Milling Wood Flooring on a Shaper
Extensive advice on how to make hardwood strip flooring on a cabinet-shop shaper. April 19, 2006
Question
Some time within the next six months to a year I plan on running about 1200 sq. ft. of solid wood flooring for my residence. I am a cabinet/furniture maker by trade and I also design and hand grind moulding knives to make moldings for my projects and for others. I use a 12" planer style molder for this type of work. The wood I plan to use is a species of white oak locally known as "Oregon White Oak". It is a very hard oak and I can buy the stock I need for $.85 a board foot. The wood is air dried 5 years or more and most of it is FAS or better, by my eye. I plan at this point to use my Freeborn stile and rail door shaper knives to run the tongue and groove with my 5 horse shaper and power feed. I plan on installing the flooring in the traditional unfinished method.
Does anyone have finer points to share with me about things like the fit of the tongue to the groove, or special tongue shapes? Has any one had any success with running flooring at less than the traditional 3/4" thickness? What about the new snap lock edges? Can they be done on solid wood without the need for V-grooves on the edges of individual flooring boards? Also, any advice about problems I may face because of the stock being air dried would be appreciated. The space the floor will be installed in is heated in winter with a free standing propane stove and circulated a ceiling fan and the movements of people.
Forum Responses
(Solid Wood Machining Forum)
From Dave Rankin, Solid Wood Machining Forum technical advisor:
A lot of the flooring that I have made over the years has a more square groove and a more round tongue. This eliminates bottoming out of the tongue. White oak is a very dense wood and most flooring tolerances will allow for a .005-.015" fit. A question that comes to mind concerning the fit tolerance is the humidity at the manufacturing plant versus the humidity at the installation location.
Another question will be how wide is the flooring and how will it be installed? Pneumatic nailers can handle a little tighter fit than manual nailers. I also recommend that you undercut the tongue and groove. This will allow the only major point of contact to be the top of the floor. There is no need for the end match unless you are installing the floor over concrete and using furring strips.
From contributor A:
Some things you should be aware of - white oak is tough on knife stock. You might want to go carbide or in-house sharpening. Another thing, especially with air dried lumber, is to stick the flooring up for at least 3 weeks in the environment in which you’re going to lay it. This will let it do all the moving its going to do before you lay it, instead of after, and after sanding is done.
I agree with Dave Rankin about the shape of the tongue and groove. As you run your stock, keep checking to ensure a good fit. The knives dull and might change a little. With sharp knives I like to have the tongue holding the bottom of the groove down. But when they wear sometimes just getting them to fit together is tough.
The last thing that comes to mind is to have relief cuts on the bottom of the planks. The style our customers like is several flutes that are about 3/8" wide, about 3/16" deep and spaced about 1" apart. Just make sure there is enough meat on each edge. The only other thing I can think of mentioning is a lot of people (up north) sand their floor too early after laying it. Then it cups after moisture from below gets to it. A sub floor will probably remedy this but just something to keep in mind about the environment in the building.
From contributor B:
Flooring cutters also have a little plow to allow for the head of the nail. The floor will show shrinkage. We primarily run quarter sawn white oak that is kiln dried and still get minimal gapping. It may be just one or two on the whole floor but there is some. I would suggest the v-groove this will hide some of the gaps but it is a dust collector. I think you should kiln dry but to each his own. We lay a 30# roof felt - this prevents squeaks and serves as a sound deadener and somewhat of a vapor barrier. When we lay floor, we start in the middle of the floor and put a spline in between the two grooves. This way the floor can expand and contract from both sides. Get a good pair of knee pads.
From the original questioner:
To Dave Rankin: Thank you for your insights. It is a good question about the floor width. The widest dimension across the grain will be about 12'. If you were questioning how wide the individual floor boards are, I am open to recommendations. The oak boards are random width in the rough and mostly 7" and wider up to 11" or so. I haven’t decided what width to make the finished floor boards.
I am not sure what you mean by "undercutting the tongue" unless you are talking about the relief grooves on the bottom face. Please clarify that point. I am surprised to hear that end matching is not required because I assumed it is always done, but now that I think about it the joints will be staggered and it would seem that the unmatched ends wouldn't be able to lift up or otherwise go anywhere (floor will be laid directly on wooden subfloor).
The shaper knives I plan on using are carbide. I am pretty much stuck with the shape of tongue the cutter will make. I knew I should probably acclimate the stock with stickers but I wasn't sure how long. If I machine the joint so that the tongue puts pressure on the lower groove cheek, does that mean the tongue side of the joint is up from the floor a little on the bottom side once the groove is driven on? And on that subject, how is it laid - first groove towards the wall and subsequent grooves driven onto nailed down tongue boards or vice-versa?
I appreciate the warning about cupping and sanding before it may cup. I have heard that some lay 15 pound felt to curtail squeaking. I wonder if the felt would also serve as a vapor barrier to prevent the problem with cups? I may have a moisture problem because the subfloor is made from individual boards and done by farmers.
From the original questioner:
To contributor B: Thank you - you got that in while I was still typing. I guess I better get a commercial flooring sample to figure out that plow/dado for the nail heads. I knew I could count on you guys to steer me straight! So, on your layment strategy, you face nail the first board? I am still trying to learn if the tongue gets blind nailed or the groove gets blind nailed or if you can blind nail either. My logic says from your method that you are nailing the tongues since you are essentially starting with a two-tongued board in the center of the floor with the added spline. Speaking of expansion, how much expansion space do you leave at the edges for a 12" wide floor with 3/4" thick base board to be installed afterwards?
From Gene Wengert, Sawing and Drying Forum technical advisor:
I would guess that 90% of the flooring problems that I am consulted about involve moisture. Remember that wood in use only shrinks of swells when the MC changes - no other reason. In the wintertime, flooring will typically achieve 6% MC; in the summertime, 8% MC. White oaks moves quite a bit - 3% MC change is 1% size change. I believe that your air dried is a bit too wet. Also, air dried oak has a risk of getting powderpost beetles if it has not been kiln dried. Once kiln dried, the risk is small of a new infection and any existing insects and eggs are killed. One other point - oak will always cup toward the bark if it dries a bit.
From contributor D:
I have some drawings of what the above post mention on my website. Go to my website, profile catalog, T&G section, put in .750" on both min & max and you will see all T & G Profiles I have in 3/4". T&G 750" - 34 has both. There are some with 1/32" relief on the bottom side and some with a nail groove to hide the head of the nail. That should give you a good look of what people are suggesting.
From the original questioner:
To Gene Wengert: Thanks - from what I have seen, powderpost beetles will eat the sap of this particular species (when moist) but not the heart. I plan on using all heart in the flooring. The planning department in my area allows the species to make ground contact if sawn box heart. It is amazingly hearty stuff. If I did take this stock to a kiln, would I need to machine it into flooring before or afterwards? I think it will take me quite a bit of my spare time to machine it from rough, random width stock through ready to install flooring and if I must machine afterwards I am not sure if I will be any better off with regards to MC. My shop is wood heat and although I don’t have any troubles with regard to moisture changes on installed product, I know my shop isn't as dry as my house and the kiln dried material will probably suck up quite a few percentage points of moisture in a couple of weeks.
From contributor A:
The plow I believe we call a nail groove -same thing. It's a little groove at the back of the tongue for the nail head to go and not interfere with the fitting. About the tongue fitting the groove, I use carbide inserts from Weinig. They have a 10 thousandths gap to allow a little play to do that and still keep everything flush. When they wear some you might have to make a slight adjustment. A nice floor sander will fix everything anyway.
Another thing to add to contributor B's reply about the v-groove or (micro bevel) is that it allows you (if machined properly) not to have to floor sand. If you want that look, just run your planks through a drum or wide belt and then nail it down. The bevel will hide the little steps that might be there as well as everything contributor B talked about.
The other thing that was mentioned was the undercut. Insert knives will do this for you. What you want is on the groove side - a little 1/32” is taken off the bottom edge of the groove to make sure the top faces fit together nice and tight. It's pretty common and works great.
From contributor E:
A couple of things from my own personal experience with the same thing - I did about 1100 square feet for my house in Ash. The material I had gave me 4" and 3" with little waste so I went with it and alternated each course. It turned out nice - something to consider if your material is in the 8" range. I did not end match my flooring and it is not a problem. In fact, if you search here at WOODWEB for end matching you will find a thread where I got some good feedback on the subject.
Also - you really may want to consider finding someone with a production moulder to run this for you. The accuracy will be much better which will make installation easier and more importantly less sanding after install. Plus it will be done in one pass. I did mine on my Weinig Proformat with great results. I would not recommend V-grooving – it’s a big dirt and dust catcher. One more thing - this may lead to some debate but I prefer a manual nailer to a pneumatic. Two hits per nail, first pulls the flooring tight and second sets nail.
From the original questioner:
To contributor A: I am with you now on the loose cheek for the bottom side to insure tightness on the face and also the nail head groove, thanks to contributor D and his online profile catalog. There won’t be any accuracy problem with a properly power fed shaper set up but I know it would be faster to have someone with a real molder run this stuff. I am a die hard do-it-yourselfer though. My wife would probably try to kill me if I installed flooring with dust and grit catching v-grooves.
From contributor A:
One last comment - it is strongly recommended to have your rough sawn lumber kiln dried and then mill it into flooring. I would hate to lay a floor vice versa!
From contributor F:
I also recommend kiln drying. Don't try to fool yourself onto thinking it is not necessary. Also, if making flooring on a shaper (I have made thousands of sq. ft. this way), make sure the width of the flooring is not allowed to vary. The best way is to use an outboard fence. This fence is mounted 2-1/4" (or whatever the flooring width is) away from the shaper cutter. The feeder is angled towards this fence, not the shaper fence. Keep the shaper table waxed and have at it. In all honesty you will be money ahead to pay a shop with a real moulder to make it for you while you work on a paying job. But, it is satisfying, gratifying, educational, and frustrating and a good exercise to do it yourself.
From contributor G:
There is a lot of great feedback in this entire string. Here are a couple of my comments based on personal experience. The felt paper will help block moisture and prevent squeaks, and is standard industry practice. The alternative - used primarily in very high-end homes - is to use a quality, thin construction adhesive to glue the floor down directly to a plywood subfloor, in addition to nailing it. Also run a narrow bead of glue below the tongue (not in it) when installing the flooring. Based on the advice about tapering the side, there will be room for the adhesive without forcing the joint apart at the top. Be sure that your plywood subfloor is glued to the floor joists, as this will minimize squeaks. Gluing is a pain - especially if you ever have to remove it, but it will minimize the impact of shrinkage or warp long term.
From what I've seen, the manual nailers will do a better job of tightening up the flooring versus the pneumatic ones. Be sure to use a mallet and a scrap piece of flooring to tighten up the planks prior to nailing. With wide boards, as Gene mentioned, you need to be cognizant of the potential for shrinkage. Realistically, with a 12" plank you should expect for some gaps to open up between the boards. A v-groove will minimize the visual impact from the gaps that open up over time, and if you want to run wide oak planks you should seriously consider the v-groove. If you don't glue it down, if possible keep some extra wood stored in the house for repair. There is a strong chance that you may encounter spot shrinkage and this will allow you to have the correct mc replacement board available. I would consider kiln drying to be mandatory, and also let the wood acclimate for 3 - 4 weeks in the house before installation (in the typical climate-controlled environment that is planned). It sounds simple, but be sure to pick out the best face prior to cutting the T&G. You can't turn the board over with a tapered side cut!
One final comment - when finishing spend the extra money and buy non-yellowing finish. Ten years down the road you'll be glad that you did.
From the original questioner:
On the machining, there really is no taper. What was being discussed was the that the bottom "cheek" on the groove side of the boards needs to be cut 1/32" shy. In other words, if you drive two pieces of flooring together so that they meet tightly on the "working face" of the boards, then flip them over, there will be a 1/32" gap between the two on the bottom side. This insures that the face side can be driven all the way tight even if the joint in the flooring falls on a hump in the floor which would cause the bottom joint to make contact first and prevent the face side joint from closing fully. You can see this machining detail if you follow contributor D’s instructions that he posted above. No worries on 12" wide floor boards. I will be machining them at about 3". I have been choosing the best face of boards for decades and it’s more than automatic for me. I also have it from a good source that if I sticker my 5 years air dried and 6 months in the heated shop machined flooring in the house where it is to be layed for a period of 30 days, and I do it in winter, that I should be in good shape as far as shrinkage.
I think a lot of folks have a phobia about woodworking with air dried stock. If you put it indoors for 6 months or longer after it is dry you won’t likely have any issues with regard to wood movement, any more than you will with kiln dried stock. I have had many years elapse from furniture and cabinet delivery dates using air dried and shop conditioned stock with out any problems.
From contributor F:
Why overlook the obvious? Check the wood with a moisture meter to find out if needs to be kiln dried. Air dried or not, phobia or not, why take a chance?
From contributor E:
A couple more things - first, and you probably already did this - check your subfloor for squeaks and pull offending nails. I replaced mine with screws where I found squeaks.
Also, I like contributor F's idea of an outside fence. This is the same concept a moulder uses when machining the tongue. I don't see the benefit of using that idea for the groove side though, maybe I'm missing something. But if you use it for the tongue (which should be machined after the groove), then if the material is too narrow, but you still have some tongue machined, you have a useable piece. Just as long as the shoulder of the board is okay, a short tongue won't hurt.
From the original questioner:
I guess I’ll be driving a lot of screws because this place squeaks like crazy. I do understand what you are saying in regard to a short tongue not showing in the finished product. The way my cutters work however, the tongue cutter is machining the shoulders simultaneously, so if the tongue is tapered, the whole floor board is tapered too.
While I don’t disagree with contributor F's outboard fence method, I don’t think I yet understand the logic for it.
To contributor F: Help me out - here are my thoughts. I can approach this issue of machining parallel boards from two ways that I have used extensively:
Step one:
I machine the edges of the rough sized floor boards with the jointer and make one edge of each straight. I then run the straightened edges against the table saw fence and saw all boards to 1/32" oversize.
Step Two, Method One:
Gang edge plane all pieces to net size while removing the saw marks and run both shaper set ups (tongue/groove) with fence settings that do not remove any material from the edges. That is, the fence setting of either of the two shaper operations does not allow a change in the overall width dimension of the flooring blanks. Thus, the finished boards are all of equal widths and have parallel edges.
Step Two, Method Two:
After ripping on the table saw, I run the saw marked edge first with a groove setup that machines 1/32" off the edge as it plows the groove. Then, run the tongue side on a set up that does not remove any material from the edge of the tongue as it is machined.
The problem I have understanding the outboard fence method is the reasoning. The only reason I can think of is if someone is concerned that the normal fence might slip or else the power feed won’t keep the material in intimate contact with the fence at all times.
If those are the reasons, why wouldn't they worry that the same two problems will occur with powerfeeding against an outboard type fence?
From contributor F:
The reason I used an outboard fence was from experience. It was my experience that no matter how careful I was in any and all the machining processes, something would still cause slight variations in width. Sawdust against a fence, board pulls away from fence for no apparent reason, board bowed, jointer not accurate, the list is endless. Believe me; some of the above will happen. The trick is to take as much room for error out of the equation. The second best way I know of is to use an out board fence on a shaper - the best being a real moulder. If you remove at least 1/32" off the width while milling each edge you will get both a cleaner and more accurate cut. As contributor E pointed out, mill the groove first. Machine set up is also a little more forgiving. The finished width does not have to be an exact width, just a consistent width. Also by removing that 1/32" your blank prep is also a little more forgiving. The outboard fence guides the cut and also perfectly sizes the flooring consistently. A shaper cutter always wants to push a workpiece away from it. The powerfeeder has to always fight against this. With an outboard fence the cutter pushes the work against the fence - a huge advantage! Some door making systems use this same idea for milling stiles and rails. To summarize- an outboard fence will save lots of time on material prep, machine set-up and actual milling. One comment on lumber MC- if you can have someone with a meter check it for you, please do so. The lumber doesn't care how it got dry, just so it is dry.
From contributor A:
I do our wide flooring on the shaper the same way. The way I look at it is if you joint one edge straight, run your groove the normal way, spin the board around, run the straight edge(grooved) toward the outfeed fence, your board can vary in width but the rough edge will be milled straight (and tongued) at the same time. Prepping both edges is recommended but in theory this is what happens. Just set your outfeed fence to the cutter with a tape measure to the desired face width and clamp it down well.
From the original questioner:
A clarification to avoid confusion to other readers - any references to "outfeed" fence in the prior post should say "outboard" fence. If you use the same terminology I do, the outfeed fence would be the left hand side of a split shaper fence as you face the spindle from the front of the shaper.
From the original questioner:
A quick question for you outboard fence guys - if the flooring blanks measure say, 3.25 inches after machining the groove edge, and then you set the outboard fence to the dimension of your desired floor face dimension from the outboard fence to the cutting edge that mills the tongues cheeks, at this point, do you set the regular shaper fence to a setting that clears the 3.25" flooring blank and still allows for dust collection or what (i.e., wider than 3.25" between the two fences)?
From the original questioner:
I’ve been out west all my working life and it has always been called "tailing off" when I heard it. Off bearing is probably new and I’m out of loop now due to working alone.
That’s one nice thing about forums if you work alone. You can still keep up on industry changes.
From contributor A:
I like to keep the first infeed fence about 1/16 away from the stock. The 2nd infeed fence I run as close as possible. I guess in theory it doesn’t matter, but whatever works best. If dust collection is an issue, play around a little. Our shop has all industrial machines so I never really had to think about that. But that seems like an issue worth looking at.
From contributor F:
The infeed fence does not have to contact the wood. As contributor A said, I also keep it fairly close. The door systems I mention earlier use a spring loaded infeed fence to push the wood against the outboard fence along with the powerfeeder.
From the original questioner:
To contributor F: The door system sounds interesting. But, I have to tell you, I must have run thousands of door stiles and rails through a sticking set up using a standard shaper fence (inboard?) and I don’t recall a single failure in the out of parallel category. I must confess though that I take my set ups more seriously than the average Joe.
From contributor F:
One huge advantage I forgot to mention. The outboard fence is one continuous fence. It is not split and does not have a gap between the infeed and outfeed fences. This is a lot safer and easier especially when milling parts as short as 3-1/2". The feeder is of course angled to this outboard fence which also helps here. If door parts are slightly out of parallel or vary slightly in width it would not be noticeable. In flooring, when pieces are butted end for end and courses are laid tightly beside each other with staggered joints, slight variations in width readily show up. Belted powerfeeders are much better at feeding short pieces.
