[trevarthan]

I have a rough interior opening for which I'd like to build a custom glazed door and frame with an arched top. Similar to this, I think, except with a three center arch instead of a segmented single center arch:

See below for a three center arch, for those who don't know. I didn't know until a few days ago, and I've been staring at them every day for three years since I moved into this house.

I'm still in the planning and research phase for this project. At the moment, I'm trying to decide what kind of wood to use. It can all be paint grade as the rest of the house's trim is painted.

I'd like the materials to be as inexpensive as possible, yet still durable and long lasting. This is my first door/frame project, so I know I'll make a ton of mistakes and might burn some waste wood.

I visited my local hardwood supplier today and looked at their pine stock. It just all seems too knotty. I could laminate it all and cut out the knots, but it seems like a pain. Still... this might be the way to go for the frame/jamb? There is one other round window in the house, in one of the upper bedrooms, and it looks to be built-up laminated pine construction. Circa 1920.

The hardwood supplier recommended poplar. She said it was often used in paint grade situations for trim and door jambs. The nice thing about poplar is I can get it in different thicknesses and widths, so I wouldn't necessarily have to laminate everything. Maybe just some scarf joints. The poplar I saw was very clear, also, which is nice.

Would poplar also make a good door in this case? Is there a better choice? 8/4 poplar is $2.75/bf and red oak is $2.85/bf, so ... eh.... the price difference isn't all that dramatic. Not that I'd use red oak for the door. I just tend to use it as my baseline for price comparisons around here.

My shop is small, but fairly complete. Table saw, 12" jointer, planer, 12" SCMS, router, router table, etc. I can handle most hardwood sizes. Sheet goods might be a bit of challenge though. It'll be interesting making the arch template.

Feel free to educate me on any idiotic assumptions I've made so far. I'm researching my ass off here, but there is surprisingly little information on the internet about building a rounded door and frame from scratch. Trade secrets, I guess. Also, any book references would be appreciated. I have "Building Doors and Gates" by Alan Bridgewater. It filled in a lot of gaps in my understanding, but it tends to be a little lacking in practical application and modern, and maybe not so modern (1920ish), techniques.

Thanks.

 

[cabinetman]

Since you're painting I would go with Poplar. There are several ways to do the project. If you can get thick stock, you could cut the curves from solid stock and scarf joint. Or, frame out the curves, use bending ply, and fill and paint, or veneer with Poplar and paint.

.

 

[4DThinker]

I'll second Poplar. It mills great. It paints great.

 

[trevarthan]

Is there any disadvantage to the scarf joint method vs laminated bending? I've got a laminated bending book. Seems like a bit of a pain.

 

[Treeoflifestairs.com]

This is the door I made for our house. The top curved rails I cut from solid stock (8/4) along with the casing (4/4). The jamb I bent and laminated though.

 

[trevarthan]

ooooo.... that's some nice work. Are the doors solid stock, or laminated to avoid warp? I've read about stave core construction, etc. I was going to ignore that stuff since this is an interior door and just use solid stock, but I can't pass up the opportunity to ask.

 

[TylerJones]

I would not attempt to laminate the bend, I'd cut from solid. If you are worried about warping you can face glue 4/4 boards together to get your thick stock. That will also strengthen the short grain sections. You could also glue up and cut from solid stock and add splines to the short grain sections.

I would check the price of soft maple, around here it's about 2.75, though poplar is a little lower than that so maybe its more where you are. Either one will work fine, poplar is just a little softer.

I wouldn't use a scarf joint for this application because over time the two pieces will expand and contract at different rates and become visible under the paint. This happens even with same species boards. You can minimize it with grain/board selection of course, and I should say that the joint itself would remain plenty strong. It's purely an appearance thing.

Actually building the door is pretty simple if you have any experience building basic frame and panel doors. I would be happy to go into that in some detail if you need, as I'm sure many others would.

 

[trevarthan]

Hrm. I guess I need to figure out how tall that arch will be then. If I'm not using built up joints (scarf) and I'm not laminating (which it's starting to sound like I need to consider again), my only other options are to bend a solid piece (not impossible, but it would add a lot of time for jigs and such, and probably a lot of attempts and expense) or cut the entire top of the jamb from one single board.

 

[cabinetman]

Hrm. I guess I need to figure out how tall that arch will be then. If I'm not using built up joints (scarf) and I'm not laminating (which it's starting to sound like I need to consider again), my only other options are to bend a solid piece (not impossible, but it would add a lot of time for jigs and such, and probably a lot of attempts and expense) or cut the entire top of the jamb from one single board.

You will get a variety of opinions and suggestions. Some craftsmen might be more successful with certain joinery methods than others.

.

 

[mmwood_1]

I would also recommend the poplar for this project. When I have made these, I make the template from 1/4" ply, making half the arc. Cut that out, use a spindle sander to get the curves smooth and then you can either use that, or use it to make a full arc template. Both halves will have the identical curve. For the thick stock, I use mortise and tenon joinery. The joints will be strong, and they won't show gaps over time. I usually do a 1" long tenon. Glue up your door stock in wider pieces than you need, lay the template on it and cut out your arc. It takes a little time but it's not that difficult.
http://markmeyerwoodworking.com/gallery/main.php?g2_itemId=61

 

[Hammer1]

Normally, you are looking at two different approaches with the door jamb and the door casings. The jamb is usually laminated around a form with thin strips. The casings are cut from segmented pieces. Resist the urge to use wide lumber for the casings, you don't want short grain on the ends after cutting the arches. One aspect you may not realize is that the arch top is not a half circle, the ends that meet the vertical casings extend so the last 2" or so straightens out. You may be able to get the top segments out of 1x6 but the lower ones will require 1x8. Hope you have a bandsaw and know how to build a circle jig. It looks weird and out of proportion if just a half circle is used.

If you click on my name you can go to my pictures. Some under "projects" show arch top windows. I made the extension jambs by laminating, the casings by segmenting. I have some detailed pictures of the build but they are on an old Mac computer and I can't seem to get them from there to my newer PC. Laminations should be thin, around 3/16" thick. Using a urea glue like Unibond will give you good open time and help prevent creep later on. I used blind splines on the segmented casings to attach each piece to the other. The strike of the jamb is generally done as a separate lamination, them affixed to the main jamb. It's very helpful to have a sandwich form for the laminations.

 

[woodnthings]

A good friend of mine owns this door company

http://troycornersdoor.com/round.html
He will not make an exterior door from poplar as it will not last. The exposure to the elements will cause the wood to deteriorate more readily than mahogany, his favorite. He build a laminated core from maple and glued with epoxy for a warp free and non-moving substructure. I have watched him do glue ups and make the doors from start to finish and he is very accomplished.

I see the bigger issue is making the arched casing, rather than the door which can be skinned over a glue up over sections. The templates he uses are huge for the laminated layers to create the arch. They are made from either particle board or plywood, heavy as all get out, and using a pattern and a flush trim bit to get identical shapes that are then stacked together.

I would suggest contacting him direct from the website and mention where you saw this post. He may have some suggestions if you are going to go ahead with the build.

FYI:

 

[cabinetman]

He will not make an exterior door from poplar as it will not last. The exposure to the elements will cause the wood to deteriorate more readily than mahogany, his favorite.

This door will be an interior door.

.

 

[Steve Neul]

I wouldn't recommend poplar for a door. There is too much wood movement and warpage with poplar. If it were me I would use soft maple or sugar pine.


As far as patterns you could just use cardboard or masonite. I'm not surprised there is little info on the internet. There probably only a handfull of shops in each major city that would attempt to make a arched door. I have my shop near Dallas and I had a customer from New Jersey have me make three sets of doors like this set for their garage.

When you purchase the lumber, buy it rough. Pick through the lumber and find the straightest flatest boards you can make the stiles. Then flatten the boards on your 12" jointer before surfacing.

How are you equipped for coping and sticking? Regardless of the joint it isn't enough to hold the door together. The stiles and rails will need 1/2" to 5/8" dowels about 5" long in the joints.

 

[trevarthan]

Poplar it is for the jamb and casing, unless soft Maple is cheaper. I'm glad that's settled. I'll look into soft Maple for the door itself. I just realized every other door in the house is stained, not painted, even though the casing is painted. I guess I wanted a white door, but it's probably better to be consistent with the theme of the house.

I have a copy of Wood Bending Made Simple by Lon Schleining. I was trying to avoid laminated bending (or bending of any kind), because it adds so much time to the project, and the one round window in the house clearly wasn't bent (gaps show in the jamb), but it sounds like laminated bending is really the only serious option for the jamb, if I want to do it right. I'll have to get that book back out and start planning my jigs and templates.

Should I make the jamb first or the door first? I was planning to make the door first, but since I find the jamb intimidating it might be better to make the jamb first and get it out of the way.

I don't suppose I need to order knife cut ply though. Table saw rips should be fine for paint grade, right?

As for how I'm equipped for coping and sticking... crap. I just have a router table. I don't have a huge shaper. I was planning to do it old school with rabbets and glazing and such. I clearly need to think about that part a little more and see how they did it on that one round window in the house.

Oh, and I plan to use my Domino XL for loose tenon joinery. It won't cut 5" deep though. Max depth is 2-3/4".

 

[Steve Neul]

Unless prices have changed poplar is about as cheap of a paint grade wood as you can get.

The simplist way to bend wood is to construct a inner and outer form and laminate multiple strips of wood. The hard part is predicting how much it will deflect back after you take it out of the forms. You kind of have to bend a smaller radius than what you need so after you take the clamps off it ends up the size you want.

I would make the jamb first, especially if you get into bending components. It will undoubtly end up with a little different shape and radius than you intended so it would be easier to make the door fit the jamb.

You could probably run coping and sticking on your router table. Because of the weight of the stiles you might construct some extensions on your router table going in and out. It's difficult enough to hold a stile down when runing it through a shaper equipped with a hold down.

I'm sorry but I don't understand this line, "I don't suppose I need to order knife cut ply though. Table saw rips should be fine for paint grade, right?
"

 

[trevarthan]

I'm sorry but I don't understand this line, "I don't suppose I need to order knife cut ply though. Table saw rips should be fine for paint grade, right?
"

Oh. My understanding is that sometimes people use knife cut veneer in bent laminations. They order it, rather than cut it themselves on the table saw. Since veneer is cut with a knife there is no kerf and the grain matches better.

I might be wrong. I just thought I read this in that book. Doesn't seem necessary to use veneer for a paint grade jamb.

 

[Steve Neul]

Oh. My understanding is that sometimes people use knife cut veneer in bent laminations. They order it, rather than cut it themselves on the table saw. Since veneer is cut with a knife there is no kerf and the grain matches better.

I might be wrong. I just thought I read this in that book. Doesn't seem necessary to use veneer for a paint grade jamb.

If it were me I would resaw or surface wood as thin as your planer will do and bend that. Veneer would tend to get lumpy. For that caliber project I think it's inapproiate but the bend could be done with bendable plywood. If you laminate two pieces 3/8" to make 3/4" it gets pretty ridgid. You might make the shape with that and laminate some better wood to the face side and finish it with veneer.

 

[trevarthan]

I thought it would be interesting and informative to add a gallery of photos of the round dormer window in the upstairs of this house. I'm guessing the quality of this window is relatively poor, though clearly fairly durable (circa 1920).


image by Trevarthan, on Flickr

The construction seems quite simple, with the exception of the tenon joints in the corners of the window frame. Those look to be machine made to me.

As you can see, it looks like the glass just sits in a rabbet and the glazing holds it against the rabbet. I had initially planned to do the same thing for my interior door, but now I'm wondering if that is appropriate.

 

[Steve Neul]

I thought it would be interesting and informative to add a gallery of photos of the round dormer window in the upstairs of this house. I'm guessing the quality of this window is relatively poor, though clearly fairly durable (circa 1920).


image by Trevarthan, on Flickr

The construction seems quite simple, with the exception of the tenon joints in the corners of the window frame. Those look to be machine made to me.

As you can see, it looks like the glass just sits in a rabbet and the glazing holds it against the rabbet. I had initially planned to do the same thing for my interior door, but now I'm wondering if that is appropriate.

The meeting rail of the window sash could be improvised using a dado set. It could be done like the sketch. The way the molding was made on a curve was done with a molder. It literally went through the machine on a radius. It could be done on a home model machine like a woodmaster or belsaw. I think grizzly also has a planer/molder that would do it. You could also do it with a shaper or router box if the wood was brushed down well. You would have to make a cradle to run the wood through on a curve.