This is a mock-up showing full-height stiles. The middle stile would probably be the same in either case.

Thanks for your comments!

There is no significant difference in strength in a door application. The stiles are typically made long so that the end grain is less visible being top and bottom

Thanks for your comments! The end-grain visibility makes sense...and I don't want them to look funny. Looks like all the weight will be hanging on two glue joints. I am thinking about gluing the left side of each panel to its adjacent stile to put the weight on the stiles.

This wood has a nice grain and is hard as a rock so it makes a great counter surface but I may need to swallow the time I have invested in the door panels and switch to glass inserts to reduce the weight.

" I am thinking about gluing the left side of each panel to its adjacent stile to put the weight on the stiles."
I would not advise this! It's very important that the panels float independently with some clearance of the stiles/rails for expansion and contraction.

What he said.

Isn't it true that most/all of the expansion will be in the cross-grain direction???

2" wide frame pieces would equal 12" of load bearing surface in the frame. It will easily support anything you use for panels.

Most movement is cross grain but not all. It will move lengthwise. You will gain nothing by gluing the panels and have everything to lose by doing so.

That wood looks a lot like a large piece of tropical hardwood I bought about 20 years ago. They called it Cabbage Bark. I made a small table with it. Wonderful stuff to work with. It had a spicy smell when fresh cut. I've never seen it for sale since.

LOL! You are just oozing with confidence! Sorry if I sound pessimistic but I have seen commercial cabinet doors fail at the joints and I know they are probably 1/3 the weight as these will be. Of course it's my hope and expectation that I will be more careful building each piece and they will be stronger anyway.

Ok, one last (??) question. What would you use as a glue at the rail/stile joint? The literature claims kumaru is an oily wood and doesn't glue well. I have had VERY good results with Gorilla glue and not so good results with Elmer's Wood glue. Gorilla glue will get messy and tend to show in the corners once I do the rail/stile joint. Just be more sparing here?

"Isn't it true that most/all of the expansion will be in the cross-grain direction???"
Yes, and if that panel is tight and glued to the stile, when it expands it can and will likely blow apart the stile/rail joint, regardless if it is doweled or mortise and tenon.
You are over thinking a basic door construction that is tried and true and been around a lot longer than either of us. While your wood may be heavy, so is a solid Oak or Walnut raised or recessed panel entry door which relies on the same basic stile and rail construction.
If you are really concerned about shear, make your bottom rail a bit wider, a very classic look on lots of old furniture and cabinetry, as well as how all passage and entry doors are made.

Ha... I'm a slow typer, just saw the other responses ;)
Gorrlliia glue is good but also messy, Titebond II works well, but if the wood is very oily like Teak, you might want to use a powdered aliphatic resin glue and mix it.
Are you doing mortise and tenon, or dowel joints?

This post was edited by ctycdm on Fri, Oct 3, 14 at 16:40

Joint failure in cope and stile is the result of a poor glue joint which comes from - poor fit, glue starved, glue set prior to install, over tightening clamps and finally inadequate joint shape. ( any glue is useless on end grain so join shape must allow long to long grain match)
If you are not going to have an edge profile on the inside of rails and stiles then you no longer have a cope and stick, but a button joint. In that case use - mortise and tenon, loose tenon, dowels, or biscuits.
Oily woods need gorilla glue, there are other ways but not as simple. Put blue tape along edges prior to glue up to protect surface. Remove excess glue with paper towel. Then before fully cured removed the glue that bubbles out when firm is but not hard, then remove the tape. Hey you picked the wood ;)

Added, wipe surface with acetone or alcohol prior to applying glue, test scrap first.

This post was edited by jakuvall on Fri, Oct 3, 14 at 16:40

Lol we're on the same page jakuvall... I was also going to edit and add "wipe the gluing surface with alcohol or acetone."
got to get faster at typing;)

Well, I don't envy you building an entire kitchen of doors. You must have some confidence, too. That's a lot of work.

You did not specify what joints you were planning. Like they said, mortise and tenon is the ultimate in strength and floating tenons are really easy to do. I know a friend who did splines instead of tenons in his frame and panel cabinet doors almost 20 years ago and his doors today are still as solid as the the day they were built.

I can't help on glue for your particular wood. I guess Google would be your best resource for that answer.

I'll echo what jakuvall said. Gorilla Glue (i.e. polyurethane) for oily woods.

If you're really concerned about the joints then you can pin the tenons, which gives you some nice decorative options. Or you could add butterfly keys. Or you can go nuts and do some form of dovetail (half-blind, through, sliding, etc).

OOps, now I'm the one who's behind! Again, thanks for all the helpful comments. It was a last-minute concern so the quick advice has been very helpful. I just finished running the stiles through the router and I'm doing mortise/tenon joints. I hadn't previously heard the tip about glue on butt ends being useless. I had wondered about taping areas I wanted to be clean so the specifics on that are also great. A couple of replies about the wood now. Yes, I guess this project would have been easier had I gone with a standard wood but I bought it from a guy who overbought. It does have a spicey aroma but I haven't seen the "Cabbage Bark" name associated with this stuff. It's not really teak but a common name is Brazillian Teak. It has "Kumaru" stamped on the back of it.

I guess the only thing I'm still curious about is gluing only a left side of the panel. Seems like that would still allow the panel to expand in three directions since only one of four sides would be held stationary. Certainly it could be a waste of time and effort but the concept seems ok with what I THINK I understand about wood at this point. I'm really not trying to be difficult, just trying to reconcile new things I'm learning with other things I thought I had learned :>)

Correction! I'm not doing mortise/tenon, it's just the typical rail/stile router bit fit....terminology is telling on me.

Chiefneil, the butterfly keys would probably look really nice just as a decorative touch....man, I already feel behind on this project as it is! I have too many hobbies and I'm just barely passing at all of them.

I've enjoyed reading this. I know about enough to follow what y'all are talking about.

So here's my dumb question:

Johnorange, if you're concerned about your joints, why aren't you doing mortise and tenon? My father (who is a woodworker) taught me that that was what held things together, and the glue is just an extra little helper.

There is nothing gained by gluing one side of the panel. The load bears on the hinges, translate to the stile, bottom joint of rail to that stile is in compression, it wants to close. Top rail is in tension, buy with hinges set down from top and bottom it gets offset, bottom compressive load compensate. Everything depends on the glue joint between rails and stiles.

Five panel doors exist that are centuries old. Trust those who have gone before, IOW don't reinvent the wheel. Just do careful work and don't over tighten your clamps. Worry about how to jig it to stay square to ensure a proper joint.

Pillog- close, a m&t has now chancel advantage such as dovetails give. But it does vastly increase the long grain to long grain gluing surface. Done properly it is stronger than the wood itself.

Some years ago a couple of fellows managed to get access to Los Alamos to stress.test joints used on house doors. M&T with wood glue against double biscuits with gorilla glue. Overall the wood failed before the joints. May still be out on the web.

jakuvall is spot on, the hinges are key in displacing the load forces of the door. Look at any passage/entry door and you will notice the top hinge is set about 7" down, while the bottom hinge is set closer to 11" up. There is a reason for this, it's called physics....
On a cabinet door, the forces are much less, so the hinge placement can be equal for aesthetics.
As was said, don't try to reinvent the wheel, and one more time, don't glue the panel to the stile!
We are just trying to help you, not argue with you :)

I remember using gorilla glue for teak cabinet doors. It is a pain to work with, but probably your best bet. I wouldn't glue the panel.. but make sure to use rubber bumpers to suspend your panel.

I saw a video title for how to do a M&T joint. I'm already into the router cuts so it's probably too late to switch, still....is a good one to bone up on. Seems that would be a booger bear to make with wood that is hard as a rock? It's been my experience with Gorilla glue that it is still intact and the wood has torn out in places where something failed. Just need plenty of surface area. Just in from watching some live music....will try to post some final photos when I get done.

Full M&T for heavy woods because the joint then is contributing to the racking strength of the structure in a way that a glue joint does not. The more weight, the more the load is concentrated on that 1/2" stub tenon and the first 1/2" of molded stile.
As the doors grow wider the racking and cantilevered load challenge becomes more acute.
Casey

Jakuvall, thanks for the explanation, and for saying it in a way that makes sense to me. :)