Originally Posted by
Mark Singer
John,
I found an article that really supports the use of biscuit joinery....They state clearly it is not the correct choice for chairs and the joint will fail without warning....see the last paragraph
[skip to quotation from article]
But know this: when biscuit joints fail, it is usually sudden and quick. When a traditional joint fails, it is usually a slow process. So biscuits are not the right joint for chairs or other furniture that is subjected to repeated and constant strain. You want a chair to become loose and wobbly before it falls apart with someone in it. But for the vast majority of other joinery applications, biscuits are an excellent choice.
Mark, that appears to be an excerpt from Jim Stack's book of biscuit joiner projects. As far as I know, Jim Stack has not participated in any quantitative analysis of biscuit joint failure. Of the articles I've read, only one article ( Wagner) reported a sudden failure of a biscuit joint. Gray's article states the opposite, when discussing the graph of load vs displacement of a biscuit joint:
It is surprising to note that although the plastic region (in which irreversible damage occurs) is relatively small, the load curve is gently rounded and the tail is long. This indicates that failure, as defined in this article, is not as sudden as others have suggested. Clearly, the joint is able to sustain significant load over a wide range of joint movement, even after irreparable damage occurs.
So who do we believe, Wagner or Gray? Well, Wagner's article describes some tests performed by Grant Taylor, who runs a custom door business. There was no mention of any expertise Mr. Taylor might have in testing failure in wood products. The fact that Mr. Taylor only tested one sample of each type of joint suggests a problem with sample size, and there may well have been other problems. The Gray article described tests performed in a laboratory facility of the Wood Science and Technology Center of the University of New Brunswick. Six samples of each joint type were tested according to guidelines established by the American Society for Testing and Materials.
Each reader can decide for himself (herself) which article is more likely to have reported reliable info. But even in Wagner's article about Taylor's "home made" experiment, Taylor found that the biscuit joint that "failed with a dramatic explosion" under 2800lb of pressure had exhibited a gap at 1700 lb. So one might wonder whether that sort of gapping would have given an audible warning (creaking) to a user of a biscuit-joined chair long before the chair actually failed.
In addition to that, we should think about how chairs fail. My stepfather is a furniture repairman who works on lots of stuff, from antiques to high-priced Scandinavian furniture. This doesn't make me an expert by any means, but I've seen quite a number of broken chairs. Those that broke at the joints were chairs that were made before the era of modern glues. The joints in those chairs failed because the glue deteriorated. Glue deterioration is a gradual process, unlike loading a joint to failure in a matter of seconds. In the latter process, the wood itself always breaks. In the case of glue deterioration, the joint loosens and begins to creak and show gaps. The creaking and gapping are the warning signs that should alert a reasonable person to the fact that one or more of the joints has failed. But even if the wood breaks or otherwise fails, the odds are slim that it will occur simultaneously in more than one joint.
However, we should keep in mind why we're even talking about using biscuits in chairs. After all, if you look at almost any mass-produced chair today, it's made with pocket screws because they make a very strong joint that is usually aesthetically acceptable because the holes and screws are hidden, and they're very cheap to make in terms of labor costs. The only reason I mentioned chairs was because you mentioned that Sam Maloof doesn't use biscuits because he had problems with them. Maloof makes a lot of chairs, so I replied that Thomas Stender (a studio furniture maker) has used biscuits in chairs for years and never experienced a failure.
But Maloof, Stender and chair-making are only tangentially relevant to this thread, because they deal with the question of how strong a biscuit joint is. The central topic of this thread is about using loose tenons to join wide boards. Specifically, loose tenons mounted in mortises whose cheeks are about 1/4" thick. Based on my limited experience and the articles I've read, I believe that a biscuit or two, set 1/4" below the surface of the joint, would be strong enough to serve as the floating tenon(s) in this type of joint. I understand that you and many people here disagree. Thank you for taking the time to engage me in a fact-based conversation about this subject. Even if I haven't changed my mind as a result of our exchange, I still think you've helped improve my understanding of woodworking.