Hi Chris

It would be wonderful to test some of the double-sided tapes as well, like the 3M9485PC which is widely used in the yachting world.

In fact, would you care to contact me (email below) about this work - it could become a very interesting full technical article.

Cheers
roger@backpackinglight.com

Edited by rcaffin on 07/15/2010 16:11:01 MDT.

"It would be wonderful to test some of the double-sided tapes as well, like the 3M9485PC which is widely used in the yachting world."

+1

It would be interesting to test some of the “heavier” cuben versions (i.e. CT2K.08 and CT5K.18). Also, the test could be done applying the forces in several directions, not just along the warp.

In addition to a full technical article, the results of these tests should be published in the BPL Wiki (as Derek Goffin did). Otherwise, in a few days/weeks they get buried in old threads.

Cheers

I'm experimenting with various test speeds and load cell settings in an effort to try to get results closer to what Cubic Tech claims. I just got a result of 25.4 pounds force from a one inch strip which is 73% of the manufacturer's claim.

Hi Chris,

I wonder if you would get closer to CubicTech's numbers
if you removed the effects of micro-tears at the edges
by taking the 1 inch strip and folding/gluing the edges over.

Al

Hmm, good idea. We'll know in about 15 minutes.

I just ran 3 6"x1" samples cut 90 degrees to my initial population because I noticed that the spectra fibers appear to be more dense in one direction that the other. These samples averaged 28.7 lbf which is 82.1% of Cubic Tech's claim. I'll try folding the edges as well...

if the edges are folded then wont the force be on a double layer and not on the single layer? I am not technical like you guys so i could be way off, but it seems to me 2 layers is stronger than one so if we're trying to find the strength of 1 layer folding it to a double layer even on the sides will distort the results won't it? I am very open to being wrong, that is just what i am thinking.

-Tim

So folding the 1" samples in thirds made it WAY easier to load it in the jaws perfectly vertical and evenly tensioned. I just got a result of 36.8 pounds force which exceeds Cubic Tech's claim. Now to run three duplicates to confirm that this isn't a spurious result...

Tim,

I'm fairly certain that folding won't change anything. The amount of material available to support the load remains unchanged so each spectra fiber shares the load equally. What folding did just do was evenly tension the entire piece simultaneously. Without folding no matter how much I fiddled with the sample there was always one edge which became taught before the other. That or there would be ripples which tells me that things aren't being tensioned uniformly despite my efforts to the contrary.

"I am not technical like you guys so i could be way off"

You may not be technical but you make bad ass quilts!

> So folding the 1" samples in thirds made it WAY easier to load it in the jaws perfectly vertical and evenly tensioned. I just got a result of 36.8 pounds force which exceeds Cubic Tech's claim.

Hi Chris,
Thanks for confirming my guess that Cubic Tech's strength results were without edge effects. Mylar is prone to catastrophic failure if a small (even micro) tear gets started. What is so great about your results is that it shows folks doing MYOG with Cuben how strength can be compromised without careful consideration of where tears can start.

So I tested a 6" peel adhesion sample this morning and the 2911 is very poor in peel adhesion. The maximum force I recorded was 1 pound force which to me is totally unacceptable.

Peel Adhesion Setup


It also appears that the solvent in the 2911 attacks the mylar film, the spectra fibers, or both. I noticed yesterday while fooling around with minimum bond overlap that the samples overwhelmingly tended to break on either side of the bond area. This was my first hint that the solvent was weakening the fabric. To confirm my suspicions I made a small population of controls and a small population that got a very thin coat of 2911. The 2911-exposed samples broke at 20 pounds force where the controls broke more in the 28-29 pound force range.

I've ordered some cuben tape from Quest Outfitters and I'll perform the same battery of tests with it. The only good things I can say about the 2911 are it has a nice long open working time (20 minutes or so), cures completely in 4 hours, has very respectable tensile shear adhesion, contributes virtually no weight (15 mg per inch @ 0.25" overlap). Down sides are a slight weakening of the fabric and miserable peel adhesion.

Edited by ChemE on 07/18/2010 07:44:09 MDT.

Great research. Interesting to see you've observed the adhesive weakening the material.

....

Edited by Steve_Evans on 07/20/2010 10:16:57 MDT.

I agree that folding a sample is an acceptable method; provided you don't change the total cross-section area of the test piece, you still have the same stress on the material. Provided, of course, the folded test piece remains uniformly loaded across the entire folded width.

As Al says, the fact that your initial tests showed early failure gives users an idea of the limitations of practical production techniques, and may demonstrate how to overcome them; for instance, ensuring all edges are folded and bonded to ensure no micro-tears are exposed to propagate catastrophic larger tears.

Good stuff.

> ensuring all edges are folded and bonded to ensure no micro-tears are exposed

Very good point!
Very different from silnylon.

Cheers

Removed!

Edited by Mountainfitter on 07/21/2010 20:22:12 MDT.

Thanks for the great information.

If I am reading you correctly what we have is a ranking of cuben seaming methods that can inform our design process.

The practical application of the results indicates that it is best to use a single piece of cuben rather than bond two pieces if possible. In second place is to use a Hysol bond with a strength loss of 45%. This is followed by well adhered taped seam with and without sewing and finally a purely sewn seam.

So the result is - use a Hysol glued seam but allow for the resultant loss of strength when specifying the fabric.

Subject to finding a better adhesive we have to live with these constraints.

“what we have is a ranking of cuben seaming methods that can inform our design process”

Different materials, different solutions. I think we have much to learn about cuben.

Perhaps we should look more seriously how sailmakers work?

Edited by marcoasn on 07/20/2010 06:52:29 MDT.

The glued bond sounds like it is not allowing transfer of
force equally along the bond in the direction of the force.
Similar consideration is needed in sewing fabric.

The direction of the seams, the length of stitch, the characteristics
of the thread, the number of rows of stitching, and the
stitch pattern, can also make a difference on seam strengths.

For example, using a stretchy thread could transfer more
force over a greater area, increasing seam strength.

Polyester thread has low stretch characteristics.