Craftsman front door build

[Mel Fulks]

Used to do something similar to that ,but mainly for wood panels. Door made with all square edges. Each opening had slot run all round center of perimeter. Spline put in slot is same thickness as panel tongue. Apply the wide moulding to one side. Turn door over ,drop in panel, install moulding. Many times panels were same thickness as door and made with two "back to back" panels.

[mreza Salav]

Those are nice doors John.
There is a lot to learn in building an exterior door (door, jamb, weather seals, lights/windows, etc). There are many people here with plenty of experience to learn from. I used that wealth of knowledge
and built our entry unit a few years ago. You can find that thread and see the process.
My suggestion is:
- do stave core instead of solid material.
- I used ready made glass units (for commercial doors) that did a special order through the door manufacturing company.
- Think carefully about the seals all around, threshold too.
- I used double floating panels with 1/2" rigid foam in between them for insulation (it gets pretty cold here in Alberta).
- I do have the big domino but still would use big hefty tenons instead (my door is 42"x96"x2.25" and guess around 200lb).
- Think about stain/pain and maintenance of it too.

[John TenEyck]

Thanks Mreza. I remember your entry door, and all the other incredible work you did building your house. Amazing work.

I agree about using stave core construction. Wayne's recommendation to only use stave core as a "last resort" makes me wonder what we're missing. Of course it takes longer to build; from that perspective one might understand doing it only as a last resort. But from a stability and conservation of materials standpoint, I see only benefits. Wayne? I recently was in The Breakers in Newport, RI. Having just finished the exterior door I showed above I looked at about every door I saw in that amazing place, as well as in a couple of the other "cottages" I went through. Everyone looked to be stave core construction as evidenced by the 3/16" - 1/4" thick veneer skins you could see when looking at the edges of the doors, including most of the entry doors. The Breakers was built in the late 1800's. I was very surprised to see that, actually. I didn't think the glues they had back then would ever survive that long on the entry doors. Of course it's possible all those entry doors were replaced at some point; I just don't know.

Of course this will be controversial but think about it before you (in the general sense, not you specifically, Mreza) react. Joinery is not a major concern if you believe glue is what carries the load. With both a cope and stick door, as well as the shallow mortise and stub tenon approach I showed, you have a lot of surface area to spread the load over. If you glue those mating surfaces then all the joinery is doing is resisting the shear load. I glue in the loose tenons I use, but my mentor doesn't even bother to glue in the dowels he uses. He's built many hundreds of doors and shipped them all over the USA w/o any joinery failures, so it's hard to argue with his approach even though I'm not going to do it with the doors I build. Through mortise and tenon joints like Jack showed sure look nice, and were and would still be critical if you didn't use glue, but are more about following tradition than being necessary in a well fitted, glued construction. It might also make for a good sales pitch. For historical restoration, sure, for other applications, I don't see the benefit.

Just my take on it.

John

[Andrew Hughes]

You guys sure go through a lot just to make a door. It must be because of your climate.Ive made several doors with clear Vg Fir. My doors have stayed flat square straight and true.I did have one that weeped pitch.I guess living here in So Cal has its advantages.

[Joe Calhoon]

My suggestion for a white oak door would be solid. As long as the door is not over thick or oversize you will be fine. 8/4 white oak if seasoned properly is very stable. Especially quartered or rift. We also build a lot of doors in solid VG fir and that is another good material. We build with cores when using unstable material or for large and thick doors. I do not care for stave core with a veneer for exterior doors because the thin edge of the veneer is to the weather. Better is a solid 3 layer construction with the center core rift or quartered. Better if all 3 layers are rift and quartered. Note – for 1 ¾” thick the 3 layer method is not good, but for 2 ¼” and up works good.
If you do stave core best is same species core with vertical and rift grain opposed. Make sure the skins are the same MC.
Here is an example of stiles on a 9 ½’ historic door that is only 1 5/8” thick. This was a bad situation because we had to follow the original for thickness without removing the jambs that were an integral part of the old store front. We did use stave here as it seemed the best choice and they stayed very straight. All rails are normally solid or 3 layer laminated to keep no glue joints to the weather.
Staves ready.jpg

Here are some other cores we use. LVL works well with unstable material like barnwood. With LVL and other man made cores you have to be careful going wide because there is no movement in the core.
Cores from our shop.jpg
Barnwood stile.jpg

[Joe Calhoon]

For joinery, I think true mortise and tenon is the best but not always the most practical if you do this for a living. We have a tenoner, sliding table shaper, Maka and a hollow chisel mortiser but still labor intensive to do true M&T on everything. We do try to use M&T for historic or special work but have built doors the last 42 years using through tenon, slot and tenon, loose tenon and dowel construction with no issues on any. With loose tenons and dowels use a rot resistant material like White Oak or Sipo. I think the large Dominos in Sipo are fine for doors.

Tenoning bottom rails.jpg
Pressing 2.jpg
Clean mortises.jpg
Maka.jpg
Stack rail assembly.jpg
Finished parts.jpg

[Mel Fulks]

Interesting stuff but I have several different takes. Don't like the look of three layer. As a mere employee I've always guaranteed every door I made. Only problems came from failure to seal bottom; stopped that with paper....stuck on reading NO waranty for anything if bottom of door is not painted. I like LVL cores,but never had any problem with door straightness whether solid or cored. For many years I made cores out of NE white pine, no problems. I've found that the best way to make straight engineered stiles is glue skins with a NO WATER GLUE. Perfectly straight cores can bow with yellow glue that is even slightly thinned with water or UF water based glue. Certainly not all wih water in glue bow but any is too many. Many of the doors I've made came to us because all competitors refused job. Largest were two painted exterior doors for a hospital used as a pair. They are under cover. Opening 84" wide and 10 feet tall solid fir. The glue joint to weather comment is hardest to understand,as I have had no problem. But I insist on hollow glue joints and have retro fitted several jointers with a calibrated outfeed table that can be quickly (as different from eventually) set for hollow joint and then perfectly reset to "straight".

[Joe Calhoon]

I agree with that Mel. A stiff glue is better for skins. Tightbond is flexible and if you stack a skinned core glued with TB horizontal between 2 horses it will bow down overnight. turn it over and it will bow the other way. That said we still use TB3 a lot with no issues. Some shops in warm climates have had problems with TB3. Especially under dark finishes.
the 3 layer is not easy to look at on edge but we use this mostly on our Euro doors that have rebates and gaskets in the edge so you don't see it. these doors get very thick up to 3" plus and solid material that thick is not easy to come by here.
I am a minority on not liking the 1/8" or so skin to the weather. Most custom builders use stave core solid or man made with 1/8" skins. I have seen some failures with this though and not real excited about this construction. Great for tall interior doors though.

[Mel Fulks]

Joe, I'm convinced that "flat gluing" ,such as gluing the skins requires a no water glue. But I also think that flat gluing requires glue on both surfaces and that can be messy and wasteful with yellow glue. No problem for a couple of doors but not really practical for quantity. I like UF glue for curved stuff and have used it ,years ago, for skins. But to over come the effects of the water we made the skins thick then re machined the stiles after a couple of days. Wasn't really practical but we did it and tried to get paid. That 800 two part UF gets good reviews but I confess I have never tried it. At times a Ive made extra stiles and used the bowed ones for top rails and intermediate rails.

[mreza Salav]

On my door cores were glued using epoxy but the skin (all Honduras Mahogany) was glued using TB3. The door bows slightly (about 3/32"-1/8" over its length) in summer and goes flat again in winter. I was surprised first but now have learned it just plays with season/sun.

[Mel Fulks]

Mreza, is it dark in color and facing sun? Or not...and you think the change is just air temp related?

[mreza Salav]

It is dark and it does get sun/heat.

n1.jpg

[John TenEyck]

That was an odd choice, Mreza. I would have used TB III for the core and epoxy or PRG to glue on the skins, and epoxy for the joinery. TB III creeps pretty badly as the temp. gets hot, and a dark colored door facing the sun does get hot. Fortunately, your door is just moving and the skins aren't coming off.

Don't put a storm door over it!

John

[peter gagliardi]

And those ^^^^^^ are just some of the reasons I too only do solid doors. I have laminated with solid lumber on one door because of unavailable stock thickness in 20 + years. 3 1/4" quartered white oak.
Every additional glue joint is a potential failure plane in exterior work.
Hickory is NOT a wood I would use for an exterior door.
Almost no rot resistance, and huge movement with humidity swings.

[Joe Calhoon]

[QUOTE=Mel Fulks;2715383]Joe, I'm convinced that "flat gluing" ,such as gluing the skins requires a no water glue. But I also think that flat gluing requires glue on both surfaces and that can be messy and wasteful with yellow glue. No problem for a couple of doors but not really practical for quantity. I like UF glue for curved stuff and have used it ,years ago, for skins. But to over come the effects of the water we made the skins thick then re machined the stiles after a couple of days. Wasn't really practical but we did it and tried to get paid. That 800 two part UF gets good reviews but I confess I have never tried it. At times a Ive made extra stiles and used the bowed ones for top rails and intermediate rails.[/


We coat both surfaces Mel. We have a pressure pot with a roller that is adjustable for flow and easy to get the right amount on. That's one of the reasons we still use TB for skins, convenience. I will use epoxy or Unibond 800 if bad to the weather. Basically any stile and rail construction is not good to the weather and try to push for plank style in that case. We do a lot of stile and rail with plank panels and developed a system for the horizontal rails to shed water. That is a big issue with panels to the weather.

We played with thick skins for a while. As you describe, glue the faces on and let them set for a couple days then face carefully and back through the four sider for final S4S. It works but hard to keep the skin thickness balanced. We went back to going for final thickness at skin glue up. And on big jobs we would have some reject stiles to turn into rails. Stave core - skin is a great method for grain matching. Cherry is a species you can build solid with but 8/4 usually ends up with sap on one face so easier to go for skins.