Looking for advice on an idea and on the sanding component of getting it done.

[rudy de haas]

I am thinking about making a bunch of closet doors from birch. These will all be quite narrow - 22" wide x 83" tall for the five largest ones, down to 17 wide and 32 tall for the two smallest. My idea is to plane some 5" x 3/4 stuff down to about 1/2", then glue up "planks" that are bit oversize for each door. I know a guy who has a CNC router that I think will go to 22" wide and I'll ask him (or find someone else if his isn't wide enough) to cut "portrait mode" images into each door - doing three on a 22 x 82 face should leave it looking like a rail and stile construct with two cross members but entirely without joints.

So:

1 - anyone know of a reason this won't work? (or things I have to do to make it work but won't think of unless someone warns me)

2- before they go to the CNC guy these pseudo slabs need to be sanded to a high degree of smooth shinyness - 400 at leasst. Birch is good for that, my sanding gear (several palm sanders plus a small Black and Decker belt sander) is not. What tools and/or methods make sense for this job?

[Adam Herman]

how thick will the final door be and what by what method? bifold type install or regular swing door?

what will the core/back be? I would worry about a warping and take that into consideration while deciding how to build this.

for sanding, i think you will do fine with a decent 1/4 sheet sander. I like to go over with a good block sanding for the final step.

[Sam Murdoch]

Do I understand correctly? You are intending 1/2" thick doors made up of edge glued boards measuring 22" wide x 83" tall and some smaller ones? I think the only way such a slab door will not warp - well, it will warp - but the only way you could control the warp, is with full length piano hinges and catches top and bottom and maybe in the middle too. And I am dubious of the success of this hanging "solution".

If you are face gluing those 1/2" pieces you are just making fire wood rather than doors. Sorry to be such a naysayer. Maybe I don't understand your plan.

Sam

[rudy de haas]

Thanks guys - I really had not thought through warping much at all.

However.. the wood is very dry and seasoned (stored here for a year) and the humidity is subject to serious short term spikes and drops, but averages a pretty consistent 55% or so outside, rather less inside. My solid hickory hardwood floor - which everyone warned me would warp - hasn't warped. We're just about to enter our wettest season (May/June) so I think I'll go make a "slab" and set aside to see what it does.

--
On review it's pretty obvious that warping is the big risk here so I made a test door yesterday and will ask Euan to do the face carving this coming week. I'll hang that on hinges for two months (our rainy season) and see what happens.

My guess is that no problems will appear in the two months - but that doesn't mean the things will be ok for many years. What I've done is alternated the grain pattern on the boards, sealed the ends with poly, and changed the large closet design to use six doors instead of five (making them narrower) and will set the face frame up so the lines between the doors when closed are a bit wider.

Anyone care to hazard a prediction for 2 months - and, if you think it will pass the 2 month test, for longer periods?

[johnny means]

I'd be surprised if these were flat after the finish dries.

[Jim Becker]

Rudy, I have to agree with the others that it's unlikely that those doors will remain stable at only a nominal half-inch thickness. Even three-quarters nominal can be touchy which is why you rarely see doors made of just a solid stock glue-up. With frame and panel, the nature of frame assembly also helps with flatness. (although it's not foolproof) Using a plywood or veneered composite product is also often a more stable solution for a slab door because of how these are made; cross grain layer on the former, for example.

I like your idea of the CNC work on the doors, but personally I'd either do a frame and panel or a veneered sheet goods for more stability as the substrate for the art.

[Alan Schwabacher]

I'd think you'd need a frame to keep the panel flat. But if you make them sliding doors, the groove they slide in may serve the same role.

[Sam Murdoch]

I'd think you'd need a frame to keep the panel flat. But if you make them sliding doors, the groove they slide in may serve the same role.

NO to the groove keeping the panel flat. :( I tried that :) I used 1" MDF panels on wheels. Not really a groove on the bottom but a metal track and a groove on the top. Still These were equally finished on both sides and the ends. I was pretty confident that this very flat substrate would stay flat but it didn't. Not a big bow and only in one of the 2 panels but enough.

OP is trying to reinvent THE DOOR. A good teaching experiment if he attempts it.

Sam

[Zac wingert]

1. OP should not try to reinvent the door. Use frame and panel construction or just use plywood. I assume there is concern with using anything other than solid wood due to the images being routed onto the door? Maybe more information about this would help.

2. I think what OP was actually asking that no one addressed is about how to efficiently sand a project like this. I am thinking a drum sander is the obvious solution, followed with palm/ros for the higher grits if you need to go to 400.

[rudy de haas]

re: Zac Wingert

"I think what OP was actually asking that no one addressed is about how to efficiently sand a project like this. I am thinking a drum sander is the obvious solution, followed with palm/ros for the higher grits if you need to go to 400."

Yes, I actually asked two questions - and this is the second one. I don't have a drum sander but will look into this option. I rough sanded (belt sander + plam sander) my trial door - not that great a result and took far too long.

[Art Mann]

Making 22" X 83" or even 17" X 32" doors out of 1/2" thick Birch planks without a much thicker frame around them will feel extremely flimsy and will almost certainly eventually warp. The situation will be worse if you use a CNC router to try to make the doors look like panel and frame construction. Maybe I am misunderstanding what your intentions are.

If you want that much area to be sanded to 400 grit (why?) you are going to have to spend a lot of time with a random orbital sander. A drum sander that will dothe job is likely to cost a good bit more than what you have budgeted for the entire project.

[Bill Carey]

I'd be surprised if these were flat after the finish dries.

2nd this. And how are you installing the doors? If bi-fold, a continuous piano hinge will help keep them aligned, and the bi-folds clips where they meet will keep the faces (almost) co-planar. If hanging with hinges, I think you'll need to bolt them on - screws in 1/2 material would probably fail. And they will warp / twist. Bypass doors will be held top and bottom, so they will bow - altho careful alignment of the grain can minimize this.

Just out of curiosity - why 1/2"? Material on hand.....

[rudy de haas]

Well, that didn't work - all of you who warned me were right.

I made a test door using the 3/4+"" stuff I had on hand -looked ok. But then just now I tried to plane it down a bit (for weight reasons). Both pieces (about 11" wide x 84 long each) started to curve after being planed down to about 1/2". I guess better aligning the grain before gluing would help, but it would be obvious where this is going even if you guys weren't telling me ahead of time. So no go.

Now, any ideas on how to achieve what I want? - solid doors, light weight, birch or maple color/grain, ranging from tall and narrow (19 x 82) to short and wide (17" x 32") all with face engraved panels of the CNC router kind. ( I don't see plywood or MDF having a role because of the engraving -really don't want to try to veneer an image)

I know I could make floating panel doors with solid panels for the engraving but those always strike me as heavy - and making the panels deep enough for 3d effects means the stiles/rails will need to be >5/8th.

Better ideas, please?

[Adam Herman]

why so much worry about weight? a large cabinet door is essentially what you are building, Millions of them exist at 3/4 or more thick and hang on 2 or 3 hinges no problem for many decades. if your image fits on a 1/2 thick panel, then why wont it fit on a 1/2 thick panel with a 3/4 frame around it?

also: belt sanding is for material removal, not finishing. if your stock is well prepared, you should be able to block sand your door in just a few min per grit. i usually start around 80 or 120 and go to 220.

[Sam Murdoch]

Like Adam - I have to ask - what is wrong with a 3/4" or 7/8 " thick door? Hanging, using most any sort of hinge, would actually be easier than with the thinner doors. My go to answer is to build a conventional frame and panel door but that negates the cool factor of having the frame and panel image done with a CNC. Honestly Rudy - you have me stumped on this one. :rolleyes: :)

Looking forward to solutions yet to be offered.

Sam

[rudy de haas]

why so much worry about weight? a large cabinet door is essentially what you are building, Millions of them exist at 3/4 or more thick and hang on 2 or 3 hinges no problem for many decades. if your image fits on a 1/2 thick panel, then why wont it fit on a 1/2 thick panel with a 3/4 frame around it?

also: belt sanding is for material removal, not finishing. if your stock is well prepared, you should be able to block sand your door in just a few min per grit. i usually start around 80 or 120 and go to 220.

The "cool factor" is important - the 118" closet is a big feature in the main bedroom.

On the other hand, reality favors 3/4" frame with 1/2"panels... :)

[Richard Verwoest]

Maybe try sawing your own veneers at .1" thick, then laminating 5 of them similar to plywood? The middle and outer 2 layers grain direction would run parallel while the other 2 horizontal.

Richard

[rudy de haas]

Maybe try sawing your own veneers at .1" thick, then laminating 5 of them similar to plywood? The middle and outer 2 layers grain direction would run parallel while the other 2 horizontal.

Richard

I can try that.

Any chance something like that is available as a commercial product?

[Richard Verwoest]

I doubt it, but I never looked either.

Richard

[Derek Cohen]

Rudy, there is a way to do this.

Imagine the door to be a table top. What you would then do is use breadboard ends to keep it from warping, etc.

If you plan to go with 1/2” thick boards, then I would advise that the breadboard ends are 3/4” thick for stiffness. Just keep the outside flush if you want this look.

Regards from Perth

Derek

[Sam Murdoch]

With respect Derek - I don't think that an 80" + tall door will stay flat - vertically - even with bread board ends.
A table uses the base to keep warp in the length somewhat under control. Maybe hinges and catches can take
on that same role as the table base - but for a 1/2" thick door I am doubtful.

Sam

[Derek Cohen]

Hi Sam

Perhaps, perhaps not. I think that it could if the breadboard ends were at least 3/4" thick. The hinges would then have to prevent twisting - I think the 1/2" thickness is vulnerable in this regard.

I don't recall Rudy saying why he wants to reduce the 3/4" boards to 1/2", so I imagine it was to save weight. I would not do it that way. I would float a 1/2" panel in a frame. It could be made to look flush. Or Shaker-style.

Regards from Perth

Derek

[Ron Selzer]

to try to answer the second question I recommend you find someone with a wide belt sander that for a price would sand all the panels for you to at least 120, then use a RO sander to step your way to whatever you want
A stroke sander (https://www.grizzly.com/products/Grizzly-Stroke-Sander/G5394) would be my next choice, drum sander would not be anywhere on the list due to previous experiences with them

[John Isgren]

If the 118” door is meant to be a feature I would submit that a well constructed, traditional frame and panel door would be far superior to trying to make a faux version with CNC. I believe the frame and panel style was developed not for aesthetics but rather to manage the wood movement inherent in a door that size.

The other advantage is the component parts can be pre-sanded and then just requiring final sanding with a RO when assembled.

[Art Mann]

Just to add to what Mr. Isgren just said, building techniques didn't just evolve randomly. Things are done a certain way because countless other ways failed to work. The original poster is trying to back up and do things in a way that craftsmen determined wouldn't work decades or centuries ago. He is going to have to use alternative techniques and material(s) besides wood if he achieves the look he wants.

[rudy de haas]

Maybe try sawing your own veneers at .1" thick, then laminating 5 of them similar to plywood? The middle and outer 2 layers grain direction would run parallel while the other 2 horizontal.

Richard


Well, I like to think I'm not lazy..

If I take a 1" birch board, plane it a bit on both sides, then use the jointer to get straight, square, edges, and finally slice it using a 1/8th blade (table saw, my bandsaw is small and leaves rough sides), I can produce three slices that are a bit better than 3/16th. If I layer these as two long sets (about 84") with a cross layer (about 20") between them I get a kind of 21/32" plywood in which the glue layer is not obvious if you just carve into it. Adding breadboard style ends top and bottom after cutting to the right lengths and widths shouldn't be hard :) .

However, It is an astounding amount of work. A small test piece just took an hour++ and that's without breadboard ends. It will get faster with practice, but the nearly right jointer work on the full plank doesn't quite work in the 3/16th thickness - and I am guessing full length pieces will be worse.

So, before I go nuts on trying this, could I please get some more feedback on whether this is likely to work?

[John Isgren]

If you are bound and determined to use the CNC why not go with traditional frame and panel but use the CNC to raise the panels?

[rudy de haas]

If the 118” door is meant to be a feature I would submit that a well constructed, traditional frame and panel door would be far superior to trying to make a faux version with CNC. I believe the frame and panel style was developed not for aesthetics but rather to manage the wood movement inherent in a door that size.

The other advantage is the component parts can be pre-sanded and then just requiring final sanding with a RO when assembled.

The wall this is going against is about 160". Of this about 40" is going to be used for a coffee station (power and water) - the remaining 120" become a closet with either five or six doors totalling 118" - not one door that's nearly 10 feet wide. It's a feature wall of doors...

[John Isgren]

I assumed you were doing ~2ft per door not one big door. Not sure you could ever stop a 120” door from moving :)