David Thatcher wrote:

So throw it all together guys and go sailing!!

Even though I am probably the slowest boatbuilder in the world, I will drink to that. 

I think one of the great things about reading books by people sailing yachts 100 years ago is to realise how unreliable they were by today's standards, and how rarely the writers got themselves into serious trouble.  I.e. died.  It really puts all our fretting about GRIB files, EPIRBs and all the other initials into perspective.  And our cheapest building goop is probably 100 times better than any glue they could lay hands on!

I think Jami has asked very good questions. With respect, there has not been much science in it, just a few crude DIY attempts at a bit of measurement and an attempt to understand something about the materials. And not agonising at all – I’d call it just another variety of “messing about in junk rigged boats” which, as our friend Ratty famously stated, “there is nothing, absolutely nothing half so much worth doing…”  I thought it was quite interesting, and I think I have learned something.

Still, David Th has made a good point, something I guess we all need to be reminded of.

Jami you are a bit ahead of me I would be interested to know how you get on making your fairing to smooth between the two diameters.

I made the final stage of one of the wood/epoxy "bandages" Coated the wood/epoxy bandage with one layer of glass and epoxy, wrapped in two layers of baking paper and slid it inside the main tube to cure - and put the heater on it. When it was nearly done, I was able to slide it out (whew!) success - and here it is, still covered with one layer of baking paper which will be sanded off tomorrow. Slight crinkling in a few places because it was a pretty tight fit. Its going to come up a near perfect fit, with very little sanding required. I'll do the other one tomorrow.

I wonder if there is too much science and agonising going on here, as seems to be the case in our modern world. What I mean by this is that our junk rigs are supposed to be very simple in the best DIY tradition. My understanding when you add an extension to an alloy tube is that the amount of bury is the secret to the two bits staying together. Annie's mast on Fantail has well and truly stood the test of time, and that was basically stuck together with some good quality goop, so that seems to be a good example of how to do things. I am also thinking about the battens on my own sail. The three alloy tube battens are 5 meters in length with a 1.5 meter cedar extension at the forward end to lengthen the battens. These were put together with a minimal amount of bury and the timber and alloy stuck together with some good quality builders adhesive sealant, (NZ $20 a tube). I do remember I cleaned up the inside of the alloy tube before sticking them together. But those battens have stood up to thousands of miles of sailing, and numerous heavy gybes, and so far they have not come apart. Remember also that in my camber panel sail there is a lot of loading on the battens. The timber extensions on the carbon tube battens are stuck together the same way.

While researching glues the other day I was perusing the website of a NZ and Australian adhesive supplier. Their recommendation for gluing alloy window frames to timber was some type of goop in a tube.

So throw it all together guys and go sailing!!

Gary, the photo in the last post was the glue join with no surface preparation. The epoxy is just what I have to hand - West System as it happens.

Thanks, looks like you are the winner by default.

Here is a photo of the aluminium tab holding 3 batteries (41kg) tail stock, transducer, transformer (total 36 kg) dumbbell, steel bar, lead weight (20 kg) lead ingots (36kg) large lead ingots (42 kg) that’s a total of about 175 kg on 4 sq cm gluing surface.

After that it broke. Here is a photo of the broken glue join. It shows adhesive failure, as opposed to cohesive failure. I did not use fine, wet sand paper – just a bit of old sandpaper I found on the workshop floor, I did not spend too much time cleaning up the joint. The surfaces were primed with epoxy (just West System, by the way) and I did not wait for it to fully cure, then dabbed some glue on one surface and laid the two surfaces together. I think the result is about 4MPa (is that right? I get mixed up with powers of ten in that wretched metric system and its 55 years since I looked at any of this.) 

Anyway it is far in excess of what I expected, and if it proves nothing else, it proves that time spent cleaning and preparing the glue surfaces makes an enormous difference. 

Summary of a research paper (below) seems to suggest 4MPa is in about the expected ball park.

https://www.researchgate.net/figure/Fracture-surfaces-of-an-aluminium-joint-bonded-by-epoxy-with-2-wt-nanosilica-filler_fig2_269308028

(By the way, this refers to a join between aluminium and a composite material - might be of interest to Marcus who has a composite topmast to bond to an aluminium main tube.)

Another paper (further below) states that the standard test for epoxy strength is conducted using an Al-Al lap joint and that high strength products can have a shear strength greater than 2500 psi, which is over 17 MPa.

https://www.epoxysetinc.com/what-is-the-strength-of-epoxy/    

I’m casting one of the “bandages” inside the main tube today, separated by 2 layers of baking paper. Cross fingers I can get it out tomorrow!

Thanks for your interest.

PS also the last paper(above) also states that the compressive strength of epoxy (typically about 10,000 lb/sq in) is about twice that of high quality concrete! This suggests that the shoulder on a conical fairing will be more than adequate to stop a composite mast "becoming a telescope".

Phil, I wish I had thought of your clever idea. I had a 5" tapered flagpole which was long enough for a mast, but we decided it was a bit too light so I cut it in half and I am using the tapered top half for a topmast. Your idea would probably have worked for me. Oh well...

I built my two bandages almost to completion yesterday, using wood strips over a base of three layers of cloth and epoxy. Here. This morning they have cleaned up nicely and I have bogged the gaps. I think this will be a little easier to sand and fit than solid epoxy, and it has been very easy to do, so far. I presume, from your description, that you let the epoxy bandages cure, then made a dry fit – and finally glued the two parts together afterwards, is that right? In any case I think it will stick permanently and Jami can be reassured too. Read on:

Arne wrote:

“Even with a compression force of 200kp (I doubt it will ever reach that on that Gallion),  the glue joint will hardly see more load than 0.2-0.4kp/cm². Now, glue two bits of aluminium together with a contact surface of about 1-2 cm². and pull them apart. You will find that it takes several kilos force.”

I tried that Arne, at the same time as investigating or comparing a couple of release agents.

I made seven aluminium tabs, each with a glue surface of 4sq cm, and glued them all to a strip of aluminium angle, like this. A liberal amount of glue was applied, to one gluing surface only (the aluminium tab) except in the case of (7). Each of the glue-coated tabs was then laid on the aluminium strip and a small length of timber placed over them, in order to provide a very light amount of pressure, and they were left to cure.

(1) CRC had been applied to the aluminium strip

(2) Baking paper sprayed with CRC had been applied to the aluminium strip

(3) Spray-on release agent had been sprayed on the aluminium strip

(4) Baking paper sprayed with release agent had been applied to the aluminium strip

(5) Baking paper on its own had been applied to the aluminium strip

(6) The aluminium tab (with glue) was laid on the aluminium strip, with no preparation to either of the gluing surfaces.

(7) The two gluing surfaces were carefully cleaned, sanded and primed with epoxy resin. When the resin was half cured, glue was applied to the aluminium tab and it was laid on the aluminium strip

 

After curing, (2) (4) (5) simply fell off. (3) required a slight bump before falling off. Each of the tabs took the glue with it. Neither CRC nor the spray-on parting agent prevented the baking paper from being glued to the tab.

(1) was able to withstand a light tug but was probably not stuck very well – I did not get to test it because it came off at some stage, I did not notice when, and it has been lost.

Conclusion: Proper spray-on parting agent is probably better than CRC but possibly not much better. Baking paper is a total success as a parting membrane. Epoxy will stick to it – but unlike printer paper (which I have tried and failed) it seems not to allow the epoxy to soak through and reach the other surface, hence it comes away cleanly taking the glue with it. See photograph of these results here.

Adhesion of epoxy to aluminium.

I hung various weights from tabs (6) and (7) with a view to testing to destruction.

(7) whose surfaces had been prepared and primed, has not yet failed. Here is a photo of it holding up three car batteries, the tailstock of a large lathe,  an anvil and a dumbbell. It was way off the scale (my scales measure up to 50 kg only – that’s slightly more than 3 car batteries). The total weight here is estimated at 75kg or more. Look at the bend in the 40mm x 20mm strip of angle. I had no more weights so I left it stuck on and turned to (6).

(6) which was the simple glue joint and no preparation, held two car batteries and when I added the anvil it held for a moment, then with a bang it let go and something went past my left eyebrow with enough velocity that I decided testing would be over for the day. (It was probably tab (1), hurled off when (6) failed and the aluminium angle strip unsprung.)

I think the glue joint (6) was holding about 40kg when it yielded.

This photograph shows the two failed glued surfaces – evidently half the glue stuck to one surface, and the other half of the glue had stuck to both surfaces. It think if glue had been applied properly to both surfaces, it might have held more.

I went back and did some further testing on (7).

Would anybody like to guess the total static weight this join withheld?

Come on, experts- if anyone is prepared to try a guess, I’ll give you an answer tomorrow.