Sharpening conspiracy!

[Eshmiel]

Perfectly flat finish would be scraped and/or sanded. It's not possible with a plane alone.

Its not possible with a scraper or sander either.

The notion of flatness as being what looks flat to the human eye isn't a complete definition of flatness by any means, even though it might be a good practical definition since that's probably flat enough for many furniture making purposes (but not all). There is a geometric flatness that can't easily (or at all) be judged by eye alone. We use engineered straight edges with tiny tolerances for a reason. And those winding sticks you mention are, effectively, amplifiers enabling the eye to notice the not-flatness known as wind more easily.

A long soled plane is possibly the best tool for achieving woodworking flatness of the kind we want when making posh cabinets with only hand tools. A scraper can easily dig dips. Sanding even more so.

The matter could be subject to objective measurement rather than to a mere human eyeballing, then. A well-engineered straight edge, long enough to span a surface in any orientation, along with a set of feeler gauges, might well indicate something dipped and bumped with far more ability than a human eye. How much is that needed? That's another question.

 

[Terry - Somerset]

The term "flat" may relate to the condition of a surface defined by:

• granularity of the surface, and
• the maximum dimension over which "flatness" is to be judged

For a plane blade the granularity may be related to the grit used to sharpen - this is a much larger dimension than that which could be defined by its pure atomic structure.

The dimension (width) over which "flatness" would be judged is typically ~60mm. Defined by a tolerance - perfectly "flat" would have zero tolerance over an absolutely straight edge. In practice a tolerance across the dimension would be expected.

It is a somewhat academic - in days gone by precision measuring equipment did not exist. Assessing the quality of a blade would have been by eye. Acceptable would have been that which the craftsman assessed produced acceptable results on the workpiece involved.

A greater level of precision is simply unnecessary!

 

[Jacob]

The term "flat" may relate to the condition of a surface defined by:

• granularity of the surface, and
• the maximum dimension over which "flatness" is to be judged

For a plane blade the granularity may be related to the grit used to sharpen - this is a much larger dimension than that which could be defined by its pure atomic structure.

The dimension (width) over which "flatness" would be judged is typically ~60mm. Defined by a tolerance - perfectly "flat" would have zero tolerance over an absolutely straight edge. In practice a tolerance across the dimension would be expected.

It is a somewhat academic - in days gone by precision measuring equipment did not exist. Assessing the quality of a blade would have been by eye. Acceptable would have been that which the craftsman assessed produced acceptable results on the workpiece involved.

A greater level of precision is simply unnecessary!

Basically you do the best you can with the kit you've got.
It's not a problem in the ordinary way of craft work and unattainable definitions beyond your means are meaningless.
You can't rely on flatness from the tools alone, such as a long plane, you have to look at the thing as you work on it!

 

[steve355]

In woodworking, nobody is getting out micrometers, but in toolmaking it’s another story. A nice fine plane shaving is a few thou thick, so our plane sole needs to be flat to that degree. And the tool we use to make the plane sole flat needs to be an order of magnitude flatter again.

I have a granite surface plate in my shed, flat to 0.00005”. That’s pretty flat. It was bought for refurbishing a surface grinder that is used for making plane irons (and ostensibly flattening the soles of planes, although I haven’t used it for that yet).

I was just curious about reference surfaces in , say , 1600.

 

[Jacob]

In woodworking, nobody is getting out micrometers, but in toolmaking it’s another story. A nice fine plane shaving is a few thou thick, so our plane sole needs to be flat to that degree. And the tool we use to make the plane sole flat needs to be an order of magnitude flatter again.

I have a granite surface plate in my shed, flat to 0.00005”.

Woodwork tools aren't rocket science!
Ordinary "kitchen table" levels of flatness are good enough for most purposes.
In fact a typical MFC slab should be good enough for flattening woodwork kit. I'd use 80 grit wet n dry, thin paper-backed, wet with white spirit which holds it down very flat, and more poured on top to float off swarf and cut faster.
2 sheets for a long plane. Take them up when finished and store them between boards to keep them flat.
I'd guess similar levels of flatness would easily be achieved even in year 1600 but with different materials.

 

[steve355]

MFC is engineered to a max tolerance of 2mm / metre.

Jacob, if you think woodworking tools aren’t precision , you’d be wrong. Unless a plane sole is flat across its length to max 2 thou, it won’t work correctly. Trust me, I’ve made enough of them.

Which is why the history of precision in toolmaking, pre Whitworth, is an interesting topic.

 

[Jacob]

MFC is engineered to a max tolerance of 2mm / metre.

Sounds good. For flattening a plane sole it'd need to be on a solid base too of course. I have actually used a lot of planes and rescued a few neglected wrecks.
The main prob with a sole is likely to be concavity along the length. A little convexity is no problem - it's kitchen table science.
With a plane blade it's the seating on the frog - it needs to be tight at each end and nipped hard by the lever cap or wedge, via the cap iron.

Which is why the history of precision in toolmaking, pre Whitworth, is an interesting topic.

Normal methods I expect. Flatten grindstones by grinding two or three together for instance.
n.b. there's lots of interesting stuff on line https://www.google.com/search?q=how...E3OTkyajBqN6gCALACAA&sourceid=chrome&ie=UTF-8
I had to google "an iterative process" !