Water absorption by epoxy.

[Tonyr]

I was surprised to read in a couple of posts recently that epoxy was said not to be fully resistant to water, at least to the extent that a properly encapsulated wood component or hull will stay essentially dry over time. For what it is worth, I am running a test for my own satisfaction, and perhaps the results may be of some interest to others. Here are details.

The test reflects my building methods. I covered the exterior of a strip built 17' by 4'6" Whitehall hull (3/16" yellow birch) with a coarse fibreglass tape (drywall tape, no less!) and then with about 1/8" of epoxy/maple sawdust mixture. This was faired to shape, coated with a low viscosity epoxy, sanded again, painted with a standard Pratt and Lambert 2-part epoxy primer, then with antifouling (below the w/l) and with Brightside poly above. The interior of the boat was finished with conventional 4" fibreglass 6 oz tape set in epoxy (no sawdust), then with the epoxy primer and finally with exterior enamel from my local chain store. I used MAS epoxy throughout because of the absence of amine blush. The boat sat along side a dock all summer, received hard use daily as a rowboat, when sailing and with a small motor. Nothing broke!

My question is, how safe is my use of an epoxy/sawdust fairing mixture as the primary exterior encapsulation. To find out, I saved the remains of about 1/2" from one mix from the bottom of a plastic container (it looks like the reverse of the inside of the bottom - of course!), weighed it carefully with a good balance, and have just re-weighed it after a 10 day's soak in room temperature water.


Results.

Initial weight 68 grams.
Final weight 68.25 grams.


Essentially, no apparent water uptake!

So far, so good. I am now going to replace the artifact back in the water, and will weigh it again in a month's time.

Anyone care to comment on any of this?

Regards, Tony.

[Don Maurer]

You covered a 3/16" strip boat with 1/8" of fiberglass/epoxy? You have a plastic boat with a wood veneer. When we talk about epoxy not being completely waterproof, we are talking about a coating of epoxy over wood, not wood particles embedded in epoxy. Epoxy will absorb a few thousandths of an inch into wood, depending on its porosity. If you coat a piece of dimension lumber with epoxy and submerge it in water, over time, it will saturate. The amount of epoxy determines how fast it will occur.

[Bob Cleek]

Yea, what Don said. Your .25 absorption was just what the sawdust soaked up. Now, think about what the wood will soak up at a ratio of 3/16" wood to 1/8" epoxy. Epoxy is not waterproof. Never was, never meant to be. Water permeates it. Epoxy slows water down a bunch, and epoxy itself doesn't absorb water, far as I know, but until you displace all the wood with epoxy, your wood is going to soak up all the moisture it can until its equilibrium is reached. Just the way it is. If you don't want water, take up building something other than boats. LOL

[thechemist]

What Don and Bob said.

If you leave it immersed long enough, it WILL eventually saturate.

I assume you saw http://media5.hypernet.com/~dick/ubb/Forum1/HTML/003460.html

and perhaps http://media5.hypernet.com/~dick/ubb/Forum1/HTML/002241.html

and perhaps some other discussions of the subject.

I am glad that you have a scale accurate enough to see a hundredth of a gram. If you plot patiently the weight every month, you will see it just continue to go up. Take the amount of wood inside that 68 gram piece, however much it is, and assume it will go to the fiber-saturation point, maybe something above thirty percent, fairly linearly. Assume you started with fifteen percent moisture content in the wood. Expect it to gain, in a fairly linear manner, another fifteen percent of its own weight in water. Now we come to the air inside the hollow wood fibers. They are big molecules and will take longer to diffuse out than the "first wave" of water molecules diffusing in.......but eventually they will go, water will completely fill up the space, and your test piece will be fully waterlogged. It will just take time...maybe a year or two.

It is not that epoxy is not fully resistant to water.....no organic material is an impermeable barrier. Water gets through it all...some stuff just faster than others.

[Tonyr]

Thanks for your replies guys.

I'll keep the test going, and we'll find out what happens. If I thought anyone really knew, I wouldn't be doing such a test in the first place. And yes, I'm not a wood purist, even though I have been a professional woodworker these last 15 years. My Harry Vail is a composite wood/plastic boat, not a wooden boat; that's the point, isn't it? What I want to know is whether the design, construction, performance and maintenance freedoms provided are worth the costs (this is only partially a comment about money).

Regards, Tony.

[NormMessinger]

Does anyone float around in an epoxy/glass/wood composite boat with raw epoxy exposed to the elements? In that case one might be concerned about the claim that epoxy is not waterproof except UV would get to it first. Seems to me to be valid for boat applications the test pieces should be finished as they would be on your boat. But then paint isn't waterproof either.

--Norm

[Tonyr]

Norm - agreed, but I thought it made sense to test the worst case, hence no coating on the epoxy/wood-flour mix. Incidentally, I used sander dust from my dust extractor bags, not saw dust, which probably matters since the particle size is very much smaller.

Regards, Tony.

[bainbridgeisland]

I don’t think a wooden boat, fiberglassed with epoxy resin, will be saturated with water in two years. It takes a lot longer.

First of all, absorption of water does not happen at a steady rate. A couple of different mechanisms can drive absorption but if we assume a good fiberglassing job, one method predominates. Diffusion is like the gatekeeper for water absorption and is approximately described by Ficks law. Ficks law = The rate of diffusion across a plane is proportional to the negative of the rate of change of concentration. I know, I know, sorry about the math. It means the rate of absorption is asymptotic. Also, the rate of absorption is related to how much water has already been absorbed.

It is even more complicated because we have two materials with different diffusion rates. It takes computer time to solve the differential equations for this one so I’m not going to try it. But here is what happens in words. The epoxy fiberglass is like a membrane where water can weep through very slowly. Assuming that the wood was dry to begin with, it’s like a vacuum cleaner, trying to suck the water through the membrane. But the vacuum pressure gets less and less as water accumulates. Now in reality, each of these would be approximated by Ficks law, but this will do for now.

You can see that initially a lot of water would come through but the diffusion rate would slow as time passed. How much time? It depends on the diffusivity of the materials (one value for fiberglass and another for wood). The diffusivity for epoxy/fiberglass ranges from about 1*10^-8 to 10*10^-8 mm^2/second. Using Ficks law, this means it would take 1 to 2 years for 1 mm thick epoxy / fiberglass to become saturated. How much water to saturate? About 7% of the epoxy weight or 3.5% of the total laminate weight will be gained by most epoxies. But we have barely started transporting moisture into the wood at this point.

The diffusivity of wood is much higher than epoxy/fiberglass. Ficks law will still work but it is not as accurate. Initially the concentration gradient goes from dry wood, 0% concentration to the wet epoxy/fiberglass, 3.5% concentration. Even though this is the highest gradient you can see, it is not a lot. Thus, absorption by the wood is relatively slow and will get a lot slower as the moisture content approaches 3.5%, when the gradient gets near zero. Remember, absorption is asymptotic, a lot at first a little later.

Is this the whole picture? Well, no. Capillary action in the wood must be accounted for and so forth. These too can be calculated using Ficks law, if you know the Diffusivity constant relating to capillary action. I don’t know the diffusivity constant for woods and besides it would be different for each direction making the problem even harder. “Flaws” in the wood will continue to collect water and that water will then diffuse into other areas of the wood.

How long to be fully saturated? I don’t know exactly but I am sure it would be longer than 2 years. I have cut into 20-year old cold-molded boats below the waterline. The wood was not as dry as a new boat. I am sure it had less than 12% moisture though. I don’t think I could judge between 0% and 12% very well anyway. I built the boat and can remember what it was like when new.

[Tonyr]

Bainbridge.

Most interesting and informative. Your numbers are much more what I had generally thought until I read the disconcerting posts implying that water would diffuse more or less rapidly and linearly through epoxy barriers over wood. It seems to me that all the material tests have to be put in the context of a seasonally wet boat (at least in this area of Canada). If a boat is wet for (say) six months, and is put under cover for the off season (suspended from the garage roof beams in my case), then all we are concerned about is the absolute absorption level reached before haulout, less the over-winter drying. That should reach some sort of approximate constant within a year or two.

Here's another experiment which might give us a feel for what is happening. If I weighed the boat carefully next spring, and at haulout next fall, that would tell us if the boat as a whole is really absorbing enough water to matter. Assuming that no wood gets wet enough to degrade in any way, then if the boat only absorbs a few pounds total over a season and loses most (all??) of it over winter, the epoxy will have done its job.

What do you think?

Regards, Tony.

[thechemist]

Tony....depending on the accuracy of your scale, that is really an excellent idea. It measures what is going on in the real world, in a way that few if any laboratory experiments ever could.

[Tonyr]

Chemist, thanks for your encouragement! All I have to do now is work out some sort of Heath Robinson contraption to weigh the boat. I do have some preliminary ideas, but they are probably pretty silly. How do other people do this? Advice welcomed.

Regards, Tony.

[Nicholas Carey]

Originally posted by bainbridgeisland:
How long to be fully saturated? I don’t know exactly but I am sure it would be longer than 2 years. I have cut into 20-year old cold-molded boats below the waterline. The wood was not as dry as a new boat. I am sure it had less than 12% moisture though. I don’t think I could judge between 0% and 12% very well anyway. I built the boat and can remember what it was like when new.

12%? That's about the EMC (Equilibrium Moisture Content) of wood in air (70 deg. Fahrenheit, 65% relative humidity). So that's probably about par for my brother's Adirondack chair. It spends every summer sitting on a porch in southern Ohio. Not exactly a hot-spot for rot, that.

That just confirms my suspicions that all this "What! Epoxy not waterproof?!" stuff is perhaps a bit compulsive.

[thechemist]

I use a Wagner model L606 moisture meter. I never have to guess. I just put the meter on the wood. I can put some amount of epoxy or polyester layup in between the meter and the wood and thus get a correction factor for the moisture content of wood below laminate. I can put a 1/32 -inch-thick piece of teflon between the meter and the substrate and get excessively wet wood or[or concrete!] back on scale so I can get a notion of how bad it might be when the meter reads over thirty percent water in wood.

[bainbridgeisland]

Tony

When you alternate periods of wet and dry, the surface moisture content fluctuates with exposure. The outer surface responds almost immediately, the inner portions more slowly. A kind of equilibrium develops in the middle portion of the material. It doesn’t change measurably with exposure to wet or dry. i.e. the moisture variation is predominately on the surface.

The total amount of weight variation depends on surface area, exposure time, temperature, diffusivity of the material and most importantly, the moisture gradient within the material.

I like your idea about weighing your boat regularly. It is a practical solution to the question of how much water is absorbed. It will give you a good idea of the moisture your particular boat picks up. Let us know what happens.

[bainbridgeisland]

Chemist

The question is, do you always measure 30% moisture content in epoxy-covered boats that are kept in the water year around?

I do not have a moisture meter. However my experience working on both modern, epoxy covered wood hulls as well as traditional carvel hulls is that the epoxy covered boats are much much drier.

[ford]

all this is fascinating, I am in the process of cold molding a carvel boat. My interest is whether to epoxy the interior and if so how much? "Curlew" was left plain but "Angelita" was epoxied, to my knowledge it made no difference to either, however to maintain equilibrium with the ambient humidity, just how much epoxy will suffice to keep water as a mass out but allow vapour to pass through, I thought that perhaps one thinned coat, a CPES type material, and one straight coat would do, anyone?

[thechemist]

I would expect that the ultimate equilibrium state would be one in which {the tendency of the cellulose of the wood to absorb and hold water} is fully satisfied, and therefore the wood would be at the Fiber-Saturation-Point [FSP], roughly 30-35%. At that point diffusion laws would govern.

As for coatings, I would put a very effective moisture-diffusion-barrier on the outside, and very little [such as CPES and, if anything more, only such paint as might be necesssary for color, or to prevent condensation from soaking into the hull before it can drain to an evaporation or bilge-pump area] on the inside.

The very light inner treatment has the advantage that if accidents happen [through-hulls do leak, caulkings are not always perfect], water can VERY easily evaporate out of the hull on the inner surface, given only adequate interior ventilation.

[ford]

Chemist, thanks,for your reply.
The exterior will have two by 1/8" laminates and a layer of triaxial glass, there will be in effect 3 epoxy barriers on the outside of the hull. Ventilation of the interior will be aggressively pursued, however water may enter through wet sails or an open hatch etc, it is the accidental entry of water that I wish to prevent from entering into the timber, hence the question of a light epoxy and paint barrier.

[Ian McColgin]

Epoxy is more vapor penetrable than water penetrable. That's why some advocate no coating beyond the epoxy on the inside - let it vent a little.

My own moisture meter is the mallet. When I hit a place that goes squish I get out the chisels and make ready for a bit of a dutchman . . .

[Tonyr]

Third weighing today. I said I would wait a month for the next one, but if there was a trend, I thought that a week might tell us something.

Weight = 68.25 grams. No detectable change from 12 November. I am a bit surprised. Zero seemed as unlikely as rapid saturation. Perhaps a well cured epoxy mix with a very high ratio of volume to surface area is essentially impermeable to water? So far, it seems that way. The sample is back in its water bath.

Completely off topic - with four grandchildren to keep up with, I have been reading the first in the Harry Potter series. It's a good "growing up" book. Recommended.

Regards, Tony.

[Tonyr]

Fourth weighing today.

68.25 grams. No detectable change from previous number.

Unless anyone feels strongly about the matter, I propose to stop this experiment. At the simple level we are working, a dense epoxy/sanderdust compositite artifact does not seem to absorb a measureable amount of water within a "resaonable" period of time. I conclude that it was quite safe to encapsulate the exterior of my strip/composite Whitehall with this compound.

I have carefuly re-read my copy of The Gougeon Brothers On Boat Construction, and while they do not have any hard data fully comparable to mine, the remarks made are in line with my findings.

Comments welcome!

Tony.

[Don Maurer]

There is a difference between a substance absorbing water and a substance being water permeable. I think the previous discussions on the forum had to do with water passing through the epoxy. The wood flour in the epoxy mix probably won't absorb much because all of the free airspace in particle that small will have been filled by the epoxy. A more relevant test would be to coat a block of wood with your mixture and floating it in water. This would give an indication of absorption minus evaporation over a period of time. On the other hand, the only real problem I see with water absorption is if the wood expands enough to break an epoxy joint. Judicial use of glued joints and fasteners should get around that problem.

[thechemist]

Originally posted by Don Maurer:
<snip>The wood flour in the epoxy mix probably won't absorb much because all of the free airspace in particle that small will have been filled by the epoxy.

Actually, the first water to be absorbed by wood is that which is chemically bound [loosely bound, but chemically nonetheless...it is "hydrated"] onto the hydroxyls [2/3 of a water molecule...like attracts like] of the big cellulose molecules of which wood fibers are made. Those are hollow, but expand a few percent in diameter and a few tenths of a percent in length as they go from fully dehydrated to fully hydrated. Beyond that point further water may find its way into the free space inside those hollow tubes. To the extent that wood flour contains cellulose, it has the potential to absorb such water. Some epoxy products associate to some degree with those hydroxyls and may affect this behavior. To the extent that the wood flour contains intact [microscopic] fibers, even short fragments, it most likely will maintain those hollow spaces inside those fibers after mixing with an epoxy system.

A more relevant test would be to coat a block of wood with your mixture and floating it in water. This would give an indication of absorption minus evaporation over a period of time.

Now, there's a really interesting idea. Encapsulate a block of wood in some epoxy product, a film of known thickness, with a scale of some sort affixed to the side. Put it in a glass cylinder, with a stopper in the top , so the air above the waterline is at 100% humidity. Monitor exactly where it floats, over the months.

On the other hand, the only real problem I see with water absorption is if the wood expands enough to break an epoxy joint. Judicial use of glued joints and fasteners should get around that problem.

It may or may not get around the problem. Wood has a certain shear strength, and it is much less than tensile. Shear or cleavage, splitting of the wood, is a failure parallel to the fibers by separation of the fibers from each other, or splitting of individual fibers, or some combination.

Imagine the wood expanding, with great force, inside a containing sleeve which is the epoxy film. That epoxy film has a thickness, and per unit length will have some area, and some amount of force will be required to elongate it some amount. That is described by a number called the elastic modulus of the particular epoxy product, and here is where these products derive descriptions as "flexible" or "hard" epoxies.

If there is no tensile failure of the epoxy film, expansive forces can crush individual wood fibers.

If the epoxy film is stiff enough and thin enough, it will crack. Once it fails, microscopically, on one small spot, the contained expansive force will split the wood, opening up the crack enough to relieve the pressure. Further water can be absorbed, and other surface checks may or may not develop, or the initial one may just get deeper. Exact behavior depends on the microstructure of the individual piece of wood.

In any case, floating a long, skinny cylinder of wood in a long skinny glass tube of water sounds like a great way to monitor this sort of thing. One would not get accuracy such as seeing ten milligrams out of 65 grams, as one could not read a floating vertical scale to an accuracy of one part in six thousand......I think that might be limited to .01" out of ten, or maybe one in a thousand , or six hundred.....still, it might be enough to do some good science.

[bainbridgeisland]

The Chemist brings up a good point about cracking of the epoxy leading to added water absorption. We have been talking about water absorption of unloaded, epoxy coated, timber. Exposure to stress, like operating your boat, big temperature changes or large changes in moisture content, will cause micro cracking.

Basically if we assume the epoxy is gluing groups of strength members together, stress concentrations occur at the end of each strength member. In this case the strength members are simply wood fibers, but they could be anything. When placed under significant load, the concentration of stress causes micro cracking to occur.

How much stress is needed to initiate micro cracking? Generally very little. The actual amount varies with the relative properties of the particular wood and epoxy you are using. You can expect quite a lot of micro cracking by the time you reach 1/3 of the ultimate strength of the epoxy and wood combination.

Oh boy, does this mean most epoxy-covered boats are sailing about with micro cracks? Yep.

One question though is: What significance are the micro cracks to the strength and longevity of the boat? The answer is, micro cracking is not terribly significant.

But why? We are talking about really tiny cracks. Cracks like this in metal would be a problem but in a wood-epoxy combination it is not so bad because the wood fibers act as crack stoppers. When the micro crack runs into a wood fiber, the fiber is resilient and stops the crack. Typically, micro cracks do not expand much further until around 80% of the ultimate stress has been reached. Also, unlike metal, micro cracks in wood-epoxy materials do not seem to affect fatigue life.

Micro cracking does, however, increase water absorption. The micro cracks act much like micro-voids in wood fibers themselves. As long as a moisture gradient exists between the micro crack and the adjacent material, water vapor will flow in or out of them. Thus as the moisture of the surrounding area increases, the micro cracks pick up moisture too.

As before, once the moisture concentration of the micro crack and the surrounding materials (i.e. epoxy) are the same, a moisture gradient no longer exists. At this point, the rate of diffusion of water into the micro crack is zero because the diffusion rate is proportional to the moisture gradient.

[NormMessinger]

Um, we're going to get the epoxy water relationship all figured out then somebody is going to introduce paint and varnish to the system and it all goes out the window. No?

And, shouldn't yaall be taking about a specific brand, if not batch, of epoxy. System Three is more flexible than WEST. Does that make one more or less subject to micro cracks, to water absorption, etc.?

Interesting discussion. Thanks.

--Norm

[Bruce Hooke]

Following up on what Don said, I think there is some confusion here about what is being tested and what needs to be tested. Tony's tests where to determine how much water is absorbed by the epoxy itself or the tiny bits of wood embedded in the epoxy as filler. I don't find it surprising that the answer is 'not much' but it is certainly nice to see that confirmed. However, as others have noted I think the much more relevant question is how much water makes it through the epoxy coating and into the coated wood, and Tony's tests do not address this issue. The floating stick in a tube sounds like a very interesting test to me. Rather than using a scale on the side of the stick I wonder if it would be more accurate to weight the sample occasionally (the stick would, of course, need to be carefully wiped dry before weighing it).

[bainbridgeisland]

The question is a little too generic. "Does enough water migrate to the wood and reside there in concentrations that promote rot?" A complete answer would require discussing may facets of epoxy and wood boatbuilding, yacht design and materials engineering.

As discussed elsewhere in the forum, epoxy products vary considerably. Some epoxies are even water-soluble, though I don't think any common boatbuilding epoxies are to any significant level. The amount of moisture that’s absorbed depends on many factors including: the particular epoxy resin and hardener selected, what fillers and extenders were added, how competently the epoxy was measured and mixed, if it was cured properly, how thick it is, the initial moisture content of the wood, any chemical interaction between the epoxy and the wood that degrades properties, the relative stiffness of the wood and epoxy when under load, the proportion of ultimate load that the structure has seen (i.e. micro cracking), how much fatigue is present, time of exposure, and temperature.

My personal experience is that epoxy and wood boat construction resists rot very well, but not 100% of the time.

I built a bunch of boats in the 1970s using two different epoxy systems, West System and Industrial Formulators. In the 1980s I built some using System Three. I see these boats from time to time and talk to the owners. Only one has had any rot at all, a small spot on the corner of the transom where the finish had been damaged for too long.

I should point out though that I also built some carvel planked boats in the same time period. They were built of Port Orford cedar on White Oak and Douglas fir on sawn fir. These boats have not had problems with rot either.

I think the primary differences between the two methods is the finish on the carvel planked boats takes a bit more maintenance and one must be more careful about providing good ventilation for the carvel boats.

One other point: Quality materials and good workmanship are important to the longevity of either method. Here in the northwest we have 75 year old cedar boats that are still sound. They are sound because the boats were reasonably well built using good quality materials and have had good maintenance.

[mtlake]

What happens to the cold molded boat hanging in the garage when it freezes? Does the water expand? Thanks Steve

[thechemist]

If the water content of the wood is below the Fiber Saturation Point [FSP] then the water should not freeze. The further below, the lower the probability, at any given sub-zero temperature. Freezing-point-depression by alcohols [including glycols] is well-known. The hydration of water by cellulose hydroxyls should function in a similar manner, if below FSP. Putting it another way, the association of water with the hydrophilic hydroxyls reduces the probability of nucleation, even though below the freezing-point. Carburetor anti-icing additives work this way.

[Tonyr]

For the sake of being thorough, I did in fact keep the test going until last night. No measureable increase in weight of the sample. Test (really) complete.

I still intend to find a way to do a set of weighings of the boat itself before and after the coming boating season.

A happy and prosperous New Year to all. This is an enjoyable and informative Forum. Thanks to the Woodenboat people for sponsoring it.

Tony.

[funson]

unrelated to submerging the epoxy, I have a question of epoxy absorbing atmospheric moisture. Specifically, on my Seacoaster, I applied two or three coats of epoxy to seal the miles of little checks in my fir plywood. In my sanding last fall before stopping boat work to get a job (I promise I'll never do it again!!),I discovered this spring that I had sanded all epoxy off the edges of some strakes and over rainy winter the exposed wood had absorbed some moisture-not unexpected- but also to epoxy around these areas seemed also to have absorbed moisture and become more translucent than the similarly sanded areas around these places. When I gently heated these areas with a heat gun the moisture seemed to dry off of the surface and it became dull and opaque agin. I have been keeping the shop warmer and the surface is getting more uniformly dry. So the question is was I really seeing water being absorbed into the sanded epoxy as well as the exposed wood? If so, why only around the sanded through areas? Anyway, things seem to be drying our again and I can get on with it. I haven't used epxoy except on this boat, but I do like it in many places. I am still trying to understand all that I "know" about this stuff.