Surface, flatness, precision.

[Eshmiel]

Another thread has had an orthogonal jaunt into the matter of woodworking flatness and precision. It needs a new thread for those of us who enjoy exploring the far end of various pharts.

One view is that woodworking flatness is that which the eye can detect and no more. Another view is that this is a limited definition albeit prevalent - some surfaces need to be flatter than that; and others don't really need to be flat in their detail as long as they're flat across a whole surface. There are no doubt several other kinds of WW flatness sought by we woodworkers.

Then there's other kinds of precision, such as that required for various kinds of joints; or for stress-bearing architecture of furniture. And for the tools used in making stuff.

* When you make a piece of furniture, what degree of flatness, joint quality and other precision-of-surfaces do you aim for?
* Does your required-precision vary with the item being made and, if so, in response to what attributes or functions of that item?
* What degree of precision do you want in your various tools and how do you ensure it?

Personally I tolerate "marks of the maker" lack of precision but feel I've been inadequate if an item somehow looks "wonked" or "bodged" in surface, joint, proportion or general appearance. I also prefer tools to have a degree of precision that can approach the ability to make changes down at the 0.1mm or 3-4 thou level. Sometimes this makes it necessary to make the tool parts themselves precise to less than those values. A plane sole and blade that can take off just one thou, for instance, is often a requirement because it enables the plane to more easily improve surface qualities and the precision of joints.

What about you?

 

[Spectric]

For me it is visual, work to what you have so putting something dead upright next to a wall with a slight lean will stand out. I think it is all to easy to forget this is a natural material and there is no point in trying to be more precise than what the material can change with the enviroment. Joints need to be tight, gappy joints look awful.

 

[Fitzroy]

As an engineer this is a rabbit hole that I have been down, and got lost in, on many occasions.

• Flatness: From a practical perspective whatever comes off of the planer thicknesser and doesn't move so much over the next few days that it causes problems in the project I'm working on. Large items can have more deviation over the total length as it's less obvious to the eye. Box components need to be dead on else the eye sees even a small error in the box shape. A gap between two components needs to not be necessarily flat-straight, but it needs to be a consistent gap. So i guess in general 'it depends'.
• Joints tight enough for a close glue line, but not so tight they are a pain during a dry fit. Ideally straight off the tool with no fettling. I historically used to saw tenons fractionally oversized and pare them to fit, but I found my ability to pare consistently across the tenon worse than my ability to set the band saw accurately!
• Tools: Oh my there is a totally different question. I like chisel ends and plane blades perpendicular to the sides, I like things sharpened to a mirror edge so that I can shave hairs on my arms. I like my planer tables perpendicular to <0.1mm across their five feet of length, my table saw blade parallel to the table slots to <.05mm across its diameter, my mitre saw cutting at 90deg with no light gap on my square across a 200mm board. But all those things are personal preferences, they are not necessary to make accurate items out of wood, that skill lies with the woodworker not the tools.

Fitz,

 

[profchris]

I make musical instruments, so I have flat surfaces that then get curved. They need to start out flat enough to meet requirement 3 below, but I might modify that to achieve requirement 1.

My requirements are:

1. It must work properly - playability, sound, etc. Wood doesn't have coherent properties even across the same board, so I might thin one section more than another to get the best sound.

2. It must feel good. The feel of an instrument is really improtant to a player - if it feels odd, or uncomfortable, they can't play it well.

3. Appearance. Ideally it will look good. All the surfaces which should have started off flat will look as if they did so, all those with curves should look evenly curved. Where pieces of wood join, ideally the naked eye will not be able to see the join.

I suspect other specialist woodworkers have their own requirements lists.

 

[Sideways]

Also visually straight and flat.
Also want joints to be tight, not gappy.
Everything relative to the tools I have to judge straight and flat. In my case these are my saw tracks, some good 600mm steel rules, a variety of engineering, combi and aluminium squares. Sometimes my spirit levels which are acceptably flat, and the cast iron top of my saw.

I'd never put a feeler gauge between a reference and a piece of wood like I would metalworking. But looking for light between the wood and the reference as we all do can show up a finer gap than you can measure with feeler gauges anyway.

Distances are tape measure, steel rules, digital caliper depending. I like to use ruler stops to get consistency across parts. Absolute dimensions are usually less important. I'd mark parts for a shed with a pencil, parts for a piece of solid wood furniture with a knife.

 

[Spectric]

I like to use ruler stops to get consistency across parts

Yes a great way to transfer a measurement as it takes out all user errors for each measurement that could sum to creat a big error.

 

[Eshmiel]

Lots of insightful stuff in the posts so far. I'm trawling them for good bits.

Feel. I can see how this will be a prime in a musical instrument made of wood but it can also be important in furniture. Our finger tips are remarkably good at feeling the quality of a surface. We have some Mouseman-style adzed furniture, in which surfaces are linear but formed of multiple adze-scallops. The feel of the scallops, as well as the look in raking light, is an important enhancement of the furniture. No surface is flat but all the joints are tight, the corners at 90 degrees and the plates don't slide off any tabletop slopes.

I've made many little boxes in which the smoothness of the sides and top seem to be important. When it's picked up, you want those surfaces to be glass-like. Warps of any kind are also immediately obvious in such small things, along with the gaps, as other posters have said.

Green wood spoons, mugs, bowls and the like - made with axes, knives and gouges ..... . Sometimes I like to leave the knife scallops as part of the look & feel of such things: "honest" work in that the tool marks signal unequivocally what tools were used. But others I like to scrape and sand to smoothness, as much for the hand as the eye. It depends on both the item's function and any decorative role it might have.

*************
Tight joints. There's tight and very tight and near-invisible (if you can match the grain of the two pieces). Gappy can even be a thing, in those greenwood items that are meant to indicate "quick, basic and probably temporary but functional".
***********
I find making curved surfaces the hardest thing to do well. Some items benefit greatly from curves that are ...... even? regular? not-bumpy? One can get away with a lot when making greenwood spoons and bowls but when the skill is acquired to make them elegant in their form, these do have an extra appeal. Despite my 14 months and >100 items made so far, I still struggle to get anywhere near elegant; but a friend who's been making such stuff for many years does it effortlessly.

 

[Sgian Dubh]

Basic solid wood edge joinery used for making wider panels is usually most successfully achieved in both the short and long term with a sprung joint, i.e., neither edge is straight with each, ideally, being concave. Slainte.

 

[Jamesc]

As another engineer I pretty much agree to the letter with Fitxroy

 

[Eshmiel]

Basic solid wood edge joinery used for making wider panels is usually most successfully achieved in both the short and long term with a sprung joint, i.e., neither edge is straight with each, ideally, being concave. Slainte.

I do spring mine but only after I've first made them both dead straight with me jointer plane. A few passes over varying lengths of the central parts with a fine-set No 4 plane makes the small gap. I prefer also to plane two boards together to make the matching straight edges, to cancel out any small cant in them from 90 degrees. I know the purist does one board at a time, varying any cant back to 90 degrees by skilful use of a cambered blade to un-camber an edge but my cack hand does for that approach.

 

[Yojevol]

@Eshmiel you've just touched in the sense of touch (pun intended). For me it's the most important aspect after visual appearance. People just want to run their hands over a visually attractive wooden piece, particularly if it's curved, and they don't want to feel imperfections be they lack of flatness or nibs remaining after a rushed finishing process.
The degree of flatness required for a table top, say, is related to it's design, it's function, it's age and it's position. Take somebody looking for a table to go in their farmhouse kitchen. They may well be attracted to an example that has all sorts of imperfections. In fact they may be attracted to a new piece that has been craftsmen made to look 'distressed'
Here's a couple of examples from my own hands that can stand different interpretations acceptable flatness:-

The elm coffee table depends on its distinctively grained surface and curvy undercarriage. The top was acceptably flat when I made it 28 years ago. Now one end is cupped due to it being a crotch but that doesn't matter because the crotch grain pattern takes precedence.

The elliptical table demands a high level of flatness because it's quality must match that of the cast aluminium pedestal. Its colour is important - (ABW) and the finish is also important due to the reflections arising from its light and airy position. Flatness is retained by virtue of its MDF core structure.

On the question of precision, here is an interesting design feature which caused a lot of discussion when I made it as part of my furniture making course all those years ago, as was the coffee table above.

The beading has several factors of precision about it-:

1. The regularity of the repeated design
2. The dimensional accuracy of the individual 'scallops'
3. Consistency of the corner mitres
4. Consistency of the horizontal beeds relative to the vertical ones

I made the basic cabinet in the college workshop but the 'pretty' bits were made at home. When I took the beedings in to be fitted my tutors and fellow students assumed I had used an indexing method to produce repeated pattern. I had to explain that no, I didn't, because any small error in an indexing jig would multiply and result in a large error in the overall length. It was the overall length that needed to be precise to fit the door frame. The scallops were dimensioned and positioned using a paper ruler printed from my CAD model. They were then machined using a router and a purpose made jig
It was important to me that the mitre joints were consistent, so these and the relative lengths of the stiles and rails meant the beedings could not all be identical, but they are near enough not to cause a visual problem.
They live up to my motto " Perfection Will Do"
Brian