Posted by Jamie Hascall (Member # 335) on 07-27-2000, 04:12 PM:

Well with much gnashing of teeth and time spent in the moaning chair, I am embarking on the recaulking of the teak over ply deck that I so carefully recaulked last September. (Yeah I know that I should just lift the toe rails and lay down fabric etc. and be done with it, but I'm just a hopeless romantic.) Last years' attempt using Boatlife one part polysulfide has failed universally, with a complex of causes probably to blame including the kind of summer that helps keep the sunny southern folks from overpopulating the Pacific Northwest. Since the gray clouds never really lifted, the planks did not shrink fully, and when the sun came out this spring the adhesion and elongation of the compound couldn't keep up with the wood movement. That said, in the removal of the old compound I can see that the adhesion sucked and probably took a very small amount of movement to fail. (The good part is that with the bond breaker I used, the old compound is very easy to get out.)
So the $64,000 question is what is the stickiest most tenacious compound you've worked with? I'm disinclined to use two parters like Detco because of the mix and fill mess; so after doing the Forum search and reading all the past posts, I'm leaning toward 3M's Teak Deck Sealer. However, in talking to the folks from Legendary Yachts (Araminta, Bounty ...) they like the silicone based Teakdecking Systems product out of Florida. I also remember some coments about using 5200, as well as the old standby of roofing tar with some sand on top. So I ask for your words of wisdom and encouragement (and will ignore those muttering about the twit going to all that trouble for a fake teak deck) while I do my best to get this sucker as dry as I can.

Posted by Srw37 (Member # 2038) on 07-27-2000, 04:27 PM:

5200 will stick, stick, stick, but I don't know how much it will flex, flex, flex...though it's pretty rubbery, and you can -- or at least used to be able to -- get it in black. Still, I wouldn't put it on unless you want to keep it there for some time, and this could spell trouble if you ever want to pull things apart to fix a leak.
[This message has been edited by Srw37 (edited 07-27-2000).]

Posted by Gary Bergman (Member # 1631) on 07-27-2000, 04:55 PM:

What did you use for a bondbreaker? I need to tackle my teak deck soon, any any labor savers are appreciated.

Posted by thechemist (Member # 1468) on 07-27-2000, 06:13 PM:

Anybody's two-component polysulfide rubber put on top of CPES the same day has a long history of sticking to clean teak...meaning you will have to fold a piece of 80 grit over itself a few times and scrub it back-and-forth to clean the seam sidewalls. Bond breakers on the seam bottom can be a narrow strip of the cheapest masking tape in town. Even after the epoxy treatment figure the adhesive will shear if the paper doesn't.
The silicones I would not recommend...they look good at first but the bond to the wood will fail.

What sort of chemistry is the 3M Teak Deck Sealer? Been too long since I was familiar with all their products...haven't kept up with all this newfangled stuff.

Posted by Steven Gould (Member # 1912) on 07-27-2000, 06:59 PM:

The problem right from the start could be that the teak needs to be preped first (especially with one part polysulphides) to ensure adhesion. A lot of restorers use 5200 on decks with great results.

Posted by Jorma Salomaa (Member # 793) on 07-27-2000, 07:55 PM:

Jamie, last winter I received very helpful thoughts on this forum on how to redo the deck seams on my 40́ Colin Archer. Since then I have removed all hardware on the deck and done some repairs including more duchmen in the covering boards than I like to think of. I guess Íll have the seams routed and cleaned in a few weeks now.
In the meantime I have tried to enlighten myself on the subject of sealants. The first thing that struck me was the vast number of sealants that are available. I have studied more closely only a few but the process has given me a fair idea of how a deck caulking compound should perform, and why.

My initial thought was to use a two part polysulfide. However, nobody sells them anymore here, so I would have had to order samples from somewhere for testing. I have concentrated on one part polyurethanes, mainly because so many are easily available and because they seem to be widely used both in construction and on boats. Also, it still bothers me that I have not been able to find out why exactly the polysulfides would be better than the polyurethanes. Moreover, in Robert. M. Evanśs book on polyurethane sealants i read that "...while polysulfides were the first high performance construction sealants, they had two faults which made them easy prey to the polyurethanes and the silicones. One of these was exessive compression and extension set caused by the unstable disulfide linkages. The other was the high material cost characteristic of a material supplied by a single supplier."

I have been making tests with ten different polyurethanes, mainly to see how they differ from each other. It only recently dawned to me how important compression and extension set must be. It́s very rare that a manufacturer gives a figure for this. But it is very easy to make a crude test in order to find out. Other criteria are hardness and glueing capacity. Sagging and stickyness are factors as well. UV resistance is important but there is no way a layman can find that out in advance.

Of the compounds you mentioned I would say 3M 5200 is too hard for this application. There is also 3M 5300 which is softer. 3M Teak Deck Sealant is still softer, about the same as the majority of comparable sealants. It is probably a good sealant, but on an old deck with less than perfect wood in the seams it is still too hard in my opinion. I dońt know Boatlife. I know something about 3M, Soudal, Sikaflex, Formflex and Tremflex.

Ím kind of happy that you put this question to us, Jamie, because there seems to be so much that we dońt know about sealants that we should know. After all, sealing leaking decks is one of the main activities of wooden boat owners. Ím looking forward to a good discussion.

Posted by Dale Harvey on 07-27-2000, 10:57 PM:

No single part product will ever work well enough to be worth the bother. If you don't have enough thickness for tar, use two part. Plastic caulk tubes and a pnuematic caulk gun will lessen the drugery. You almost have to have two people to work it, and three is better. I would try Smith's products. You know that his caulk will be compatible with his CPES. The only failures I ever had with two part were in overly moist and unprimed seams. CPES should cure that. Two part polysulfide is the only stuff that will permanently seal window glass in a wood frame. Never worried with that bond breaker crap. NOTHING that says silicon anywhere on the label should ever be used on or near wood. It is always trouble, and silicon oils are near impossible to get back off.

Posted by thechemist (Member # 1468) on 07-27-2000, 11:58 PM:

There are traces of silicone oils (meaning low-molecular-weight silicone compounds which are not part of the cured polymer) in silicone rubber sealants. The controled -volatility space grade stuff used in satellites has that removed by vacuum distillation. The consumer stuff does not. It is not ordinarily a problem except when failed sealant is removed from wood. There will be some of that residue near the wood surface. Sanding or routing of seams is vital, to remove the 1/32 of an inch of possibly contaminated wood.
Any time a sealant of any sort in a teak seam has failed, the wood must be cleaned of the old crud on the surface, otherwise the best glue in the world will only end up glued onto crud, not to teak.

The two-component polysulfides have lost much market share, it is true. When manufacturers of formulated products tried to make one-part or two-part sealants with only twenty-five to thirty percent thiokol polymer in them, and did not formulate with the right materials, they produced poor materials that had poor compresion-set and biodegraded easily. Boat Life was famous thirty years ago for shrinking, getting hard, and coming loose. Might still be, although I think they went to a more expensive plasticizer that did not bleed into the wood so fast, leaving just the polysulfide and mineral extenders.

Properly formulated polysulfides have 45-65% polymer (more is not necessarily better, once you use enough) and do not have compression-set problems that render them unsuitable for boats. Remember , Boeing glues all the riveted seams on a typical 747 jet with 500 gallons of that stuff, and it is used in a similar manner in the riveted seams that hold the skin to the airframe of most military jets. The disulfide linkages are not inherently unstable. That is false data. Most people in the USA do not know how to formulate them any more. Even the chemists at Thiokol cannot offer adequate applications assistance to new formulators, sadly. Thiokol sold their sealants division to a company who wanted the coatings side of it and trashed the commercial sealants business.

Other competitive pressures also enter into it. In the recent ten years or so, there have been considerable strides made in the technology of urethane prepolymers, and now single -part sealants with urethane backbones, inherently less expensive to make than the polysulfides, are widely available. They have captured many market applications formerly held by polysulfides as top-of-the-line rubber chemically cured sealants, even thought the one-part caulks are moisture-cured. For many applications it is viable. Obtaining adhesion with those is sometimes dicey, although in certain applications it can easily be done. Don Dannenberg has excellent results using 3M 5200 on wood impregnated with CPES. Likely black 5200 would stick to teak primed with CPES, although whether the sealant should be put on the same day ( as with the 2-part polysulfide) or after the CPES is cured, I do not know. Someone should do experiments. I would guess it might not matter, that either would work.

A properly formulated 2-part polysulfide has extraordinary weather resistance, being essentially unafffected after 20-30 years unprotected exposure. Biodegradation is something not often talked about or appreciated , but is the main reason most organic things fail, aside from ultraviolet degradation of clear finishes. Bacteria even live in the failed gloss (microscopic pits) of a paint, eating away at it. The 2-part polysulfides are extraordinarily resistant to that failure mechanism because of the polymer itself as well as other major ingredients, all of which are inherently toxic. Not very toxic, just slightly. Just enough to discourage small life. Things that are absolutely totally nontoxic to anything end up being food for something. Always.

Posted by Jamie Hascall (Member # 335) on 07-28-2000, 12:18 AM:

As always, this group is the place to get pertinant commentary. I'm glad to hear the comments on the silicone compounds as they bring to mind how all silicones were banned from the instrument shop I worked in as just the hint in the air caused all sorts of problems in the finishing booth. I no longer am giving them any serious consideration.
When I did this last year, I followed all instructions and have had this confirmed by Boatlife's tech people. The one thing that raised a flag with them is that as I told them about washing the grooves with acetone to clean off the oils, they remarked that much of the acetone available now is recycled and actually may have contaminants such as oils in it.(they said they had just found out about this last February.)
The bondbreaker was 3M 1/8" Fineline tape which was easy to used and worked well. Using a bondbreaker is one of those procedures that doesn't make total sense to me, but I've read it in the recommendations of every manufacturer, so I will assume that they do have a good reason for specifying it. It does have the side benefit of making the reeving out of the seams very easy. However, if the water gets in at one point, doesn't it just travel down the seam?
The 3M Teak Deck Sealer is a polyurethane base, but I don't know what else is in it. I'll try to get my hands on some further specs and post them. Their primer is labeled as an epoxy lacquer. When I mentioned the acetone question to the 3M tech person, she said they never recommend acetone for this purpose, but rather their own cleaner instead and to wipe it off with a rag being rotated frequently to maximize the cleaning effect. This does feel like a way to get a greater degree of quality control.
Jormaa, I'm glad you're still working on this too. I went back and read the thread you had started from last fall and it was a good discussion but I felt we needed to beat the horse a little more.
Dale, I know in my heart that what you say is true, but the thought of mixing goop and loading tubes in addition to the rest of the work feels pretty daunting. However, I truly want to never have to do this again so I may opt for two part. The compound that remained of the original caulking was probably two part polysuphide and I was amazed at it's flexibility and strength after 30 years.
Thanks for all the thoughts,

Posted by thechemist (Member # 1468) on 07-28-2000, 12:23 PM:

Epoxy Lacquer....I weep for the language.....
A lacquer, by definition, is something that may be left when its solvent solution evaporates, but which will redissolve into the same solvent solution, unchanged. Lacquer drying and redissolving is a completely reversible reaction.

A two-component epoxy product is not a lacquer. One might make a lacquer by polymerizing an epoxy with something else to make a resin, soluble in some solvent (and, indeed, there are such, brittle solids), that could be applied in such a manner, but the properties would not be adequate for this application.

The assertion that much of the acetone nowadays is recycled and may have oils in it, I find ludicrous. My company uses industrial quantities of such solvents and the legitimate chemical distributors all handle only virgin material. A chemical manufacturer who used acetone as a process solvent would save money by having a commercial distiller redistill the material for him and reusing it, so contaminated acetone would not end up in commercial distribution channels. If one distributor of acetone to Boatlife inadvertently mixed some oil with a batch of acetone they shipped to Boatlife, I can understand how they could get a contaminated batch, but as a general thing it ain't happening.

[This message has been edited by thechemist (edited 07-28-2000).]

Posted by Jamie Hascall (Member # 335) on 07-28-2000, 01:27 PM:

Thanks for your comments on both these points. The "epoxy lacquer" term surprised me when I read it and I begin to wonder if the marketing people are writing specs instead of the product engineers. Sadly, I have always felt good about 3M products and this bit of language twisting has shaken that opinion.
I also had concerns about the acetone story as it would seem that nothing is safe if a can of solvent from a widely distributed manufacturer could commonly have contamination problems. I had wondered about seeking out reagent grade solvents from my art conservation suppliers, but that seemed a bit much for boat work. I think I can more interpret it as corporate excuse making for whay products don't perform quite as advertised.
The mention has been made of Smiths having a two part caulk, but there is no mention of it on the website. Can anyone give me further info on this product? This path seems to be leading inevitably to a two part compound, so now it's a matter of finding the best one and setting up jigs and equipment for assembly line mixing and filling of tubes.


Posted by thechemist (Member # 1468) on 07-28-2000, 02:03 PM:

They have a general web site, It does not seem to be a "buy-it-here" sort of site, so I would say just email them, or phone, or whatever.

Posted by Jorma Salomaa (Member # 793) on 07-28-2000, 05:00 PM:

Smith sells a two part polysulfide with lead peroxide as curing agent. Probably a good product. If memory serves the price was $35 a quart.

Posted by Jamie Hascall (Member # 335) on 07-28-2000, 06:35 PM:

Well, I just got off the phone with Steve Smith and was impressed enough by his product line and the usage tips he was able to give me that I've decided to go with the Smith and Co. two part polysulfide. I have no doubt that it will be a lot of work and a messy job, but I won't have to do it again. Of course, while we talked about using CPES to seal the wood before caulking, he inquired about the brightwork and ended up selling me a batch of their Five Year Clear polyurethane for the toe rail. Well, at least this part of the project will be finished soon. Thanks to all for your help in flogging this subject for a little longer. I'll try to post some pictures of the final result a little later.
BTW, when I mentioned the concept of "Epoxy Lacquer" to Steve, he went on a good three minute rant. Always great fun to hear a well informed rant. He's truly worth consulting with on any of these applications and is very free with his advice and information

[This message has been edited by Jamie Hascall (edited 07-28-2000).]

Posted by Jorma Salomaa (Member # 793) on 07-28-2000, 07:02 PM:

Chemist, your post on polysulfides was interesting reading. Boeing certainly is a convincing testimonial. But there must be different formulations for different applications. The hardness when cured should be an important factor - I believe a wooden deck needs a rather soft compound. The manufacturers normally give the hardness as measured on the Shore A scale. 3M 5200 is Shore A 60 (which practically makes it a glue not a sealant), 3M 5300 is Shore 50 and 3M Teak Deck Sealant is Shore 40 which is typical of an all-round sealant. The old polysulfides I have seen on peoples decks I would estimate to be around Shore 30. The first thing I would find out when choosing a sealant is its hardness.
Another thing I came to think of is two part polyurethane sealants. Two part polysulfides are frequently mentioned but one part polysufides hardly ever, even though they seem to exist. Is there something in the two part goops - apart from the more controlled curing process - that make them superior? So what about a two part polyurethane?

Posted by thechemist (Member # 1468) on 07-28-2000, 10:40 PM:

Well, Jorma, the two-part polysulfides were superior simply becasue the one-part products that were manufactured for the consumer marketplace in this country 10-20-30 years ago were such c r a p. really. Starting with a resin that cost $25/gallon, One needed to end up with a product in a caulking tube that could retail for $5/tube (back then), which, at 12 tubes (the 10.8 fluid ounce caulking tubes) to the gallon meant a retail price of $60/gallon.
The facts of life in a distribution chain here are that the distribution chain would need to buy it for 40% of retail, maybe less. that is $24/gallon. Take out a few dollars for caulking tubes, cardboard boxes, freight and whatever and figure the manufacturer gets $20/gallon, has to remain in business and be viable.

The way this is done is to add microscopic glass bubbles (microspheres) to the mix, to double the volume. Only now it does not stretch too much, being now rather hard. So, add an inert oil, called a plasticizer, to the mix, to make it softer and more stretchy. Now you have doubled the volume again. The plasticizer oils cost a quarter or less of what the polysulfide resin costs, and the microspheres cost about a tenth of what the resin costs. Formulation raw materials are now down in the zone and area of eight to ten dollars a gallon, and the guy can mix it and package it and sell it for a profit.

Unfortunately, with age the stuff shrinks as the plasticizers bleed out, diffusing into the wood and being biodegraded. The remaining stuff of course comes loose from the wood. At the surface ultraviolet degrades the rubber-wood edge a bit, allowing a small crack. Fungi and bacteria get into that crack, eat the crack deeper, and teach their young. The stuff slowly lets go, the shrinkage opens up the crack further, and it all slides down that slippery slope of no return.

The issue really revolves around the length of a molecular chain that could be manufactured as a resin which could be further reacted to make a useful solid material, and the key factor is what percentage of the molecular chains had a reactive group on both ends. Length of the molecular chain (the distance between junctions after curing ) has a great deal to do with cured properties. Polysulfide technology, it turned out, inherently gave molecules about a thousand atoms long with a reactive group on each end of most, maybe 98% of them. Urethane or epoxy technology was not so well developed, and the raw materials could not be made in chains that long with most of them having a reactive group on each end. So, polysulfides made the better elastomers. Besides, the polysulfide resins of that big size were liquids, where those for epoxies were not. By way of comparison, making urethane raw materials with molecules that big, maybe two out of three had a reactive group on both ends, the balance having only a reactive group on one end. What this means is that the length of the molecules in the cured polymer could be probably only a few or so times as long as the length of the starting materials. This does not yield good mechanical rpoperties.

One needs molecular chains in the cured polymer maybe a thousand times the length of those polysulfide resin molecules to get good properties, and those were the only raw materials that could do it, just because such a high percentage of them had a reactive group on each end. It was a consequence of the nature of things and the technology of the time. Eventually the starting materials for urethane polymers were developed to a similar (call it purity for a simple term) degree and then the inherently less-expensive-to-make-by-a- factor-of-five urethane caulks began to capture market share from polysulfides. Then the guys making one-part high-quality polysulfides sold off their businesses or gave up, and here we are.

You raised the issue of how flexible the deck elastomer should be, and it is now 7:30 and that needs to be dealt with in a separate unit of time. Wasn't there a thread...something about teak decks...that touched on this? I cannot remember...and I think at home there is still a bit of Talisker left in the bottom of the bottle. This program will reboot at another time.

[This message has been edited by thechemist (edited 07-30-2000).]

Posted by Jorma Salomaa (Member # 793) on 07-29-2000, 04:15 PM:

I then assume that if I want to use polysulfide I have to find a manufacturer whós serious and not only beating the last dollars out of a potentially dying product line. By the way, I have noted that many manufacturers offer both polysulfide and polyurethane sealants. Íve also noted that they are much more keen on selling the polyurethanes with a large variety of color, hardness, sagging characteristics etc. Some of them are introducing what they call hybrid sealants. Could it be then that the polyurethanes of to-day are as good, or rapidly becoming as good as the best polysulfides?
My simple "shear tests" indicate that the polurethanes glue very well, though some may be a little better than others. But when it comes to compression set there are remarkable differences. Ím not yet through with my compression tests but the specimens I have tested hint in the direction of the softer compounds having more compression set, i.e. less re-dress. The reason could be in a larger amount of plasticisers that dońt contribute to the main task of the sealant.

What I would like to find is a sealant that, apart from having an excellent glueing capacity, has a hardness in the range of 20 to 30 Shore A, stays that way for twenty years, and is a one component product. I could́nt find any earlier thread with a discussion on softness but I noted that many forumites had found 3M 5200 to work well as a deck sealant. This is interesting, since wére essentially talking about a glue here with a Shore 60 hardness. 3M 5200 is nearly identical to 5230 which is marketed as a glue and is also of Shore 60 hardness. The difference is that 5230 cures in 3 days whereas 5200 takes 7 days. To me, the requirement of softness is logical since the softer the compound the more forgiving it is on the glue line, all other things being equal. Well, maybe they arént equal - maybe a polyurethane just glues better the harder it is, which may be the reason why 5200 may work as a deck sealant at all.

Anyway, I asked one reputable polyurethane sealant manufacturer why their deck sealant is so hard. The answer was that of course a softer one would be better. They actually marketed a Shore 20 product in the early nineties but had to withdraw it because when it got warm in the sun it became sticky and the gals in their white pants walked away with black stripes on their butts! It́s not a simple world.

Posted by thechemist (Member # 1468) on 07-29-2000, 07:34 PM:

Okay, Jorma, the more firm polysulfides will run fifty to sixty Shore A, and those are made with the thiokol polymer that has two percent cross-linking. The softer material, twenty to thirty Shore A, is made from the half-percent-cross-linking version. Thiokol sells them both. Any manufacturer who is making two-part polysulfide sealants with one could make it with the other, given only enough customers to justify a minimum batch. These manufacturers have mixing equipment of certain sizes, and one does not want to make a very small batch, or the cost goes up disproportionately.
The softer one-part sealant that imprinted gals in their white pants, those guys tried to make it softer by using too much plasticizer and not enough of the expensive polysulfide resin, and possibly also not mixing it properly. Dispersing carbon into liquid resins is not an easy nor a quick task, and if they were trying to cut corners in one area they would likely have been cutting them everywhere. There is not reason you could not get what you need....and if your boat is big enough and you are laying or relaying an entire deck and the seams are big enough, you might find a small manufacturer willing to make you a batch. How big is your boat, anyway?

On the third hand, the elongation needed I think depends on the relative motion expected, which depends on how wide the teak planks are and how stable is the substrate to which they are attached. Some boats have everything screwed and epoxy glued to a very mechanically stable GRP deck, some have four-orsix-inch wide planks sitting directly on ribs, others have a plywood surface, etcetera. Mechanical stability varies, and thus softness of the sealant varies. More motion anticipated (compliance required) with a minimal stress on the glued joint requires a softer sealant.

The nomenclature of softness or hardness is not exactly the correct one, but is close. One actually should use the term low-modulus to describe an elastomer which requires little force to stretch it a given amount. There is a number, called the modulus of elasticity, which is the measurement of force required for an increment of elongation. A high-modulus elastomer is one which has a higher modulus of elasticity, requiring more force for the same elongation. The force manifests at the wood/rubber interface. If the bond is stronger than the tensile strength of the rubber, then the rubber will tear at failure. If not, the rubber peels from the wood as an adhesion failure. there is also an elongation capability or limit, and an elastomer may fail when its elongation capability is exceeded.

The Polyurethane products have better profit margins and manufacturers of them will push them over polysulfides. The polyurethanes can be introduced into a market where the high-performance polysulfides commanded high prices. If the polyurethane costs a third to make and is offered at two-thirds the price, which product does the manufacturer push?

A polyurethane sealant could be softer, but the distance between cross-links has to be increased (unless we mix in plasticizers). Increasing molecular weight (distance between the ends) with material that has mostly reactive groups on both ends results in needing a more expensive raw material. Softness or hardness may be a consequence of the profit margin the manufacturer needs to be viable and pay their stockholders, etcetera. It may also be determined by the people in their marketing department, who pretend to know what everyone needs and then tells manufacturing what one product they should make, and designs an advertising program intended to convince everyone that they need this one product in their application. Ultimately, it's marketing. The bigger the company, the more they listen to Marketing and the less to their customers. The smaller companies can still see their customers. Not always, but for most.

[This message has been edited by thechemist (edited 07-29-2000).]

Posted by Jorma Salomaa (Member # 793) on 07-30-2000, 07:21 PM:

Chemist, that was a great post again.
Ím interested in the polyurethanes mainly because they are easily available and because therés such a large number of different one part sealants for different applications. Ím not even considering one part polysulfides, for the same reasons. Ím completely satisfied with the adhesion capabilities of the polyurethanes I have tested. The desired modulus of elasticity is available. The only big question mark is longevity. Will the desired properties be retained for twenty years?

I know, when in doubt use what has proven to last for 20 years and buy a two part polysulfide from a reputable company. In the end, I might go that way, but in the meantime I have decided to spend some time investigating the polyurethanes. I sort of enjoy it as you can see.

The boat Ím talking about is 25 years old gaff ketch, 40́LOD, 13́4" beam. The deck is 4" by 2" teak, covering boards 8" by 2" oak. All is vertical grain. The deck seams are 6 mm (1/4") wide, for the covering boards 7 mm (I cańt compute this to inches, sorry). All seams are 6 mm deep, as far as the sealant goes. The total length of the seams is about 250 meters, roughly 250 yards.

The moisture of the teak will probably vary between 10 and 15 % and the oak between 12 and 17%. Theoretically, the width of the teak plank will vary 0.75 %, but let́s say 1%. The same for the oak plank will be 1.5%.
In practice, it will not happen exactly like this, but this is based on what has been written on the subject.

When this is converted into millimeters and percentage of seam contraction and expansion, we come to the following: The 6 mm deck seams will have a movement of 1 mm, or 17% of the seam width. The 7 mm seam between teak and oak will move 2 mm or 28%. Now, it́s a matter of judgement how much stress you think the interface between sealant and wood can take before the adhesion is lost. Based on what I read and hear and think, a sustained elongation of 17% requires a sealant with a modulus of elasticity normally found with a hardness of maximum 40 Shore A. In the case of 28% elongation I would feel safe with Shore 20. By simple calculation, if I wanted to have the 17% elongation also for the 7 mm seam between oak and teak, I would have to widen the seam to 12 mm, which is plain ugly.

A thing in favor is that the seams are now practically as wide as they will ever get, but an allowance has to be made for some inevitable compression set in the sealant.

The main thing I have learnt in the process is that you cańt just pour in any goop and expect it to hold on. The wood/sealant interface is seldom optimal for good adhesion on older wood, so we have to help it by all available means. CPES is going to be used liberally in my seams as a primer. I have tried it with three different sealants (on Scots Pine) and all have better adhesion with CPES. Whether the CPES is partly or totally cured seems to make no difference.

It has been said that seam caulking aińt rocket science. But it is science to the sealant manufacturer, at least until the marketing department is let loose. And a little bit of science wońt hurt us either.

[This message has been edited by Jorma Salomaa (edited 07-30-2000).]

Posted by thechemist (Member # 1468) on 07-30-2000, 08:44 PM:

Excellent data, Jorma.
Now for an experiment: we apply a force to a seam and measure how far it stretches. Since measuring a single seam stretching is not so easy with simple tools, make a series of ten seams, gluing together eleven blocks of wood. Make the seams the same dimension as on your deck, make the wood blocks any convenient thickness, and when cured saw-cut to trim to a nice even number (an inch? two centimeters?). Then run a screw-eye into the first and the last blocks, centering them in line with half the depth of the seams, hang one from a skyhook and hang a bucket of water from the one below. Add water until you have the desired 28% elongation (ten seams times 1/4 inch equals 2.5 inches times 28% equals 0.7 inches additional, about 17.5 mm more. Measure the water and you have found the tensile stress per unit length of the seam at 28 % elongation.

Now comes the fun part. I have a sign that says "real engineers learn by breaking things" and we are now gonna be engineers. Add more water until the thing comes apart. [ Make sure everything was fully cured before you do this...] See what the ultimate elongation was, and whether the cpes/wood interface tore or the urethane elastomer peeled off the epoxy which maybe remained stuck to the wood. Edge effects may dominate, with actual physical properties being better. Adhesive bonded joints usually fail from an edge inwards. Remember that thing I posted a while ago about joint design? Now, with your experiment done, we know what the safety factor is. Write your report and turn it in.

[This message has been edited by thechemist (edited 07-30-2000).]

Posted by Jamie Hascall (Member # 335) on 07-31-2000, 12:46 AM:

The last few posts have brought up some other questions in my mind. In testing the compound for its' failure point, I have to wonder if there is an acute failure point and a different chronic failure point in the bond? I feel that bonds that are stressed to an amount less than the point of acute failure will over time slowly start to creep and fail. I have no real evidence for this except for the behavior of other types of adhesive bonds such as tapes which of course are not formed by a reaction that has a curing effect. Is this possible or am I just transfering an effect from one realm to another unrelated but similar looking situation?
The other thing that puzzles me is Jormas' comment on "compression set". I take it that as the cured compound is squeezed by the expansion of the planks as they gain moisture, it is compressed to a smaller dimension and when the wood dries again the compound will not fully regain its' original dimension. Does this lead to the conclusion that the best time to caulk is when the wood is at its' mean dimension so that the compound does not exist under as high compression when things get wet and then have a more difficult time regaining dimension when the deck drys during a prolonged sunny period? This especially concerns me as I have been trying to dry the deck planks as far as I could so that the caulking compound would spend as little time as possible under tension, thus lessening the possibility of the bond breaking.
I was very interested in your experimental conclusion that it seemed to make little difference in bond strength (of the polyurethanes) whether the CPES was fully cured or not. This would make it possible to prime all the rabbets with CPES and then move on to the caulking stage instead of trying to do both within a proscribed time frame. I definitely want to follow the procedure that will give the best adhesion, but if this seems to make little difference, it will make the flow of work run much more smoothly to prime seams en masse.
As always, thanks for this discussion. These are the threads that make me carve out time to paticipate when there's other work calling.

Posted by thechemist (Member # 1468) on 07-31-2000, 01:06 PM:

from your post a bit above I gather you are using a two-part polysulfide, whereas Jorma is using a one-part polyurethane. The chemical groups on the ends of the two kinds of molecules are NOT the same, and priming procedures for different chemistries are different. Do whatever Smith said to do for his stuff. Jorma did his experiments with a material that may have a backbone containing urethane linkages, but the reactive groups are definitely different from the reactive groups on the ends of the polysulfide resin molecules. His chemical results as regards priming are not to be yours.

As for compression set and tensile stress, you do raise a salient point. I do not have simple answers for you. I am not sure there are any simple answers for the general situation, since we are dealing with structures made of natural materials (meaning every piece is different) combined with chemical glues applied by humans (you know abut them.....everything you have heard is true) and then exposed to different weather cycles, and all the while everything under attack by small life.

It seems sensible that having the planks as dry as possible before caulking would minimize tensile stress in the final cured elastomer, cycling over the ensuing weather extremes. Soaking the top surfaces of the wood planks with CPES should reduce the expansion/contraction cycles of the wood, somewhat. Maybe not much, but it probably would do more good than harm, and being hydrophobic when cured should slightly reduce the water takeup and wood expansion. There was an interesting experiment in that paper at that suggests such a conclusion.

As regards creep failure, this does not happen in materials that are well-cured, with everything turned into one giant molecule of infinite molecular weight, below its glass-transition temperature and stressed within its elastic limits (joke...that's not the real world). To some degree there will be such issues. I don't know what the time frame of such failures would be, but from the lack of evidence that such things are happening a lot, it would appear at first blush that other failure mechanisms dominate the scene, such as extreme stress, weather and small life.

Posted by Suprcargo (Member # 1313) on 07-31-2000, 03:26 PM:

Grand Banks, Built in Singapore and Malaysia use the material, on ALL their new boats, from Teak Decking Systems in Sarasota Florida.
[This message has been edited by Suprcargo (edited 08-27-2000).]

Posted by Jorma Salomaa (Member # 793) on 07-31-2000, 04:04 PM:

Thanks, Chemist. The water bucket sounds like a wonderfully simple and practical way of doing the measuring. The rest of your advice also sounds OK. The outcome then obviously is a force per some area of caulk/wood interface. Íll report when Ím done, but it may take a while...
Jamie, Ím afraid that what happens after the compression set is there, the next shrinking phase of the plank will start with a narrower seam, as it were (from the sealant́s viewpoint). Ím in the same situation as you are. This is why I consider the redress of the sealant to be so important. This is quite easy to check by laying a bead of the sealant on a piece of wood, let it cure for a week, measure its thickness, squeeze it to 50% of its thickness in a vice, wait three days, take it out and measure again. Then draw your conclusions.

When considering your seam failure of last autumn one would probably hold the seam walls and their preparation as the first suspect. Moisture, pollution? Next comes the elastomer itself. Bad batch that cured too hard? Product too old (Íve heard it happen with a sealant the date of which had expired). Then maybe wrong type of sealant (too hard for the seam dimensions in question), but that is improbable since your seams must be something like 1/6" by 1/6".
What do you think? Small life could be the culprit, but in less than a year and iśnt one part polysulfide quite toxic? Maybe other people could weigh in here and tell their stories of seam failure.

[This message has been edited by Jorma Salomaa (edited 07-31-2000).]

Posted by Jamie Hascall (Member # 335) on 07-31-2000, 05:34 PM:

Thanks for confirming to me that I was comparing apples and oranges in my concerns over creep type failure and also on taking Jormas' test results with polyurethane and applying them to the two part polysulfide that I have ordered. I a conversation with Smith & Co., I was told that the best priming for their two part product was to prime the groove with CPES and apply the caulk immediately while the CPES is still wet. When I inquired about how this should be integrated with the application of the bondbreaker, I was told that especially for seams of my size (1/4" deep x 3/16" wide) none was needed, and that they had never recommended a bondbreaker as part of their procedure. The only other thing he mentioned was to be very careful with messing with the filled seams for a couple of weeks as the CPES is much slower to cure than the polysulfide and until it has fully cured it is possible to pull the string of cured compound right out of the groove.
Jorma, as far as last years failure, I think it owes itself to the obvious complexity of real life. I had worked hard to prepare the grooves by cutting them deeper and wider with a circular saw so I had fresh wood all around, but I'm sure that sometimes contamination still occurred. When I wiped down the grooves with acetone, I brushed it on, but did not then wipe it off before it evaporated, therefore probably redepositing the oils it had dissolved on the walls. This all couples with the fact that it was a moist summer in the northwest and the planks had further to shrink than I thought possible , and that the planks are old and are screwed down at their center line and probably have more vertical flex at the edges than is ideal. However, the fact that the failure was truly universal tells me that the Boatlife just didn't have the tenacity or elasticity to make good no matter what I could have done better.
As for concerns about compression set, I will still work to dry out the deck as far as possible as I have to believe that if the compound does take a compression set, the tension forces generated when the wood shrinks again will still be as low as is possible with the compound being under at least light compression at all other times and thus well sealed even if adhesion is compromised.

Supercargo, thanks for the info about the TDS products. I had also heard good things from the people at Legendary yachts.


[This message has been edited by Jamie Hascall (edited 07-31-2000).]

Posted by Don Braymer (Member # 388) on 07-31-2000, 06:53 PM:

Small tale of woe, and you are right to try to get this all straightened out AT THE BEGINNING.
When we redid the Dirigo deck, we determined that we were going to use a polysufide based caulk.

Well, my project foreman went to Fisherman's Supply and got some stuff in tubes and it was one part, and he told me what a good deal it was for "polysulfide one part".

So we reefed and corked and reefed and corked some more, put in the Detco polysulfide primer/sealer, and then squeezed in the stuff, and everything was going great guns.

Ran out of the stuff, went back to Fishermans Supply, with an exhausted tube, and they sold us more of the same.

EXCEPT, went we started putting putting in the second batch, it cured a different color grey, and after a few days, we noticed that if you got a good hold of the dryed bead, you could often pull it completely out of the seam for several feet.

A few trips back to Seattle, and a bit of yelling, and it turns out we are now using silicon, and the batch that was supposed to be polysulfide is a "bad batch".

So in the middle of the boat, we begin doing Detco the two part method.

Well in Port Townsend, the wind, temperature and hunidity can all change 5 times in one minute, and I ended up doing everything except eating the polysufide. I have one interesting patch in the cockpit where rain drops sunk into the polymer and then evaporated leaving little dents; "distressed polysulfide" I guess.

It appears that the Detco primer/sealer makes a good release agent for the silicon, as the one part no primer silicon is intended for raw wood, which is a pourous material, and the whole idea of the polysulfide primer/sealer, is to well, seal the wood.

At least it doesn't rain in Southern California, and the corking is mostly all sealed up under the sealer. But I do have several places where caulking separation from the wood is evident.

Do all the experimenting and testing that you can. You are going to spend hundreds if not thousands of hours doing this thing on your knees, and you don't need to find out that something is wrong with the system 1/2 way into the project.

I will very seriously consider using some kind of high melting temperature natural pitch kinda thing next time.


Posted by thechemist (Member # 1468) on 07-31-2000, 09:26 PM:

Hi, Don B
That was a great concatenation of disasters....a really good demonstration of why you need to drill a procedure with something until you know it works and then use THAT and do it THAT WAY.

One of the reasons I do not like paste Sealants (did I mention that I do not like paste sealants?) for deck seams is that they do not necessarily wet the side walls of the seam, and so even if the manufacturer's primer actually worked in the real world, the sealant never really made intimate molecular contact with it, in spite of being squeezed into the seam. It is a paste, a thixotropic gel. By definition it wants to hold its own shape. A liquid FLOWS and WETS the side walls of the seam, and so makes contact with the primer on a molecule-by-molecule basis. It is impossible for a paste to do that unless you smear it on each sidewall of the seam with a miniature spatula, and THEN feed it into the seam and knife it level.

You might ask, then, why do not the manufacturers of one-part polysulfide rubber sealants (which work so well in the laboratory) make a liquid version of their one-part sealants, for these applications?

Because they are humanoids from planet Earth, and they cannot see the obvious, and it has always been done that way, and you have to put only pastes in caulking tubes because of all the problems Manufacturing would have if it was a liquid and the customers would be sure to reject it because it could not be used on the garboard seam without running out and our customers want one product they can use everywhere and we are already making it this way and Marketing says this is what we should make because their marketing program is already geared to the present product and the labels would have to be changed and the stores would not give us more shelf space for more different products and if it lasted longer then sales would drop.

That's why.

Posted by ACB (Member # 1213) on 08-01-2000, 12:28 AM:

Thanks to everyone who has posted here - what a mine of information this thread is!
I have a question, of course. Since I Coelan-ed the decks I have stopped worrying about them, but the seams that I forgot are the seams where the dreaded wooden bulwark rail stanchions pass through the covering board.

The covering board is teak, as are the stanchions, which are not extensions of the grown frame futtocks but are clenched to the top three strakes of planking. They were originally oak, but I replaced them with teak about 15 years ago, for obvious (rot at deck level!) reasons.

This gloop has benefitted from being varnished, but it looks like it has now died. I can't remember what I used. It is awkward to rake out these little seams, but how clean should I get them and what should I put back? Same goes for seams round bitts and quarter posts of course.


Posted by Jorma Salomaa (Member # 793) on 08-01-2000, 06:19 AM:

Don, your story confirms that you can be hit by a "bad batch" anytime. Testing in advance every batch you use is therefore a logical precaution. But it is a lot asked in a job where every day may have a price. However, I believe anybody who has been hit wońt let it happen again. I have contemplated the "back to nature" approach myself sometimes. Pitch is straightforward to apply, servicing the seams is simple (but frequent), it́s cheap and the stuff itself lasts forever. The draw backs, however, are obvious.
To Jamie: The seam dimensions of 3/16 wide and 1/4 deep go against what EVERY elastomer manufacturer prescribes. Of course, they have to prescribe something that is easy on the adhesion and they may at times overdo it. The depth of 1/4" is the minimum they normally ask for on a porous material. They would say OK to a width of 1/4" but more is recommended and anything less is a no-no. Considering how a sealant behaves in tension, this makes sense, I think.

Chemist, Ím glad you mentioned flow and wetting. I have noticed that it is virtually impossible to guarantee that the seam is 100% filled with most of the non-sag elastomers. I absolutely agree that to do it right you would have to actually smear it on the sidewalls. I have found only one polyurethane manufacturer that has taken this into account, namely the Sika Company of Swizerland. Their deck caulk Sikaflex 290 DC has more flow than the others, certainly just for this reason. They say it can be used on a seam up to 5 degrees from horizontal. When you apply a bead on a piece of wood you can actually notice the "wetting" happen. I think this is one reason why it adheres so well. I haveńt seen any other goop glue better, but many perform worse. It has a hardness of Shore A 35. Needless to say, it́s my favorite polyurethane at this time. Sika also has a Shore 25 product, Sikaflex 298 which has the same characteristics. It is intended to be applied under the teak planks on plank-on-plywood decks and other similar installations but there seems to be no reason why it could not be used in the seams. Ím not familiar with the flow and wetting characteristics of the polysufides, except that Formflex, a two part polysulfide which behaves exactly like Sika 290 DC. However, the batch was very old and created a lot of minute bubbles which led to a lot of compression set.

To ACB: To excavate your old stanchion seams I would say a very, very sharp, narrow chisel works wonders. If you happen to have the Fein oscillating tool with the seam cutter yoúll save a lot of time. Once yoúre at them seams, it makes sense to do a proper job and remove all of the old compound. At the same time you might consider widening the seams a bit. For longevity, the seams can practically never be too wide. But of course you want to have something in reserve for the next seam renewal. By the sealant having gotten "tired" I assume you mean it has become hard. It would be interesting to know what product it was and especially when it was applied.

Posted by Don Braymer (Member # 388) on 08-01-2000, 07:09 PM:

Per wetting the seam side walls:
We taped the deck, filled the tubes, squeezed them out into the seams, and then ran a putty knife over the positively filled seams. This captured a bit extra which we trowelled into additional open seams with the putty knife.
With this method, we are pretty sure of a good sidewall bond (on the aft 1/2 of my deck), but it took two guys running around pretty fast to do a gallon of Detco, get the tubes all squeezed out, and get all the seams trowelled in.

Working with this gooey sticky stinking mess is not a task for the faint hearted.

Get plenty of paper towells, spares for all equipment, because if it gets covered with the slime, there is no easy way to clean it off.


Posted by Jamie Hascall (Member # 335) on 08-01-2000, 10:07 PM:

Sadly I know your words make sense and that I need to rev my little circular saw up again and widen my rabbetts. Their depth is somewhat controlled by the thickness of the teak only being 7/16" when new and probably a maximum 3/8" now. My choice of groove dimension was a function of the width of the previous seam widened and deepened slightly. I feel that I don't want to cut all the way through so that there is a bottom to the rabbett that is attached to the adjacent plank. The 1/4"x1/8" dimension actually matched the recommended seam width in the literature from Boatlife. However, we know how well both their advice and their product served me. Of course a wider seam will have a greater capacity to stretch and running the saw down the seams to widen them will have the added benefit of again cleaning the wall surfaces to fresh uncontaminated wood. It feels inevitable that the project will expand to fill the time I had allotted for cruising. So it goes.

Thanks again,

Posted by Jorma Salomaa (Member # 793) on 08-02-2000, 04:16 PM:

Sorry Jamie if I spoiled your party. The manufacturers have to be conservative with their guidelines, that́s understandable. But of course we dońt HAVE to believe them. On the other hand, as you said, it makes sense. By the way, we cańt disregard the fact that your seams actually had made it for many years with those dimensions.
ACB, I forgot to comment on the goop to put into your stanchion seams. Well, I wish I knew. You could probably use a two part polysulfide which the manufacturer has intended for wood deck caulking. That́s the safe way when one listens to those who have done a lot of seam work. As for my own seams, I still intend to use one part polyurethane, and probably two different kinds. Unless someone tells me not to, and gives a reason. Any thoughts and especially experience will be welcome.

The one-part/two-part question somehow intrigues me. There are both polysulfide, polyurethane and even silicone two part sealants I believe. I have understood that the only benefits of a two part elastomer are simple production (thus cheaper), and rapid cure. I assume that the curing mechanism in a two part product is such that the integrity of the sealant bead is not that much affected by movement during the curing process. There is a standing warning against even slight seam movement during the curing of a one part sealant. Do the two-parts have any other merits?

The above reminds me of Jamiés seam problem. Is it possible that after a long period of high moisture the air suddenly dried dramatically and the planks continued to shrink while the sealant was still curing? It could be manifested by the seams more in shadow being better than the rest. Just a thought.

Posted by thechemist (Member # 1468) on 08-02-2000, 08:37 PM:

The chemistry of the one-part and the two-part things are not the same. The one-part polysulfides are cured with zinc or calcium peroxide which is moisture-activated and generates an oxygen which reacts with the mercaptans (sulfur-hydrogen) groups on the ends of the polysulfide resin molecules to create more disulfide linkages. The residue of the curing agent is zinc oxide (a mild fungicide, but only effective if there is one to two percent by weight in it, I think) or calcium hydroxide which will eventually go to calcium carbonate with the carbon dioxide in the air. The two-part polysulfides create the same disulfide linkages as the one-parts (which are peroxide-cured) but use either manganese peroxide or lead peroxide. Only the lead-cured systems leave a lead oxide residue (maybe five to six percent, an adequate amount) in the rubber to stabilize it against long-term attack by small life. These materials are not known to migrate out of the material with time, so there is not an environmental toxicity issue.
There may be silane-terminated polysulfides, I do not know. The urethane caulks are silane terminated, and there are two or three reactive groups on each silane ( a compound of silicon, not a silicone but similar, only these stick.). These things react with the moisture in the air and hook onto each other. Thus, there may be molecular junction points where several of these things come together. Unfortunately, not all of these reactive groups can react with each other, and so there are some reactive groups left over which are unreacted. These hold water molecularly, and that water can, long-term, re-attack and break the bonds between the other molecules at that point. Long-term mechanical stress would not be good for this kind of situation.

There can also be isocyanate-terminated urethane caulks, and these would react in a different way. With the cellulose of wood they would create urethane linkages by reacting with the hydroxyls of cellulose, although water is hydrated on those hydroxyls and water plus isocyanate gives an amine which will not stick to much of anything on the wood. Enough isocyanates would dehydrate the wood surface and then some isocyanates will bond, and those bonds are quite stable. On cured or uncured epoxy resins the isocyanates will react with the amines or hydroxyls left after the epoxy curing reaction, and the urethane or urea linkages are quite stable. These types of elastomers do not inherently have stability against small life, although things can be added. It is unlikely, since they are expensive.

I am aware of silicone "bathtub and shower" caulk that has some mildewicide added, but to my knowledge no sealant manufacturer promotes such properties, so I think they do not go out of their way to make it more biologically stable. What the long-term biological stability of urethane elastomers is I do not know, although I think it pretty good. I do know there is a bacterium that lives in the pits of aged paint, even urethanes, but the attack rate is slow compared to the rate of fungal rot of wood.

Posted by Jamie Hascall (Member # 335) on 08-02-2000, 09:23 PM:

In my discussions of this subject with Steve Smith, he made reference to the high cost of the raw polysulfide compounds and that the economics of selling tubes of compound demand that there be a smaller percentage of the polysulfide and more fillers be in a tube of single part than is generally in a kit of two part polysufide, and thus it is inherently less elastic and strong. I think this is inevitably true since the majority of people will only tend to choose one type of product over another if the prices are competitive, and so the best compromise of product and profit will be struck.
Jorma, Don' t worry about spoiling anything, the truth and a dose of common sense is always worthwhile. The real benefit is that by widening the seams, I will get the clean wood that I need to make this really fly and it will actually be easier than trying to sand the seams to clean the walls. I feel this discussion is really helping this process fall into place.

I have been wondering what your thoughts were on the dispensing with the bondbreaker and the caulking directly onto the wet CPES? It feels like a good procedure to me (especially the wet CPES) but I still wonder why so many manufacturers specify a bondbreaker. I know that miles of seams were done with no bondbreaker in the past, and lasted quite well, so where does one get the greater benefit.


Posted by ACB (Member # 1213) on 08-03-2000, 03:39 AM:

I would use a bond breaker; if only because it makes it easier to get the goop out when you next have ro repeat the process. Did mine 3 times in 12 years before I gave up and Coelan-ed and painted them, but admittedly mine were 1 1/4" teak; I was therefore caulking over cotton in a V shaped seam, so the situation is different.

The last two lots were Boat Life, followed by Sikaflex.

Thinking about it I'm pretty sure that the stuff round the stanchions is Boat Life. It is cracking and pulling away from the sides a bit. Looks "perished".
I think the arguments for using a bond breaker are convincing.

Posted by Jorma Salomaa (Member # 793) on 08-03-2000, 08:07 AM:

Thanks for another healthy dose of chemistry, Chemist. I think your posts in general are extremely valuable and provide the rest of us with tools for better thinking. I keep them in my mental tool bag. Sometimes though, I dońt exactly know how to use them. Would you please help me to use your last post in deciding what the practical benefits of a two part caulk over a one part caulk are?
As to bond brakers, imagine we have a thick rubber string with a rectangular cross section. Let́s say it is 10 cm long. To extend it by 20% or 2 cm we need a certain force. Then we take two planks, place them side by side and glue the rubber string down to both planks at 90 degrees. Then we pull the planks apart the same 2 cm. If there is a gap of 2 mm between the planks, the string bridging that gap will now be extended by 1000% and the force exerted on the rubber will be ten-fold compared to the non-glued string. Not quite, because the upper surface of the string is not glued to anything but the stiffer the rubber, the more evenly this new force will be distributed across the strinǵs section. In principle, the same happens to a caulk bead that is glued to the seaḿs bottom. Hence the bond braker whós mission is to keep the rubber unglued like our initial rubber string. Cotton in a V shaped seam is a bond braker in that it will give with the sealant it is glued to.

About caulking on fresh CPES: My earlier trial was on CPES that had been applied 3 hours earlier. After a month, I could see no difference in adhesion to another bead of sealant that had been laid on CPES that had been allowed a curing time of three days. Now we have to remember that my testing techniques are not very sophisticated and this trial was not repeated. Thus, it should not be considered as conclusive. Also, we have to bear in mind what Chemist said earlier of different types of sealant reacting with the primer. Just yesterday I laid a bead of Sika 290 DC on oak that had been treated with CPES barely 15 minutes earlier. I have to wait a few days before tearing it off. Actually, I should probably tear off only half of the bead after three days and the rest later on in order to give the eventual reactions the time they need. The same bead continues over oak that was treated with CPES 3 weeks ago and not even wiped off after that. Another part of the substrate (with 3 weeks old CPES) I wiped with acetone and a third section of the same I lightly sanded. A fifth section of the substrate (which is the actual covering board of my boat) is untreated oak. As a control I put a bead of 3M 5230 beside the Sikaflex.

Unfortunately, Ím only a boat owner. If I were a scientist my testing would be better. But maybe I would́nt have a boat. I dońt know which would be better.

In your place Jamie, I would wait for the Smith sealant to arrive and in the meantime treat a selected plank on your deck with CPES every day in 4" passes marking the date beside it on the plank. Then yoúll have in 10 days 10 pieces of substrate, each with a different curing time for the CPES, and may be another primer, and maybe still another with only acetone, and maybe another with no treatment at all. The same can be done with the bond braking tape, although I think the only possible problem there would come from non evaporated solvents damaging the tape.

This way, Jamie, wéll be wise men by Christmas with decks still waiting to be caulked.

[This message has been edited by Jorma Salomaa (edited 08-03-2000).]

Posted by Jamie Hascall (Member # 335) on 08-03-2000, 12:48 PM:

Your testing is encouraging and I would like to replicate much of your work with the materials I'm working with. However, I have been given the great gift of a period of continuous sunny weather which here in Seattle is truly extraordinary. Thus, I need to make hay (or lay caulk) while the sun shines. There will be a point where I will have to take the combined wisdom of this forum and use my best judgement and just do it. I'm one of those people that loves to work on the boat, but really has it for the purpose of getting away. I really don't want to squander this years vacation since I didn't go cruising last year because I was working on my decks (I can hear the Canadian Gulf Islands calling me).
The role of the bondbreaker has always made sense in looking at the elongation issue. I'm afraid my reluctance at using one this time around was the grasping for any way to make the job have a few less steps, so when someone I trusted said it was unnecessary, I wanted to air the question and get the reaction of the forum. I truly have found how much easier the bondbreaker made the stripping out of the old compound, but using it for that purpose is a little like getting married with the number of a good Reno divorce lawyer in your pocket. If I'm going to go to the trouble of installing it, it's on the assumption that it will materially effect the longevity of the caulking and will keep me from having to do this again. My question now is what the best order is for applying it. Since my grooves are of bare minimum depth due to plank thickness, I use 3M Fineline tape for a bondbreaker. Last time I primed the seam and then installed the tape, but if the Smiths representative is correct, it would be better (and probably less messy) to apply the tape before priming with the CPES. Assuming that the CPES has little or no adhesive qualities, or bonds poorly with the tape, this seems like a viable procedure. I truly would like to get the highest level of elongation out of my compound because as Steve Smith had pointed out, the real job is to avoid failure of the bond. Once that has happened, it's all downhill.

As always, thanks again,


Posted by thechemist (Member # 1468) on 08-03-2000, 04:00 PM:

So many little time...
A bondbreaker strip of some narrow tape in the bottom of the seam cannot hurt, it seems to me. This appears to be a mechanical issue rather than chemical. The amount of movement of the seam is unknown. Some seams will be lightly stressed and the bondbreaker would not be needed. If an area happens to be highly stressed by some movement beyond the average, it could be useful.

What a bondbreaker tape does is prevent a high-strength bond to the bottom of the seam, allowing it only on the sidewalls. Thus, there is no local stress in the bottom of the seam in the corner where the two separate pieces of teak come together and move apart. Without that , the sealant would tear maybe a little right there in the corner, but not further in. There would be a region that would be stressed to its elastic limit, possibly peeling from the seam sidewall in a small area. That peeling stress could propagate with repeated stress cycling. Thus, it should make a more reliable joint.

Sheet metal rain gutters are sealed in a similar way. The separate pieces of sheet metal are pop-riveted together, then a piece of tape perhaps an inch wide laid over the metal lap and rivets, and then a couple of inches or more of an elastomeric sealant painted over the whole area. The theory is that the gutter may move one way or another a sixteenth of an inch, but it does not leak because that amount of movement is spread over the inch of sealant floating on the bondbreaking tape and it, stressed lightly, lasts forever.

Jorma, you wonder how to apply that last dissertation about one- or two-part sealants...and it is not an easy thing to answer. I will try. Depending on the chemistry of the primer (if any) on the wood, and depending on the chemistry of the sealant as regards how it cures and develops adhesion,, there will be a better bond (or maybe none) if these things have some points of chemical compatibility. The manufacturers will not tell us such desirable details, so I guess the main lesson to learn is that there are more than a few things going on, and there can be more than one reason for adhesion failure in these applications.

The most reliable adhesion-promotion mechanisms in single-part sealants, I think, would be the isocyanate-terminated resins (sometimes called prepolymers) because they should develop the most reliable bonds to epoxy primers. I doubt you will find any single-part caulk salesperson who will even know. The MSDS may or may not reveal such deep, dark secrets. Two-part sealants would be a different story. Hope that helps.

Posted by Jorma Salomaa (Member # 793) on 08-04-2000, 04:42 PM:

What Ím trying to figure out, Chemist, is whether there is a good reason for me to stay away from ANY one-part sealant. A possible reason would be for instance that, with the present know-how, it is not known to be technically POSSIBLE, at any cost, to produce a sealant as good as the best two part polysulfides. "Good" to me is a sealant that retains its elastic properties for 20 years outdoors and unprotected. I have seen that some two part polysulfides can do it. So the question is do you think it can be done?
Because if it can, then someone has already done it, or will do it very soon. Then it is possible that the best one part sealants already are that good. In that case, we only have to find them.

My impression is that by sheer volume, the silicone sealants hold the first place in the building industry to-day. But that refers mostly to glazing applications (windows). Then come one part polyurethane sealants wich are modified in many ways for different uses. A further impression of mine is that two part sealants are used only when their typical curing characteristics are needed, e.g. when joint movement is expected during curing. Thus, a one part polyurethane is what normally is used when it is not about glazing.

Ím speculating here that the best one part polyurethanes (or polysulfides, for that matter) may already outperform the competition without our knowing it. Do you think this is possible, Chemist?

Jamie, you were considering installing the bond braking tape before applying the primer, in this case CPES. In that case there will inevitably be some primer on top of the tape. If it doeśnt adhere to the tape much, then no problem. But if it does (which I doubt) then the tape is not going to do what it is expected to. Thing is, I dońt know. We have to try it.

Posted by Jamie Hascall (Member # 335) on 08-04-2000, 08:00 PM:

Jorma, when I have stripped out the one part polysulfide from my failed attempt, the one thing I noticed was that the tape was often left behind at the bottom of the groove. This leads me to believe that the surface is so smooth that there is little opportunity for a bond to form with it. Even with the sealer between, I don't think there is enough adhesive capability to the CPES to matter. However, I have just received my boxes of compound from Smiths and will do a few experiments before trying it on the boat. I'll let you know what I find out.

Posted by thechemist (Member # 1468) on 08-04-2000, 09:52 PM:

Jorma, you ask a good question and I do not know how to answer it. Being cynical, and seeing how so much of modern production is not quality-motivated, but rather for the maximum short-term profit at the lowest possible quality that the customer will just barely tolerate and still come back and buy it again, or so the people in the Marketing department calculate, I wonder if there is some way to tell based on the cost of the product. With urethane sealants, (having pondered while running through some of I conclude that I do not see how to equate longevity with cost. Biological stabilizers add little to cost, and so one cannot infer the presence or absence from the cost difference of eight or ten dollars U. S. per tube.... The stuff is opaque, so UV absorbers and their cost do not enter into it....
Perhaps if they added a lot of inert plasticizers to cheapen it we could detect that. Tell ya what. Get a sheet of brand-new clear or black four-mil (100 micron or so) polyethylene plastic sheet. Spread out about 1/2 meter x 1/2 meter. Squeeze out an entire tube of the caulk, and spread it out to a thickness of about 1/8 inch, 3mm or so. Let it cure a week. peel it off the plastic should release from NEW plastic. cut it up into six-inch (15 cm) squares, and weigh them on a handy gram scale, one that reads to the nearest tenth of a gram. You may have to weigh them in two portions. The whole tube should weigh around 250-400 grams. Then spread them out and bake in an oven, about 100 degrees C (212 deg. F) for a week, weighing daily. (take out, allow to cool, weigh, write down weight and time, put back in oven, all spread out.) After a week, draw a graph.

Good quality stuff should not contain much volatiles. A few percent of weight loss quickly indicates normal reaction by-products, while a slow and steady weight loss is a bad sign. Even the two-component polysulfide has a certain amount of plasticizer, but it is mostly a type that does not come out readily. The urethane caulks should, to my mind, not need any plasticizers. Mechanical properties should be fully controllable by the molecular weight and cross-linking profile.

One-part polysulfides are another matter. One can use too much crud in the formula due to the inherent high price of polysulfide resins due to the only viable commercial synthetic method developed over 50 years just is really expensive to do. There, you could guess a high-quality one-part polysulfide caulk might retail for U. S. $20 per tube (the underlying assumption is that a distribution chain needs to buy it for 40 cents to have it retail in the store for a buck, and a manufacturer wants to spend twelve to twenty cents for raw materials if he is to get only forty cents for it) . I am not aware of any one-part silicone sealants that have performed well long-term on wood. There may be some, it may depend on surface prep and priming, but I have heard a few horror stories and found no positive evidence of products that do last on , for instance, teak decks, although I have not gone to any length to research silicone sealant applications on wood. I think the tasks can be done other ways than with silicones, and I do not work for Dow Corning or G. E. Silicones and so have no access to affordable silicone raw materials. Guess I'll never know about that possibility......

Posted by Jorma Salomaa (Member # 793) on 08-05-2000, 04:48 AM:

Thanks Chemist, wére moving in the right direction. Keeping the isocyanates etc. in the oven for a week does need approval from SWMBO. But that will tell us the amount of volatiles. Some of those are presumably needed to get the stuff out of the tube in the first place and some maybe serve purely as fillers. Is it then that the plasticisers are also volatile and can be detected this way? From what you say I infer that plasticisers are, at least to a great part, used to do the things that really should be accomplished by molecular weight and and the crosslinking profile. Maybe poor redress also is a sign of a high content of plasticisers? My vice in the garage tells me that the more elastic sealants tend to have more compression set (but there are interesting exceptions like e.g. Sikaflex 298 with a low modulus and still with good redress).
My machine did not load the page you referred to but Íll try at another time. And now dear fellow forumites, Ím going sailing for a day. I hope wéll hear from your trials with the Smith polysulfide shortly, Jamie.

[This message has been edited by Jorma Salomaa (edited 08-05-2000).]

[This message has been edited by Jorma Salomaa (edited 08-05-2000).]

Posted by Kel (Member # 1992) on 08-05-2000, 10:38 AM:

This is a very interesting and informitave thread especialy since I am about to recaulk the decks of my 63'motorsailer, 1" teak over 3/8"plywood. The decks were recaulked by the previous owner with Detco 2 part caulk about 6 years ago and are letting go.Bad prep?? I was planing on using Teak Decking Systems one part silicone blend deck caulk? This product is very widely used here in south Florida and the east coast with very good success from what I have heard. I would sure like to here about any experiences and/or problems with this product before I use it

Thanks, Greg

Posted by Kel (Member # 1992) on 08-05-2000, 10:39 AM:

This is a very interesting and informitave thread especialy since I am about to recaulk the decks of my 63'motorsailer, 1" teak over 3/8"plywood. The decks were recaulked by the previous owner with Detco 2 part caulk about 6 years ago and are letting go.Bad prep?? I was planing on using Teak Decking Systems one part silicone blend deck caulk? This product is very widely used here in south Florida and the east coast with very good success from what I have heard. I would sure like to here about any experiences and/or problems with this product before I use it

Thanks, Greg

Posted by thechemist (Member # 1468) on 08-05-2000, 11:36 AM:

Use some other oven than the one you cook food in! Here in the lab I have ovens...and when I think of baking something in the oven it is one of the laboratory ovens. I forget that in your house there is only one oven.......

Go to someplace where they sell recycled, used electric ovens torn out of a kitchen, get one of those, and set it up in your garage. drill a dozen or so quarter-inch holes in the bottom, and the top, so outside air can convect in and out, but not so much that the heating element cannot keep up with it. The holes will likely go through fiberglass insulation, so sleeve them with quarter-inch copper tubing to keep the fiberglass insulation from clogging the holes. (Convert the quarter-inch dimension to something nearby in your metric system.)

Don't bake chemical stuff, adhesives or glues of any sort in a food preparation oven.

Posted by thechemist (Member # 1468) on 08-05-2000, 12:01 PM:

Back to business.....
The volatiles are there for more than one reason. A certain amount of volatile solvent is often needed, as Jorma surmises, to get the stuff out of the tube more easily, as higher molecular weight resins are more viscous, and as one loads the formulation with mineral fillers beyond the amount needed for better mechanical properties, one gets a stiffer, more viscous product with unacceptable flow. Both volatile solvents and "non-volatile" plasticizers are used to reduce the viscosity to get enough flow to make the material workable. On a long-term basis, more plasticizers are necessary where a formulation has been overloaded with mineral extenders as part of a design-to-cost philosophy. The result needs to have certain elongation properties. The plasticizers can get that result. The irony is that cheap plasticizers will migrate out with age. Expensive plasticizers will not , for all practical purposes.

Compression-set is an indication that the bond between the sealant and primer/wood sidewall may not be that permanent, and may fail under stress, but an over plasitcized formulation will be less likely to exert a constant stress on the adhesive bond, as the opposite of compression-set is cold-flow, and that will relieve the stress.

Constant tensile stress on an adhesive bond between an elastomer and any wood, with any primer, I would expect to eventually lead to failure, as small life will eventually find its way into the parallel wood fibers at the interface and eat them, giving rise to failure where the wood/sealant interface is exposed to the air and the elements, thus providing nourishment for small life.

Even protected completely from the elements, I would expect a teak-elastomer joint under constant stress to fail, as the natural oils of the teak would migrate around in the wood and eventually find their way to the stressed interfacial region.

Excessive "teak oil" (a trigyyceride, similar to other vegetable oils), applied to a caulked teak deck in hot weather does chemically attack polysulfide rubber, likely at the disulfide linkages, causing depolymerisation and bond failure. I have seen this happen and verified in the lab oven by accelerated ageing that it does happen. This is not something that happens with casual use, but a LOT of that stuff applied over a long, hot summer can do damage.

It follows that the natural oils of teak, over many years and to a joint under continuous tensile stress, would do the same thing. It folllows from that, that joint and caulking design should be such as to leave the joint under neutral to compressive net stress, with only occasional tensile stress.

[This message has been edited by thechemist (edited 08-05-2000).]

Posted by Jorma Salomaa (Member # 793) on 08-06-2000, 05:46 PM:

OK,Chemist. I waśnt going to storm into the kitchen with my pancakes yet. Maybe I have to furnish a kitchen of my own in the garage as per your instructions. Especially as it has been mentioned that my bed might be moved there soon...
I think we are gaining some understanding of how deck caulks work and what should be expected of them and that́s thanks to you Chemist. However, wére only sailors and part time caulkers and your posts may be often cryptic to us. By digesting them carefully each of us can draw some conclusions. There iśnt a lot we can do ourselves to assess whether a particular sealant is suitable for the intended use. Íll venture to propose a few things we can do, with some question marks:

1. Even if the sealants normally show good adhesion we can stumble on a product that performs worse than we expected. A crude adhesion test is simple, and especially enlightening when several brands are compared. The result is short term and long term adhesion performance is complicated. Is there any way of assessing it long term?

2. Elasticity is also easy to check but to be useful we need to have an idea of how much of it we want. The long term capability of the sealant to retain it́s elasticity is maybe the most important factor of all but alas, therés no way we can predict it. Or is there?

3. Redress from compression or extension is important and easy to check out. Especially compression set. But what is the proper way to perform the test?

These three properties come to my mind as the most important among those that we are able to assess. Is there more?

Posted by thechemist (Member # 1468) on 08-06-2000, 07:07 PM:

Jorma, I am really trying to keep it simple. If I seem confusing please ask me for the exact thing that needs to be clarified. It will usually be just where it was not confusing, and then became confusing.
You have used the word redress, which may have the meaning in English of repayment for damages. One seeks redress for damaged property. In your context I think the word is recovery, where something stretched a distance and released should recover to the original dimension.

I will comment about your points. Accelerated Ageing is used to predict in the short term what will happen in the long term. It is based on the idea that failures are the result of chemical reactions, and the rate of a chemical reaction doubles roughly every ten centigrade degrees (that's another one of those Chemist's Rules of's sorta close most of the time). the rule only holds above the Activation Energy for that particular reaction, and below that energy the reaction essentially doesn't happen. different reactions have different activation energies, and the temperature of atoms and molecules is an expression of their average energy. At absolute zero they theoretically have no energy (well, technically they have the ground-state energy but that is quantum mechanics and I don't want to hurt you so let's just pretend we didn't go there) (fine line between pleasure and pain, isn't it?).

Anyway, we can do things similar to what we expect to see long term for the sealant, and then do it hot and observe what happens. To do it right, we would age at different temperatures and look at the failure rate. As long as the failure rate was linear with temperature (well, on logarithmic graph paper it looks linear...) we would not have activated any new failure mechanisms which would falsely accelerate the failure rate.

Elasticity is hard to predict (in english we can stretch things to make puns) unless it is due to things that should not be there and that is something we can do easily. Volatile plasticizers should not be in sealants, and to my mind no plasticizers at all should be in one-part polyurethane sealants.

Another test for things that are unreacted and will migrate out: Cure a 3 mm thick pad of sealant a week, then put between two sheets of plain white paper. Put a couple of blocks of wood on either side, clamp to squeeze a bit, leave over-night. Inspect for stains on the paper. If none, put in the oven for a day, then three, then a week at 75 deg. C. Howzat?

[This message has been edited by thechemist (edited 08-07-2000).]

Posted by Jorma Salomaa (Member # 793) on 08-07-2000, 05:56 PM:

Yes Chemist, recovery is what I meant by redress. As for your "cryptic" posts, please make no mistake, they are highly appreciated. It́s just a pity that Ím not always capable of translating the facts you describe into practical conclusions.
Accelerated ageing, as you describe it, is then probably our only means of predicting what is maybe the most important property of a deck caulk - the way it retains its elasticity. As we dońt know how many heating cycles correspond with any number of years, we then obviously should just compare a number of brands. Do you think it is feasible to make test specimens by glueing the sealants to wood, measure the elongation as per your water bucket method, heat, let cool and measure again until you can see differences? Maybe recovery from elongation can be monitored in the same go. The paper test also sounds useful.

In the meantime here, I think Kel deserves an answer to his question. I dońt know anything about the one part silicone blend deck caulk you describe. I assume it cańt be a normal silicone sealant. Anyone know?

Posted by thechemist (Member # 1468) on 08-07-2000, 08:33 PM:

Jorma, I cannot help Kel with his question. Perhaps someone else can, and Kel should also post his question on usenet newsgroups that have to do with boats. The usenet was here before the web, and actually seems to lie under it, although it now seems to be squished off into a corner. Lacking pictures (except for the alt.binaries stuff, where the naughty pictures are, so I hear) , it is well suited for slow modem connections around the world and thus there will be a wider spectrum of reader.
As for the accelerated ageing tests, I doubt it would be practical to set up with heating and cooling and stretching and so forth for many cycles, but as a static load for a day or a week, that may be practical. Unfortunately, many epoxy primers (such as CPES) would be expected to get weaker with temperature, so that may be a limiting factor rather than the elastomer properties. Simpler and faster is to screen your candidates first by looking for the presence of something that should not be there, and is easy to detect. Try the paper test.

Posted by Kel (Member # 1992) on 08-07-2000, 08:45 PM:

Thanks for the response, you guys are very thoughtful. I'll keep researching and do some testing of my own
thanks, Greg

Posted by Jamie Hascall (Member # 335) on 08-07-2000, 09:36 PM:

In my looking for a single part solution to this caulking problem, I talked to the representative from Legendary Yachts of Washougal Wa. who are making amazing modern versions of classic Herreschoff designs. He said that they had settled on the Teak Decking Systems product for their production after having failure with other unspecified caulking compounds. We were standing beside their ketch Araminta which had beautiful decks made of Port Orford Cedar. He then confided that even this caulking had started to have some small bits of separation after a couple of years in place.
I cauked a hatch cover that was new wood construction and had been built indoors with the Boatlife Sandable Silicone which they also market for deck caulking, and have generally been satisfied with the result. However, I have just noticed spot failure on two seams and am thus a little concerned. Of course these are two different companies and products so I cannot compare them with any veracity.
I feel that if we are addressing the issue of renewing the caulking in an existing deck, we have to assume that there is no way to bring the moisture content of the wood down to the levels of new indoor construction. Therefore the shrinking that will inevitably happen on a very hot sunny day is going to truly test both the elasticity and tenacity of the compound. The experience I have had with the silicone compound that I have used and Legendary's experiece with TDS leads me to believe that they may both be generally good products, but I so not have a good feeling about the margin of safety that they have to give. Especially since I have already done this job once and know how much work I am in for to do it again, I feel that going for the proven highest elasticity and bond is the only prudent path. I also have to take very seriously the issue of silicone contamination that these other compounds present, which was articulated earlier by the Chemist.
The testimonial for TDS by the Grand banks owner was definitely persuasive and I feel that it may be a great product, but I have truly been bitten once, and I'm now very shy of anything that I can't truly feel certain about.

Good luck with your project,

Posted by Jorma Salomaa (Member # 793) on 08-08-2000, 04:59 PM:

Íve understood that the silicone sealants generally perform excellently where UV and weathering in general is a factor, but according to others and also my own findings, adhesion is a weak point. Which is not to say that there could not be a formulation with good adhesion. I just havént ever seen a silicone that would glue as well as the polyurethanes and the polysulfides that Íve tried.
To Chemist: Thanks for your comments on accelerated ageing. Ím going to do the paper test. I also have another question to you. Íve noted that some sealants are more springy than others when cured, i.e. they recover immediately when I press them with my finger nail while others take a few seconds. Generally the high modulus sealants seem to recover fast, but Íve found one low modulus, 25 Shore A polyurethane product to be exceptionally springy. It also has better recovery from a three days 50% compression test (after a week of recovery) than the other low modulus products. Can it be an indication of something we should know about?

Posted by thechemist (Member # 1468) on 08-08-2000, 07:59 PM:

Very interesting, very interesting.......
What is that stuff?

Posted by Jorma Salomaa (Member # 793) on 08-09-2000, 05:59 PM:

It́s Sikaflex 298. The information I had was from the Sika Canada web pages. Now I read in a brochure that it́s 35 Shore A and I think Íll believe that. It is sold as a bedding compound. It́s very similar to Sikaflex 290 DC (also Shore 35) which is the deck caulking stuff but it has a higher reading for elongation at break (600% versus 250%) and a lower reading for tensile strength (1.0 versus 1.3 N/mm2). This is in line with my water bucket test results. My paper test shows noticable and similar leaching for both after one day in the vice. Recovery from compression is remarkably good for both. They share the same kind of "springyness" that other polyurethane sealants dońt seem to have. According to the factory, the 298 needs a somewhat higher application temperature and can take a slightly lower service temperature. Interestingly, the 290 DC is said to have a shrinkage of 7%.
To me both look like promising candidates for a deck caulk, although I would prefer a still lower modulus for my oak plank seams.

By the way, I asked the representative here what they tell a client when he asks about the expected service life of a Sika sealant. The answer was 15 years, at least.

I would appreciate to hear your thoughts on this, Chemist.

Posted by thechemist (Member # 1468) on 08-09-2000, 09:36 PM:

well, the elastic modulus is the amount of force required for a given amount of elongation. Pick the one with the lower modulus, cross your fingers and hope. Do the paper test and expect when it is fully cured that nothing bleeds.
Your durometer numbers, 35 Shore A and so forth, are hardness, not elastic modulus. They may be related, but elastic modulus is separately specified.

No deck seam is going to elongate six hundred percent, or even two hundred. Vote for the one that stresses the primer/wood interface the least, by having the lower elastic modulus. If they all have the same modulus, then go for the higher elongation one. good luck.

[This message has been edited by thechemist (edited 08-09-2000).]

Posted by Jorma Salomaa (Member # 793) on 08-10-2000, 04:25 PM:

Thanks Chemist. This has been an interesting topic.

Posted by Jamie Hascall (Member # 335) on 08-14-2000, 01:03 AM:

Well I've spent the better part of a gorgeous weekend cleaning the grooves in my deck with my trusty little 4 1/2" Porter Cable saw and bringing the dimensions up to a 1/4" x 1/4" size, but decided the weekend wouldn't be complete without flogging this equine just a little more.
I've also been talking to some locals who have done a bit of two part polysulfide caulking, and to a man they all questioned the use of the bond breaker. The chief concern,(and one which I have pondered) is not the issue of failure due to restricting the elongation (the logic of which did not escape them), but that the detached lower surface provides a conduit for the movement of water when spot failure inevitably occurs over time. The ability of water to travel horizontally through the deck is of real concern as it leads to the spread of small life and the possibility of rot occuring at points away from the site of the leakage. Does anyone out there who has used Detco or other two part compounds recall the recommendations for those products or any long term experiences with them? I am going to call Steve Smith and Detco in the morning to try to get some further clarification on their procedures.

The other thing that I have found is that there were still a few sections of seam that contained the original 30 year old caulk. When I compare the resiliency and strength of this to the Boatlife that I used last year, it is obvious why the Boatlife failed. There is so little strength to that compound that even if the adhesion had been there, the compound would have torn under low levels of stress. When the old sections of seam were in spots hidden from wear and scrubbing, the 1/8" wide x 1/4" deep seams were almost always intact. I'll let you know what I find from the manufacturers, and I should be doing some experiments in the next day or so. Now, if I could only put my job on hold for a few days...


Posted by Jorma Salomaa (Member # 793) on 08-14-2000, 05:27 PM:

The horse seems to be breathing...
Jamie, the old polysulfide seams Íve seen never had a bond braking tape. If therés cotton in the seam, then the cotton is a bond breaker. Cotton is also a good water conduit because it works like the wick in an oil lamp. Íve seen polysulfided seams that did not have any cotton in them, and they did́nt have any bond breaking tapes either. I guess the whole idea comes from the more recent one component sealant manufacturers because these compounds can use all the help they can get in order to keep the glue lines unharmed. I may be wrong but I think this is because they are of a higher modulus than they really should, and that again is because a low modulus one part sealant can remain sticky and it may not be sandable.

Most of the two part polysulfides I have seen have a very low modulus of elasticity. If I would have to estimate their hardness I would say as soft as Shore A 20. Interestingly, they apparently were not sticky and they probably were sandable. If this can be generalized then I think here is the reason why they worked without any bond brakers. Probably these sealants are becoming obsolete for the reasons that Chemist very well explained earlier. If Smith́s polysulfide sealant is that type of product, and I think it is, then therés a good chance that you can forget about the bond breakers.

Posted by thechemist (Member # 1468) on 08-14-2000, 09:53 PM:

The horse that would not die......
The two general types of polysulfide resins have a cross-linking difference of a factor of four. That would give Shore A durometers of perhaps 25 and 45, although either of those could be run up or down ten or fifteen durometer points by filler content and plasticizer content. For economic reasons commercially affordable materials sold through distribution will have to be over-plasticiized or extremely pricey, as I went through previously. Jorma, if you know of a particular two-component polysulfide in a seam with a 25 Shore A durometer, see if you can do the paper test on it, with a weighted sandbag or something, even overnight. I find it here in the lab to be a decent test for gross plasticizer migration, and that is what we would be looking for. I do not think anyone made two-component seam sealants that soft in this country, but they might have in Europe.

This is interesting, because that might have been the market opportunity that opened the door for the one-part urethane seam sealants in the U. S., even though they shot themselves in the foot by making them pastes instead of liquids. That, and inadequate seam priming, led to the failure of one-part sealants. An excessively cheap formulation, leading to shrinkage and hardening, plus being a paste and inadequate seam priming, led to the legendary earlier failures of one-part polysulfide seam sealants. The widespread use of "teak oil" products exacerbated the failures of any and all seam sealants, with some formulations more sensitive than others. I do not know of ANY sealant manufacturer who specifically tests their seam sealant/primer/teak interfacial bond for "teak oil" resistance. That's a really scary thought, having to do that with everyone else's uncontrolled formulations.

It follows that a market-competitive seam sealant should be fairly low-durometer (for low-stress elongation and so a bond-breaker at the bottom is not necessary, since that allows water migration from a point failure to larger regions) and not have any migrating plasticizers, and be a liquid so it wets the seam sidewalls, and somehow address the wood priming issue, either inherently or with a separate primer. "Teak oil" resistance can be dealt with by a non-warranty if it cannot be dealt with chemically.

I think I will write up a proposal to management to authorize a new product development program (you would not believe how much paperwork is involved in something that was not the idea of the Marketing Department). This looks like an interesting opportunity........capturing back market share........

Posted by Don Danenberg (Member # 946) on 08-14-2000, 11:40 PM:

Ya know..., when I was a kid (13-14), it was my job, after school, as second oldest of ten, to ride the yearlings as far away from the barn as I could get before I was thrown, rolled off in the Alfalfa, or scraped off on a tree in the orchard; It never occured to me then, that I was expendable.
I just realized that a 'bond breaker' might actually be usefull in deck seams if one is planking 'pretend' planks over plywood.
Is this why the 'bond-breaker'? Don

Posted by thechemist (Member # 1468) on 08-15-2000, 01:54 PM:

Now I'm confused. Why put bond-breaking tape in a fake seam bottom? If the seams are only pretend, then the sealant will bulge up and down with thermal expansion and contraction, and the sidewalls of the "seam" never move because it's only one piece. In that case the bond-breaker tape in the bottom would be a liability because it would not contribute to holding the rubber strip in the "seam" after small life had degraded the sidewall bond. Indeed, if excessive "teak oil" products or other strange stuff were used, such that the sidewall bond failed at some point, a bond across the bottom of the fake seam would be all that remained to hold the rubber in place. I am beginning to think that bond-breaker tapes at the bottom of seams are more liability than asset, on boat decks.

Posted by Jamie Hascall (Member # 335) on 08-15-2000, 02:58 PM:

****, this is one tough old horse.
Today I've talked to the folks from Smiths again, and I also found the Detco Marine web site which had full application instructions for their two part polysulfide caulk. Once again it was reiterated from Smiths that they saw no need for a bondbreaker for their product. Likewise Detco made no mention of the use of one in their instructions. In a phone conversation with one of their tech staff, they did not recommend the use of a bondbreaker in a teak over ply style deck as the compund was not only acting as a filler to the groove, but also a sealer to all routes of water penetration. He did emphasize the need for a primer to be used. He was chagrined to hear I had already chosen the Smiths compound, but he had to admit that the Clear Penetrating Epoxy Sealer was a very good product and that he didn't know much about their caulk. After researching both Smiths and Detco, I believe they are both good products that have a good history to prove them.

I think I have to agree with Jorma that the use of the bondbreaker feels like a result of trying to give marginal products a fighting chance. I have decided to forego the use of one in my caulking project as I feel that is much more important to use all the adhesion surface possible to prevent the entry of moisture into the deck structure. By working hard to dry the deck and shrink the planks as much as possible, I feel the tension forces that the bondbreaker would be addressing will be low enough to make stress concentration problems a moot point.

I also inquired with Smiths about the use of acetone for cleaning the seams before priming and they assured me that the CPES would not have any problem with the oils in the teak and that the alcohols in it would do the job of drying the moisture out of the surfaces. He also recommended mixing the caulking compound by hand (when mixing small amounts) to lessen the introduction of air bubbles and since it is very difficult to see how well mixed the black compound is, a primary mixing should be done in one bucket, and then transferred to a second bucket and mixed again.

It's amazing how once again we seem to be coming around to the fact that the way it's been done for years is probably still the best choice.


[This message has been edited by Jamie Hascall (edited 08-15-2000).]

Posted by Jorma Salomaa (Member # 793) on 08-15-2000, 05:53 PM:

Good luck with your product developement, Chemist. Or maybe luck should have no place there, except with the marketing department. When you say "no plasticizers", do you mean that a low modulus, say 20 durometer reading sealant (one or two part) can be achieved without them?
"Pretend" planking: I think what Don refers to is something that looks like a plank deck but isńt. Meaning there are rather thin teak planks glued and screwed to a plywood or GRP deck. This does not mean the seams are "pretend" because the planks still move with moisture content like any planks do.

I had a discussion with a gentleman who does seam work for a living in building construction. He uses only polyurethane sealants nowadays (mostly Sika and Tremco), mainly because he is a sub-contractor and he uses what hés told to use. He said the problem with the polysulfide sealants he used earlier was that one could never be sure the batch that was received would cure normally. Then one could́nt trust the workers did the mixing properly etc. The final blow came with the poison hysteria which means those sealants are now practically banned from use in buildings. He was quite happy with the sealants he normally uses and said he would expect them to last 15 years.

Alcohol: The polyurethane people say dońt clean the seams with alcohol because the sealant doeśnt like it. If CPES has alcohol in it, does it mean that it should not be used as a primer? My own tests show good adhesion to CPES, but that is short term of course. Any thoughts?

Posted by thechemist (Member # 1468) on 08-15-2000, 09:41 PM:

....the night of the living horse....
Jorma, A low modulus sealant can be achieved without plasticizers given the right chemistry. The important thing is that the plasticizers not migrate appreciably, not that there be none. One needs these things in many rubber formulations. Think of them as something like the grease in a bearing. It helps the molecules slide past each other. Just as in a sealed bearing, one wants the grease to stay inside the bearing. Similarly with plasticizers. One needs certain types, and while a little of something migrating out may not be harmful, and sometimes one may need to use small amounts of those things in the formulation for technical reasons, most of the ingredients should be in the formulation permanently, since those determine the long-term properties. Given the available raw materials, some chemistries built on some raw materials may or may not need them. It is often easier to control mechanical properties consistently with a higher level of cross-linking and a stable plasticizer than a very low but uncontrollable level of crosslinking. An important factor is that the production batches be consistent, and easy to make consistent. That is where I get to tell Marketing how it will be.

I think rather thin teak planks glued and screwed to plywood will have some movement, although less that heavy planks supported on ribs. I have seen too many teak decks slapped onto a GRP deck with epoxy (or screws only)which leaked like a sieve or came loose later (due to thermal coefficient of expansion mismatch between the wood and GRP ..... on steel decks it is even worse) to trust seam stability at all. You have to use a relatively thick elastomer to glue the teak to GRP or steel, and you have to limit the length of the teak to what the underlying glue line can absorb in shear, or else the mess comes loose from the substrate and failing seam sidewall bonds are incidental.

In a temperature cycling test of teak I measured a change in a six inch length of about .010 inches max, and no drift over ten cycles, where in width the change was ten to twenty mils, and over ten cycles drifted sixty mils. Steel, by comparison, had a three mil change. In previous tests I found the coefficient of thermal expansion of fir in the direction of the grain to be very similar to steel, while in the other two dimensions perpendicular to the grain, to be about eight times more.

A mismatch of three mils versus ten mils, over six inches, becomes thirty mils versus a hundred mils over sixty inches. In other words, a five-foot teak strip glued to a steel deck will have a differential growth or shrinkage of .070 inches for a 100 F degree change. This is why one should not glue teak (more than perhaps a foot long) to steel with any epoxy. They have not enough give and will pop loose. Similar calculations can be made for other materials. Plywood has a thermal coefficient of expansion on the average half way between the two extremes, becasue it is made of perpendicular layers of both orientations. Of the various materials, it is probably closest, on the average, to teak.

The two-part sealants do have the disadvantage of uncontrolled temperature storage conditions at the distributor and at the contractor, and this can give working times (pot life) unpredictably longer or shorter. Further, incomplete mixing gives batches with various properties, and sometimes spots cured here and sticky there, then other spots cured and half-cured and so forth. The decay of literacy in our society coupled with the inability of sealant manufacturers to appreciate this and provide instructions with pictures only, not words, all this contributes to the lack of viability of this type of product. I bought one of those Bose came with an instruction sheet that was pictures only. I was impressed. At least someone understands what society has become. The one-part sealants bypass all that. Just point and squeeze and it all works, or so one hopes.

The polyurethane versus alcohol issue...Two-component isocyanate-cured polyurethanes may not like alcohol, but (first) it may well have all evaporated...(second)even CPES is known to work with 5200, a urethane caulk, so it can't be a problem, third, the adhesion-promoters in these things give off a bit of alcohol themselves (unless they are isocyanate-terminated, which is possible) and that does not bother anything, (fourth)Just because it is a urethane caulk may only mean lots of urethane linkages in the backbone but all that is cured already and so is inert to the solvents even if there were any left, and fifth I think the "polyurethane people" think urethane equals urethane equals everything else, and in that respect are misinformed. Unless, of course, their caulks contain isocyanate terminations. Even so, the reaction with alcohols is slow, and other reactions could bind the caulk to the wood via an effective primer even in the presence of alcohol....what the heck, CPES is known to work anyway with 3M 5200, a urethane caulk, so the statement cannot possibly be generally true. Ask the "polyurethane people" to specify the chemical functional groups (meaning reactive molecular structure) of the caulk molecules, and lets's see what they have to say.

[This message has been edited by thechemist (edited 08-15-2000).]

Posted by Jorma Salomaa (Member # 793) on 08-16-2000, 06:15 PM:

Chemist, your explanation of how the plasticizers work was great. Ím visualizing a bowl of spaghetti (the tangle of long molecules) and the effect of olive oil mixed into it (the plasticizer).
The alcohol issue caught my attention when I read the data sheets of various products. I checked them again, and actually it was mentioned only with the Sika products, not all of them but with many. It read "avoid contact with alcohol and other solvent cleaners during cure" or "...with alcohol or alcohol containing solvents...". The Sika polyurethane sealants as a rule have isocyanate.

Your description of teaḱs dimensional changes due to temperature was striking. I never thought temperature would have that big an effect. We tend to think that only moisture has to be considered. Even then wére taught the effect along the grain is negligible. In fact it can be 0.25% from green to oven dry in teak. Much less of course in the moisture range wére normally dealing with. Probably the two fenomena together should be considered when glueing down a long strip of teak.

Posted by Don Danenberg (Member # 946) on 08-16-2000, 08:16 PM: pucky...,
Seriously guys, this has to be the most in-depth, educated study of these materials and their uses. Kudos for the depth and breadth of your studies and tests. WoodenBoat magazine should certainly consider an article on this subject by just copying the debate and the lengths to which it has gone and will continue to go.
On the practical side, however, I must ask how and why the 'bond-breaker' aspect entered and affected your intent.
I have caulked many decks from Thiokol to Detco to 3M-5200 and 3M-101, in the last 30 years and am somewhat reluctant to admit that I have never even heard of a "bond-breaker" of any kind, for any purpose, let alone in something as important as a deck seam; until this diatribe.
Have you come up with any purpose for it?

Posted by thechemist (Member # 1468) on 08-16-2000, 09:48 PM:

Diatribe? I'm shocked! Truly, I'm shocked! See what happens after we resurrect a dead horse? Others of the undead arise.....
The "bond breaker" is some of the folk wisdom of our boat-repairing-society and some sealant manufacturers and/or sealant applicators, and I have only a slight idea how the practice got into boats. I have heard of it but never gave it a lot of thought until we got to discussing failure mechanisms and so forth. I have been told by a building contractor, an installer of sheet metal rain gutters, of the use of a strip of bond-breaker tape over the sheet metal lap joints before inside sealant application. He also told me of the use in some joint designs in tilt-up buildings. Maybe some architect invented it and specified it, and/or some contractor did it on the job site, and took some sealant to the dock to use on his boat and did it the same way and later wrote it up in an article for Wooden Boat #12, and it has been like that ever since.

It comes down to the Classic Homer Simpson defense, " I didn't do it and it was like that when I got here."

Don, it looks as if you had no problems with any of the decks you did because, among other reasons, you evidently never used a bond-breaker tape on the bottom of the seam. Another reason to not do that. Any idea what was the durometer of the sealants you used (my guess from what you say is that they ranged all the way from soft, 25 Shore A to firm, 50 Shore A), and the typical construction of your decks?

[This message has been edited by thechemist (edited 08-16-2000).]

Posted by Miller, Robert W. (Member # 48) on 08-17-2000, 01:13 AM:

Has anyone figured out that the Nimitz has deck leaks? So to, the USS Constitution leaked on deck when she slid down the ways. Even the the frozen snot of utopia leaks! Get over the yuppie puppie idea that you can have it ALLLLLLL!!!!! tain't so, Joe. The water leaking on your head can be directed, retarded, and slowed. Not stoped for ever as in the how to books. There are systems that have come down to us through the years, dealing with wooden boats, that slow water as much as can be counted upon. The hall marks of this system are to lay the deck of wood for the job. Lay the planks in a mannor to do the trade proud. This done, the edge of each plank holds it's own. Covered by naught, or pitch, or oil based seam compound,splines, and of course, water on a daily basies. I just wish we all could git passed the idea of the utopia boat where things are done once and never done again!
Most of the time when I see lARGE numbers of replys to a question, it indicates mostly smoke and very little fire. Sorry to be so blunt. R.W.M.

Posted by ACB (Member # 1213) on 08-17-2000, 04:35 AM:

Well now, when we come to the "teak on steel deck" business of differential expansions and leaks, the Royal Navy, back in the days of big gun battleships, went in for teak decks, and many and varied were the crew complaints that resulted, because a battleship is very "wet" at sea (to keep the C of G where you want it) and these ships carried crews of 2000 or so who slung their hammocks in every inch of available space. The entire Grand Fleet and its sucessors to the end of WW2 leaked like sieves onto their miserable crews.
And have you ever looked underneath the teak decks of a passenger me, you won't want to!

Posted by TomRobb (Member # 1216) on 08-17-2000, 10:15 AM:

I wonder if Los Angeles, SSN whatever, and her sisters leak? But who has that sort of budget?

Posted by Jamie Hascall (Member # 335) on 08-17-2000, 03:17 PM:

Well, I guess since I picked up the stick and started whacking this thing first, I guess I should come to the defense of the monster that arose.
When it's obvious that your 30 year old deck no longer holds the water out but has not yet started to rot the overall structure that supports it, it seems like a good time to recaulk. I don't think that doing major maintenance to try and keep from having to do major repairs on deck beams and sills is a misplaced idea. Moreover, being a liveaboard in the Pacific NW makes a tight deck seem like a pretty neat idea. If I can have a dry bunk, why not ask the questions to try to find out the best way to do it. (Call me a yuppie if you want, I just wish I had the Beemer to go with the moniker).

Last year I did the survey of the articles and the manufacturers and this idea of the bondbreaker came up a lot. Not having been a shipwright for umpteen years or knowing one with the specific information that I trusted, I did as my research told me and as the manufacturers recommended. After the universal failure of the compound I used, I decided I needed some less corporate and more experienced advice, so I appealed to the Forum. Well yes, there is a certain amount of the blind leading the blind here, but in the process, a whole lot of very interesting facts have come to light and I think we're all a bit more informed about the workings of the industries that give us crap that doesn't really perform well, whether installed right or not. In the end we concluded that bondbreakers are logical, but not actually workable; a fact that all the guys that have been laying deck seams for years might have told us earlier if they'd felt so inclined.

In the process of looking for an answer, we often wander around a large circle and eventually find ourselves right next to where we started, but with the knowledge that without taking the journey, we could not have known that we were at the right spot. Some circles are bigger and more interesting than others. This one turned out to be a doozy. Thanks especially to Jorma and the Masked Chemist , whoever he may be, for making it such an interesting thread. I think it's time to ride this horse off to the yard so I can shoot a couple of miles of black gooey stuff into seams before it starts raining again.

Adios Amigos

Posted by Jorma Salomaa (Member # 793) on 08-17-2000, 04:46 PM:

My gawd, theýre going to shoot the horse!
It́s quite possible that the "it́ll leak anyway" attitude is widespread and it may be a part of our naval heritage. It is understandable when the deck is a traditional one with thick planks, cotton and pitch. You can see where it leaks. Just light the fire, warm the pitch and go fix it. And what the heck, the ship waśnt designed to last forever anyway.

Now most of the boats are plastic, and as a rule the decks dońt leak. At least they should́nt, and the wooden boat owners (most of them, I probably should add) have been smitten by the same idea.

Everybody used to know how pitch was going to behave in the seam. I believe the two part polysulfide sealants were first used as a modern equivalent of pitch: pour it in and see what happens. Mostly nothing happened, because we were blessed with a sealant that was so forgiving that we did́nt have to understand anything much about joint design and elastomers. I think that three things have happened since those golden times. First, competition resulted in too many inferior products and the users did not bother to learn how to tell the good from the bad. Second, the good products were declared poison. Third, the users were led to believe they dońt have to mess with two components.

When even the boat manufacturers use one part stuff, and mostly polyurethane, it must be OK for us too, right? Yes, if we use it like they do. They study the subject, pick a few contenders, take them to a testing facility, make the choice and negotiate the price. Then they design the joints so that the goop has a chance to survive, with bond brakers and all. I dońt think they will go back to two part polysulfides.

We still have the choice but I dońt know for how long. In the meantime we could educate ourselves on elastomers and joint design. I think that is what is keeping the horse alive.

Posted by thechemist (Member # 1468) on 08-17-2000, 11:05 PM:

Yeah, Jorma, that and the fact that if we do accidentally come full circle we can all stand around and admire the circle.
Some race horses were exercised by running in a big circle in a big round barn. They got enough exercise the year around that they were winners at the races.

Posted by Jorma Salomaa (Member # 793) on 08-18-2000, 04:37 PM:

Right, Chemist. It́s a full circle. We must ask Jamie if we can let the horse go. Jamie are you there?

Posted by Jamie Hascall (Member # 335) on 08-18-2000, 05:16 PM:

Yup. This horse deserves to gallop off into the sunset.
See ya later pardners.