Special tooling conception to remove an aluminium mandrel, seeking advices & ideas

[Erwan]

Hi Everybody,
In order to make a unique seamless carbone tube, The cost effective solution (autoclave compatible) could be to wrap UD around a 90 mm diameter aluminium tube with 2mm thickness and 19 feet long. The "mandrel"

Both ends of this aluminium "mandrel" will be morphed into an elliptical shape, so impossible to remove the mandrel unless you can cut it in 4 parts along the longitunal axis, by the inside.

So I am thinking of making a micro circular saw with a old Dremel electric motor, or equivalent.
The max wide should be under 54 mm and max height 86 mm.

Any idea for this circular saw, or alternative process to cut the aluminium from the inside as above-mentionned (may be electro-erosion?? ) I don't know that is why all suggestions are welcome

Fair winds

EK

 

[waikikin]

Maybe you could first make a carbon tube then split it off the aluminium then continue your carbon layup without the alu inside. Or make the alu tube a female mold and expand the carbon laminate within using a bladder. Or just use the alu tube for the application... whatever that might be & apply some carbon styled wrap ?
Jeff

 

[Deering]

Both ends of this aluminium "mandrel" will be morphed into an elliptical shape, so impossible to remove the mandrel

Can you describe this shape further? I’m having difficulty picturing it. A sketch would be helpful.

 

[wet feet]

Maybe you could first make a carbon tube then split it off the aluminium then continue your carbon layup without the alu inside. Or make the alu tube a female mold and expand the carbon laminate within using a bladder. Or just use the alu tube for the application... whatever that might be & apply some carbon styled wrap ?
Jeff

This is by far the most logical approach.You don't even need a bladder-a mandrel covered by a tubular vacuum bag will work and can be sealed onto the end flanges of the female mould.The mandrel can be foam and needn't weigh much.

 

[KeithO]

Is it not feasible to consider making the special tube in 2 halves ? if the halves are mirrored, then you only need 1 tool. If you can add a draft angle to the circular section, it will release more easily, or else make the mandrel in 2 pieces so you can split the eliptical section from the round section, then give each the appropriate draft angle, this would increase the diameter of the tube towards the center where it probably needs additional cross section anyway. Then taper the lamination at the edges to be bonded (can be ground after laminating) and lay up the bonding laminate at the center, joining the 2 sections together.

 

[upchurchmr]

Are you going to wrap the uni's over the elliptical ends of the mandrel?
Why?
If not there is no issue.
Cure the tube under heat.
The aluminum expands, carbon does not with temperature.
When you cool the cured tube/ mandrel, the mandrel shrinks.
A little force to break the tube free and the mandrel just pulls out (assuming it is straight).
Or actually cool the mandrel to get more shrinkage.

What are you actually trying to do?

 

[wet feet]

I didn't give any consideration to the ends being wrapped as that would mean you couldn't possibly extract the mandrel.It seemed easier to believe that if closed ends were needed it would be simplest to bond some end caps in place after making the tube.The OP also says that both ends will be elliptical,should we assume symmetry about the mid point?

 

[Erwan]

Thank you All for taking time to write.

This tube will have to stand different kind of loads: , Flexion ,Compression (bukling), and significant Torsion.
So to address torsion, wraping 45° Carbon UD seems to be appropriate, and that is why seamless is much better, so make the tube in 2 halves is not optimum, and reliable bonding of the 2 halves can be challenging too.
The point is a 20 feet long aluminium tube cost +// 100$,
and autoclave curing can add to the good mechanical property of the tube, as long as it is cured on an aluminium mandrel which expands under heat, creating some "good" tensions in the carbon UD.

That is why the aluminium tube solves some critical issues, but brings some new ones.

The shape of the tube: To make it simple the aluminium tube is 20 feet long, 90 mm diameter and 2 mm thickness.
The section will remain circular in the middle, and both ends will be "shaped" gradually along a 3 feet lenght from the 90 mm circular section to a 120mm x 60 mm elliptical section at each tip.

The alternative could be to have a tapered tube both sides, but as the circumference will change it will be necessary to cut the carbon UD accordingly.

Instead, if your section changes from circular to elliptical, as long as the circumference remains constant, you can wrap UD at 45° easily. Like a "developable area".
The 60 mm elliptical small section at both tips improves the aerodynamics, and more important, provides a better base to attach torsion arms than a circular tip.

Please find this link to get more details about this project.

Beam calculator Free software ? https://www.boatdesign.net/threads/beam-calculator-free-software.56244/

Again, thank you very much to all of you for reading and writing.

Cheers & Fair winds

EK

 

[Deering]

Why not just leave the aluminum tube inside? It only adds about 20 lbs to the piece.

Looking at your sketches it seems like you’re making the ends unnecessarily complicated. As I understand it, the elliptical shape is to prevent the spar from twisting loose from the arm as you rotate it. A simpler and stronger solution may be to epoxy the spar to the arm with fiberglass reinforcements. Or if you want it removable, form a flange on the end using epoxy/glass that can be bolted to the arm. The large load you’re concerned about is deflecting the tube. The twisting load appears to be fairly moderate.

By eliminating the ellipses, you can make the tube circular the full length which may make it possible to extract the aluminum tube if it was lubricated and wrapped with plastic sheeting before laminating. You may have to remove it prior to post cure depending on the temperature. Extend the aluminum beyond the end of the composite tube and you should be able to twist it free. Perhaps do a short trial run first.

 

[Erwan]

Thanks Deering,

In fact the target weight is around 4 kg, so letting the alu tube inside is not an option.

I should not have written:
The 60 mm elliptical small section at both tips improves the aerodynamics, and more important, provides a better base to attach torsion arms than a circular tip.

Instead I should have written:

The elliptical tips are necessary for aerodynamic reasons, and as a side effect it makes it easier to attach a twist arm at each end.

But it is not necessary to go below 60mm width at the tips, and the 90 mm diameter is what "was" required with 2007 technology (prepeg+alu mandrel+autoclave) to get the stiffness.

I just get a new information which could be a "problem changer" for my tube:

Rafale II carbon/epoxy C-Class hydrofoil catamaran http://www.jeccomposites.com/knowledge/international-composites-news/rafale-ii-carbonepoxy-c-class-hydrofoil-catamaran?utm_source=SalesForceMarketingCloud&utm_medium=email&utm_campaign=JEC+Composites+Market+News+N.+812

In this above-mentionned link, the carbon tube used for this big wing is 120 mm diameter, while 10 years ago for the same kind of catamaran, the carbone tube was 200mm diameter.

So for my tiny wing, may be the 90 mm diameter is not required anymore with today's technology, I have to check.
But if it is possible to go as low as ie: 75 mm, it would make things much simpler.

Regards

EK

 

[Barry]

A couple of options
Form the shape from foam with a density of 2lb/cubic feet. This would add about 2 1/2 pounds to the weight, cover it, you could then enclose the ends at the same time and leave the foam
in place

Alternatively
Form the inside mandrel from strips of cedar or soft wood. Make the block large enough at the ends to form the 120mm flare but round in the middle. Add some extra dimension to allow
for the kerf of 4 cuts with a bandsaw so when finished longitudinal cuts, you will still have the pre requisite 120 mm at the ends. I have provided a sketch. The plan is to, when sawn, wrap the strips together to get the shape back with the only difference being the slight discontinuity of fairness at the kerfs. Wrap the entire piece with stretch wrap to hold it together, clamp it along the way to make it straight. Take say 3 feet of wrapping off, sand the discontinuities, wrap. Remove another 3 feet of wrap, sand this area, then wrap, and continue it until you have the mandrel made up wrapped strips.
Carbon warp the piece. After curing, drive the inside center strips out from the middle. Drop the remaining strips into the void one after another until all the strips are gone. Enclose the ends as required

 

[upchurchmr]

Can you describe how an elliptical end helps put on a torsion arm?
I think that is causing a lot of "ideas" from people who have never laid up anything.

 

[Erwan]

Thanks for your suggestion Barry, foam shaping on a smaller aluminium tube was my first option, but autoclave cannot be used with foam, for the wood I don't know, may be the simplest and cost effective solution, I have to check. Thanks.

Upchurchmr, Just because it looks easier than pure bonding on a circular section. But the mini thickness (60 mm)at both ends is the goal.

After all your suggestions I sussed out what seems to me the simplest approach.

The last 3 feet of the top will change from 90mm circular section to 90mm x 60 mm elliptical section

The bottom section will be circular in order to extract the mandrel/alu tube, but the lenght will be 1 foot shorter, and the mast foot will be conical with 1 foot lenght + 20 cm to be plug into the main tube, a bit like a windsurf mast foot.

So the issues will be
1-to cut correctly the UD for the last 3 feet at the top
2-To bond the torsion arm on a circular section at the bottom.

If the last foot at the bottom is conical from 90 mm section to 60 mm it will be enough, so not necessary to have the main tube tapered both side.
the control arm for the torsion will be bonded 1 foot higher on the circular main tube, and if the mast foot is heavy, it is not an issue as it is low, it will have to address only flexion and compression, no torsion.
Thanks all for helping me to make the appropriate trade-offs,

Fair winds & Cheers
EK

 

[gonzo]

You should look at a manual on casting. You issue has been solved a long time ago. All that is needed is a core that is made in sections. There are several ways of doing it. One is to have a central section that one it slides out, lets the two sides move away from the part and can be removed. Another is to have two wedges that can be pulled out from the ends. As the move they also separate.

 

[upchurchmr]

Erwan,

Every daysailing catamaran I have ever seen has a control arm bolted on.
Simple, effective, and you can figure out the load of the bolts on the mast, relatively easily.

Bonding has the problem that you can see if a bond is formed, but there are times where you cannot determine what the strength of the bond is.
You can always do a test after bonding, but you will not know how much excess strength margin you have.