Louvered Door Assembly Techniques

Here's an extended discussion about systematizing and speeding up the fussy, complicated process of assembling big louvered doors. October 11, 2012

I build doors, however what I haven't done a lot of is louvered doors. They've all been fun, weird shaped projects, but few and far between - probably less than a hundred in 15 years. Anyway, machining the parts is no problem, but a large order of huge doors for a golf course clubhouse is testing my patience.

When you have over sixty slats and three rails to fit tight into their mortises, how do you keep everything in alignment? Some sort of jig? It seems a bit complicated and would take almost as much time as the door itself. We've always just slightly tapered the tip ends of the slats to aid in getting them started, then go from one end of the stile to the other, fussing with each slat until everything is started before clamping. There has to be a better way.

Forum Responses
(Architectural Woodworking Forum)
From contributor M:
We have cut jigs on our CNC router and pulled them together lying flat on a table. Can you run an extra piece with the grooves in it and split in half for an exact jig? Like you, we don't do a whole lot of them.

From the original questioner:
I was overthinking this, looking at the jig as a 5 axis surface problem to machine out. After lunch today I made a few extra stiles out of scrap, and ripped off a couple inches, essentially leaving 2" stiles with though mortises. I flipped these 90 degrees and ripped them in half, and glued a bunch together, leaving a washboard jig. Hopefully this will work, I haven't tried it yet. Sometimes the obvious is staring right at you.

From contributor O:
I am about to give away a little trade secret I learned in a shutter shop in the mid 70's. I know you will be able to appreciate it, so I can't think of a better recipient. We cut the slats square ended, but the slat mortises were about 1/64 longer in one stile than the other. We had Festo equipment, and this was easy to do. In fact was done once about a century ago, and never changed.

The slats were all just under 1/4" and matched the mortises pretty well in thickness but not at all tight - would fall out in fact were it not for the width of the slat. The slats were fixed in width so that the tight mortises (1/64 less) held the slats firmly, and the 'loose' mortises (relative term) held them looser. The slats were all put into the tight side stile and then glue applied to the rail mortises and then the assembly laid down on risers on the bench.

The risers kept the work flat, about 2" above the bench for both clamps and hands to get underneath. The tight side mortises kept all the slats in place, in the bottom of their mortises as it laid down. Starting at one end, a clamp was tightened to bring the slats to their mating mortises on the loose stile. One hand would rock or push the slat down, then the other hand would push it back up until it mated into the mortise. This rocking down and back would pull the slat out of the tight mortise just enough - 1/16" or more - to slide into the loose mortise and stop as it hit the mortise edge. Then repeat on the next slat, then a few more. Tighten a clamp, rock the slats, tighten clamp, and then you have them all in place and can screw the clamps down completely. I would then tap the stile faces along their length with a mallet to insure the slats are all seated and sight the stiles to insure they are straight. Pin them and set aside to dry.

If both mortises/stile are tight, then it is too hard to get the slat to rock into the mortise - you rock like crazy, but the hands tire, the slats don't come out much, or won't fit into the mortises, and the glue dries. If too loose, then they fall out and you are playing what we called 'pick up sticks'. Now, the doors with two panels in width of louvers were a whole different story.

When touring the old Morgan factory, they had ancient wooden jigs that would hold slats at the right angle while the stiles were pulled up hydraulically. The slat ends were all tapered, so no fitting was required. The Schumacher shop crimped all their ends by crushing them. Then as they aged, and picked up moisture, they were supposed to self-tighten.

Do a dry run or two and learn to size the slats correctly - it took me years, it seems, to get it right. Once you get it, it is a beautiful thing. Coincidentally, I have a series of louvers to do this on next week, so I'll be rocking with you.

From the original questioner:
Great stuff Contributor O. Very interesting about the crimping part as well as a looser fit in one stile. I had been ever so slightly adjusting the fit, and got it too loose on one door - it was an easier assembly but after it was clamped up I slapped the slats and everything was a bit too rattled for my taste. Back to tight and sanding a taper on the end. I should note they were tight but no glue was used on the slats. I'm thinking with one side loose, but with a little blob of glue in the bottom of each mortise might be the way to go.

I'm going in this morning to work by and I plan to put a few left side stiles up on the CNC and buzz out the mortises just a hair larger. This, combined with my new jig should make putting these together a much more efficient and enjoyable task.

I briefly entertained the crimping idea and how to implement it. By not having the equipment I thought I might be able to improvise by laying a bunch of slats on the floor, placing a 12" wide piece of steel an inch from the ends and then running over the steel with a forklift. It could be quicker than sanding tapers. If it was too heavy and crushed the ends, I could just cut off the crushed part and nothing would be ruined. Anyway, this run is not big enough to justify all the business.

From contributor O:
I have never used glue on the slats - no time for that. They are cut to about 1/64 less than the opening/mortise bottoms, and fit fairly tight to width, less so in thickness. On tall runs of slats with no mid rails, I will pin a few slats on each end to keep the stiles from spreading and dropping slats. I would think that is more effective than glue. Everything is way too tight for rattling - I don't think I have ever had that happen.

Once my panels are assembled, after beating on the stiles to seat all the slats, I would also give all the slats a good flex one way then the other with my hand to just to insure everything was relaxed and where it should be. Occasionally there is an errant slat that is bowed out of line, and it must be replaced. The fixed louvers are easy to replace, but the operable louvers are another story.

From contributor U:
Contributor O is giving excellent advice on the assembly technique. I have always found that a jig on operable and fixed louvers for assembly just adds to the problem and headaches. I like the idea of using risers to get the panel off the table surface. We have been placing supports under the top and bottom rail to allow for getting our hand under the louvers. We have always cut out louver slots in the stiles first, and then run the louvers slightly thin. We also mould our louver width slightly less than the length of the slot cut into the stile.

By doing both it really makes assembly much easier. With practice, running your clamps in as you position the louvers will get easier as well. I have never glued louvers, just the rail/stile joints. As far as any louver movement is concerned, we find our finish application usually offsets the allowances made for assembly. I have always believed that movement is preferable to everything being too tight, especially when or if the product gets placed in areas where moisture related expansion will be taking place.

From the original questioner:
I had a better day with one side loose and the jig was discarded in the first five minutes. These doors are hefty, slats being a stiff 1/2" x 2 1/8" quatersawn sapele, so I guess I'm in trouble if I ever need to replace one. It may be possible to sandwich two 1/4" thick pieces quarter rounded I suppose.

I am mostly machining the parts for others to assemble so I probably won't build up the skill level that comes with years of practice. I'm just trying to make the process easier. I still haven't trashed the jig - it seems like it would work if the slat ends were significantly tapered, as is some slats were just slightly off using it, preventing a straight up immediate clamping. It was made with the same tight fit as my original mortises, so there was no room to wiggle anything. Almost makes me think I should be concentrating on a router table jig to taper the ends. In the past we always just sanded the ends a bit - not much more than a chamfer really, but still time consuming.

From contributor U:
I never could figure out a simple way to do this with the router. The way we taper our ends is to rotate them on an inverted pad sander while holding them at something like a 45 degree angle. To make it easier, we made a stand that holds a 1/4 pad sander with 120x paper.

Buy holding at the correct angle, you can sand down the last 3/16" or so. We try to taper the edge more than we do the flat sides, and that seems to work best for us. Also, if you sand a little too much off, it is more forgiving to the appearance if the edge is overdone versus the flat area being oversanded. With practice, you can walk through a lot of louvers pretty quickly, at least fast enough to make it a fair trade in the time saved in assembling panels (not to mention the savings in not getting so frustrated). We usually size our louver length so at least 5/16" on each end is in the slot, so our tapers are pretty well hidden.

From contributor O:
Examining my old habits, I can add a bit more: The end tapering always stopped me since it was another step, with jigs, time, and its own faults and merits. Like the jig, I see it as a hindrance to the goal of assembly. You end up working for that goal instead of the end goal.

The fit (the tension) is in the width instead of thickness to allow easy placement into the tight stile. The rocking motion is used here also, with one edge going in tight to the rounded mortise end, and rocking down into the bottom. This same motion then allows the slat to come back out to slide width-wise into the loose stile and stop at the rounded end of that mortise, lining it up perfectly for the coming clamp closure.

Also, the slats mortise in about 5/16", but the rail mortises were a minimum of 1", allowing the rails to be started into their respective mortises so you could focus on the slats.

I should ad that I built fixed and moveable slat shutters for 4 years, every day, with a man that had been doing it for 40 years, and he had a trick for everything, so we could really crank them out. A 40" tall panel of louvers would assemble in about 5 minutes - slat placement, glue all mortises, place rails, add second stile, align and start rail mortises, then the slat mortises, then full clamp and pin. The one hand over the other slat wiggling was so quick as to be almost unnoticeable, and absolutely no motion was exerted without purpose.

From the original questioner:

I'd love to get to that five minute mark. You guys had that down like some kind of pit stop in a NASCAR race. We probably average a half hour, sometimes with two of them at it, grumpy and stressed trying to get it clamped before the glue dries (this is with all tight fits). Yesterday I clamped up a couple by myself that took maybe that long but I was definitely not in a hurry. I was looking at things trying to figure out a better way. It was quite relaxed actually, compared to many of my other glue-ups in the past. I must admit though I like the way doors feel when the slats are rock solid tight (which these were not quite) but passable after a few hidden micropin shots, which in the end is fine. The customer doesn't know they're there.

From contributor N:
I donít have any tips on direct setting louver in to the stiles, however what we sometimes do is slot a 1/2" thick frame that is slightly thinner than the door to create a deliberate reveal instead of slotting the stile. The door gets built as completely open. Slats are a little loose and easy to set in a thinner frame when set up on a bench or table. Lower the frame now with blades set in to the door opening, spread and nail. Set a matching size trim piece at top and bottom. There are some pros to this method, but ultimately it's just picking where you want your bottle neck to be.