Hey-
I’m making a small, square, tabletop out of QS oak. In order to avoid end grain on two sides of the table’s edge I’ve cut four. 8″ wide by 18″ pieces, with a 45 degree between the long and short sides on each end. (I think the technical term for each piece would be an isosceles trapezoid.) This will net an 18″ square tabletop. A small square of maple will float in the middle via t&g) I’ll be using long splines to join the four angled corners.
Where should I put the glue to secure the splines? One small application in the middle? Or should I use two different splines (one at each end of the groove) and apply the glue there?
The obvious problem will be movement. I’m just not sure, in this configuration, how much there will be or which way it will go. Thanks.
Replies
Hi, Christopher ...
The technical term for what you're doing with those ends is mitering them just, for example, as you might find on a picture frame. Splines are a competent solution for the weakness inherent in a simple butt miter.
SPLINES, full-length or blind, are inserted into mating grooves along mitered ends with short "length" grain running perpendicular to the joint. Spline width runs parallel to the joint line.
KEYS may be made in several configurations, but are characteristically cut to allow greater with-grain length and are installed at 90 degrees to the joint line.
Your Full-Length Splines:
Because each joint is comprised of two pieces of equal cross-section of the same wood, there will be no issues of stress induced by differential cross-grain movement. Your splines can be full length, and should be glued full-length to maximize joint strength.
Be careful not to let the inner ends project into the top's dado. Most elegant would be to chisel-cut the inner ends to chevrons so as to straddle the dado, then let the outer ends (of the deliberately overlong splines) project at assembly, subject to final trimming.
Cut the splines "the wrong way" to ensure that their long grain is perpendicular to the joint. That is, they're several inches wide and perhaps 1/2" (or more) long, if "long" follows the grain.
Proper fit for splines would be tight when hand-inserted ... that is, not driven in, but needing moderate hand pressure to seat. Too tight will invite splitting, and too loose will add little strength. Technically, desired clearance for perfect glue performance will have 0.003"-0.005" clearance at each mating surface (which in practice usually equates about to a snug hand fit.) Width needs to be near-perfect to minimize gap at the visible end. Procedure is to cut the spline dados, then adjust tablesaw to cut well-fitted splines. Dry fit to be sure everything works before committing to glue.
Think about where excess glue can go before gluing things up. If excess can't escape, the joints won't stay closed, or may not close at all. Glue will begin to set very quickly in a well-made joint. It may be worth your while to dry-fit all but one joint, and glue the first two one-at-a-time until the routine feels familiar to you for the last pair (which must be done simultaneously.)
Worry card:
If glue is extruded into the top's dado, the top will be firmly glued and could split. With no access to the inner end of the spline after assembly, you may find a blind spline to be a better choice. It's just like the one you're envisioning, but a little shorter on each end, and the dados in each butt are more like mortises - a slot stopped before it runs out the end of the joint (cut with your router or mortise chisels.) To give excess glue a place to go, you can drill a small hole in the underside of each piece at the bottom of the mortise, and plug or ignore them later at your option. Another option: make the splines shorter than the mortises and chamfer spline ends so as to force excess glue into a hollow recess at each side of the spline.
Oak is extremely porous, especially on end grain (exposed at the spline ends.) If staining, sealing the end grain first will prevent its becoming overly dark.
Another option is to capitalize on the natural differences in appearance between the oak and maple, and make the splines of maple so as to echo the differences visible in the top.
--------------------------------
ALTERNATE METHODS:
A Keyed Miter Joint is assembled first as a simple glued butt miter. (Weak, but strong enough to hold together as you finish the work.) The assembled frame is placed vertically in a jig to hold it at 45 degrees above the tablesaw top, and a deep slot cut is made through the joint. A long key, similar to the spline, is then inserted so as to fill the entire slot. It's a BIG, strong reinforcement extending far beyond the width of a spline. Thickness of either spline or key might be equal to about 1/3 the stock thickness. This key looks something like a "spline run the other way."
The feather key is a third, and perhaps overly sophisticated, method wherein the miter is butt-glued, and a pair of skinny slots cut at each corner perpendicular to the joint line. Those slots are then filled with keys (feathers) made of veneer.
Others:
Half-lapped corner, half-lapped miter, mitered slip, blind spline, face key, dovetail key, butterfly key.
==========================================
Don't worry about the *right* way. All are used, and all are suitable. Do what suits your design fancy and is most comfortable for you to accomplish.
As always, lotsa words for a simple picture. Hope it's helped a little.
Regards,
---John
================================
---Pericles
John-
Thanks so much for the fantastic answer. I'm going to cut splines from both oak and maple and take a look at what works best visually. You've also got me taking a hard look at the way I glue up near the interior dados (dadoes?).
I have high hopes for the piece -- inspired by the Joseph Hoffman table in Mark Rodel's "Arts & Crafts Furniture" book. Maybe I'll pop a picture into the gallery when it is complete.
Thanks again
Chris
Hi again, Chris - You are most welcome - glad you found something useful in it. Some object to "overlong" answers, but it's sort of hard to divine exactly what it is that's going to be of help!Yes, our impossibly convoluted and beautiful language terms them "dados" *or* "dadoes."Purist's language:
A DADO is housed on both sides and runs perpendicular to the grain.
A GROOVE is housed on both sides and runs parallel to the grain.
A RABBET is housed on one edge only, and runs either cross OR parallel to grain.Your slots for the splines are dados, but the same slots for the tabletop lips are grooves. You may not run a groove at any other angle because it's just too confusing to name. And da foot bone connect to da ankle bone.Suggestion, if blind splines are a problem. You could pre-glue short spline fillers at the inner corners to create dams against intrusion at assembly - they'd never be seen, and would not compromise the joint. If you ensure that there's also an alternate bleed pathway, then no glue will migrate into unintended areas. You can even trim the edges of the splines to be a tad narrow, leaving room for glue, save but for the last 1/4" or so at the outside corners.Make a few in trial joints in scrap pine or the like, and fiddle until you're confident that it's under control. Saw 'em up to see how the joints came together. Then YOU can write the expert advice for the rest of us ! Regards, ---John
====================================="What you leave behind is not what is engraved in stone monuments, but what is woven into the lives of others."
---Pericles
"Some object to "overlong" answers" You'll never get any complaints from moi for a detailed, descriptive answer! Beats the heck out of
with no other info, especially if more info is obviously needed by the OP.forestgirl -- you can take the girl out of the forest, but you can't take the forest out of the girl ;-)
Hi, F.G. -Nice to see your smiling face. Thanks much for the kind post.---John
============================================="What you leave behind is not what is engraved in stone monuments, but what is woven into the lives of others."
---Pericles
love this forum.....copied your post for future.one other question, when building a table top with a "picture frame" sor t of configuration, any tips on the expansion of boards infilling the frame? i intend to tongue and groove them, but am concerned about the loosening up of the joint between the "frame" and the field. also, does it make a difference that i will be using "Chunkier" than usual boards (8/4 cherry)?thanks,
john
Hi, John ...Good questions all. How come you get to use so few words, and I gotta respond with so many...?? As with any frame-and-panel construction, you must allow room for the panel to expand inside the frame. Lengthwise will be little, but cross-grain may be up to 1/4" per foot. That may seem like a lot, but I'm a believer.... I witnessed my daughter-in-law's gorgeous hand made maple and walnut frame/panel headboard self-destruct. (No, I didn't make it.) Would never have been a problem but for the fact that house exchange necessitated storage for a few months. Money being tight, they put it in a "storage unit." Nice and dry, well-padded, responsible decisions, space in joints, etc. - - but the rains came and went and humidity made a wild swing, and it emerged badly damaged for lack of sufficient expansion room. Cause? Allowance for "standard" (whatever that is) expansion room, such as for cabinet door panels...about 1/3 of what was needed. Good sense suggests that we hope for the best but design for the worst, and fer pete's sake account for actual conditions and dimensions.Short answer - you MUST allow a gap between the frame and the field if you expect the assembly not to explode some day. Workaround: flat plywood fill panel - they don't move. The trick lies in imagining what will happen as the panel expands and contracts, and then managing resulting appearance. You simply have to accept that you and I aren't in charge of whether wood has a natural reaction to change in moisture content. All we can do is acknowledge that as fact and then attempt to design with some modicum of good sense.
===========================================
A main key is to allow for expansion PER FOOT (or inch) of cross-grain width. You can't say that some number - maybe 1/4" - is good. It has to relate to the overall width of the piece. Naturally, you'd probably choose to split the total allowance with half at each side.Allow about 10% of that for linear expansion - it's tiny and for most people-sized furniture ventures not even an issue. That implies differing groove depths. Actually, you could groove all of your frame stock with one setting, then fill the rail grooves as needed to support the panel at "vertical" center =/- 10%. (Rails are the short sides.)Try to use well-dried and stable wood with moisture content under 8%. Don't trust "Oh, it's kiln-dried -- wonderful stuff." Kiln-dried to exactly what moisture level? Rule of thumb for air-drying is 1 year per net inch of thickness. If you're not sure, set kiln-dried aside for at least half that long and let 'er naturalize. Keep it in the shop for not less than 2 weeks per inch before working. Moisture meter's a good tool to have. Kiln-dried too fast, and it will have internal tearing. If you spot that problem, get another piece of wood. Roast a marshmallow with the old one.OK - so now we have TONS of room, the joint is sufficiently sloppy, and what's next? You can use the little foam balls made for the purpose to keep the panel centered in its opening, or you can put a spot of glue at the backside center of the rails bordering the cross-grain edges so as to retain the panel at center. If you use that method, you'll want to assemble with the panel horizontal, and move it for perfect centering before the glue sets.As the panel expands and contracts, lo and behold the unfinished portion at the edges shows up in dry weather. Ratz. Finish the panel before assembly, but for that little spot at edge center for the glue dot. At final finish, mask it off. Pay especial attention to the panel/frame joint. If you glop in too much finish it will bridge the gap, and then a line of finish will snap loose and show as the panel moves. Many builders will add chamfers, beads, dancing girls, or whatever to help mask the fact that the joint is changing a bit ... that is, to catch the eye with detail so the mundane stuff fades into the background.If the center panel is comprised of several tongue & groove joints, then their total free crossgrain room needs to be equal to that required for a single panel. A nice result is that the distributed joint movement will be less noticeable. A disadvantage is that finishing is likely to glue'em all together. Also, the closed nature of the joints will capture dust, which may in an extreme case somehow absorb fairy glue (a spilled drink, perhaps?) and become hard enough to interfere with expansion. One solution is to stain and apply ONE coat of finish to the tongues and edges and to completely finish the faces before assembly (mask off the glue dot reserve.) Then mask off the lot for final finishing after assembly. No help for fairy glue.You could REALLY make yourself crazy and determine the center point of each piece and use a glue dot for each one. Huh? Figure the total expansion room and divide by the number of panels to get the value "spacing." Cut the tongues just shy so that panel edges can actually abut. At assembly, insert loose shims (pieces of playing cards, e.g.) above the tongues of thickness equal to "spacing" to space the panels apart evenly - and apart from the frame edges - before glue dots set. If panelpiece widths vary, then distribute joint spacing proportionally. You'll still need tongue/groove depths to allow for inclusion of foam balls to maintain equal spacing among the pieces.Making sense? If you've a bunch of pieces, then you need springs (foam balls) between them to keep them centered. Using spacers at assembly will ensure that no line of balls collapses too far. Alternate - center-glue each piece in its rail.
----------------------
Stock thickness shouldn't be a factor, provided that you're starting with properly conditioned wood. If you really need that much thickness for strength, then I'd suggest fairly thick tongues (about 1/3 of stock thickness) so that bending of any single panelpiece doesn't easily snap its tongue.Quartersawn for the field will be more stable dimensionally than flatsawn or riftsawn.Don't know the size of the finished table, nor its intended service. If the field is large, you might want to consider underside battens to help ensure that the individual parts don't go too far astray vertically. Should you use them, remember that they must allow for movement - one tight screw, and the remainder in slotted holes.If I may ask, why would you use such an abundance of premium wood to make such a thick top?Once again, too many words. Hope there's someting in there that you actually needed.--- John=======================================================
"What you leave behind is not what is engraved in stone
monuments, but what is woven into the lives of others."
---Pericles
so, can we say plywood with veneer for the field OR wide joints in the field, but not tight boards (in paralell) in the field, quartersawn notwithstanding?
your posts are awesome! thank you, looking up your other threads.
Ummm - I think I understand what you're generalizing. Tight boards in the field are just the same as one big panel. And that's OK until you bind it up with a frame, two of whose sides are perpendicular to the field's grain. That's when you get into expansion issues. If there were no end crosspieces (e.g., no frame), then the whole thing would be fine.If the frame were designed to allow for expansion, that would be OK. If the ends were made as breadboard ends, that would be OK. But a tight frame around any solid wood filler is gonna be disastrous without expansion allowances. Plywood is an excellent solution to the expansion issue. Were I doing it, I'd be likely to rout or plane a tiny chamfer around a veneer infill just to differentiate it visually from the frame.Blashphemy! Plywood! Hang him! They had no plywood in 1731!
Give me a break. But beware. It's hard to find good plywood with no core voids. Baltic birch is a bit pedestrian, but is virtually void-free and exceptionally stable - a great veneer substrate. You may be able to find good hardwood plywood, but ask lots of questions and ask to see a sample.If the field is some tiny thing, then all this goes away - differentials won't be large enough to break it.Was that in any way a useful answer?Regards,---John=======================================================
"What you leave behind is not what is engraved in stone
monuments, but what is woven into the lives of others."
---Pericles
blasphemy....curses!
with a breadboard end, say it was a dado on the end with a splice or tongue, would you pin it in the middle then...so it could expand? i guess i'm asking how to pin the breadboard to the rest.
you're going to be sorry to answer, i'm gonna learn everything i can!
btw, the thick stock is just asthetic.
thanks so much
Breadboard:
Ends of center board(s) have a long tongue which eventually is cut to leave 3 long tenons - one at center and each end - plus a short stub tenon in between the long ones. It's like a long tenon all the way across the ends which is then cut shorter in 2 center areas so as to leave 3 actual full tenons plus wide stubs stickin' out in between. That arrangement will minimize weakening of the end board (by not mortising overly deep over its full length), but use the stub tenons to fill void between elements. Splice not as good - not as strong. Take ya longer to make than just cutting the danged tenons as above.Personally, I'd glue the center section into a single full-width panel. Splines or T/G are OK, but will add nothing - a full-length glue joint will be stonger than the wood it joins. You can still chamfer edges before glueup if you prefer the grooved detail. Greatest advantage is that you'll not have the slightest chance of any of them deciding to bow or twist out of alignment with the rest. Once it's glued up, plane it flat and parallel.If your stock is not quartersawn, then look closely at its ring pattern. An edge-to-edge glueup of multiple boards should alternate orientation - bark up, bark down, bark up, bark down, etc. Also arrange them for best grain match on the visible side, then mark the underside so you can reassemble in order when you're gluing it up.The end board should have the three mortises cut, and then have a shallow groove cut full length to house the stubs. Be sure that each end mortise is wide enough to absorb half the designed expansion when its tenon moves. The center tenon is pinned and glued. The end tenons have slots cut rather than pin holes, and are fitted with pins through the slots so the top's edge can move sideways, but the end board can't distort away from its intended horizontal line. They have no glue.You can hide pins via bottom insertion, or you can make them decorative, e.g., of contrasting wood, and insert them from the top. If not fully through the wood, drill their holes within a skosh of emergence - make them as long as possible.Note that a breadboard end automatically shows the groove when viewed from the side. You can add a decorative spline to close the groove, use a stopped groove, or just let it all hang out.Plan to see up to 1/8" expansion and contraction per foot of tabletop width. Allow that amount on each side of the two outside mortises, and you'll have 3 or 4 times what's likely to be needed. Let's say that you decide to use a stopped (or filled) groove, and the top is 24" wide. That's 2x1/8 = 1/4" expected movement. Then you'd allow 1/4" of clearance to each side of the two end tenons. You're probably on the low end of water content, assuming that you've properly dried the lumber. In that case, the top is likely to expand if anything, but not shrink noticeably. Now you have choices to make as to how much you want the breadboard end to overhang past the edges of the top. To keep it from looking like an error, you might want to deliberately have 1/8" or so hanging out at assembly - e.g., end board length 24 + 1/8 + 1/8 = 24-1/4". You decide what you really want.If hiding the joint, allow at least 1/4" for the mortise "end cap" - the portion of the original board left untouched by mortising. Allow another 1/4" for clearance. Then the outside tongue should extend sideways no closer than 5/8" from the edge of the top - 1/8 o'hang + 1/4 fill + 1/4 for meovement. The stub tenon does not go past the end tenon. There is no mortise exposed.If the top were to expand to the full potential of 1/4" at each edge (unlikely), then the breadboard end would now be 1/8" inside the top's edge, but there would still be no mortise or tenon showing. Pay attention to how all of your final dimensions add up, and prefinish any normally hidden spots that may occasionally be exposed.You can band, carve, or otherwise detail the ends of the end board, so that when the ends change position relative to the edge of the top, it still looks as though it's supposed to. Sensible? [ Plain stick looks like it's too long, then top expands and it looks too short. If the ends are carved or detailed to resemble a cap or band perhaps an inch from each end, then it looks as though the cap is added and the long/short appearance seems to be more deliberate. ]Aesthetics and weight of top:
Hmmm - No harm in the heavy stuff. You could probably fake the thickness where it's visible, and use less expensive and thinner stock for the bulk. How much more visual height/depth will the apron(?) and legs add? Experience says to draw it to scale and massage for most pleasing effect long before making sawdust! I think you probably want "WOW !", not "CLUNK !" No criticism intended - you may already have the aesthetics completely in hand.Regards,---John
=======================================================
"What you leave behind is not what is engraved in stone
monuments, but what is woven into the lives of others."
---Pericles
yeah, i got the aesthetics figured...i'm one of those artsy fartsy guys."Allow that amount on each side of the two outside mortises, and you'll have 3 or 4 times what's likely to be needed."that's what does it, i was thinking that the mortises without room would split out the endboard as the endboard's grain is perpendicular to the field.ding!...that's the bell going off in my noggin'
thanks so much.john
Movement indeed will be a problem with long miters, and splines won't help. The miters will open up during drier weather, just like mitered casing does around the house (and why you want to cope crown, not just miter it).
I don't have my shrink-o-meter cheat card handy, but for a shrink factor of 0.003, for a 5% change in wood moisture over 8 inches, we're talking 1/8." You'll probably experience less, but even 1/16" opening will be very obvious in a table top of this design.
I hope I'm wrong, but...
Pondfish
Thanks for the info. As part of the design(and as a hedge against the movement) I'm actually leaving about a 1/16" reveal at the four miters. I'm hoping this will guard against the problem you mentioned and maintain an even "reveal" (albeit changing) through the seasons. Time will tell!
Chris
Hi ...Please trust that I mean no derision nor other unpleasantry in this response. Truth is, ya almost got it right.Yes, a solid hardwood panel may expand in the range of 1/8" to 1/4" per foot [[ across its width ]] per 5% humidity change. That's the key. Lengthwise movement is virtually negligible.Mitered cases of a single material will not have a problem - both sides of both elements will shrink cross-grain exactly the same amount, and there will be no stress imposed on the joint. Similarly, if opposing sides are the same length (as they must be for miters), then there will be no differential expansion/contraction in length. If that weren't true, there wouldn't be an intact picture frame in the known universe ...Doesn't have to be 90 degree miters. Any regular polygonal shape with appropriately mitered joints will be similarly stable.The key lies in DIFFERENTIAL movement - if it all moves together, then it could go to Hong Kong and back without showing a mark. If it moves differentially without wiggle room, then it's toothpicks.There can exist some problems with case construction using 90 degree elements joined with tenons. If a tenon exceeds roughly 3" in width (or length ... !), there's likely to be trouble somewhere in the future. That's because the joined pieces are at 90 degrees, and do not move in the same directions - a differential. The wood can by nature absorb some of the differential movement, but there's a limit - er - about 3 inches of tenon width and/or length.Same for a captured panel - without expansion room, the panel will expand across its width (your rough 0.003 rule) and crack the frame. Christopher will most definitely need to leave room in his frame/panel design for the panel's cross-grain movement. Similarly, planning will be needed with leg/apron attachment to keep the thing from self-destruction. Typical solid tabletop attachment methods will suffice, but even that may be unnecessary (if there's a mitered apron attached to the mitered top frame, then there can be no differential expansion.)For a 45 (or other) degree miter, all movement is identically mirrored in the two halves of the joint, and there is no differential tension. If the joint employed dissimilar materials, there might be a problem because the two would move at different rates, regardless of dimensional match.
------------------------
Home decorative trim is a completely different set of circumstances with differing rules and consequences. Dwelling walls are a hodgepodge of materials assembled with myriad movement directions and characteristics, and subject to both temperature and humidity gradients. Wallboard doesn't move so much but casing does, yielding immediate differential and joint movement. And so on ... but ...I don't cope house trim joints (I admit that's living life the hard way) and never have 'em open up. If lengthwise differential movement were really the culprit, then coped joints would open also. It's because framing ain't cabinetmaking. Coping just makes it a bit easier to make good-looking joints when walls and ceilings aren't perfectly plumb/level/square ... or trustworthy, loyal, brave, clean and reverent.If a new house is constructed of poorly dried wood, then joints will open and nailheads will pop and cracks will appear and some of the sky will fall - most of which IS due to shrinkage upon loss of moisture. However, that huge differential won't return - it was just because the framing lumber wasn't properly kiln-dried. I *do* allow purchased trim to dry for a week or two in-locus before cobbling it up. Fresh-from-the-mill trim may need months to come to equilibrium.Regards,---John=======================================================
"What you leave behind is not what is engraved in stone
monuments, but what is woven into the lives of others."
---Pericles
Your geometry doesn't seem right to me. Bear with me while I use a little trig (and finish carpentry experience) to show why the miters will open when the seasons change.
If you consider just one part of the miter and examine the angles and lengths: when first cut at 45 degrees, you have a miter that is w wide (say, 8") and whose length in the grain direction is also w; the length along the miter cut is 1.414*w.
When the board shrinks, the shrinkage, as you point out, is tangential to the grain, and negligible along the grain. This means our piece of wood will now be w-1/8" wide, but still w long. This will no longer form a 45 degree angle; it will be smaller.
Repeat this for all joints, and you get miters opening on the heel side of the cut. If the board were to expand wider than w, your miter angle will now be greater than 45 degrees.
This happens to every painting I have in my home where the frame is wider than 3", as well as to wide casings I have in the home. And wood movement is the reason that good crown molding installs are coped: it does a better job of hiding the seasonal wood movement. It has nothing to do with cheap materials or construction techniques. The problem is that when your miter gets wider/longer than a few inches, either the heel will open when the wood dries, or the toe will open when the wood expands.
If this isn't clearer, let me know and I'll draw a picture instead of posting another 1000 words...Recommending the use of "Hide Signatures" option under "My Preferences" since 2005
Good explanation.
Hi, Paul ..Good comments and examples. We essentially agree, but there's more to examine. This is not an argument - please be patient and read on.
-------------------------
That's a major concern for outdoor installations. Indoors isn't generally as great a problem unless the wood wasn't sufficiently dry to begin with - a common problem with lumber supplier materials. (My countertop experience below.) That's not "inferior material" - it just isn't dried to furniture-grade specs, and isn't suitable as-bought for furniture construction or joinery.
Here's a related reference - there are others:
http://www.garymkatz.com/TrimTechniques/OpenMiters.htmYour math is correct, but there are conditional variations. In practice, a reinforced joint will withstand considerably more stress than a simple butt joint. Interior joints are likely to see less moisture-induced stress than exterior. Quartersawn lumber moves half as much. Is he safe? Maybe. There are plenty of examples of its working, just as there are plenty of failure.I've numerous heavy picture frames in my home in the range of net 4-6+" wide. All have resided in various U.S. states over 35 years and none have exhibited any difficulty whatsoever. Witness as well the museum pieces, some centuries old, with frames a foot and more in width. I agree that if they get sopping wet, a problem will undoubtedly show up. On the other hand, if my furniture got that wet, I'd expect to have to replace it all anyway.There's no doubt that you've installed MUCH more than my several hundred linear feet of crown. However, I've installed crown to about 5" wide with never a joint problem. Door casings to 6" nominal without problems, though I always reinforce such joints.I've no argument whatsoever with using the cope method to ensure that construction doesn't resolve itself into callbacks. However, I can't agree that it's mechanically superior - it's just safer, in my opinion and experience. I also think that installing reinforced joints for every stick of residential trimwork would be silly, overkill, and economically ruinous....same as framing isn't rationally cabinetwork. That's not compromise - it's just sensible differences in methods for differing purposes and economic realities.And who's gonna store trim for a year to get it halfway dry? Trim ain't furniture. Oudoors, you can make yourself nuts with it or redesign using smaller cross-sections - no amount of protest is going to change the way wood works, and we're in perfect agreement there. I would never install an extra-wide tight miter in a fancy deck surface, for instance.For a quartersawn board, the shrinkage factor diminishes by 1/2 - another reason that furniture joints may be less subject to opening. And ... there's a warning for Chris and one which should have occurred to me earlier ... if the oak is plainsawn, it may be unwise or at least marginal to miter such wide joints.Similar to pine, his oak will shrink about 4-5% radially - that is, if it's quartersawn. Over an 8" width, that's about 1/32nd diff from dead-green to about 7%. The resulting stress from 1/2 that moisture variation because of atmospheric changes will be easily accommodated by a reinforced joint. By way of similar example, I've made reinforced oak miters up to 7+" with no problem. BUT the self-described "cabinetmaker" who installed a nominal 8" mitered oak divider top in our new home's kitchen just went'n got 'im a hunka flatsawn from the local yard, then milled, finished, and installed it butt-jointed. Too green. Not reinforced. Yup - heel opened, and now I have it to fix.As an engineer with over 40 years experience, and with 50 years considerable (not professional) experience in framing, trim, and cabinetry, I am quite confident that the experience - both good and aw shucks - is valid, though most certainly not all-encompassing. I've considerable confidence that the miter would work if well-managed.I genuinely appreciate your having defended yourself, as this has been (IMHO) a fruitful and informative exchange. Thanks for making me think.Regards,---John
=======================================================
"What you leave behind is not what is engraved in stone
monuments, but what is woven into the lives of others."
---Pericles
Pond,
You have it exactly right. It is well-nigh impossible to restrain the movement of 8" wide boards across a miter, especially in a wood that moves as much as oak. As shrinkage takes place, it effectively changes the miter angle; the length of the board at the toe and heel of the miter remains the same, but the distance between the toe and heel gets smaller, making a more acute angle. Putting four of these joints together in a frame just makes things worse. A single joint might remain intact, by pulling the two "legs" extending from the miter, closer together.
With narrower stock, this effect is less, plus the narrower stock may be able to spring a bit in the center of its length to accomodate the change in angles, without compromising the joint.
Regards,
Ray Pine
Thanks Ray for pointing out that an isolated miter might hold up better; I'd neglected that point. Miters separated by about 29-31" (like for a door casing) might manage to stay together if the case allows for the sides to bow in or out. But I think we agree that an 8" miter in an enclosed piece is highly unlikely to hold up in most households during the change of seasons. Maybe in the controlled, constant temp and humidity of an art museum you will not see problems. But in the real world, ya sure will.
I guess it's too late now, but I would recommend that the original poster rethink their idea about how to build the table. If end grain is definitely out of the picture, use a frame & panel with a frame less than 3" wide. I tend to like end grain on QS white oak, so I would go the breadboard route, but it isn't my project or tastes.Recommending the use of "Hide Signatures" option under "My Preferences" since 2005
This forum post is now archived. Commenting has been disabled