What tolerances do you work to?

[Tom Maple]

After reading some of the discussions on different calipers and micrometers and how they can be used I started to wonder to what tolerances do most woodworkers strive. Are most cabinetmakers and furniture makers satisfied with 1/100 of an inch? Or is it necessary to go to 5/1000 or even tighter? Where does the effect of wood expanding and contracting with differing humidity preclude obtaining any closer tolerances?
Pardon me if this has been discussed, as I'm new to this forum.
Thanks,
Tom

[Mike Henderson]

Since I build "one off" furniture, I cut to fit and am not really concerned about the exact measurement. If I have to fit two things together, I'll measure to make the cut but once I cut, I do a trial fit and see how they look. If they look okay, I proceed to the next step. If not, I trim until the fit is what I want.

I could never do an exact duplicate of anything I build - nor would I want to. I always find things (usually design things) I could have improved.

Mike

[Gary Keedwell]

You are right. There have been numerous postings about tolerances in many forums, this one included. There is a faction out there that insists that tight tolerances are unnecessary because wood "moves". There is another side that says that people who use that "move" excuse, are just sloppy woodworkers.:)
The only thing I can say is that I have ( as an example) picture frames that were glued together seamlessly and after many years they still look good and there are no "gaps" in the joints. My guess is that all the parts move together as a whole.
Anyways, regardless if the wood moves or not, if you put the wood together with gaps, your still going to have those same gaps down the road., I reckon.;)
Gary K.

[Jim Becker]

My answer is as tight as possible...which is possible because I tend to work like Mike, building to fit. I use story sticks, etc., for layout and use stops and other aids to insure similar components come out exactly the same length. When I do measure, I try very hard to avoid a tape measure and also try to use the same measuring tool for everything if at all possible...this avoids variances from scales that are "off" between measuring tools.

[pat warner]

I rout to ~ .001" or less when making tools (for example (http://patwarner.com/bev_straight_edge.html)).
Joinery, for good glue lines, is on the order of 1-3 mils slop depending on the connection and its utility, (KD for example can be tighter than a joint that gets glued).
Parts, on the other hand, (legs, drawer & door panels) can swing all over the place as long matched/paired parts are equal.
Stuff that fits inside of stuff (a flush inset door e.g.) has its own set of rules for sizing & fitting. Seasonality, humidity, type of cut (flatsawn e.g.), must be accounted for. No easy answer, always some conditionality.

[glenn bradley]

I'm with Mike at this stage in my learning. However I do try to make my joinery appropriate based on visibility, material movement, etc. There are times you don't want things super tight in June . .. they'll play hell on each other come September.

I like my exposed dados near-invisible, same for through tenons and any locking joinery like drawer lock joints. Cabinet back panels I give room to breath as I do raised panels, table tops and drawer bottoms. I guess for me, if it looks good and is structurally sound, I'm OK with it.

[Andrew Williams]

I work to "dead on", miss it, and plane until it looks good :P

[Mike Henderson]

From the late 17th Century to the early 19th Century, our ancestors built some of the most beautiful, precision furniture ever - and they probably didn't have one set of calipers between all of them. You can be very precise by eye. When grinding a lens or mirror, for example, which must be accurate to a quarter wavelength of light, the measurements were made by eye, looking at interference fringes.

The only reason you need very accurate measurements are for mass production - where any piece must fit the product.

Mike

[Bruce Wrenn]

Measure it with a micrometer, mark it with a piece of chaulk, then cut it with an axe! Precision work! I use dial calipers to check wood out of the planer. Use a modified HF plastic digital caliper to set bit height and depths on router work. Tape on fence of TS is used to set cuts. To set hair line on TS fence, I rip a short piece of wood (less than 4" wide). Then measure with a caliper and set fence accordingly. Beats the heck of trying to figure exactly when the fence is just against the blade.

[Richard Butler]

After reading some of the discussions on different calipers and micrometers and how they can be used I started to wonder to what tolerances do most woodworkers strive. Are most cabinetmakers and furniture makers satisfied with 1/100 of an inch? Or is it necessary to go to 5/1000 or even tighter? Where does the effect of wood expanding and contracting with differing humidity preclude obtaining any closer tolerances?
Pardon me if this has been discussed, as I'm new to this forum.
Thanks,
Tom

It depends on where, when, and in what material. One can't make a blanket statement that .01 or .001 or .0001 is the accuracy best held in woodworking.
Sometimes +0" -1/8" is ok. Other times +0 -.001 is held.

The short answer is "It depends".

[Joe Jensen]

After reading some of the discussions on different calipers and micrometers and how they can be used I started to wonder to what tolerances do most woodworkers strive. Are most cabinetmakers and furniture makers satisfied with 1/100 of an inch? Or is it necessary to go to 5/1000 or even tighter? Where does the effect of wood expanding and contracting with differing humidity preclude obtaining any closer tolerances?
Pardon me if this has been discussed, as I'm new to this forum.
Thanks,
Tom

I never use plans, and I alter along the way as needed to improve the design or to cover mistakes. I do however make sure all joints are tight, boards are straight, etc. I'd waste a ton of wood if I used plans :eek:

[Todd Jensen]

When is a gap a gap? :)
Are you building deck furniture or a violin? I like my stairs to be pretty near seamless - the filled gaps are increased per the size of the time whip, but usally a 1/64 is big. I appreciate the reference to infinite smallness, aka the 1/500000 inch gap, but apologetically think its meaningless. Wood is mushy to some degree. Is your super teeny gap a gap at 15% humidity or 22%? does it become a super tinier bigger at 5%? :)
Long story short, my answer would be to make it as flawless as your time and skills allow. Anything else is a moot point.

[Cliff Rohrabacher]

That all depends.

I may only work to eye if there is no greater need. I have on occasion held tolerances as close as .002" to .003" but expecting the wood to remain within those tolerances is a fool's game because moisture in the air and temperature will play havoc with tight tolerances.

However, if you have a whole lot of little parts that all depend intimately on uniform consistency to very tight tolerances then you can expect to benefit from working to very tight tolerances because as the parts swell and shrink they will be doing it in unison maintaining the relationships.

[J.R. Rutter]

The more standardized you are, the tighter the tolerances need to be. I build cabinet doors, and the only places where I need to be very precise are in width of stiles (+/- 0.005"), length of rails (+/- 0.01") because errors here add up fast. Even within these tolerances, a door can end up 1/50" over or under, but I only guarantee 1/32". Similar thing for thickness of panels, location of grooves / toungues. The closer to perfect alignment you get, the easier it is to assemble and sand.

We only use tape measures for rough cuts. Everything gets easier and faster if you have at least some standards, even for certain types of one-off pieces. I'm not saying that you have to have a caliper, a story stick will be more accurate over time than a tape that gets dropped or replaced, even if you calibrate it by bending the tab in/out.

[Don Bullock]

The INCRA TS-LS lead screw technology I just installed on my new table saw allows me to position my work instantly and automatically to within 0.002. Will I ever need that. No, not on my table saw, but perhaps when I use it with a router table. Even on the router I doubt that I will use that level of accuracy.

That being said, I wanted a fence system that would cut wood to the size I want, every time I use it. With the Incra, I get that kind of accuracy. (you must understand that before getting my new saw and fence I had a saw and fence system that was a struggle just to get the fence parallel to the blade, much less be an accurate didtance from the blade -- as a result I bought the Incra so that I wouldn't have those problems with my new saw -- Overkill? Perhaps, but I'm a happy woodworker)

Most of the time I work like Mike and Jim, cutting pieces to fit. My accuracy is measured on how well the piece fits.

[Bruce Benjamin]

The two paragraphs below seem to contradict each other. Why is it not a, "Fools game" to work to very tight tolerances under the conditions described in your second paragraph if it is a, "Fools game" under the circumstances of the first paragraph?

People frequently bring up wood changing with humidity and how this wreaks havoc on tight joints built to close tolerances. This has rarely been a problem for me. I'm not saying I haven't seen the effects of wood movement due to humidity but the wood usually moves together. When I make a perfect miter or other joint I expect it to stay that way for a long time and for the most part they do for me. If a joint has a gap because it wasn't made to tight tolerances it will have a gap down the road and vice versa. At least that has been my experiences. That's why I use calipers and try to mill my wood to precise tolerances. In the building process I also use my eyes and finger tips but measuring with a dial or digital caliper helps me get there.


That all depends.

I may only work to eye if there is no greater need. I have on occasion held tolerances as close as .002" to .003" but expecting the wood to remain within those tolerances is a fool's game because moisture in the air and temperature will play havoc with tight tolerances.

However, if you have a whole lot of little parts that all depend intimately on uniform consistency to very tight tolerances then you can expect to benefit from working to very tight tolerances because as the parts swell and shrink they will be doing it in unison maintaining the relationships.

[Art Mann]

Trying to answer your question is like trying to answer the question of how accurate a thermometer needs to be. If you just want to use it in the oven to bake a cake, +/- 20 degrees might be ok. If you want to know whether to wear a short or long sleeved shirt, you might want to know say +/- 5 degrees. If you are making some kind of candy with melted sugar, then you might want to know +/- 2 degrees. If you are trying to control a difficult chemical reaction, the +/- 0.5 degrees might not be good enough. The answer depends entirely on what you are trying to do. I strive for +/- 0.003" planing door frames. That is about the thickness of notebook paper and the human finger can easily detect a mismatch of that much. I do not enjoy sanding. If I am building a cabinet carcass, +/- 0.062" is probably just fine. If I am framing in a door opening in a house, a 0.125" error or more will not normally cause me to cut another piece. Most often, I have found that being able to make a number of pieces the same size, or match the length of one piece to the width of another, is much more important than absolute accuracy. In general, the more I have learned, the less I depend on precision measurements.

[Cliff Rohrabacher]

The two paragraphs below seem to contradict each other.

They, in fact, are not at all contradictory. I thought the qualifier I put in the second was pretty clear.


Why is it not a, "Fools game" to work to very tight tolerances under the conditions described in your second paragraph if it is a, "Fools game" under the circumstances of the first paragraph?

Because (as I stated) parts that work together will move together but the relationships remain constant. So if you have a thousand pieces all moving together they will not pose issues in that the ratios will be consistent so long as dimensioning was highly consistent.



However, when one is producing component parts over a period to be assembled at some later time the act of depending on wood to have retained its dimensional stability is simply error. Each piece made at a different time will have been made at a different. humidity and temperature. Nothing stays the same. There are plenty of other examples of why tight tolerances in wood is a game one plays on one's own mind. But that doesn't mean that there are also not plenty of reasons to try for the best closest dimensioning one can achieve.




After all, even where the wood will be moving it'll be less off the mark (after it moves) if it was produced closer to the mark in the first place.







People frequently bring up wood changing with humidity and how this wreaks havoc on tight joints built to close tolerances.

That wasn’t my proposition though.

[Brian Dormer]

Tom,

I think you might be confusing the tolerances used to align machines (TS, etc.) .vs. tolerances for measuring and cutting stock.

Alignment tolerances are generally in the .002-.003 and .1 degree range. As for marking and cutting stock... I can only say what I do...

I mark as accuratly as possible. Sometimes with a tape or ruler, other times with a story pole. Prefereably by marking off of whatever I'm trying to match up to (the best measurement is no measurment). If I'm "making it up" - like say, cutting a panel for a cabinet that needs to be 18" deep - I'm way more concerned with both sides of the cabinet being the same rather than being dead-on 18". Straight, square and the same is what I'm after. A 64th or 32nd more or less isn't going to hurt as long as everything matches up in the final assembly. So I will try to set up my TS and cut everything that needs to match up without changing the setup.

If a piece has to fit a given space - leave a little "slop" factor to scribe it to the wall or whatever it has to go up against. Again - beign a 64th off isn't going to make a difference - you'll make it up in the scribing.

Sometimes woodworking isn't so much about doing it right as doing it so the mistakes don't matter (or show).

[Dave Falkenstein]

I'm thinking along the lines with Brian, just above. I can and will set up tools to tight tolerances, measured in thousandths. But when measuring and cutting wood, whatever level of detail I can see on my rule and line up with my saws seems to work well enough, at least for me. Gang cutting matching pieces works well. However, when cutting dados, I like to measure the stock with a digital caliper and set up the cut to be just slightly wider - a few thousandths wider, if possible.

I'm a mechanical engineer, and tend to be overly concerned with accuracy, but when working with wood there is a degree of artistic expression that becomes acceptable. Don't ask me what that means, please. Smile.

[Gary Keedwell]

:)
I'm thinking along the lines with Brian, just above. I can and will set up tools to tight tolerances, measured in thousandths. But when measuring and cutting wood, whatever level of detail I can see on my rule and line up with my saws seems to work well enough, at least for me. Gang cutting matching pieces works well. However, when cutting dados, I like to measure the stock with a digital caliper and set up the cut to be just slightly wider - a few thousandths wider, if possible.

I'm a mechanical engineer, and tend to be overly concerned with accuracy, but when working with wood there is a degree of artistic expression that becomes acceptable. Don't ask me what that means, please. Smile.
I have a whole pile of "artistic expressions" that is also called fuel for my stove.:D
Gary K.;)

[Dan Mages]

I do the very best I can to make the tightest tolerances possible given my skill level. Given that, I do not guarantee anything...

Dan

[Howard Rosenberg]

1 - is it square
2 - does it fit

Anything else is a conceit.

After all, it's WOOD!!
It moves!

My 2C.

Howard

[Richard Butler]

Years ago I thought I was doing good to get my stuff to within 1/32 of an inch. I spent a lot of time at the bench sanders. Now I have my cuts down to about -0 +.004 and don't use the bench sanders at all.

[Brad Naylor]

1 - is it square
2 - does it fit

Anything else is a conceit.

After all, it's WOOD!!
It moves!

My 2C.

Howard

Absolutely!

Tolerences of 1/1000's of a inch are essential in engineering but in woodworking irrelevent.

If it doesn't fit get a bigger mallet!

[Rich Engelhardt]

Hello,

What tolerances do you work to?

I'm pretty much still at the "painted" stage.
"painted" as in - the caulk used to fill the gaps doesn't show quite as bad.

Having said that - I consider this a quantum leap from where I was a year ago.

A year ago I was at the "all things get downsized" stage.

"all things get downsized" as in - Everything I made started out large and got progressivly smaller as I removed a bit here, a bit there to make it fit. Which of course threw the opposit side off, so I had to remove a bit here and a bit there to make that side fit, which of course threw the other side off..and so on and so on - until "shop cabinets" turned into lopsided birdhouses..

My immediate goal is - a "natural finish" project.
"natural finish" as in - very unforgiving since no caulk can be used to hide the gaps.

Side note -

Tom,
Where in Summit are you? I'm over in Stow.

[Steve Schoene]

The big difference is how tolerances are achieved. A properly fitted mortise goes together with firm hand pressure. Well fitted edge joints don't pass light. I think in each case, if one were to achieve this my engineering measurement, as might be required for mass production, it would call for very small tolerances. But I don't know, and never need to know the difference in thousands between too tight and too loose, or what size gap will show light between the edges of boards.

[Michael Schwartz]

I achieve tollerences upwards of .001 all the time but I dont' try, because if I did I would be wasting time. I am a Woodworker, not a Machinist. The fact of the matter is good equipment setup right and used with skill will deliver very tight tolerances. The key with woodworking is for your parts to be exactly the same, and the actual dimensions off the plans don't matter as much.

For a tape mesure I generally use a Fastcap Flat Back Tape, or a Lufkin Centerfinding Tape, mostly to set fences where I need the ridgid blade.

Keep the tapes in your woodshop separate from your jobsite tapes. Check the tapes in your shop often for accuracy.

I have a 6" hooked rule from Veritas I use for measuring anything up to 6" It is marked in 8ths and 16ths on one side, and 1/32 and 1/64 on the other and is very easy to read. Woodcraft sells one but they only go up to 1/32.

My 6" starret combo square is marked up to 1/64 on one side and 1/100 on the other.

At the planer I use my Starret combo square and a finger. The human finger can feel a difference as small as 40 microns, or 0.001574" Every bit as good as as the reading you will get from a caliper.

[David DeCristoforo]

This is a trick question! What are you talking about? A Mitre joint? Well in that case the "tolerance" would be zero. It either fits or it doesn't. A drawer side? Not so critical right? A rail that needs to slide easily in a groove wants to be "looser" than a shelf fixed in a dado. But you don't want to have to use an eight pound sledge to get the shelf in either. Simple common sense prevails.

[glenn bradley]

I have a variety of measuring tools and use most of them regularly. I use them as 'tools' to help me, not regulate me. I build one-off stuff so once the general sizes have been achieved the calipers etc. help guide me in fitting my joinery. I would never be able to make 4 or 5 different parts based on measurements alone and expect them to go together as planned.

In a nutshell; I work to get close as possible but not so tight there's no room for the glue.

[Lee Schierer]

I use the measure with micrometers, mark it with chalk and cut it with an axe method of sizing. :D

Closer is better, but no sense in working with micrometers when you are making a dresser. +/- 1/32 should be good in most instances. Use jigs for repeat cut lengths to get them all exactly the same.

[Jim Becker]

Although I've already posted in this thread, let me expand slightly...

I will "measure" some things with a rule or even a tape if it's a large enough dimension, but in general, I use the project, itself, to measure and calculate what needs to be cut to what exact length. Story sticks, in other words. So I go from a specific dimension, such as a 1 7/8" square leg stock and worry about the exact distance between two of them by physically marking a simple scrap stick of wood that is in turn used to mark and make an initial, slightly oversize cut of that component. A quick check of the component on the actual project will confirm what adjustment needs to be made with tools. In that manner...the accuracy is about as exact as one can get. No ruler is going to do better, IMHO. And this also works no matter what measuring scale you prefer to work in. It also insures that identical component are, well....identical.

[Gary Keedwell]

I measure with steel rule when I can. STOPS are the secret for consistency. When I have multiple parts that have to be exactly the same...I use stops.
Stop...that's my story, and I'm sticking to it!!!:p

Gary K.

[Andrew Williams]

I like stops too, and sometimes will just gang-crosscut boards if there are not many, like a drawer for instance. The pain is when I measure correctly and cut correctly and then the wood cups or bows or something and the drawer binds. Plane to fit.... I actually have a few drawers in a shop cabinet that changed too much since I fabricated them and have to be planed again now. Should have made them looser in the winter.

[Michael Weber]

After reading some of the discussions on different calipers and micrometers and how they can be used I started to wonder to what tolerances do most woodworkers strive. Are most cabinetmakers and furniture makers satisfied with 1/100 of an inch? Or is it necessary to go to 5/1000 or even tighter? Where does the effect of wood expanding and contracting with differing humidity preclude obtaining any closer tolerances?
Pardon me if this has been discussed, as I'm new to this forum.
Thanks,
Tom
I generally work to my tolarance for frustration or exhaustion. Then I quit. :D