Wooden Entry Door

[crokett™]

I'm thinking of replacing our steel entry door with a wood one.  The door is both north facing and there is a deep covered porch so it won't see a lot of sun exposure.  I don't have a storm door.  I am currently thinking walnut and it will be rail/style construction.  My current plan is the panels will be edge-glued pieces with chamfered edges to create grooves on the panels.   Questions...    How much smaller should the door be than the opening to allow for expansion?  I am thinking I should allow the panels to float though they will be fairly thick.   Should I put caulk in the grooves that the panels go into before assembly?

[Tapper]

I'm thinking of replacing our steel entry door with a wood one.  The door is both north facing and there is a deep covered porch so it won't see a lot of sun exposure.  I don't have a storm door.  I am currently thinking walnut and it will be rail/style construction.  My current plan is the panels will be edge-glued pieces with chamfered edges to create grooves on the panels.   Questions...    How much smaller should the door be than the opening to allow for expansion?  I am thinking I should allow the panels to float though they will be fairly thick.   Should I put caulk in the grooves that the panels go into before assembly?

Calling John T., the door expert!

Doug

[jteneyck]

Well, certainly no expert but I have built a few exterior doors and know enough to keep myself out of trouble most of the time.  First, no storm door is a good thing with a wood door.  Your door faces N, so it likely wouldn't matter if there were a storm door, but for doors that get direct sun exposure a storm door spells disaster unless they are vented and the glass has heat reflective film.  

OK, now for your questions:  

The reveal around the door should be 3/32", same as it is for interior doors.  The lock edge needs a bevel of about 3°, too.  Some folks put a small bevel on the hinge side, too, to prevent binding if the jamb is a little twisted but I never have.  

I advise you NOT to use TB III for edge gluing your panels, or any part of the door for that matter.  TB III has really poor strength when it gets hot, and doors that get much sun can get really hot, panels in particular because they are relatively thin.  If they get trapped in the frame just a little, from cupping a little, for example, the TB III will fail when it gets hot and one or more joint will open up.  I use System Three T-88 structural epoxy for exterior doors.  Polyurethane glue (yes, Gorilla Glue, though I like Titebond's PU better) and Plastic Resin Glue also work well.  

No, don't put caulk in the grooves when you install the panels.  Moreover, your panels will survive longer if you make them as two, half thickness panels.  That way, each side can better adapt to the seasonal expansion contraction.  It will be really dry inside your house in the Winter compared to the RH outside.  One piece panels often crack under those conditions.  I put a sheet of aluminum foil between the two panels to act as a vapor barrier.  The panels need to float in the door frame so they can expand/contract, but that presents a draft issue.  With panels held in by applied moldings I overcome this challenge by using silicone caulk between the inside molding and panel.  The outside panel is pin nailed to the inside panel at the top and bottom at midspan.  For panels captured in the frame you don't have any really good option.  Backer rod might be a viable option but I haven't tried it.  In any case, you need to let the panels float.  That's one reason I don't build exterior doors with captured panels.  The biggest reason, however, is it's nearly impossible to replace the panel if it cracks.   

Finish the panels completely on all sides/edges before you install them.  This is easy on panels captured by moldings; less so with captured panels.  If you use captured panels, it's imperative not to glue them into the frame during finishing.  You can see lots of examples of doors with cracked panels because they have been glued to the frame after someone repainted the door.  They slather the paint on to seal the panel into to the frame to stop drafts not considering that they have just doomed it to cracking in the future.  


You can get the look of captured panels, even with cope and stick construction, by using moldings on the inside.  You run the cope and stick joinery, dry assemble the door, and then route away the sticking on the inside where the panels go.  This gives you a rabbet to capture the panels and added moldings hold them in place.

If you use solid stock for the door stiles and rails, it's best to use quarter or rift sawn stock.  KD and well conditioned to your shop, too.  Walnut isn't very pretty in quarter or rift sawn
grain, however, so I'd probably use stave core construction with 1/8" plain sawn walnut veneer.   

I hope this helps.  Good luck. 

John

[crokett™]

thanks John. I like the suggestion of 2 half thickness panels. I was already thinking of a rabbett on one side and some nailed on molding holding the panels in. I will also remember not to use Titebond III.

[lift mechanic]

I did not know that heat and Titebond III. I have used it on everything I have built in the last 10 years or so. I guess I have been lucky nothing has gotten warm enough to cause a problem................so far. Thanks for the heads up.

[jteneyck]

I did not know that heat and  Titebond III. I have used it on everything I have built in the last 10 years or so. I guess I have been lucky nothing has gotten warm enough to cause a problem................so far. Thanks for the heads up.

If you look up the tech. data on Titebond's website it's pretty shocking how low the strength of TB III, and TB II to a lesser extent, is after 150F.  Worse, it does not recover, although they don't tell you that.  So after a few temperature cycles joints start separating if there is any stress on them.  The stuff I've built with TB II and III are fine, but most live indoors.  I built a couple of exterior projects with TB III, too, and they are OK so far so it can work.  But I know some pro door makers who have documented failures of TB III on glued up door panels.  There are glues with better performance for that application so it just makes sense to use one of them given how much time it takes to make a door and the pain involved with fixing a problem.  

What I really like about epoxy is how easy it makes gluing up the door itself.  It has a really long open time so you aren't fighting to get it together before the glue sets, and it acts like a lubricant and the joints just slide together.  Epoxy cleans up lickety split with white vinegar, too.  

John

[Cooler]

They do make "Super Tite-Bond" glue, but I've only seen it in large sizes (and only online).  It is available in 5 gallon and 55 gallon sizes.

http://www.titebond.com/product/glues/0c...eaf96f8e71



SUPER TITEBOND GLUE

Super Titebond Wood Glue is a high-quality, professional woodworking glue that offers superior sandability and solvent-resistance. It develops a bond stronger than wood and provides a longer working time than traditional aliphatic resin glues. Super Titebond resists stress imposed by heat and moisture during the production of panel...


Addendum:  I see it is for "interior use only".