Question
I want to create an edge for my kitchen tile counter. I want to use ipe wood to join the inner and outer corners when assembled. What are your recommendations - Gorilla, e2000?
Forum Responses
(Adhesive Forum)
From contributor A:
Gorilla Glue (and all other brands of poly glue) were outed last year by Fine Woodworking as poor wood glues in any circumstance. Countertops are typically glued with epoxy resin. Epoxy resin prefers to have a sanded surface, unlike conventional wood glues which prefer freshly jointed edges.
1. Sand with 80 grit paper.
2. Wash the dust and oils (ipe) off with a rag soaked with alcohol or acetone.
3. Use any brand of epoxy resin (West System, MAS, System 3 , etc.). Coat both edges with the resin. Allow it to soak in for 10 minutes
4. Recoat a second time.
5. Clamp.
6. Cleanup anything with rubbing alcohol.
It's a good idea to use masking tape and plastic to protect anything around the joint.
Regarding epoxy, it is a tricky adhesive requiring careful mixing, complete avoidance of having it touch your skin, adequate vapor removal, and so on. Also, it takes 24 hours to develop full strength. Squeeze out is hard to remove.
Gene, the PUR adhesives (Gorilla type) are by far the most expensive wood glues used today. They are better used when bonding different materials together.
Epoxy resins are likewise expensive, but are far more useful in woodshops. The newer resin systems like MAS and System3 are mixed in 2:1 ratios versus the old school West System 5:1 ratio. Mixing takes seconds and is easy to measure or use pumps that come with quarts/gallons. MAS epoxy has been my favorite for about 8 years. They are the only brand that has two viscosities of resin - thin and thick. System3 has thin and West System has one that is more thin than thick.
The MAS Flag (thick) resin is ideal for bonding wood. No need to do the first soak-in coat that most people using West System are familiar with. Also their hardeners and resins are all compatible, giving one infinite control over resin viscosity and dry times. Their prices are also lower.
As I stated, hide glue will in fact outperform any type of adhesive as to longevity and ease of repair. All adhesives will eventually fail, even epoxy. Hide glue cannot fill gaps well or be submersed in water. As to total waterproofness, it is by far superior to PURs and PVAs if used properly. The long history of its use from ancient times until now supports that. There is no long history of modern adhesives, so it will be our children or grandchildren that will have to comment in the future on this matter.
Meanwhile you can use it with confidence on oily/waxy woods as long as you prep the surface. 5% phosphoric acid, wiped immediately with methanol, then 95% ethyl alcohol. This will remove all the surface contaminates from interfering with the glue's ability to adhere properly, whether it's PVA, PUR, or hide glue, epoxy or other. My understanding of the original post is a molding is being applied to a kitchen counter edge. Unless I'm mistaken, I suspect the edge to be wood edge, not an edge covered with tile. If I misunderstood, I concede the use of epoxies or PUR adhesives for such a limited task as edge bonding to ceramic and wood combo.
While hide glue may be quite useful for repairing antique furniture, it is not a viable glue for the majority of the wood industry. Comparing the relative strengths of a water resistant glue such as Titebond 3 and a standard adhesive such as Elmer's is also a pointless exercise. If you don't need water resistance, why pay for it? Just because an adhesive is water resistant doesn't mean it's a better or stronger adhesive.
Adding the PUR and the epoxy to the mix just adds to the confusion. Both have their place, but neither would be right for the joint construction in question. That said, I do think the epoxy might be a good choice for the countertop in question. Ipe can be quite difficult to glue, especially if it is not freshly machined (something the FW article neglected to mention). Another factor that makes epoxy a good choice for the countertop is that it's waterproof and somewhat gap filling. All of these factors and others should be taken into account when deciding which adhesive to use. There is one factor which probably shouldn't be used; the article in Fine Woodworking.
Here we are talking of a straight wood to wood flat surface to be joined. Epoxy will do just fine, PUR will work also, but I know from experience hide glues will do just as well and can be made virtually waterproof by adding alum. I don't care what you use; I was just offering another very viable glue product to the list of contenders. A good fast paced production glue? No! A small glue up job like this application? No problem.
I thought the FW glue test was done well with good choices in joint construction, material, and adhesives.
I've personally seen many PUR glue joints fail. Especially in exterior situations like decks details. Too many, by too many craftsmen, to be clinically insignificant. The FW article lends some insight into why these failures occur.
Don't misunderstand my points. I have used PUR extensively in the past, rarely at present. The PUR glue guns are a different story and are used for much different applications. If I can solve 99% of glue joints with Titebond 1 or epoxy resin, why purchase all of the other glues? For other folks, Titebond 3 will even replace epoxy in most exterior joinery.
When I consider the postings here I do it from a production point of view. In a production situation there is a prescribed method for choosing the right adhesive. Cost is definitely a factor in this decision and that's why one wouldn't necessarily use a glue like Titebond 3 for all the applications in a factory.
As for the hide glue... Steinway does use hide glue in a very small part of their operation, but they haven't used it for laminating the piano cases for many years. This kind of construction has been done with UF resins since their invention around WWII. They have been tested and retested and are a far better choice in most cases than hide glue. Not to knock hide glue, it's just not really a factor in any real production setting today.
Production glue for modern use in fast pace assembly? No. As good or better than PVA? Yes! Hide glue does not creep over time like PVAs have been proven to do. Don't relegate hide to a time gone by - it is capable of competing with all the other adhesives out there except epoxies. Just not in a fast paced environment. And for you that have used prepared liquid hide glues, this is not what I'm referring to. I mean ground hide glue that you have to soak and heat before using. The bottled liquids have nowhere near the strength of fresh prepared hide glues. Hides come in gram strengths of 35 to over 500 bloom grams and can be used in so many ways and on so may types of materials there is no room to list them here. Can you say the same for the others?
Another plus - it's green.
The joint itself is a sawn edge. This is not the best surface to glue. There is no mention of the freshness of the joint. With ipe, 10 minutes is a good maximum aging time. The adhesives tested should be designed for a shear type joint. I agree with Jeff that this test is not the best way to evaluate an adhesive, without using other tests as well.
Pressure in an epoxy joint is a bad thing (a tight joint). Epoxy adhesive is not good when tightly squeezed. Although they talk about tight and loose joints (prior to gluing), they do not indicate how much pressure was used even with the tight joint.
Wood MC is very important.
When looking at the data on shear, the numbers are all quite high and I believe that no adhesive failed the test... In fact, much of the variations of their numbers is due to the wood and not the adhesives.
When we make a joint in wood, often the strength we need to keep the joint from breaking is 50% of the ultimate strength. Hence, in the shear test, if one test is 1800 pounds and the next is 1400 pounds, both should do well in woodworking projects (but maybe not for laminated beams). Again, a shear test is not what I would have chosen to represent the suitability of an adhesive.
Elmer's glue is over 50 years old. What is this information about adhesives aging over many years? There is no long term chemical change in an adhesive over time. Joints fail because the wood is moving in our modern environments.
Creep is the long term deformation of a material (in the context of adhesive joints in wood) when it is under stress. A mortise joint in a drawer has no creep because there is essentially no stress on the joint over a long time. A laminated beam, however, does have stress. Further, the beam is always moving with the wind, snow loads, shrinkage and swelling due to MC variations annually, and so on.
It is not as easy as edge gluing a couple of boards together and putting a little pressure on the joint. As you have noted several times in this thread, all of the glues are stronger than the wood itself. A good sawn joint is a better representation of most joints used in woodworking in my opinion. Only in edge gluing using PVA glues do we prefer freshly jointed edges.
Most mortise/tenon joinery is sawn or shaped. A lot of the joints are end grain/side grain or miters.
The best structural testing of wood joinery that I've read is in The Gougen Brothers Book of Boat Construction (West System). They had extensive testing done on epoxy/wood joints used in wind turbine blades. They compared butt joints/different ratio scarf joints, etc. Also many of the common physical tests as well.
At the end of the day very few of our joints will ever be stressed to those loads. I can't remember where I just read (maybe FW as well) that the old wive's tale about squeezing the glue out of joints and starving them is just plain false. With the exception of epoxy joints - as Gene noted, they need a little more room.
In production, edge gluing is almost always a sawn joint. A starved joint because of excessive pressure in a small shop is rare indeed, but it is possible in higher production.
A dull sawblade can lead to poor adhesion. A sawblade in dense wood will often heat the wood and destroy the gluability. A poor sawblade that creates some fuzzing can lead to poor adhesion with some adhesives. For this reason, a saw is a poor choice for preparing a surface to test for gluing. (As a special note, remember that it is the sides of the blade's teeth that touch and prepare the surface. Hence, the type of sawblade used and how it is sharpened are very important for testing adhesion and in production.)
There are many standard tests for adhesive joints. Two basic types are shear and tension. I would prefer tension over shear for a general evaluation test, but both are important together. We like to use a standard test so we can be repeatable and various factors that affect the results can be controlled. All tests are an attempt to simulate real-life conditions.
The US Forest Products Lab has an excellent booklet on adhesives. They have more info and experience than probably any other source.
I tested TBII once with ipe, even using an acetone wipe to clear off any oils or contamination. After 24 hours in the clamp, the joint was easily broken by hand, leaving a nice, shiny glue patch on one piece and nothing on the other.
Comment from contributor S:
Gorilla glue has its failings but should be strong enough for a counter edge that will not be highly stressed. Other stronger glues that might have superior numbers will also give you a less repairable joint. Some of the antiques that still exist only exist because they were in fact repairable. I am interested in the statement about hide glue not creeping. I haven't heard that before. I would be real interested in more info on that particular subject.
