Finger Jointing Plywood: Discussion Thread

I agree Ilan, my circular saw scarfer can do over 8' long scarfs if I desire. I posted an image of it a few years ago, but I suspect it's buried in the archives of previous posts. It is just as you describe, a saw sliding on a angled guide rail. Very simple, no moving parts except the saw.

http://www.boatdesign.net/forums/attachments/wooden-boat-building-restoration/23115d1215302702-plywood-scarph-joint-ratio-scarfjig.jpg

Found it.

 

It seems to me that the key is increased joint length - it shouldn’t matter how this is done.

The scarf joint is so wide it doesn’t require a directional change. In a finger joint where the joint width is only the thickness of the ply a straight joint would break along its length, so at least one direction change is required. Where the direction is changed, there is a weak point as it curves around; this is not so if the change is a point not a curve, as in the birdsbeak joint.

That is key. The plain butt joint has 2 problems. The end grain absorbs the glue leaving a starved joint, and neither the bond nor glue strength matches the strength of the wood along the grain. The scarf joint avoids both problems. First it increases the area of the joint. Secondly, by cutting the end grain at an acute angle, it reduces glue wicking into the grain.

The finger joint also does both these things. If the cut direction changes is a curve, as it must be when cut with a router, part of the joint becomes a butt with all the problems that introduces.



Why bother? Good question:

My main complaint with the scarf joint is, it tends to be easily visible in plywood. In relatively narrower solid wood I easily achieve joints that vanish under varnish so I can’t find them myself. Not so with ply: slight imperfections when cutting the joint faces are magnified in proportion to the slope of the cut so that the feather edge at the end of one piece of ply does not coincide precisely with the start of the cut on the piece underneath. This requires surface material to be removed to get a flat surface across the joint. The visible thickness of the glueline is also magnified by the slope of the cut. All of which tends to leave the finished joint easily spotted under varnish.

No doubt a more skilled craftsman would achieve a better looking joint than I am able to produce. However, I am limited to my own capabilities and not entirely satisfied with the results so far. I want to be able to make a bright-finished ply boat with joints that even I have trouble relocating. I haven’t got there yet.

I have made stepped scarf joints that were virtually invisible, however if the joint is across the face grain, as is generally the case, these are weaker than a regular scarfed joint. Also they are more difficult to make and only good one side.

 

Use a "nibbed" scarf Terry. This places much more "meat" at the feather end and you can accurately machine a perfect fit. Setup is more difficult and alignment must be perfect, but it's the traditional way of getting scarf to look good.

 

I tried one of those once and it looked great, I think it would have been hard to find once finished, but I was worried about the loss of strength in 3 mm ply: I made the nib about 0.02" which was the thinnest I could do but even that reduces the strength 30% or so, unless the face grain is parallel with the joint of course. It's much less of a problem in thicker ply. In a boat with a handy frame such a tiny strength loss can be made up by placing the joint over a frame, if it is not bent enough to creat a bump. It was a bit fiddley, I suspect the finger joint might be less of a bother but again the alignment must be perfect. Life's tough ...

 

Finger jointed plywood joining

I have great results using finger jointed plywood for curved shapes. Lasercut Fingerjointed plywood construction is beautiful, easy, awesomely strong and durable! The fit is always pefect, and most applications don't require glue or fasteners.

Just visit my gallery www.cutyourway.com for a show of my current capabilities, or visit KIDO SHOE in Austin TX. The really complicated stuff was designed with the following tools, in the following order:

Pen, Paper, Rhino3D, Autodesk Inventor 2010, Table Saw, Epilog 36EXT

Lasercut design offers the following advantages in fingerjointed construction:
1. Automated cutting.
2. Infinitely variable finger pitch. Even random finger jointing is possible.
2a. Computer Generated fingers means fewer design headaches!!!!
3. Instant Square (and non-square) holes, allowing finger joining across surfaces; also cubby holes, keyholes, lightening holes and glory holes.
4. Non-linear bender boards are possible. This means the bender board can be formed to any (non-zero Gaussian!) geometry without the use of steam, and or press.
5. By controlling laser etching power, a bender board's K-factor can be determined or varied. This means any finger jointed design will fit perfectly every time, using only a soft hammer.
5b. No Specialized Tools.
6. Is the design not strong enough? Put another rib anywhere you want. Or specify a thicker material.

For the boat application, I would use approximately the same design workflow as I did for the store display fixture shown at:

http://www.cutyourway.com/gallery/large/2 - Display sections.jpg

Note that each bender board is socketed into the ribs, showing what looks like a flashing, or lip, on the assembled display. The lip allows allowed high-speed, casual assembly without concern of accidental release. For shipwrighting, this lip could quickly be sanded away after gluing.

. The design approach I have in mind for shipwrighting is:
1. Draw the hull geometry and compartments
2. Design the ribs and trusses (skeleton), with fingers about the hull interface perimeter
3. Sheath the skeleton with lasercut, finger-jointed benderboard. Triangle-wave of puzzlepiece splice each sheath panel, finger-joint splice at the bow and other sharp-edged features.
4. Paint the sheath with epoxy, fiberglass, polyurethane, or laminate with veneer, or wrap in plastic or sheet metal. It woodn't look very wooden, I suppose, unless veneer is used, but veneers are irrepairable; what are you going to do when you scratch it, much less run aground?

An ideal boat hull is a "non-zero gaussian surface;" of course shipwrights have been approximating this ever since milled lumber. I haven't learned how to design non-zero gaussian plywood parts yet, but I am interested.

Also, I haven't attempted non-perpendicular fingerjoining yet. We have fingerjointed about skew angles, but each part still make a perpendicular intersection. For Example:

http://www.cutyourway.com/gallery/large/Custom Display.jpg

Skew intersection is possible, even between bender boards, but the fingerjoints would need filling and sanding.

My cutting table is only 36 inches long, but I can make high-performance splices, or order a job shop to cut from 8ft panel.

 

The worst problem with puzzle-joints for most of us (especially the amateurs) is access to CNC equipment, since some of us prefer not to work from kits. It's not just the cost of the service, since we do not possess the equipment, it's transporting the full-sized sheets to the site (not so bad coming back) and having software that can generate the compatible CNC path or command file.

The butt joint is easiest but ugly on one side, the scarf joint is ho-hum under a bright finish, the nibbed scarf joint is OK on one one side but does involve a slight loss of strength, and if there is a rebate on both sides the strength loss is magnified - and it demands an accuracy that few can achieve. There are also various jointing methods that rely on glassing which I do not use personally.

The puzzle-joint does not, in theory at least, deliver the full strength of the ply if the ends of the cut are rounded, only if a triangle wave cut is used. Depth of cut is critical so it cannot be employed for very narrow planks. Has anyone actually made finger joints with either curved or triangular cuts then broke them to see if the joint is as strong as the plywood? That is the criterion of acceptability for methods I am willing to use.

I am still working on ways to achieve sheet-to-sheet joints, as well as butted chine joints - which of course cannot use the finger-jointing method.

 

http://www.westmarine.com/1/1/17746--875-scarfer.html

 

Software is easy. Plenty of free CAD exist, but the right stuff are Rhino3D and Google Sketchup; both are cheap, efficient, easy, and powerful. Rhino3D has are large shipwrighting community too. For example:

http://blog.rhino3d.com/2010/11/ship-hull-design-training-with-t.html
http://www.youtube.com/watch?v=-LjHR8nclpE

Almost any CAD software is crazzy cheap if you hold a student ID card to any kind of educational institution. I'm talking 90% off, and full-power Autodesk products are completely free!!!! So go take a gym class at a college 100 miles away and drop out, if that's what it takes.

You don't really need CAM software, especially for 2D cuts. And we laser cutters know our kerf. In any case, designing for kerf is usually a case of typing "offset," then clicking a few times.

And what's so back about transporting raw material? We're typically cutting Lauan and Baltic Birch, and we'll probably switch to Finnish and Italian. For most cutting jobs, the significant costs are in design time, cutting time, and assembly time.

 

"ancient kayaker" ::

What type of failure test(s) are relevant here? Shear, compression, and tension tests are easy to pass. but moment tests are difficult because the central axis of the splice acts as a fulcrum, where the glue boundary can crack and fracture.

A proper glue joint is always stronger than the wood. The problem lies mostly in difference of materials. Part "A", Part "B", and the glue will always have slightly varying material properties.

According to http://www.inspectapedia.com/BestPractices/Exterior_Trim.htm , , the trouble with splices, is that over time, the splice appearance will "telegraph through the paint," leading to protectant failure and wood damage at the splice. This kind of problem is resolved through careful surface treatment, not mechanical stress testing.

Even the highest spec'd sheet materials have problems with different material properties. I remember in welding college, my instructor strictly advised my classmate to specify "matching heat and lot numbers" for his future aluminum skinned trawler project, or he would be chasing misfit plates from bow to stern. And that's before maiden voyage.

 

Yes indeed Cutaway. Fiberglass butt splices are the worst. They will always print thru.

The reason a good boat builder is familiar with several joining techniques is that each has its own use. Butt joints use the full length of stock, scarphs are easiest to hide and bend fair. Puzzle joints are the odd ball...I only ever see them for interior work were the puzzle joint is a design detail for your eye to appreciate...How did they cut that ?

 

8 foot long sheet, 6 foot long van!

Shear, compression, and tension tests require forces too great for me to apply, not having hydraulics, so I apply a rough and ready moment test, measuring the force required to snap a jointed ply sample of standard length and width, and repeat the test on a control ply sample without a joint. I test assymmetrical joints like butt block style in both directions.

I have noticed material variations, not merely from sheet to sheet but within the same sheet of ply, supposedly BS1088. Temperature and humidity have an effect too. I get fairly repeatable test results with butt block joints and scarf joints with angles finer than 1:6 but coarser scarf angles are highly variable: I have had 1:3 joints match the ply control sample at some times, and fail well short of the control samples break point at other times.

Joints did not appear to be starved, but the "poor joint" results are consistent with adhesion only to the grain parallel with the joint so presumably there was wicking taking place. But not always! The strange thing is, the best results came from 5-minute epoxy which isn't even a true epoxy. I am starting to appreciate why experienced boaatbuilders tend to be so conservative!

 

Well you could take the back door off your van, or build a trailer. You will need the trailer later for your boat.
The original Saab 99/900 hatchbacks can conveniently fit plywood with the hatch slightly ajar. The 2nd generation cannot, unfortunately. That was a significant design fault for my family.

Those results seem about right; funny thing about the epoxy. The woodworking magazines at HD/Lowe's have some good articles about glue and splice performance, check their back issues. What are you using to apply the moment?

 

- so far they have all been car toppers.

I use a spring gauge; the test sample is clamped lightly in a vice between wood blocks with rounded edges, and the spring gauge has a guide to keep it steady during the pull. I try to make the time to break the same for each test, as I have observed that the results can be time-dependent, presumably due to fatigue.

I undertook testing because I found there aren't a lot of published resources for wood strength, they seem to have different data done in different ways for various species, the data themselves are not self-consistent, and there is considerable variation between authorities.

I am beginning to understand why this is. In my tests, I can get a set of figures with perhaps 5% spread across a number of identical samples. The same size samples, cut a few inches away from the same plywood sheet and tested the same way can give values up to 20% different a few days later. Or one sample will fail in a totally different manner from the rest. So the measured values are rather meaningless, and I do not trust published figures very much as a result. There is no way one can base a calculation on such unreliable data. The only reliable way to judge a joint is by whether it or the wood fails first, nothing else is meaningful with such a fickle medium as wood.

 

has anybody heard of a fiberglass ply (fiberply) in 4'x8' sheets ? carbon fiber sheets (carbply) would even be better but SO expensive. I have been many years ago at a seminar where bamboo ply was the subject. Pre scarfed edges would be an asset wih synthetic plys.
thank you.

 

Carbply ? Dont know. In the past Bruynzeel fabricated a kevlar Plywood panel. You might do some googling