Order of Operations in Solid Wood Panel Assembly
Is it better to dress lumber before gluing up wood panels, or glue up first and then sand the whole panel? People do it both ways; here, they discuss why. December 6, 2006
Question
For years, I've been face jointing, planing, edge jointing and ripping stock to get it flat and close to final thickness before gluing panels with it. Seems like many (most) production shops just gang rip S2S lumber and go right to panel clamps. Then thickness using a planer and/or wide belt to get it to final thickness.
Many of my solid wood panels go into shaker looking doors with a reverse raised surface flush to the back of the door. Every other job seems to have some panels wider than can be practically flattened after gluing up. My panels are flat right out of the clamps and just need the glue squeeze out plus maybe 1/32" total wood thickness removed before they are ready for shaping. What do you see as the pros and cons of these two different approaches?
Forum Responses
(Cabinetmaking Forum)
From contributor F:
I work my solid stock the same way you do. I think in my case it comes from having a background in furniture making. I know that companies that make strictly cabinet doors certainly do not flatten stock on the jointer and I think most of them glue up edges that are sawn rather than machined.
The methods they use certainly seem to work okay. I have never seen any open joints between the panel boards and the panels are reasonably flat. I would assume that, like we do, they start with 4/4 rough or 15/16" hit or miss so that misaligned boards will be taken care of in the wide belt sander and still net at .75".
I think if you have the machinery, it's a profitable way to make door panels. At minimum, you would need a wide belt and carousel panel clamps. A good sharp 60 tooth ATB saw blade should saw a glue quality joint even on a common 10" table saw.
It would take some getting used to for me. As I can guess you probably know, once you learn to flatten and straighten your stock and you get used to doing so, trying to work a piece of wood that has not been dressed yet just doesn't feel right. But in the instance of panel work, gluing up boards straight from the stack can make a quality panel from a structural standpoint. If I was to do it, I would probably reverse the bow in mating boards to aid in straightness.
Then the tradeoff will be in the grain and color match, since certain boards would have to be placed in the panel according to their shape instead of just their appearance. All that said, there is nothing like gluing up a panel with flattened and straightened members that has joints so flush that a little glue scraping and a light sanding is all that stands in the way of the finish.
From contributor K:
We mill our solid stock per your directions also, but we edge joint both edges prior to glue up. We manufacture frameless cabinets, and a truly flat door is a must. We also flat joint and plane each piece of rail and stile stock as well. This process is the only way to achieve a flat, stable product. I know that this is not the most expedient way to manufacture doors and drawer fronts, but we are committed to producing hand built quality at a production price point.
From contributor J:
If you are paid for all the extra time you spend making your panels, that is great. We used to buy our lumber s2s, then rip it with a glue joint blade. When gluing, we alternated the growth rings and removed the glue squeeze out while it was still wet. After removing from the clamps, we planed the panels to thickness with about 4 passes through the planer, taking light cuts. Finally we ran them through the wide belt sander. These panels were always flat.
From contributor R:
I take 13/16'' and do as contributor J did.
From contributor I:
I've done it all different ways in the factories and custom shops. We had the revolving clamp carriers and wide belt sanders to finish thickness. In my home shop, I face and finish plane to the final thickness. By using curved cauls, my panels only need to scrape the glue off and finish sand. The cauls give perfect alignment and can be used in other case work glue-ups. The extra attention in prep milling saves having to fix problems after the glue up.
From the original questioner:
I should have said up front that my panel process is fairly quick because I have dedicated facing and edging jointers, both with power feeders. One other advantage to pre-milling is that I can use twisted and/or crooked lumber for panels and improve my yield. I guess it is a balancing act. I'll be looking into a hybrid method where the lumber gets some flattening treatment before glue-up, but not quite to the level that I have been. Thanks for the input.
From contributor R:
I know I'm nutty, but isn't planing one wide panel in a wide planer faster than planing two boards, then gluing them together, then drum or wide belt sanding the wide panel? It just sounds like a few people are doing the best they can with small planers and sanders. Now edge joining them on the jointer is a good thing, but not an absolute to get a good panel.
From contributor F:
Yes, if the individual boards were already flat and free of twist, cup and bow, it would be pointless. What is being discussed is that some makers face joint individual panel boards for the purpose of flattening them before gluing them to the mating boards in the panel. When you flatten the face of a board on a jointer, the second step is to run that flattened face down on the planer bed to transfer that flatness and straightness to the opposite face of the board. Although this practice is time consuming and definitely not a production technique, it is amazing how flush you are able to glue up the boards in a multi board panel when they have been dressed out this way. I am often able to do a little glue scraping and light finish sanding to have such a panel ready for finish.
From the original questioner:
I get what you're saying. For some of the species that I use, panel first, plane second would definitely be possible (and faster). Seems like 20% of the lumber that I get is too "non-flat" to use for frame parts, but does work for panels if it is cut to rough length and flattened first. With dedicated facing setup, this is not a huge slowdown.
From contributor R:
I start out with 13/16'', sometimes gluing up 8 ft long boards. Sometimes I rough cut to panel length. I either plane to 5/8'' or back cut them if I get them flat before 5/8, and I'm talking raised panel. I did not mean to knock small machinery, but I have learned the larger and more costly the machine, the easier it is and more money it produces.
From the original questioner:
I agree - the right production equipment is a must for making money. I'm looking at going from my 20" shelix planer to a 24" model, which is one reason I was asking about panels. All of my doors finish at 13/16", so that is another factor. It is one more thing that sets me apart to the customer. It's a subtle perceived quality thing, but it does mean that I have less wiggle room in my stock flatness. I'm starting with H&M 15/16" for most everything.
From contributor R:
I have the 20'' straight knife, and was hoping to move to the Byrd schelical head. Does your shelical head work as they claim? The bad thing is that doors have bottomed out in price. The only way they can get any lower in cost is that they come from China. I guess you would need more than 20'' if doing frameless boxes. But I'm getting by with my 20'' on face frame cabinets and building doors 24'' with 1/2'' overlay. Most door manufacturers are calling 22'' the normal, and anything over 22'' is an upgrade. I would really like a 37'' wide belt so I could build 36'' entry doors, but the truth is I can get an oak or mahogany from South America for about what the lumber and glass would cost me, so I can't see dropping that dough.
From the original questioner:
The shelix head does work well, like they say. Minimal tear-out (although I did get maybe 5% of a recent 1200 BF ribbon grain African mahogany run get bad tear out on grain reversals). Seems to like more horsepower than straight knives. Quiet. Forget about "reading" the grain - just shove stuff through. I just ordered a custom head for my facing jointer.
The comments below were added after this Forum discussion was archived as a Knowledge Base article (add your comment).
Comment from contributor E:
All of this starts at the stock selection and unfortunately that selection is getting poorer by the year. Younger trees, compression wood (trees that grew at a marked slant in the forest) that would never have been harvested in the past and accelerated drying schedules all contribute negatively to lumber that is just inferior in quality. In the beginning I rejected and sent back to my supplier more than 20% of the stock I get. Now they know I do that and select it with a keener eye for me so the number is closer to 5%. Does every shop get that treatment? I doubt it, but I do because I insist on it. I also ask them to skip-dress my stock at least two days before it is delivered so that I can note any twisting that results in removing the shell (first dress) before it hits my rack. It's easier to reject a piece on the truck than do the paperwork to send it back later.
As to the machinery: let's not forget that a planer or a widebelt cannot and does not flatten anything. They smooth one side only. Yes, there is a certain amount of flattening done but the simple fact is that if we do not choose our stock carefully, joint it (without power feeders, they won't take any real twist out of a board-just cup) with either a jointer (if you buy a big one get a really big one - say 20" and set the fence at a 15 degree slant to the blade. It kicks the crap out of any helical blade I have ever used because it slices the wood like a knife to get the absolute best results very quickly. Not enough throughput for your shop? Get a double sided planer with a carpet or pin feeder. It is the only planer system that will actually flatten a board or a panel on both sides in one pass. I think they are available up to 30" and the smaller ones can be had a less than $30k.
New whiz-bang machinery is great and can make us all more money in volume, but let's not get all that far from the basics of wood mechanics. After all, these methods have been worked out over centuries and have some powerful truth attached to them. Wide belt sanders, huge capacity planers and other machinery can make it seem we have succeeded in producing a quality product but came back to an entry door that you fabricated for a New England beach house ten years later and tell me how straight it still is, or whether it is still sound at the joints - that's our test.
