I’m building a workbench top out of 3/4″ x 3″ wide rift and quarter sawn maple glued up face to face to get a 3″ thick top. The top is 34″ wide. I’ve installed 3″ x 3″ “breadboard” ends the width of the top which are attached with massive sliding dovetails. I want to attach the end boards to the ends of the glued up top with bolts in the middle and each end. Am I safe to ignore the effect of expansion and contraction on this joint?
Discussion Forum
Get It All!
UNLIMITED Membership is like taking a master class in woodworking for less than $10 a month.
Start Your Free TrialCategories
Discussion Forum
Digital Plans Library
Member exclusive! – Plans for everyone – from beginners to experts – right at your fingertips.
Highlights
-
Shape Your Skills
when you sign up for our emails
This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply. -
Shop Talk Live Podcast -
Our favorite articles and videos -
E-Learning Courses from Fine Woodworking
-
-
Replies
If you have done as you described, you have missed the goal a little.
Most wood moves twice as much tangentially as it does radially, which is flat or plain sawn and quarter-sawn respectively.
What you should have done is start with flat-sawn lumber which will be quarter sawn relative to its edge, so when you glue up this big stack of boards, the long face of the top will end up being QS and thus have 1/2 of the movement across the 34". It sounds like you did opposite of what you should have.
Now as for bolts through the dovetail breadboard ends. I would not use more than a couple in tight holes over maybe 3 - 4" off of the center-line. If you feel that there needs to be some out near the ends of the BB, then you should have long slot like holes through the BB end parts which will allow the middle to move back and forth.
Without knowing what the MC of the wood started out at, I can not tell you how much to expect.
My guess is that you have forty something glue-lines, and if you used a water-based glue, even if you lumber was kiln-dried, then your top will likely shrink some as the moisture from your glue eventually escapes. I would suggest that you have about 3/4" of slot toward the inside of any bolts located near the front and back edges.
< Am I safe to ignore the effect of expansion and contraction on this joint?> If you choose to ignore the advice that you ask for, then I would like to ask. How will you feel about a 1/4" splits between the bolts down the length of the top?
Good Luck, K
I order for my top to expand/contract in the 34" direction, the individual boards have to expand/contract. Since there would be very little additional shrinkage occuring in the thickness of the individual boards and no shrinkage in the length of the bread boards, I was hopeful that most of the movement would be gone. The original moisture content before glue up was about 9%, the entire assembly has been in an air conditioned shop now for 6 months and the humidity in the shop runs about 37% at 75 degrees, about ideal. The moisture content of the wood now is about 7%.
I suppose that if you ran your air conditioning 24/7/365, and it never failed, then you won't have to worry about expansion and contraction. Of course if you moved, the bench would have to stay behind in the old shop, since it would self destruct either in the moving van or in your new shop if it weren't so precisely climate controlled.
Your bench, and any other furniture you make, if they are well made, could easily be around a century from now, though of course they probably won't still be in your shop, so you have to design for all possibilities, not just the ideal climate you have today in your shop.
I always design furniture to handle a wide relative humidity range of 15% to 80%, which is what the actual range is here in New England, and is just about as bad as it gets. In that humidity range, the moisture content of the wood in your bench top would change about 12% between summer and winter. Running the numbers for maple, this means that your top will be almost 1 1/2 inches wider in a moist location than a dry one. If you don't plan for this, your top will someday, somewhere, have some serious problems.
John W.
I understand about moving the bench. Where are you getting the formula for the 1 1/2". Is that is a table somewhere on the different species of wood?
I use the "Wood Movement Reference Guide" published by Lee Valley. It comes with an excellent small booklet that covers most of the basics of wood movement. The guide is cheap around $6.00. The information included with the calculator has one error in a section marked "Coping with wood movement", they state that wood covered with urethane, paint, or varnish won't move nearly as much, which isn't true.
Other sources for wood movement information are Bruce Hoadley's "Understanding Wood", a must read in my opinion, a USDA book titled "The Wood Handbook" that's available from a number of sources, and there is a wood movement calculator on the web, but I can't recall the address.
John W.
Edited 8/30/2005 2:43 pm ET by JohnWW
The wood handbook is to be found here: http://www.fpl.fs.fed.us/documnts/fplgtr/fplgtr113/fplgtr113.htmIt is free, but Hoadley is easier to read. Both are full of useful info.Chapter 3 has the info on moisture and wood, with table 3-5 listing properties of various species.
Hi John,
You and I have argued about the expansion of wood before, and I think you might be reading the table wrong. A 34" wide top will not expand an 1 1/2", no matter what you do to it, short of perhaps boiling it for a week. I'd say if the maple was dry to begin with, as it apparently was, then an extreme expansion,contraction would be in the neighborhood of 1/2". I know a flat sawn 12" panel can move aroud 1/4" seasonally. Of course this movement can be destructive and should be planned for, but if we had to compensate at the rate you list, houses would not stay on their foundations, a 7 foot tall highboy would gain 5 inches periodically, and not a single old wood door would work.
Steve
edit to add this link: http://www.woodbin.com/calcs/shrinkulator.htm
Edited 8/31/2005 12:02 am ET by dirtstirrer
Steve,
According to the USDA "Wood Handbook", a 1 inch wide flatsawn sample of hard maple expands .0035" with a 1% change in moisture content. If the moisture content changes 12% and the panel is 34" wide the equation becomes:
.0035 x 12 x 34 = 1.43"
This is pretty much a worse case scenario but I have seen it happen and design furniture for to cover such situations, I have to in New England.
As to houses and highboys, most furniture isn't made of continuous slabs of wood, panel construction divides the wood movement into smaller segments where the movement is handled by the grooves in the frame, so the overall piece isn't affected. By the way, a large number of the classic, museum quality pieces of furniture, have cracked quite badly because the makers of a century or more ago hadn't planned on the drying effects of modern heating systems.
House framing is set up so that for the most part the expansion and contraction of the frame doesn't add up to large movements of the walls, and sheetrock and plywood don't expand and contract at all.
However, I have seen a too tightly laid, very dry wood flooring, laid without a gap around the perimeter, push out a brick wall a couple of inches when the humidity rose to steamy levels the following summer. If memory serves, Bruce Hoadley, in "Understanding Wood" runs the calculation for how much a house will shrink in height as it dries out, mostly the floor joists, and it is several inches.
John W.
John, I do not want to cast any doubt on the math, but I think you may have greater expectations as to how fast thick wood can take up that much moisture under normal conditions. I am also sure that I would fall short of ever convincing you that there may be a little misunderstanding on your part, since I do not have a Ph.D. degree in plant physiology or other closely related fields, so what I would like to suggest would be for you to copy and past your comments to the sawing and drying section of the woodweb forum to see how Dr Wengert responds to it.Here is a link if you are interested.
http://www.woodweb.com/cgi-bin/forums/sawdry.plI am not saying you are wrong, but he will have information that will tell what volume of water and RH over X period of time etc. to change that much. I am looking forward to seeing what we can learn from his remarks. Cheers K
Keith,
I did as you suggested and sent a rather lengthy e-mail to Dr. Wengert asking his opinion, hopefully he will have a chance to repond soon.
While I was on the Woodweb site, I used their calculator and it came up with the same answer I posted, in an environment with wide humidity swings, the top would move about 1.5".
John W.
John, I went over expecting to see your post, but I did not see it there. Now I see that you emailed your question. If he responds to your email, will you post it along with your question here for all to read?
Keith,
I did send the question as an e-mail directly to Dr. Wengert, thought we were more likely to get a prompt response if I sent the question to him directly Of course I'll post his reply, even if he says I'm wrong.
John W.
Edited 8/31/2005 5:53 pm ET by JohnWW
He has not answered the last two times that I emailed him. I think you may have more luck posting in the forum.
I am very keen to see what Doc Wengert has to say about this ...11/2 inches on 34inches-that's about 4,41%...while John's application of the theory and his arithmetic are correct I am having trouble seeing how this will happen in practice, unless the timber is submerged in warm water for a long time. This is not likely to happen in a gentlemans workshop-Katrina aside, and then there will be other more important things to worry about. I think Keith is barking up the right tree-the reaction of timber to moisture changes is not immediate and the rate is affected by various other factors.
Going back to that 4,41% : if this gets out some folk are going to make a lot of money just buying and selling timber at the right time.
Sorry , unable to take things seriously right now-if folk think I am being silly I will keep quiet.Philip Marcou
Philip,
No the response isn't immediate, but after the heat has been on for three or four months even a thick piece of wood in a home in the north country will be down around 4 or 5% moisture content. By the end of August the moisture content will be back up around 13 to 14%. I've been watching and measuring this for 25 years, it really does happen.
John W.
Please forgive my math error in my post about floors; it was corrected by another poster. But I still stand by my analogy.
If your example is correct, that a 34" wide top can expand to almost 1 1/2, then what about quartersawn or flatsawn oak floors?
A narrow 10' + wide oak floor(or whatever the specie in question is) would expand by 15"!!
This is preposterous.Standard hardwood floor construction for rooms even larger than this is to allow 1/2" side to side for expansion, that's all.
Yes, there is only 1/2" left at the edge of a hardwood floor, but what about between the individual tongue and groove boards? They are nailed to the joists and subfloor (which expand and contract very little along the grain), but can individually expand and contract since they are not glued edge to edge. Gluing floorboards would be a very bad idea, for the reason you indicate. We may value wide boards today, but why do you think narrow floorboards were used for higher quality work in the past? Presumably because the narrow floorboards open smaller, less noticeable gaps between them from expansion and contraction.You can wait for your expert of choice to chime in, or you can simply look at any or all of the sources listed above. It may seem surprising, but wood does move. And you'll also notice (from any of the sources) that one of the benefits of quarter sawn wood is that it expands roughly half as much in width as plainsawn. Ancient Egyptians knew about wood expansion. Here's a quote from a website ( http://nefertiti.iwebland.com/timelines/topics/stonework.htm#rem1 ) describing quarrying of granite near Aswan:
"Holes were cut into the rock, wooden wedges driven into the slots and moistened. The expanding wood cracked the rock."
I don't understand your post; you've added nothing new here. I already understand and appreciate wood movement. I've been a professional woodworker for many years.
If a 34" wide top expands 1.5 inches then a 10 foot floor (120 inches) could expand 120/34 * 1.5 = 5.3 inches.
But I would not lay a floor if the wood were at 4% knowing that it might go to 16% (12% difference). I would do it mid year when the wood was around 10%. Then it would expand or contract only 2.6 inches total from its "neutral" position. If you allow for half of that on each side of the floor it's about 1.3 inches. That's theory. Since the floor has a finish the maximum or minimum humidity will never be reached except under highly unusual conditions (it's likely that the humidity will increase or decrease well before the floor has fully acclimated to the higher or lower humidity). And, as a previous poster mentioned due to the tongue and grooves you could also allow for less than 1.3 inches and be okay.
you're right, I made a mistake. 5.3 " is more like it.
As you say, that's theory. Because flooring installers don't pick and choose which months they'll work in; they need to eat. Which is why allowing 1/2" on each side is more than enough to allow for expansion and contraction. It's when this practice is not followed that floors buckle and problems arise.
Just in case anyone cares, the bench top has shrink about 1/16" in about 2 years. It has been in the 35 to 40 percent humidity shop the entire time.
Hey Chester, Thanks for remembering. Glad to hear that the movement stayed so small. As long as it stays in a controlled atmosphere, obviously it is stable. What I always try to do, and that John WW stated, it to understand and that out work may not always live it's life in a perfect atmosphere. As I looked back over this thread, I would like to state that I did not have any argument with the math that he was using, but had doubt about the speed that the wood especially relative to density, and thickness. I hope you have been pounding out some treasures on that monolithic top.
So, if we do the math, the moisture content of the top dropped about 1/2% over the two years. I wonder whether that was the glue giving up its moisture. Now, you have to leave the bench outside all summer in Georgia, say, so we can see what happens during prolonged high humidity.
The great thing in this thread is that everybody can walk away saying "I knew it." The interesting post was from Dr. Wengert through John White (thank you John), as he introduced (to me at least) applying the technical idea of hysteresis in absorption, and also that few boards are perfectly flatsawn. Indirectly, he also mentioned surface to volume considerations. Many on this board have heard the notion of air drying boards one year per inch of thickness. Apply all these thoughts to the 3" benchtop, and the hysteresis effect is different than it would be on a dining table where the top is 3/4 solid wood. Hence, making fine furniture will show errors in moisture judgement sooner than making a fine workbench with the same judgement errors.
Thank you for posting a follow-up.
BTW, I didn't post an original prediction, but given all the info in the thread(especially post #40), my prediction would have been that the benchtop would have about 0.03" shrinkage per year the first few years ;)
I guess in the name of science...... I could put it outside for a while...... nawwww. Thanks for the interest.
I received a response from Dr. Wengert, it is posted below in full:
Hi John,
You are close. The RH in a home in the wintertime is probably around
20% RH on the average on very cold days. However, this extreme exists
for a short time, so the wood would not respond to this low value.
Therefore, I suspect that the lowest MC we would see in the wintertime
in a heated location without humidification would be 5% MC. In the
summertime outside, the RH averages about 65% RH or 12% MC. In a home
or shop, the RH average would be lower...perhaps 50% RH (9% MC), or in
a few cases, 56% RH (10% MC). Again there might be a few extremes,
but they are short in duration and not important. So, I would give an
opinion that the total MC change is 5% to 9%, or 4% MC with a few
cases being 5% MC change.
The flatsawn value is for 100% perfectly flatsawn. As we likely have
not quite perfectly flatsawn, you can moderate the value somewhat...I
would suggest 0.003. So, 0.003 x 4 x 34 = .41".
Now, because of the size, the piece will actually swell and shrink
about two-thirds to one-half of the calculated value. The Handbook
values are for thin shavings. Further, they do not factor in that a
piece that is losing MC and then gains MC will have a lag (hysteresis
effect). So, although the change is 4% MC, the actual will be 3% MC.
And there is a lag in changing from shrinking to swelling too. So, I
would easily drop the value calculated by 50%.
OK?
Eugene Wengert, Emeritus Professor and Extension Specialist of Wood
Processing
President, The Wood Doctor's Rx, LLC
-----------------------------------------------
I would have to disagree, from actual experience, with his assertion that the extremes of humidity are brief and therefore can be discounted, where I live the bone dry winter weather goes on steadily for months and even large planks fully dry out by March and then fully recover their moisture by late August in an average summer.
Based on his advice, combined with my own experience, I may in the future figure on an 8% moisture range instead of 12% as a worst case scenario, but I'll run some experiments with sample blocks starting now while the weather is still humid and running through the driest part of the winter.
John W.
Edited 9/1/2005 5:44 pm ET by JohnWW
John, thanks for keeping us posted. I hope we all have learned something here.
Maybe Chester can keep a monthly record and keep us abreast of what actually happens with this piece. This subject comes up a lot here and the answers are all over the map. I for one would like to know if I am giving good advice or not.If Chester will do that we can all add his information to our knowledge base and be wiser for it. I would like to be wise someday. K
This was only about the second time I posted a question on here..... I didn't know folks would get so involved. I will pledge to measure the top on one month intervals. The problem is that my shop is heated with a gas furnace and cooled with a heat pump. The RH in the shop this summer has been around 37% at 75degrees. Don't know about the winter yet because I just got set up in the spring. So.... since I don't intend to open the doors to check the hypothosis, there probably won't be much movement.
Chester and others,
Next time you join up wide tops/panels doing what I do may be useful for those who like to observe timber movement:-
It's usual to to have extra to spare on the length (one reason being so that you can glue back the edge grain to the ends if you want the top to look thicker as in a bar top).
Another reason is that you can make a movement recorder with the end grain offcuts by taking two lengths of this and attaching one end only of each to a stable board so that each can move in opposing direction. Just make sure they are free to slide and cannot curl upward. If you start by having them level with each other when they shrink the gap increases and they slide past each other as they expand. Whenever I use an unfamiliar timber I try to have one of these-gives a good indication of stability and how quickly the timber responds . I just record a date and the size of the gap when I see a movement.
Sorry for the long winded explanation-but I'm sure folks can get the picture and improve on it.Philip Marcou
John and all,
The diagnosis from Doc Wenge was very interesting-thanks John . The word "hysteresis" as applied to timber has now been added to my vocabulary. I am pondering on all those factors that affect timber hysteresis, and how they would be relevant to Chester's Maple top.
So to revert to the original question of how best to attach his bread board ends: I don't see the need for a sliding dovetail there, especially as I would be more inclined to do a large tongue and groove, making the bb slightly dished so that by pulling up the one bolt only that I would put in the center the ends would pull tight, and the whole lot can move as much as it wants (as long as it is not more than about 5/8 inch in my books!)
On a lighter note, if Chester is overwhelmed by the need to have the dove tailed look he can cheat by planting false tails at the ends- not forgetting ofcourse to leave a gap of (?!) between them and the end of the t and g. When there is movement people will be wandering why the dove tails are remaining flush with the b/b ends....Philip Marcou
John,
Check your math John, you figured the movement for a 34 foot wide area. It should be:
.0035 (per inch) x 34 inches=.119"/% increse in moisture content
You multiplied by 12 but everything is in inches. If you multiplied by 12 assuming a 12% increase in moisture content, I challange you to get a moisture meter and do some tests. You will not gain 12% moisture from dry stock, unless your furniture is in New Orleans.
Steve
Steve,
The formula is: chart movement value (.0035" for hard maple), times panel width (34 inches for this example) times moisture content fluctuation, (12%): .0035 x 34 x 12 =1.43".
The 12% change in moisture content is perfectly reasonable over the course of a year in New England, or would occur in a piece built in new Mexico that later wound up in Florida, something of a worse case scenario, but not completely unreasonable. The last time I noticed, furniture I had built, mostly in Vermont, is in at least a dozen states outside of New England.
The Lee Valley chart says that a piece of furniture in Boston could experience a 10% change in moisture content over the course of a year and it would be worse inland where the winters are incredibly dry when the air temperatures go well below zero.
Also, it doesn't take more than very minor changes in a design to make it handle a large moisture swing rather than a narrower range, and the conservative approach allows me to sleep better at night.
John W.
Edited 8/31/2005 1:34 pm ET by JohnWW
Edited 8/31/2005 2:18 pm ET by JohnWW
John,
While I usually agree with your posts on many topics, in this one, the 12% moisture change should expressed not as '12' but as .12 or 1.12
I believe this is where the error lies.
I'm actually looking at a textbook example of how this equation is used, the moisture content, in the book example, is entered into the equation as a simple number, a 12% moisture variation is entered into the equation as the number 12, as I wrote it.
My numbers, I am quite sure, are correct, and the calculator on the Woodweb site gives the same result.
John W.
John,
Here is the same calculation made paralleling the
example given on pg. 77 of Bruce Hoadley's book "Understanding
Wood" change in D = (34")(0.099)(0.12)/(0.28) = 1.44 inchthe 0.12 is the 12% change in mosture content, and the
0.28 is the fiber saturation point (avg. value 28%). Hoadley
states this a formula for shringage from the green condition,
and its use from partially dry lumber results in a 5% error.
When using his revised formula for partially dry wood, from
a mosture content of 16% to 4% the formula gives a result
of 1.50 in. My workbench is 18 inches wide, finished on all surfaces,
never seems to move more than 1/8 inch. fall to spring. In the
winter the shop is kept at least to 50 deg with outside air
mostly near 20 deg. One would expect more movement, since
summers are fairly humid. Hoadly has maps for average moisture
content for interior wood in January and July on page 86.
In Madison, WI the change is from 6.25% to 10%. The revised formula suggests change of 0.11 inch. Seems right to me. Can we ever see a 12% change in mosture content of interior
wood? From his map, if a piece went from San Diego
one July to Phoenix, the next July there would likely be a 9%
change in mosture content.
John,
The designing of furniture to withstand the movement is what makes this so interesting. Did you look into that link I posted? It give the radial movement for a 10% change in moisture content (which I think is high) for maple at 1/2", for 32" wide. One of our two numbers is wrong.
Steve
Steve,
The radial shrinkage in hard maple is slightly less than half as much as the same wood flatsawn, .0017" for radial, .0035" for flatsawn.
Because of the way the benchtop is glued up, with the quartersawn set on edge, the end result is actually a flatsawn panel 34 inches wide, this was explained in someone else's posting, the second or third message on this topic.
Using the correct higher expansion figure for flatsawn would double your result, to around an inch. Add the additional movement for a 34" wide panel, versus your 32", and a 12% moisture swing compared to your 10% and you'll get my result.
John W.
Edited 8/31/2005 3:08 pm ET by JohnWW
What you should have done is not glue ANYTHING! My 'GOOD' bench..Just 'sticks' in a Horizontal pile.. NO GLUE!.. About every 6 inches I drilled a hole for a threaded rod (3/4 inch) to go through.. If somethin' happens.. I loosen the bolts and hammer it flat! And tighten the bolts up again!EDIT:: NOT really.. Just a thought I had!
Edited 8/30/2005 12:15 pm ET by WillGeorge
Chester1,
Some suggest you glue the first few inches on the breadboard end where it meets the front apron. That way as it expands and contracts the breadboard stays flush with the front to support the vise...if that's relavent.
You don't say where you live but using an annual season average range of relative humidity of 40% to 70%, a 34" hard maple top will expand/contract 1/3" if quartersawn or almost 3/4" if flatsawn You can hardly ignore that.
You can determine the expansion/contraction of almost any wood by going to http://www.woodbin.com and clicking on the "Shrinkulator". Input the species, range of RH or moisture content, width and the output is the expansion/contraction.
I'm in TN. I guess I just hate putting bolts in my maple top, seems almost sacraligious. I got on the USDA government site and it calculated 1/8 to 1/4". I think I will glue the ends on the front of the top for about the first 6" and then slot a hole in the middle and opposite end. Any ideas to avoid the bolts?
Screw from underneath and pre-drill elongated ovals to allow for expansion. Either that or use old fashioned figure -eight fasteners.
This forum post is now archived. Commenting has been disabled