1348

Item 1348

DESIGN: UniCopter ~ Rotor - Blade - NACA 00xx - Torque Tube

Questions to be worked out:

How are tows wrapped at the tip end and all lie in the same plane; for the quill?

How are these tows (quills) held in place after being wrapped around the spar?

Have the quill holder etc. become an integral part of the blade tip.

If the tow cures at room temperature the winding procedure must take place at cold temperatures.

How well is the torque tube resisting twisting?

Overview:

Function of Torque Tube: To transmit the commanded tip pitch from the tip-pitch horn out to the tip of the blade.

The tip end of the blade will be rotated by this tapered composite tube, which will be located inside the spar of the blade. This tube will only require torsional rigidity. It will be constructed out of carbon thread, at a wind angle of close to 45º.

Picture of Mockup:

Done to develop winding profile.

Change to picture ~ There will be 2 quill tip bars on each side. ?

Notes:

Notes re Root End:

The radial pins at the root end are used to wrap the tow around so that the new tow path will lie adjacent to the previous tow path.

The number of pins must be evenly divisible into the number of strands of tow; at every layer.

For a picture of the use of pins see [FWCS ~ Figure 1-6]

The pins will be probably be cropped after winding and curing. Or, they will be drilled out for other pins to lock the pitch holder in place. See; 1353.html and 1355.html.

It may be an advantage for the pins to have a slight axial bend away from the tip end, to facilitate the wrapping of the tow around them. Alternatively, they might have nubs on the ends.

Notes re Tip End - Inner Layers: Also see; Rotor - Blade - NACA 00xx - Tip End Layout

The first layer consists of a single tape of 1 or more tows at a wrap angle probably greater than 45º.

Notes re Tip End - Outer Layers: Also see; Rotor - Blade - NACA 00xx - Tip End Layout

The first layer consists of a single tape of 1 or more tows at a wrap angle probably greater than 45º.

Use quills to assist with the tip end rotation since the strength of the epoxy will probably be enough.

All following layers consist of a band of tows and the band consists of 2 tapes. One tape creates one quill and the other tape creates a second quill at 180º about the X-axis (rotation axis) from the first. Subsequent layers will each produce additional pairs of quills and each pair will be located approximately 3.5" closer to the root end than the former pair.

The radius diameter of the torque tube, where the tape leaves the tube to become the quill, is to be the blade thickness less one hoop wrap (around the spar) and two times the skin thickness at this location.

The quills pass over (or under) the spar, then wrap around the leading (or trailing) surface of the spar, then back toward the torque tube. They are then bound to the spar tip by the single hoop wind, which passes up the full span of the blade.

Perhaps the clamp that aligns the quills should consist of a separate clamp for each pair of layers. This will allow the clamping to be done soon after the wrapping of that layer. It also sets the clamp up for the attachment of the next layer's clamp.

This note was for the previous spar tip design. Since the wedges are 2.1478" wide [ref. 0932.html] each layer on the torque tube i.e. each pair of quills) will be approx. 2.1478" shorter than the previous.

What about attaching the end of each tow at the tip end with fast acting Krazy clue.

Notes re Outermost Two Layers:

Consider winding the last 2 layers with only 1 tow for every 2 or 3 tow locations. do this for the +45º layer and the underlying -45º layer. This will create a checkered grid for assisting with the holding of the elastomer.

Notes re Core:

Consider burning the center hole through the Styrofoam block
Consider dissolving the core after completing the torque tube. This will eliminate any tendency of the core to resist the twist and remove some unnecessary weight.

General:

The torque tubes for one craft should consist of an equal number of items with a CW initial winding and with a CCW initial winding. The CW ones will be used on one rotor and the CCW ones will be used on the other rotor. This is so that the positive pitch that is applied to all 6 blades is equally transmitted up the torque tubes, should the torque tubes not have absolute torsional rigidity.

Consecutive tows must be laid down in the direction of the root. This is necessary because of the wraping of the quill over the quill tip arm. This necessitates a long X-axis travel because at the root end, once the tow has passed over the pin the eyelet must move further to assure that the tow will not slip off the pin when it is wrapped around the pin. It should be noted that the eyelet is already quite a bit ahead of the wrapping point because of the wrap angle and eyelet distance from the torque tube.

Tow Wind Angle:

"... it is known that torsional stiffness is maximized when the filament lay angle (wind angle) is relatively large, but in such event, the bending stiffness is relatively low. At relatively small lay angles, the bending stiffness increases, but the torsional stiffness decreases." ~ Patent US 4,863,414 Lord Corporation. It looks like the optimum angle will be approximately 60º, or even slightly greater.

If I recall correctly, 30 deg off radial is optimum for filament winding. [Wind angle of 60º] ~ from rnd2

For maximum torsional stiffness +/-45 degrees is best because that is the direction of the torsional loads. A +/-45 degree tube can be used to transmit torque through a 90 degree bend. ~ from Compositepro

The large wind angle means that there can be only one tow wrap around each pin at the root end.

For additional comment on Wind Angle see [Tow Wind Angle] below.

Winding:

It appears that the tow can be laid with only the use of the X and the Y machine axes.
It should be noted that there is no crossing under&over of the tow in the layers.
For additional information see [Winding] below.

Thoughts:

The ID of the root end might be wound in the shape of a radiused triangle, so that a radiused triangular pin on the pitch horn can slip fit into it. See additional info on the pitch arm, such as it ability to take the tip-pitch bearings, somewhere else. The area of the triangle must be greater than that of the root of the torque tube so that the mandrel can be removed.
As an alternative, the pins might be used for attaching to the pitch horn; if the pins do not stop the mandrel from being removed. Could the pins be plugs for the wrapping and curing?
The diameter at the root end should, and could, probably be smaller then initially considered. Particularly if the tip-pitch horn is bearing mounted. This diameter will probably be determined by the desired thickness of the elastomeric.
If the fabricated torque tube has too much twist, more material could be put on the inside. For instance, what about a light weigh core that has its surface hardened just enough to take the 1-2 pound tension of the winding. This light weight core need not be removed.

Considerations:

The diameter of the root end has been reduced and this reduces the taper.

This will cause the wind angle to be closer to 45º at the root end.
The torque tube taper will be shallower than the ID taper of the spar. This will allow for a varying elastomer thickness, which will probably be proportional to the desired active twist at each blade element.
The root end can be supported by the post on the pitch horn.
There may be no need for additional layers beyond those used to provide quills This is because the increasing diameter will provide the additional rotational strength that is transmitted through the elastomer.

To provide pins to wrap the tow at the tip end of shorter outer layers consider one or more extensions to the tip wrap bar device that rest over the laid plys and is rotated by the tip wrap bar device.

Notes:

Have the elastomer, which is located between the torque tube and the spar, transfer a large percentage of the torque to the spar along the full span of the blade, before reaching the tip. This will significantly reduce the amount of torque that must be transferred at the tip. The torque tube is thicker and has a larger diameter toward the root, which is ideal for a progressive transfer of the twist. By having the thickness of the elastomer greater at the root then at the tip, the transfer of torque will be equal along the span, but the actual rotation of the blade will increase linearly along the span. I.e. The thicker wall on the elastomer at the root will accommodated the greater discrepancy between the rotation of the torque-tube and the rotation of the spar at that point on the span.

Have the OD of the torque tube small enough at the root end so that it can be pushed to the side and the blade's attachment bolts replaced. This means that the only thing between torque tube and the inside wall of the spar will be a few layers of elastomer and shims. Alternatively, stop the torque tube a couple of inches back from the end of the spar.

Re Multiple Threads; To start the winding, consider bonding or clamping the leading end of all the threads to a portion of the paddle that will be cut off and some point in the manufacturing of the blade.

Re Single Tow; Wind single thread back and forth along mandrel, loop it over 'bar' at each end.

The lesser the bias means the more wraps, the fewer strands to be looped at the root ~ Good; the fewer tows in the quills ~ Bad.

The final layer may not experience the pressure that the inner layers do in compacting the thread/epoxy.

Tow Wind Angle: Working on now.

Which is better; > tows & < angle ~or~ < tows & > angle?? I think that more tows will be better since the over-layer will be ~ 90º and so is the first layer

Design so that the elastomer transmits a LOT of the torque:

The area between the quills will be used to tie the two halves of the spar together at the tip end.

Additional Thoughts on Wind Angle:

Each successive layer will be wound over the previous and will therefor have a larger diameter. In addition, each layer starts a little closer to the root, which has a larger diameter than the tip. If each layer starts with a 45º wind angle, the wind angle at the root will improve for each layer. The final few layers will have much shorter spans and this will result in a good bias at the root.
Each layer will have 1 or 2 more threads than the previous if every start is at the same wind angle.
Consider using the same number of threads on every layer and vary the wind angle. This will then work with the fixed number of pins at the root end.

Winding:

The sketch, which is related to the winding program in Access CNC Winding Generator Database is on Web page; OTHER: Composite - CNC Work Station - Generator (winding to g-code)

The initial inputs are the desired angle of the quills (tow wind angle at the tip end) and the length of the torque tube. The program then generates the other values, some of which, such as tows in a tape are fractional values. The next step is to select the step increment size and substitute integer values for the tow count and the number of wraps.

The purchased carbon tube over the cable has an OD of 0.170" ID + (2 x 0.025" wall thickness = 0.220"

If 12K yarn is used then each tow is 1"/42 = 0.0238" wide by 0.0238" thick. The epoxy may add to this dimension. For more information see; OTHER: Composite - General - Tow Or better yet see Access database FORM: Tow where the tow width, c/w epoxy, is currently shown to be 0.02786216"

The diameter of the centerline of the initial layer of thread is 0.220" + 0.0238" = 0.2438". This has a circumference of (22/7) * 0.2438 = 0.766"

The root end has 8 pins for wrapping the tow around, therefore the number of tows must be a multiple or a devisable of 8. The FORM: Torque Tube - Layer is used to work this out.

Not only is the diameter greater at the root but, perhaps, the thickness should also be slightly greater. This additional thickness means that a few of the windings must start at varing distances away from the tip and if they start at a 45º bias they will have a greater angle at the root than the full length wrappings.

Transferring some of the torque along the length of the blade via an elastomeric 'sleeve' decreases the amount of torque that must be transmitted all the way to the tip.

The following is just quick thinking. to be relocated into code or web page.

Calculations:

The outside diameter of the mandrel tip is 0.170" ID x .025" Wall Thickness = 0.22"

12K tow width c/w epoxy is 0.02786216"; from FORM: Tow

The diameter of the center of the layer of tow at the mandrel tip is 0.22" + 0.02786216" = 0.24786"

The circumference of the center of the layer of tow at the mandrel tip is PI * 0.24786" = 0.7790"

At a bias of 45º the advancement of a 360º band along the mandrel is 0.7790" (this is slightly too much due to the taper of the mandrel).

There are two tapes per layer and one band per layer.

This database has a profile for producing both the follow tables directly in html; for copying to this page.

FORM: Torque Tube - Mandrel:

RootDia	TipDia	Span	TaperDeg	TaperTan
1.0000	0.1250	117.0000	0.214181	0.003739

TABLE: Torque Tube - Layer:
The current coding in the related Form (and the following table) assumes that the taper is constant from the start to the end of each layer. It must be modified. The only layers which is tapered from start to end is the top (last) layer.

There may be fewer than 16 layers.

Layer	Tip Bias	Root Bias	Tows / Tape	Span	Start	Finish	Nbr of Wraps	Step Angle	Quill	Rotation	Tow Quantity
1	54.9811	85.8065	4	116.774	0		500	30	Yes	Yes	602.5838
2	0	0	0	0	0	0	0	0	No	No	0
3	0	0	0	0	0	0	0	0	No	No	0
4	0	0	0	0	0	0	0	0	No	No	0
5	0	0	0	0	0	0	0	0	No	No	0
6	0	0	0	0	0	0	0	0	No	No	0
7	0	0	0	0	0	0	0	0	No	No	0
8	0	0	0	0	0	0	0	0	No	No	0
9	0	0	0	0	0	0	0	0	No	No	0
10	0	0	0	0	0	0	0	0	No	No	0
11	0	0	0	0	0	0	0	0	No	No	0
12	0	0	0	0	0	0	0	0	No	No	0
13	0	0	0	0	0	0	0	0	No	No	0
14	64.2602	81.6184	8	88.9350	0	0	188	30	Yes	No	542.7169

Related Patents:

	US 4,863,416	Misalignment accommodating composite shaft. [See also ~ FWCS, Page 12 - 18] ~ Have hard copy
	US 4,381,960	Method of manufacturing a filament wound article. [See also ~ FWCS, Page 12 - 19]
	US 4,118,262	Longitudinal load carrying method for fiber reinforced filament wound structures. [See also ~ FWCS, Figure 12 - 35]

Related Pages ~ This Site:

	DESIGN: UniCopter ~ Rotor - Blade - NACA 00xx - Tip End Layout
	DESIGN: UniCopter ~ Rotor - Blade - NACA 00xx - Root End Layout ?? To come ??
	MAKE: UniCopter ~ Rotor - Blade - NACA 00xx - Torque Tube
	MAKE: UniCopter ~ Special Tool - Rotor - Blade - Torque Tube - Mandrel
	OTHER: Composite - CNC Work Station - Generator (winding to g-code)
	OTHER: Composite - Special Tool - Thread Winder

Initially displayed: June, 2004 ~ Last Revised: January 12, 2007

The above utility invention is openly and publicly disclosed on the Internet to negate an entity from patenting it, to the exclusion of all others whom may wish to use it. ~ Reference patent law 35 U.S.C. 102 A person shall be entitled to a patent unless - (a) the invention was known ... by others in this country, ..., before the invention thereof by the applicant for patent.

Old previous stuff is located below. It will eventually be tossed out.

Earlier Mockup Picture of Tip End:

Concerns:

What do shorter outer layers wrap around before heading back to the root?.

Sketch of Root End:
This is probably obsolete. See concern below. The new picture and method appear to eliminate the problems mentioned in this picture

'Button' should read 'Bobbin'
Concerns re Sketch:

The buttons sticks up above the overall diameter and will interfere with the bearing.
The wrap 180º later looks like it will interfere with the button.

Notes re Sketch:

The objective of the following is to allow the CNC filament winder, for torque tube, to only use the X-axis traverse and A-axis rotation.
A single axle passes through the mandrel at the root end.
A thin deep Bobbin is slid over each end of the axle.
All the threads from one tape of one layer will pass around one bobbin. All the threads from the other tape in the band will pass around the other bobbin.
The final result will be two 'stacks' of bobbins, one for each layer.
The start of the original tow will be at the root end. This is so that all layers will start at the root end.
Description of operation;

The mandrel rotates (CW) and the tow dispenser traverses (left to right) until the single tow has reached the tip end. The tow will be wraped over the angular tang and the back to the manderal to be wound (CCW) back to the root. At the root, the tow will be wrapped around the bobbin.
This method, for better (or worse) means that there is no 'weaving' of the tows. I.e. the tows in one layer all wrap in the same direction.

Winding:
This is applicable to the now obsolete stuff below the end of the standard page. There is an updated version of this above.

The sketch, which is related to the winding program in Access CNC Winding Generator Database is on Web page; OTHER: Composite - CNC Work Station - Generator (winding to g-code)

The purchased carbon tube over the cable has an OD of 0.170" ID + (2 x 0.025" wall thickness = 0.220"

If 12K yarn is used then each tow is 1"/42 = 0.0238" wide by 0.0238" thick. The epoxy may add to this dimension. For more information see; OTHER: Composite - General - Tow Or better yet see Access database FORM: Tow where the tow width, c/w epoxy, is currently shown to be 0.02786216"

The diameter of the centerline of the initial layer of thread is 0.220" + 0.0238" = 0.2438". This has a circumference of (22/7) * 0.2438 = 0.766"

If the tow is wound at a bias of 45º at the tip then the band width is sin(45) * 0.766"= 0.542".

If 12K yarn is used, the number of strands in the initial band width is 0.542" / 0.0238" = 22.8. This means that 11 tow will go to one bar and 11 tow to the other bar.

If the OD of the torque tube at the tip end is 0.5" and the purchased carbon tube is 0.220" then the wall thickness at this end is (0.5 - 0.22) / 2 = 0.14" The number of wraps is therefore 0.14 / 0.0238 = 5.88. The actual number will probably be 6 since the number of threads in each band is slightly under the requirement. How much area will the epoxy take up?

Assuming that the skin consists of two layers the 0.5" diameter location of the torque tube (tip end with only quills protruding closer to the tip.) will be located at r = 0.90 approx. See drawing on DESIGN: UniCopter ~ Rotor - Blade - NACA 00xx - Planform & Thrust

At the root end the diameter of the mandrel is 1.8124".

At the root end the diameter of the centerline of the initial layer of thread is 1.8124" + 0.0238" = 1.8362". This has a circumference of (22/7) * 1.8362 = 5.7709"

Band width divided by circumference is 0.542 / 5.7709 = 0.0939, therefore the bias (band angle) angle = 5.388º

Not only is the diameter greater at the root but the thickness should also be slightly thicker. This additional thickness means that a few of the windings must start at varing distances away from the tip and if they start at a 45º bias they will have a greater angle at the root than the full length wrappings. Transferring some of the torque along the length of the blade also decreases the amount of torque that must be transmitted all the way to the tip.

The following is just quick thinking. to be relocated into code or web page

Calculations:

The outside diameter of the mandrel tip is 0.170" ID x .025" Wall Thickness = 0.22"

12K tow width c/w epoxy is 0.02786216" from Access

The diameter of the center of the layer of tow at the mandrel tip is 0.22" + 0.02786216" = 0.24786"

The circumference of the center of the layer of tow at the mandrel tip is PI * 0.24786" = 0.7790"

At a bias of 45º the advancement of a 360º band along the mandrel is 0.7790" (this is slightly too much due to the taper of the mandrel).

There are two tapes per layer.

The width of the band is 0.7790" * Sin(45) = 0.550836"

The width of the tape is 0.550836" / 2 = 0.275418"

The hypothetical number of tows in a tape is 0.275418" / 0.02786216" = 9.8851

The actual number of tows in a tape is 9.8851 rounded down to 9. Down due to a flattening of the tow during the winding).

___________________________

This is the first recursive of the above, just to see what the difference is :

The tangent of the taper of the mandrel is 0.00585902777777778

Since the deltaX of one 360º wrap along the mandrel is 0.7790" The mean diameter of the mandrel under this band is 0.22" + (0.00585902777777778 * 0.7790" * 2) = 0.22" + 0.00913" = 0.22913"

12K tow width c/w epoxy is 0.02786216" from Access

The diameter of the center of the layer of tow at the mandrel tip is 0.22913" + 0.02786216" = 0.25699"

The circumference of the center of the layer of tow at the mandrel tip is PI * 0.25699" = 0.8077"

At a bias of 45º the advancement of a 360º wrap along the mandrel is 0.8077" There are two tapes per layer.

The width of the band is 0.8077" * Sin(45) = 0.57113"

The width of the tape is 0.57113" / 2 = 0.28556"

The hypothetical number of tows in a tape is 0.28556" / 0.02786216" = 10.24921

The actual number of tows in a tape is 10.24921rounded down to 10. Down due to a flattening of the tow during the winding). Interesting difference.