Teardrops n Tiny Travel Trailers

I'm starting this thread to document some foam tests I am in the progress of making. I doubt these are terribly unique, but I share them none-the-less. For me, this is mostly and exercise in getting acquainted with some of these materials and techniques (of which I have no previous experience) prior to building.

Currently I have 6 "boards" drying/curing. They are roughly 2.5 feet long by 8 inches wide. Each one has been prepared differently, and I intend to see at which point each fails under load. I will do this by securing a small portion (perhaps 1/4 or 1/3 of the total length) of the board to a table, and then pile weight on the free end (hanging in space) until they break.

The materials were mostly selected because I had them on hand.

A description of the boards follows (all boards are EPS foam)...

(A) 1" thick foam, medium weight cotton, attached with TBII.
(B) 1" thick foam, rip stop nylon attached with TBII
(C) 1" thick foam, plastic rodent screen (looks like plastic window screen, but probably not as nice) attached with gripper paint.
(D) 1" thick foam, metal rodent screen attached with gripper paint.
(E) Plastic rodent screen sandwiched between two sheets of 1/2" foam secured with Gorilla Glue
(F) Metal window screen sandwiched between two sheets of 1/2" foam secured with Gorilla Glue.

I expect to run the tests in the next week or two. I'd be happy to hear thoughts/concerns before I proceed.

Sounds great. A couple things to also test for: weight each piece to see which options are lightest. I'd also be interested in stiffness, how much force does it take to bend each piece. And finally dent resistance. Not sure how to run that test other than dropping stuff on each piece from different heights.

I like the idea of a stiffness test, which may be easier, and more useful than a load/failure test. I could place the board on a scale and press down on it, measuring the force required to deflect it some distance from horizontal.

Good idea about weighing them. I'm not sure I've done the preparations fairly enough to draw strong conclusions. For example, the metal screen was a little creased, and I gobbed lots of gripper on in a few places to get it to stay down, and then added staples. Also, I decided I wasn't very interested in that method, and kind of short-changed it, rushing the work more than I did for other methods (fabric, plastic screen). The results will be more anecdote than science at this point.

As the foam provides the compressive strength, and I'm using the same foam, I was mostly operating with the idea of deciding what material provides the best bond/tensile strength combination. Consequently, I haven't given much thought to something like measuring dents. Perhaps a puncture test, dropping a dull knife like a dart would give some indication.

I'm interested in any observations you want to chart.
All that information will be useful.

2011 foam dent test …

Looking forward to the results.

Intro

Here are the results of my foam tests. They're mostly arbitrary and not well controlled, but I learned a few things.

Recall.... Below is a list of the different samples. Since my first post, I changed only (E), which I decided to cover with a cotton sheet. The reasoning for this change goes like this: I Since I have no access to 2" foam, I was curious if making 2" foam out of 1" foam sheets glued together would be stronger if you put a ply of material between them. Since my tests are 1" foam, I did this with a sandwich of 1/2" foam sheets around plastic screen. Obviously, in order to make a fair comparison, it needed to be covered on one side just like the 1" foam.

(A) 1" thick foam, medium weight cotton, attached with TBII.
(B) 1" thick foam, rip stop nylon attached with TBII
(C) 1" thick foam, plastic rodent screen (looks like plastic window screen, but probably not as nice) attached with gripper paint.
(D) 1" thick foam, metal rodent screen attached with gripper paint.
(E) Cotton covered (one-side) and Plastic rodent screen sandwiched between two sheets of 1/2" foam secured with Gorilla Glue
(F) Metal window screen sandwiched between two sheets of 1/2" foam secured with Gorilla Glue.

Here is a picture of (E)...



Methods

Since I was testing with the same foam (ESP), this is mostly a test of tensile and bonding strength. I did three things. (1) I placed bricks on the boards. (2) I bent the boards by hand (3) I dropped a kitchen knife into the boards.

Here are some pictures to give you a sense of what I was doing...



Results

Well, almost nothing broke. (F) broke with 2 bricks. The rest all bent until the bricks slid off, and then I could force the board to the ground with by hand, no breaking. However, (E) could hold 3 bricks, while everything else could only hold 2 bricks (before bending so much that the bricks slid off). So it would seem that (E) was stiffer.

When bending by hand, it was apparent that (E) was stronger than the others, and it was almost impossible to tell the difference between A,B,C,D.

Stab tests were almost indistinguishable. The board with the metal screen (D) did slightly better, but I'm not sure that it was significant. It's likely that at this scale, the differences aren't big enough to be detectable.

When bending the boards opposite the tensile strength, the boards broke almost immediately (like ~10 degrees at most). Except (D), which I think didn't break. I'll have to go double check, but it definitely bent much further.

Conclusion

So nothing too new here. The fabric/coating on foam really is amazing. It turns out that I really couldn't find any difference in the tensile strength between metal screen, plastic screen, cotton, and ripstop nylon.

The ripstop nylon didn't shrink, and was well bonded. The cotton shrunk, and bowed the foam ever so slightly. I really didn't expect the nylon to bond well, but it did. I would be curious about the long term waterproof results of nylon vs cotton/canvas. However, it seems like a nice choice for interior shelves.

Ultimately (E) was the strongest (and stiffest). This was the board with the cotton on one side, and the plastic screen sandwiched between two 1/2" boards. It's unfortunate that I didn't test a 1" board with cotton on both sides, or a sandwich with cotton in it, because I haven't really determined if having the plastic inside a foam sandwich does any good if you have skin on both the outside and the inside.

The puncture tests were unproductive, but my daughter really liked dropping knives into the foam, so it was a success in that regard.

I guess my big takeaway is that anything with tensile strength will keep the foam from breaking. Obviously there are other important factors that went untested, such as waterproofing, and lifetime durability.

Now on to building a fort, so I can learn more before building my trailer! Thanks for listening.

Thanks for the testing !!!

I suspect the sandwich with skin outside and in the middle wasn't much stronger in bending than it would have been w/o the second unskinned layer of foam. Had you put the middle skin on the outside of the second foam ply, instead of in the middle, it would have been multiple times stiffer (I (capital 'eye') to the ^4 power, moment of inertia from static material analysis).

It's why the drive shaft in your car is hollow, it's why I-beams have flanges separated by skinny webs. Put the material doing the work farthest apart for the most bending resistance.

KCStudly wrote:I suspect the sandwich with skin outside and in the middle wasn't much stronger in bending than it would have been w/o the second unskinned layer of foam. Had you put the middle skin on the outside of the second foam ply, instead of in the middle, it would have been multiple times stiffer (I (capital 'eye') to the ^4 power, moment of inertia from static material analysis).

The sandwiched version was definitely stronger than the non-sandwich version. But I agree, it would have been even stiffer had it been on the opposite side of the board. However, I'm looking at having to sandwich foam due to supply issues, and I wanted to see if it would be worth it to put the screen in the middle.

Some things to consider are rotating the weave so that the middle layer's bias is orthogonal to the outside weaves (which optimally would have the warp weave parallel to the longest panel axis). This may add strength when the sheet experiences a twisting. Another is that the plastic has not just tensile strength, but compressive strength in the opposite direction. I believe that this is really where the extra strength/stiffness came from. Foam with fabric on the inside of a bend (concave side) is no stronger than foam without fabric, while plastic screen will give some strength.

Can't locate an insulation supplier near you?

Around here the 2" xps isn't in the big box stores, but the insulation suppliers, roofing suppliers, contractors have the stuff.

I'm not sure if the eps will ever be strong enough, compared to the xps.

The FoamStream is EPS… And as we mentioned so many times , when in doubt make it THICKER … mine is 3.5” thick ... EPS , isn’t as easy to work with ( all those beads ) , and doesn’t sand as well … but it’s cheap and Light and seems available in many places … and it also heals itself in the Sun.

Even in warping the strength comes from the outer most 'fibers' (be they grains of foam, grains of metal, strands of cotton, strands of plastic, strands of glass, etc.). Whether a uniform material or composite, the applied stresses in bending or torque are greatest at the fibers that are the farthest away from the center of mass of a given cross section. In bending (or torque) one "skin" (or boundary layer of fibers) will see the most compression and the other will see the most tension; on a moment diagram, in the middle of the section (almost always dead center) the forces pass through zero. So adding strength to the middle of the sandwich is wasteful.

There are lots of examples of the math theory behind this (Google centroid moment of inertia), but they tend to get muddy with calculus and vector analysis. The bottom line is that moment of inertia, the ability or efficiency of a given cross section to resist bending, is expressed as a unit to the forth power (I^4) based on how the material is distributed around its center, so the further the outermost fibers are, the bending resistance goes up exponentially. If the fibers are stronger (tension will govern here), the bending resistance also goes up.

So long as the compression strength of the foam can hold those outer fibers out, and the strands can stand the tensile load, the panel will only bend to the acceptable limits of yield. This is where the juggling match of engineering lives; picking materials and methods that achieve the design goal, while remaining practical to build within the restraints of cost and time management.

By putting the strong fibers in the middle, you are wasting time, material and money. Just saying.

KCStudly wrote:By putting the strong fibers in the middle, you are wasting time, material and money. Just saying.

KCStudly wrote:...So adding strength to the middle of the sandwich is wasteful....

I've only quoted the above, but I appreciate all the points you have made, and think it's good that they are here for others to read. I certainly don't advocate going through the trouble of glueing panels together in order to have a sandwich of material. Hopefully your informed knowledge about the matter will keep people from wasting their time and money.

However, I will again mention that while my tests revealed that there was no real detectable difference to me between skins of nylon, cotton, plastic mesh or wire screen, there was a very real increase in the strength of the board with the plastic mesh in the middle. Your real world arguments make sense (although you seemed to imply that a hollow driveshaft is stronger, but really isn't it just that it has a better strength to weight ratio, it is in fact weaker?). Here is how I think about this, and perhaps I'm far afield in doing this, but: The panels are layers which when bent, change length. If the length becomes longer (outer edge of bend), tensile strength comes into play, if the length becomes shorter (concave side), compressive strength is required. So, if we're dealing with a panel, that is bent in an textbook physics way, then the middle layer's length doesn't change, and it adds nothing to the strength of the board. My experience is that this middle layer adds strength, and so I am curious as to why. Wouldn't one expect changes in length at places within mesh layer due torsion? Or perhaps the change in length of these imaginary layers (as one moves from the outer-edge to the inner-edge) is non-linear and so this imaginary layer with no change in length is offset from the physical center (where the mesh layer is located). I suggest the non-linear effect after thinking about how the shear strength of this material is not large (as it is in the example you site of metal i-beams, etc) and thus am reminded of the non-linear regions of Mohr's circle. While a slight offset from center due to this non-linearity shouldn't provide a ton of strength, it would add some, and perhaps a noticeable amount if it's adding to the tensile strength in the direction of the outer skin's bias direction.

It's of course possible that I've mislead myself by poor testing. That would mean that I made the 5 non-sandwich test boards weaker than they were supposed to be, and the one mesh sandwich board stronger through some error in how I put things together. I did make a sandwiched board with no mesh in it, and it was noticeably weaker than the one with mesh as well.

Again, I won't be going out of my way to put material in the middle of my foam boards (that could be diminishing returns indeed), but since I have foam boards (leftovers) that I will be glueing together anyway, I will definitely be putting in some mesh after my experience. I think these larger panels experience multiple torques (just look at what happens to them when you move them around!) that in some way engages the strength (both tensional and compressive in this case) of the plastic mesh. I'm only guided by intuition and empirical results here, which obviously can be wrong.