tolerances in frame construction

[hayton1960]

I am keen to know how accurate a frame should be.
Realistically, what tolerances are to be expected in the joints in a timber frame. I've seen all sorts, from chippendale-type cigarette paper tight to, shall we say, "loose" (1/4" to 1/2" gaps here and there, and that on new repair work). I know you get shrinkage across the grain, wear and tear from weather expansion and contraction etc; but when the joints are cut and fit up, as a guide, what degree of acuracy should you be aiming to work to?
Your input and comments would be very welcome
Cheers Jonathan ;)
PS apologies if this was covered already; I tried word search but references to tolerances were mostly to do with sawmills _grin

[Jim_Rogers]

When you cut a tenon to length it should be shorter than the depth of the mortise. So that the tenon won't bottom out in the mortise, when the timbers shrink.
Tenon sides and edges should fit precisely to the side of the mortise, no gap at all.
We use a framing square as a mortise checker (um, do I have a photo of this on file? Maybe, I'll have to look. and I'll post it later if I do.)
To size the tenon we use some outside calipers set to the outside of the framing square. I know I've posted a picture of setting the calipers to the square, it should be in the story about cutting a tenon.
This insures that both the mortise and tenon are cut the the same dimension, whether it be 2" or 1 1/2".
To make sure we are right on, we measure are layouts and draw them with pencil. Then we knife the lines with a straight edge and measure to the next joint from the knife line to the tape measurement. This insures that the width of the pencil line isn't added or subtracted from the overall lengths.

You should see in the story "Tools for Timber framing" a timber framer named Jack Hayden using a knife to score a line and profile of a tenon.

At a compound joinery drawing workshop the instructor suggested that we figure our timber lengths to four places beyond the decimal point. Someone asked why? He said if you're making a 40' long building and your measurements were off 1/1000 over 40' (480" / 1000 = .480") That means in 40 feet your stick is either short or too long by 1/2".  What that means is your last bent is going to stand proud of your foundation location, or be 1/2" too short. It can add up to an unacceptable error.

Yes, a timber will shrink but not usually much in length.

Joint layout has to be accurate.

Mortise housings can be scooped out to compensate for expected shrinkage.

Good question......

Jim Rogers

[hayton1960]

At a compound joinery drawing workshop the instructor suggested that we figure our timber lengths to four places beyond the decimal point. Someone asked why? He said if you're making a 40' long building and your measurements were off 1/1000 over 40' (480" / 1000 = .480") That means in 40 feet your stick is either short or too long by 1/2". What that means is your last bent is going to stand proud of your foundation location, or be 1/2" too short. It can add up to an unacceptable error.

When you describe it like that Jim, it shows how crucial is acurate measuring and maths etc! I was trained (regular joinery) to transfer sizes rather than measure and replicate them, where possible (pinch rods, scribing (coping) etc). I know that if I have a full scale drawn rod and keep referring to that & take my sizes off it its harder to make a mistake. I reckon thats one bonus of drawing the frame full size first like in France ("scribe rule"?). In that case the drawing is the dimensions, not a representation of them?

I expect its embarassing though to have the crane lift your last bent in place and the whole crowd of folk see it sticking in or out.  :-[ Or find a mortise in the wrong location :-[:-[ Is there an acceptable margin of error in creating a frame, or does a master framer get it perfect every time  :D ??

Thanks for your advice Jim ;D

[Jim_Rogers]

Over here I think what you call a rod we call a story stick. This stick (or rod) would have all the locations for all the joints or other measurements on it for, lets say, a post.

We do use story sticks and this is an acceptable method of laying out a timber as long as it is accurate and you double check your layout on your post with a tape after you have laid it out and that you make sure you're not accumulating an error from post to post, or layout to layout.

There is lots of information here and you'll just have to read and see if it will help you.

Jim Rogers

[Jim Haslip]

A technique which I use is to begin measuring from one end of the timber  or frame only. Kind of like the Storey pole, the various Layout lines are not 8 " from the last layout line, but 48 " from the end of the timber or 48 ft from the outside reference face at the end of the building. I also layout Raftes and Joists on stick frames the same way. Pick an end and start there. Too much room for error to measure 24 " centres along the way if you measure each one at 24 " from the last one. Start at zero and measure them all from a single point. By the time you are at the end, you might be out a pencil width only... Good enough for this guy.

[hayton1960]

Hi Jim R
Thanks for your reply
Your right, here a story pole is a rod, but a rod can also be a full scale plan that you take you dimensions and positions off of. Its one thing I reckon doing parts for a garden gate that way, but an 18 foot 12"x8" as part of a bent??? Must be the case that the less you have to handle and move them big timbers around, the better?
I'm studying the back issues in the archives. Its like a living book :P

Jim H
Thanks for sharing that tip. I like that technique. Great common sense. The best acumulated error avoiding ones usually are! I hope you dont mind me borrowing that ;)

Is there any place for gang layout in framing? Like when you set out (eg) four table legs, marking the mortices positions all at the same time? Or when roofers set out and cut all the birdsmouths in 2 angled passes with the rafters strapped together and held level?
Cheers Jonathan ;D

[ARKANSAWYER]

   

  The most I have ever learned was taking down old frames.   Old frames that lasted had to have been done right to begin with to last so long.  I have never taken down a old frame that was real sloppy.   I have seen some very creative ways of putting things together.
  I like joints tap together tight.  Since I do all my layouts and cutting mostly solo I try to always do the whole frame with the same square and tape.

  One time we got a frame up and a beam would not fit into it's mortise because it was to short by 3 inches.  We braced up with 2x6's and sent the crane home as we were going to have to cut a new beam for that side.   Cut it by the specs and lifted it up in place and it too was 3 inches short.
  The metal plates the post sit in was measured and it was found that the slab was off 3 inches to long in that corner and the guy setting the post did not measure from the base corner he just went off the slab.  Talked to the owner and after some debate the house is now 3 inches longer on one side and no one notices.  :   But all the joints are tight.
  Looks are every thing and most people do not see plumb and level any way.  I would bet that in a 40 ft long run of a house no one would notice 1/2 and inch.  I have seen that much in a 12 ft bay and if not told will never notice.   A good framer lays out well and with care but, scribe rule flows together so well most folks will never notice.

[hayton1960]

Hi Arkansaswyer, thanks for your comments

By tap in tight, do you mean where if you fit the joint home, theres just enough friction for it to hold together if you lift it up and it doesnt fall apart?
Is that tapping in with your chiselling mallet, or a heavier beetle or mell?

Looks are every thing and most people do not see plumb and level any way.  I would bet that in a 40 ft long run of a house no one would notice 1/2 and inch.  I have seen that much in a 12 ft bay and if not told will never notice.   A good framer lays out well and with care but, scribe rule flows together so well most folks will never notice.

Especially a drunken man on a fast horse!!  :D
Some of the old frames you see in places like York, Suffolk, Herefordshire etc look like they 12 " out in 12 foot, never mind 1/2" in 40 foot!! I know some of them are maybe 6 or 7 hundred years old. I dont know if they were that much out when they were built, or if its down to expantion/contraction/settling/shrinkaging/effects of repairs of differing quality etc over the years. Maybe because the frames have always been exposed rather than covered with siding?? I like the look of wonky wood anyways (up to a point!) I imagine the standards were high in past days, especially on important public buildings, or is that a false assumption? Like you said Ark, the well made frames were the ones that lasted....

I agree with Jim though; I still think its always good to aim for the best perfection your capable of though, rather than accept sloppy standards to begin with. Then the box pack kitchen fitter wont throw a tantrum if the kitchens off square or plumb :
Cheers, Jonathan ;)

[hardworker]

Here's 2 cents from a new builder.
First the accuracy required depends upon the joint. The fit of a plate to post isn't as important as the fit of a braces. Unlike platform construction methods, where the sheathing acts to prevent racking, siding in a TF may not help barely at all. For example with vertical siding, you nail in the middle, and this doesn't add so much to prevent racking as sheathing would in a platform constucted house.
As an avid woodworker, I don't see how one can use the same method of ganging up members to cut them.  I do think it makes sense to follow a process. Once you have a good brace, cut out  a bunch more using the first as an example.
I built the Sobon structure. I cut and fit one complete bent. I then used this as the model and cut out the remaining pieces. Same for the rafters. Built two, and tested them and then stamped out the rest. 
Jonathan

[Tony_T]

Old scribe rule buidings had the compass settings scribed into one of the principal timbers.  The settings were used to layout the whole building and were a reference if repairs were needed later on.  As long as you used the exact reference settings on the compass you should be able to duplicate the joinery.

[ARKANSAWYER]

  By tap tight I mean that you are just unable to slide it into the joint.  You need to bump it with a mallet.  I am not talking about having to hit every thing with a 20 lb commander to get it together.  I draw bore most my joints so I strap my bent together to mark for the tenon hole.  The strap keeps the joints tight and pulled flush to each other.  Hard pounding to make a joint fit is just as bad as one that is to loose.
  But if you are out there with a micrometer and trying to get a tenon to 0.0005 of the mortise you are wasting your time.  Wood moves for ever as the seasons and time pass.  I have taken frames apart that have 1/4 inch of clearance and some where with even the pegs out you can not knock the joints apart easily. 
  If you get a chance to tour old buildings and bridges you should.  There is lots to be learned from the men of old.  I looked at two bridges this summer that had the same strange joint in them.  I took a photo but it did not come out clear.  I have been looking in books for another like it but no luck so far.

[Jim_Rogers]

Needing to pound every joint together may not be the best methods.
A fat tenon can split the mortised piece.
A tenon should slide together into a mortise without a lot of force. A mallet maybe but not usually, and not a 20 lbs commander for sure.
Correct is correct, not oversize or undersized. That's why we use a framing square body as a mortise checker, and set of calipers set to the square as a tenon checker. The steel framing square does not change from season to season, and should be used as a constant measuring tool.
Accuracy is needed at every joint, regardless of position. The whole is only as strong as the weakest link.
Everyone has their own methods, and opinions on whether or not their methods are right or wrong.
An individual should try and create the skill of using his hand tools to the best of his ability and polish that skill set as he cuts each joint. The last joint cut should be better then the first. With skill comes accuracy, with accuracy comes speed.
I too, thought that figuring your lengths to four places beyond the decimal point was too extreme but when you look at 1/2" in 40', it is my opinion that this is too much to be out.
Dealing with the out of square foundations is a common problem with all types of construction.
As timber framers we should try and set an example of a level of accuracy that other could try to achieve.

Jim Rogers

[mark davidson]

as far as tolerance goes, i've learned to like 1/8"
if i put something together and the measure is within and eigth either way, I go forwards. if i'm more than an eigth out, I fix it.
personally, i don't like joints to be overly tight, unless they are tiebeam connections.
to check tenons, i made a guage from 2 old framing squares. clamp them together so you have a 2" space between them(the space should be around 13" long), using your main framing square as a guide, then, when the blades are parallel drill and bolt the blades together. Once they are together, cut off the extra 1 1/2 blade sticking out and use these peices to help you layout stuff.
I have a 2" and 1 1/2 inch, someday will make a 2 1/2" for the 10x10's
nice to find another tf forum,
-mark in ontario.

[hayton1960]

to check tenons, i made a guage from 2 old framing squares. clamp them together so you have a 2" space between them(the space should be around 13" long), using your main framing square as a guide, then, when the blades are parallel drill and bolt the blades together. Once they are together, cut off the extra 1 1/2 blade sticking out and use these peices to help you layout stuff.
I have a 2" and 1 1/2 inch, someday will make a 2 1/2" for the 10x10's
nice to find another tf forum,
-mark in ontario.

When I read Jims suggestion of using the calipers to check tenon width I thought what an exellent idea. It reminded me of the guages that woodturners use to quickly find any given diameter while doing production runs, I even heard of one turner who uses various spanners (wrenches) for that purpose! I then thought why not get a guage made, like a tuning fork with a 2" opening; you already made it Mark! ;)
Cheers, Jonathan

[Jim_Rogers]

A tenon checker like Mark has described is useful, at times.

You have to understand that you first create a face/side/cheek of a tenon a certain distance off the layout face, for example 2". Then you roll the timber over and cut away the other side until the tenon is 2" thick.
What this means is once you've established that your first tenon side is parallel to the layout face and is truly 2" off that face it can't be trimmed again. Or you'll change the offset.
So you can only trim the other side to make it 2" thick.
Using a caliper on softwood leaves a mark or scratch where the tenon is too fat. Remove these scratches with your chisel, slick, hand plane and the tenon should be the correct thickness.

There are step by step photographs showing this procedure in another story on this forum.
I'll have to find the name and post a link here, later, if needed.

Jim Rogers

[hayton1960]

Hi Jim, Thanks for your advice
I understand what your saying about the tenon "top cheek" must be exactly 2" from reference face, I expect you use the framing square to check for flush depth. Or possibly the borman device?? Or even an exact 2" thick chalked board about 6 or 7 inches square, place it on the tenon, rub it, remove high spots & adjust till the block lies exactly level and flush with the referance face (check with a straight edge such as a spirit level)? Then do the second tenon cheek. I understand what your saying about checking the thickness of the tenon so it is exactly 2" thick. But if you use the caliper and rely on the scratch, do you keep adjusting with your slick or smoothing plane on the "bottom cheek" of the tenon, and keep "re-scratching" until it wont scratch no more? Is that how it works? The method you describe reminds me of those depth guages you can make with a simple wood block and a screw through it that you set to your required depth and "rout" to find fat or high spots.
When you think about it a 2" tuning fork shape would easily get bent out of shape and be useleess anyway :
Cheers, Jonathan

[beetle]

I just use my combo square and a caliper ?? Check the face depth off the layout face, check all faces for being parallel ( or square ) off the layout face, give the tenon a slight taper, chamfer the edges and make sure the caliper passes over the tenon with little or slight drag.

Final check...slip it in a mortise, if it fits nice  8) if not  :(  :P

[Jim_Rogers]

Jonathan:

You asked: "and keep "re-scratching" until it wont scratch no more? Is that how it works?"

Yes just keep trimming until all scratches are gone and then as mentioned try it into your mortise.

The standard wood working procedure is to make the mortise first and then trim the tenon to fit. If you do that just remember to only trim the non reference face side or you'll have an incorrect offset......


You asked: "Or possibly the borman device??" A big Al Borneman layout tool is usually just for layouts, but can in some positions be used to check surfaces for offset.

A chalked board could work if that's what you want to do. But that means you'll have to create this board, insure it's truly 2" thick, carry it with you including chalk. Adding more tools and procedures to your methods. Stick with metal tools and you can't go wrong.....


Jim Rogers

[hayton1960]

Hello badmojo 

I've wondered the same thing myself, about "great joinery" going iffy when the wood shrinks over time etc.
But imho, I think great joinery isnt just having joints where you cant fit a cigareet paper into, that can be a superficial image thing, (not always though); Its also correctly positioning frame members in a sound structural relationship, using apropriate joints arranged in correct proportions, a good foundation, arranging the timbers to even out structural stress etc. And of course any joinery, decent or crappy, will soon fail if the structure isnt well maintained. In every case I saw (too many sadly), the barns and colombage houses in Normandy that were collapsing into ruin, was down to either:- poor roof maintenance, poor gutter maintenece, allowing weeds/bushes/trees to grow too close to the building (I even saw trees growing inside the buildings sometimes!! :) The actual frames were no worse than others nearby, possibly even built by the same men, that were in good condition. The torchis (clay infill) would drop off bit by bit till there was just a silver grey skeleton left, that one day someone would flatten with a tractor, and burn for fuel : Perhaps theres a market for reclaimed frames? ;)!!

[Jim_Rogers]

How do the tolerances change after the wood shrinks or twists or checks or bows or bellies?  Will all great joinery go to crap after the wood cures?

Tolerances don't change, you must maintain as accurate as possible cuts as you are making your joinery. Once your structure is erected if the wood shrinks the gaps may show, they may not. It will depend on your joinery and whether your not you've cut your joinery to compensate for that shrinkage.
Every frame is going to check, that's a given unless you're using dried timbers. These shouldn't effect the strength of the timber.
Having erected your frame with green timbers the amount it bows should be reduced as the timbers are held in place by the joinery.
I'm not sure what you mean by "bellies".

No, good joinery won't go to crap.
But bad or loose joinery will be even worst or looser.
This is why you should try and make the best you can make.

Housing joints helps to hide shrinkage gaps.

Hayton1960 is right about having a good frame design and maintaining the building properly for years to come.


Jim Rogers

[beetle]

I beleive that if you cut your joints properly, keeping in mind how timbers shrink when you cut the joints and compenstate for this shrinkage that the joinery can actually become stronger as the frame ages.

Just like a fine Wine and good Scotch ;D

[beenthere]

Is 'crap' a defined word in the TF dictionary of terms, or is it just used loosly in the vocabulary?   ???

How about a definition for the forum?   

[beetle]

Didn't mean to rub ya the wrong way. I did not realize we we talking crappy timbers, I thought we were talking joinery and joint tolerances. My mistake :-[

[beenthere]

Not trying to cause a problem here, but just trying to understand what a 'crappy' joint is, and when does the joint 'fit' into this category. Maybe there isn't a definition and/or shouldn't be.
I've heard 'tight' and 'loose' joint, but is it in the mind of the beholder?    or is there some point at which it is a bad or unexceptable joint? 

Not 'rubbing' me the wrong way, nor do I want to rub you guys the wrong way either.    Just trying to 'keep up'.

[hayton1960]

One of the most useful things I was taught was how to get around the vagaries of having to use less than ideal wood. Its part of our tradition. In fact I was sometimes given duff timber just to learn this. Clear straight grained timber with few knots is very hard to come by. In some ways thats why us Englishmen and europeans had to find ways to use any old stuff. I've got 3 ash trees up to 9" diam in my front yard, nicely pollarded. I've deliberately stopped side growth and restricted top growth. I'll probably get about 40 good riven tool handle blanks (24-30") out of one tree. Some builders used coppiced and riven small wood for rafters, fencing etc. (eg chestnut, ash, oke) They used pre-bent knees from branches for braces etc. (you guys already knew that anyway) You end up with an organic looking vernacular architecture style where true plumb or level seems sort of a thing of the past. I dont know wether or not there was any urgency to fit the structure within a strict level plumb and square  scenario (at least in domestic houses anyway) As I already said before, I dont know if the old buildings were made crooked, twisty etc or if time, settling, shrinkage has acheived it. Either way its great joinery, it still stands, and still looks great!  Some of the houses from the 70's (1970's) now look tired and dated, plain uninteresting!!

PS When you think of the inuit canoe builders having to wait for the tides to wash up the timber to make their fantastic boats, it makes me wonder what I have to complain about  :

[Jim_Rogers]

I guess by "belly" you mean sag.

When a beam has deformed over years of load and has began to 'sag', that it has grown a "belly"?

To me a "belly" protrudes out.

I don't recall every seeing a timber grow something that protrudes out, other than branches on the tree when it was alive.

To discuss in timber framing terms one should try and use common terms which mean the direction of the distortion, as with standard lumber.

The grade rule books list definitions of warp, bow, crook, cup and twist. And other defects such as shakes, checks and splits.

A timber framer should understand these terms and how they effect your wood.

[hayton1960]

I was just looking at a back issyue of the tf journal, the one where they describe the repair of charlton court barn. Those 600 year old mortices and tenons were a bit on the sloppy side-1 1/2 " tenons into 1 3/4" mortises, cut up with a spoon bit and twybill (mortice axe) That structure held up a 100+ ton stone roof for all those years It was foundation subsidence that caused the major structural failures.

[Engineer]

I have helped two really talented guys cut and erect two frames in the past six months - my own house, and a 40x40 pavilion for the local Rotary club.   One lays out, one cuts, and five or six helpers and a crane put it up.  So far, every joint, when first put together, has either been a snug push fit or a slight tap or two with a framing mallet.  Pegs are all octagon hickory in drawbores.   You can't put a playing card in ANY of the joints once they have been pegged.    I wish I could have spent more time observing and participating in the layout process.    I have pictures of a couple of frames put together by another local "framer", and they are sloppy to the point of scary.  Some of the joints depend entirely on the strength of the pegs to hold them together - not good.

I'll be doing another frame next year, a 24 x 32 shop and office building, and I hope I can be part of a third frame raising that went together as well as the last two.  The house frame has been up for five months and some of the joints have minor gaps due to drying and shrinkage, but most are still airtight.

[Jim Haslip]

Goos Stuff. Keep it coming.