30 years ago, I had a Graupner plane named Der Kleine UHU ;)



DITTO !
Hence my forum name,
some people have thought this was because of me building foamies, with Uhu Por glue,
while this is also true, it is the glider that I used for my name.

*********************************************

I think I now understand the use of epoxy in your drawing :--
it is to match the splines on the servo output shaft.

The problem of the servo parts being made of weak plastic, is difficult to get around.

Some of the indoor foamy flyers have been gluing small carbon rods at 90 degrees,
through the servo output shaft, instead of using the supplied servo arm.

Several months ago, I posted some ideas on how to use a servo arm
and some pushrod anchors to drill this hole more easily and accurately.
I posted these in the INDOOR PATTERN forum on RC Groups.

Instead of putting a carbon rod through the servo shaft,
you could use the hole to fix 1 or 2 set screws/grub screws in,
so preventing the RDS coupling rotating on the servo shaft.

Using the coupling in your drawing, you could add 1 or 2 small screws
at 90 degrees into the servo shaft to prevent rotation.
The screw(s) would go in as far as the servo arm screw.

*******************************************

If the internal diameter of the RDS tube
matched the external diameter of the "cap" part of the servo arm,
then it would be possible to slide the RDS tube over the "cap",
and secure the RDS to the servo shaft with a small rod or screw,
(that went through the tube, and the "cap", and the servo shaft),
at 90 degrees across the tube/cap/shaft.

The problem with this arrangement is the lack of rigidity in alignment along the tube.

*******************************************

From following the links you posted,
it looks like you could buy some RDS couplings off-the-shelf.
These look of decent quality and design,
but I am not clear if they would match the requirements of your installation.
(I post a photo of this coupling below, in this post.)

*******************************************

I particularly like the Fork type coupling, shown in your link.
It fits the servo without modification,
and it allows for some mis-alignment in the setup.

(It is very similar to the design of some marine drive couplings.)

The coupling shown is high-quality, and looks a pain to make.
I think it may be possible to produce a similar coupling using only nuts and bolts,
without needing welding or soldering.

EG "pegs" for servo arm simply 2 very small bolts held to coupling with nut(s).
The tube end of coupling is replaced by a bolt that fits through the coupling plate,
and is held in place by a nut.
The end of the bolt then screws into the end of the RDS tube,
and can be locked place with a set screw etc.

The weakness of this arrangement is trying to stop the bolt from rotating
inside the plate that holds the 2 servo "pegs".
The usual locking methods can be tried here.

****************************************

You could also try simply drilling a small hole across the RDS tube,
insert a piece of piano wire the same size as the holes in the servo arm,
and bending the wire at 90 degrees twice,
to form a sort of a fork.
So that the 2 ends of the Fork fitted into 2 holes in the servo arm.

The problem of this coupling is the lack of rigidity of the wire in the tube,
this would be best fixed by welding/soldering/gluing etc.
The wire might move across the tube, as well as rotating around inside it.
The across motion could be fixed with washers, a nut or 2, sleeving, etc.

BETTER STILL :

DRILL RDS TUBE FOR WIRE THE DIAMETER OF THE SERVO HOLES,
INSERT PIANO WIRE INTO HOLE,
BEND PIANO WIRE AGAINST SIDES OF RDS TUBE ON EITHER SIDE,
BEND PIANO WIRE OUTWARDS TO MATCH WIDTH BETWEEN SERVO HOLES,
BEND PIANO WIRE INWARDS SO BOTH ENDS ARE PARALLEL AND FIT INTO SERVO HOLES.

Piano wire should be held against sides of RDS tube using sleeve,
wound wire/thread etc.
(heat-shrink-tubing may even be strong enough??)


Looks something like :


----
.....|
......---------
......=======.=========================
..................|
......=======.=========================
......---------
.....|
----

[Mister UHU]

I found the RC Groups photo showing,
a small linkage like RDS.

It uses a bent metal rod,
and some heat shrink tubing fixing it to the servo arm.

[John O'Sullivan]

I have read this thread with interest. As one who has been using RDS type actuators for more than 20 years and who had long term correspondence with Harley Michaelis on the development of his RDS, I would like to make some points.
The key advantages of the RDS system are
1. Simplicity
2. positive action
3. lack of stuff hanging out in the breeze
4. Once installed, works flawlessly without fiddly adjustments
5. reliability

In the mid 80's I started to use the crank system with a box in the aileron operating in a box in the aileron (elevator, Rudder). I was surprised when Harley published a similar system in two models published in RCM. (one was the Orca).
He was, at the time using drilled out automoible metal valve caps with rods soldered in and the cavity filled with epoxy to match the servo spline output. I was using DuBro Collets with a 5/16" aluminium sleeve to bridge the gap between the collet and sleeve. I also used epoxy to match up to the servo spline. (there is no need to add glass fibres to the eopxy mix I have not had one fail yet and the fibres only get in the way of a proper fit). This allowed me to use the set screw of the collet to lock to the drive shaft. I also had tried the yoke system plugged into the servo arm, but found it was harder to build accurately.
Harley finalized the RDS and Kimbrough produced them.
They are in my opinion the most positive actuators for operating control surfaces with short drive distances.

On my smalller, lighter models, I use 0.039" wire (a bit thicker than 1/32"). For power models up to 50 inch and gliders to 100", 1/16" wire works well and over this 3/32" or for really heavy duty 1/8" wire does the trick. Using tubular drive shafts works well, but adds to the complexity of the setup.
In my Ohmbrew 36" sports electric model, I have used two splayed out shafts from the same servo to drive two ailerons. (note the wire thickness is 0.039" rather than the 1/32" on the drawings and photo)
http://www.windandwavemodels.com/DoubleRDS.html
http://www.windandwavemodels.com/Ohmbrew400.pdf

Installation in Claim Jumper 60" slope racer
http://www.windandwavemodels.com/RDS.html

The use of tubular drive shafts is overkill on the average model - 1/16" or 3/32" rods will satisfy 99.9% of needs depending on the model size and flying speed.
Above all, dont be tempted to use carbon shafts or tubes. Most of these are fabricated from unidirectional fibre and although having excellent tensional strength, are next to useless in torsional stiffness (unless specifically made using oriented fibres).

Forget the fancy stuff and keep it simple.
John

[Mister UHU]

The drawing in post #8 is somewhat accurate,
in that the HS125 has a shaft diameter of about 4.5 mm,
which is about the same size as the external diameter of the aluminium tube shown.

Another possible variantion is eg:

.......---
...........|.========================
............-------
.....................O|
............-------
...........|.========================
.......---


.....................^ is a bolt with nut holding the wire in place,
........................fitting inside a hole drilled through the tube.

At the end of the tube, 2 notches/a small slot are/is cut
to hold the wire, and stop it from moving.


The tube can be filed/ground to make 2 pegs into the end eg:

..------------------------------
=
..------------------------------


=-----------------------
.|
=-----------------------


So that the pegs fit directly into the holes in the servo arm.


If the wire diameter is the same size as the wall thickness of the tube,
then this arrangement can be made:



---
....|
..||.--------===============
..||\\\\\\\\\\|
..||.--------===============
....|
---

Where a bolt is screwed into the end of the tube to hold the wire in place.
A slot having been cut into the end of the tube to accommodate the wire.

A short length of tube can fit over the wire and RDS tube to hold the wire,
and stop the fork end from expanding as load is applied.
Alternatively a nut can be screwed over the end of the RDS tube to keep the wire in place.

If a steerable nose leg of the correct size can be found,
then it might be possible to fix it to the end of the RDS tube.


..........===
...........||
.-------...-------
|.0..0....O....0..0.|
.----------------

...........^hole for nose leg (RDS tube), set-screw locks in place.


A prop-saver of 3/16" internal diameter would fit over the RDS tube eg:


.........---
[]===(.O.)===[]
.........---



So the locking screws could be fixed to the servo arm eg with heat shrink tubing.


The arrangement shown in post #13,
could be modified so the wire has a "Z" bend in it,
to fix to the servo more securely.

The wire could also be bent eg:


.----
|.....|
|.....|
.-------------------------------
|----|----
|.....|
.----

In a sort of "8" figure, so the wire could attach to both sides of the servo arm.
The "return" extra length of wire could be held to the longer length of wire with heat shrink tubing etc.

If both lengths of wire are long, then they can be used to move
2 control surfaces eg 2 ailerons, by spliting them, so they curve in 2 different directions.
(as shown in the John O'Sullivan plan)

As separate "]" fork of wire can be fixed to the figure-of-eight "8" wire,
so the ends of the fork fit directly into holes either side of the servo arm.eg:

------
........|
........|
........|
........|
------

The "8" wire can also be bent eg:


.------
|........|
.--......|
...|.....|
....-------------------------
...|----|-----
.--......|
|........|
.------


Where the holes in the servo arm have been enlarged into slots,
so that the wire can fits into these slots.


Obviously all these double-sided servo arm arrangements
only fit into a thin wing if the arm is "horizontal" and "flat".


If only a wire is used and not a tube,
then the wire can be inserted into the servo shaft hole,
and held in place with 2 screws at 90 degrees to the wire.
Fitting the "cap" from the servo arm gives the screws extra material to fix with.eg:




---------..||
............|..----
............|---|.|
..........================================
............|---|.|
............|..----
............|..||
............|
............|
---------



I hope these suggestions help anyone trying to make an RDS.
The previous work in this area should obviously be studied first.

I wasn't aware of this servo interface before this thread,
I am only looking into various theoretical options,
none has been tried in an actual model, or even made.

I am surprised Hitec, Futaba, Graupner, DuBro, Sullivan, etc,
don't make parts for this linkage already.
It would be very easy for them to do.

[markdrela]

Here are some techniques for making pockets and carbon RDS shafts and yokes for DLGs:
http://www.dlg-supplies.co.uk/
click "Construction Techniques"
click "RDS Set-ups"

[jfrickmann]

That's pretty neat with that teflon tape - is "pocket friction" generally a problem?

I am a little concerned if I can fit the 3/16" alu tube in the aileron pockets. It's a "Supra inspired" 3m e-glider with AG40d - AG43d airfoils. Do you think that e.g. the soft thin wall 1/8" alu tube with a couple of carbon socks over would work better, or should I just get some stainless hypo tubing, such as you recommend for the Supra?

Thanks,
Jesper

[Mister UHU]

As well as the "figure-of-eight" wire,
there is the simpler (and more obvious?) "loop" eg:


.----
|.....|
|......------------------
|......------,,,,,,,,,,,,,,,,,optional second "output"
|.....|
.----


Which can also used in the various combinations as the "8".

[Mister UHU]

Small helicopter tail rotor drive shaft couplings may also be of use for RDSs ??

[jfrickmann]

So maybe I should buy that Dragonous Heli before I go ahead with this build ;) Nah, I will see what I can do with the alu tubing I just bought. I just blocked out the foam for the main wing, so it will a while before I get to the RDS... I also need to build a Joe Woodworker vacuum system first.

Thank you for the many good suggestions!

Jesper

[Mister UHU]

It is one of those problems that is stuck in my head,
I won't be happy until I have gone through every possible possibility.

Here are a few more:

"The Question Mark" eg:

.---
|...|
|...|------------------------
|
|

"Open"


||
||-------------------------
|
|

"Closed"


|\
|..\-------------------------
|
|

"Diagonal"


Also see photos:

The collets are optional, and add weight.

The wire should be bent to fit the hole nearest the output shaft,
(with the servo arm shown, an extra hole further in could be drilled).

The bolts should fit the holes in the arm as closely as possible,
(they can even screw into the soft plastic).
The bolts should be as close in to the shaft as possible,
for the most compact layout.
The bolts (or bolt heads) must clear the servo case, and not rub against it.

In the layout showing 2 servo arms,
one arm is screwed onto the servo spline,
then the other is slid over the 2 bolts, and the 2 nuts are tightened.

Obviously if the servo arm is orientated so it is "flat"/horizontal
so it is in line with the servo case it will take up less space,
and allow for a thinner wing.

[John O'Sullivan]

Why not just RDS couplers? They are simple and effective. There's no need to make anything more complicated than it need be.
Many things work on the bench, but I've never seen a bench fly.
John

[Mister UHU]

If a tube is to be joined directly to the servo shaft, then this is a very simple way to do it.
The screw fixing from within an electrical terminal block.

Flats can be put on the servo shaft and/or the tube,
or even all the way through as holes,
to make more secure and prevent rotation.

Obviously the screw can be replaced by a headless grub screw,
and cut as short as possible to save weight, and provide more external clearance.