Making A Wooden Straight Edge

UKworkshop.co.uk

Help Support UKworkshop.co.uk:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

custard

Established Member
Joined
20 Aug 2008
Messages
7,170
Reaction score
672
Location
Hampshire
There's been a few mentions recently about wooden straight edges, every time the topic comes up I remember that mine is getting bit long in the tooth and should be replaced. I had a spare half hour this morning so made one up.

For shorter straight edges, say 150mm or 300mm I tend to use an engineer's quality steel ruler, they're a bit thicker than normal metal rulers so have the stiffness you need for a straight edge. For a longer straight edge, say 2 metres, it's easy to rip off the factory edge from a sheet of MDF and use that. Where wooden straight edges really come into their own is in that middle ground, say from 0.6 to 1.5 meters. The metal equivalents are too heavy and expensive for everyday use, plus they're unwieldy so you can easily damage your furniture with them.

If you're going to make one there are three hurdles that you'll have to cross. Finding a suitable piece of wood, drawing a fair curve on it, and planing a really accurate straight edge. I'm sure there are loads of different solutions to these three problems, but here is how I set about it.

Unless you can get hold of the right piece of timber it's a waste of time making a wooden straight edge, a gnarly bit of wet stuff from a skip isn't going to deliver the stability you need. What you want is a straight grained, quarter sawn or rift sawn piece of completely dry hardwood. That's not so easy as almost all PAR and sawn kilned hardwoods are flat sawn. That's because the market usually wants cathedral grain in the centre of the board, tapering out to rift sawn edges. This is the configuration that delivers most visual grain interest in the centre of the boards, along with straighter grain at the edges that makes it easy to seamlessly joint boards together to make them wider. The ideal solution is to visit a proper timber yard that has through sawn, waney edged boards re-stacked into the original boule or flitch. You can then dig out a straight grained, quarter sawn board and you're pretty much sorted.

But if that's not possible then you still have some options. Oak is pretty much the only timber that's commercially quarter sawn (so as to maximise the distinctive Oak ray figure), so there's a good chance you can get quarter sawn Oak. Alternatively, if you get a board that's 75mm or thicker you can rip off a thin piece from the edge, which is likely to then be rift or quarter sawn. Finally, if all else fails and you're near the south coast in the Dorset/Hampshire area then PM me and I'll sort you out a suitable off-cut!

Here's the board I used,

Straight-Edge-1.jpg


It's quarter sawn Mahogany, machined to about 1000mm x 80mm x 10mm. Any width from about 65-90mm is suitable, but you want to keep the thickness in the range 9-12mm. Any thinner and it's too whippy for a straight edge, any thicker and it's too heavy and too thick for an accurate read.

The bow shaped item next to the board is how I draw fair curves . It's a lath about 40mm x 15mm that's about 1600mm long. The string runs through two holes at one end and is secured with the same arrangement you might use for tensioning a tent's guy rope.

Straight-Edge-2.jpg


Normally in a commercial workshop you get someone to help, between the two of you bend a long flexible steel rule to form the curve you want and then pencil it onto the workpiece. But on your own you'll find you're at least one hand short of pulling this off, which is why I use this bow arrangement. The resulting curve is fine for curved aprons, most furniture components, and for the back of a straight edge. But even though it's fair to the eye it's not an accurate section of a circle, the radius is actually very slightly smaller at the ends than in the middle. Absolutely not an issue for this or many other applications, as I said, visually it's fully acceptable. Here's the resulting curve (look closely and you'll see the red line),

Straight-Edge-3.jpg


Then it's off to the bandsaw to cut down to the line,

Straight-Edge-4.jpg


Then over to the bench to spokeshave it smooth and knock of the arrises with a nicely spokeshaved round, smoothed off with 120 grit.

Straight-Edge-5.jpg
 

Attachments

  • Straight-Edge-1.jpg
    Straight-Edge-1.jpg
    79.6 KB
  • Straight-Edge-2.jpg
    Straight-Edge-2.jpg
    88.9 KB
  • Straight-Edge-3.jpg
    Straight-Edge-3.jpg
    72.9 KB
  • Straight-Edge-4.jpg
    Straight-Edge-4.jpg
    75.7 KB
  • Straight-Edge-5.jpg
    Straight-Edge-5.jpg
    88.8 KB
I've heard that putting on a curve makes the straight edge react faster to changes in humidity by exposing more end grain. I can't vouch for that, but what I can say is that a curve stops you accidentally using the piece for another project in the mistaken belief is just a plain board, and it also makes it clear which edge really is the straight edge!

The next job is the most important, planing a precise and accurate straight edge.

You could simply run it across a power planer/thicknesser, but every furniture maker I know prefers to hand plane for ultimate accuracy. I'd never joint up boards straight from a planer/thicknesser, so neither would I prepare a straight edge on a machine.

The problem is that, at 10mm thick, this board is right on the limit of what you can deal with by placing the workpiece in a vice. It's really on the verge of being just too thin to balance a plane on. Plus that balance problem is made much worse because it's likely to flex and twist under the pressure of the plane. So to stand any chance you need to stiffen it up, here's how I go about that.

This is my bench,

Straight-Edge-7.jpg


The trays in the tool well can be lifted out like this,

Straight-Edge-8.jpg


Which then makes it possible to run a sash cramp across to tightly secure the other end of the workpiece.

Straight-Edge-9.jpg


If your bench won't allow this then you could G-cramp the workpiece to a larger flat board and secure this in your vice. You need to follow one of these two routes because this is a tricky planing challenge to begin with, if the ends of the workpiece are flopping around then there's zero chance of pulling it off accurately.

With the workpiece fastened down you can go through the normal procedure of taking stopped shavings (starting and stopping 15mm away from each end) until the plane won't cut any more, then taking two or three through shavings with your plane set fine. This should yield a dead straight edge which you can test, either with another straight edge or on a known flat surface like this,

Straight-Edge-6.jpg


But the edge doesn't only need to be straight, it must also be at 90 degrees to both faces. Actually it could be off slightly provided it was off by precisely the same amount throughout the entire length of the edge. However, we have convenient checkers for 90 degrees in the shape of small squares, where as we're unlikely to have a checker at 88.3 degrees or whatever angle the plane might have been tipped over at, so I'll aim for 90 degrees!

It's much harder to achieve this on a 10mm wide workpiece than on a 20mm wide one, I can just about pull this off but it's right at the limit of my hand tool skills.

Luckily there's a much easier alternative.
 

Attachments

  • Straight-Edge-7.jpg
    Straight-Edge-7.jpg
    80.5 KB
  • Straight-Edge-8.jpg
    Straight-Edge-8.jpg
    78.7 KB
  • Straight-Edge-9.jpg
    Straight-Edge-9.jpg
    89.3 KB
  • Straight-Edge-6.jpg
    Straight-Edge-6.jpg
    74.1 KB
The easier option for precision planing is basically to improvise a kind of a shooting board type arrangement.

This is the way I plane the edges of saw cut veneers before jointing them up, or how I'd prepare the edges of 8mm thick drawer bottoms, it's a great technique to have in your armoury as, with a bit of care, it pretty much guarantees excellent quality results even if you're a relative newcomer to woodworking.

This what you're aiming for,

Straight-Edge-10.jpg


I start by removing several of the lift out trays from the bench tool tray, and laying down a piece of scrap 15mm or 18mm MDF that's a bit longer than the workpiece and about 300mm wide. This forms a bed for the plane to run along, exactly as with a shooting board, but if your bench is absolutely dead flat then you don't really need it. My bench is pretty flat but I still prefer to have this base board as a plane runs slicker on MDF plus it keeps the bench surface clean. On top of this you place another piece of ply or MDF, it could be any thickness from 6mm to 25mm (as long as it's thick enough to allow your plane blade to reach the workpiece), and ideally would be about 200mm wide. On top of this you place the workpiece itself, overhanging the 200mm wide piece by just one or two mill, and finally you place a packer board on top and then secure it all with F cramps or G cramps.

Straight-Edge-11.jpg


This is what the "package" should look like,

Straight-Edge-12.jpg


Actually, the top packer board is only strictly required for veneer cutting, but as that's what I mainly use it for I'm used to having it there!

Set your plane's lateral adjustment as level as you can, or if the plane iron is cambered set it so the apex of the camber hits the centre of the workpiece. Incidentally, I find the lateral lever on virtually all planes is too coarse for really fine adjustments, and lightly tapping the iron with a small hammer is much more accurate.

After that you run through the same procedure as before, stopped shaving until the plane won't cut any more, then two or three very fine through shavings until you get one perfect and full length through shaving from the entire length of the workpiece. At this point hopefully you should be bang on. If I'm working on pieces that will be edge jointed together I normally add one or two more full length fine shavings to add a minute hollow for a sprung joint, but that's not something you want with a straight edge.

Again, test the edge against a known flat surface,

Straight-Edge-6.jpg


Or you can lay it flat on a board and run a pencil down it, then flip it over and check a pencil line drawn from the other side verifies it as straight (this doubles the error by the way). However, this method of checking only works if you're 100% confident that the edge is at 90 degrees as well as straight. This sketch shows why,

Straight-Edge-Sketch.jpg


There are actually two completely different straight corner edges on a straight edge, and flipping it over means you're testing one against another. That's why the second, shooting board style, planing technique works much better unless you're completely confident in your hand planing technique, chiefly because it guarantees consistency in that angle. Ideally it's 90 degrees, but even if it's very slightly out then it'll be out by the same amount all along the edge.

You shouldn't get too obsessive about achieving an absolutely perfect edge, for woodwork rather than engineering there is a small acceptable tolerance. But the beauty of a wooden straight edge is that you can improve the precision of the edge over time, finessing it with some light sanding to remove a high spot, or re-planing the edge if it gets bashed, and each time you go through the process you'll get that little bit more proficient at it.

Good luck!
 

Attachments

  • Straight-Edge-10.jpg
    Straight-Edge-10.jpg
    81.2 KB
  • Straight-Edge-11.jpg
    Straight-Edge-11.jpg
    69 KB
  • Straight-Edge-12.jpg
    Straight-Edge-12.jpg
    73.7 KB
  • Straight-Edge-6.jpg
    Straight-Edge-6.jpg
    74.1 KB
  • Straight-Edge-Sketch.jpg
    Straight-Edge-Sketch.jpg
    13.5 KB
custard":17tvx3gy said:
Which then makes it possible to run a sash cramp across to tightly secure the other end of the workpiece.

file.php


If your bench won't allow this then you could G-cramp the workpiece to a larger flat board and secure this in your vice. You need to follow one of these two routes because this is a tricky planing challenge to begin with, if the ends of the workpiece are flopping around then there's zero chance of pulling it off accurately.

Robert Wearing provides an alternative method, which (like yours) involves no modification to the bench.

post503172.html?hilit=%20Robert%20Wearing%20#p503172

BugBear
 
Thanks Custard - I think I might have a suitable offcut to make one of those!
 
The classic way to guarantee straightness is to make three. If all three can be paired up and show a perfect fit then they must be straight. Any two could be off but not show a gap if a convex shape fits a concave shape. A third one will reveal all; i.e. it can't fit both if any of them are off.
 
Jacob":23x909sw said:
The classic way to guarantee straightness is to make three. If all three can be paired up and show a perfect fit then they must be straight. Any two could be off but not show a gap if a convex shape fits a concave shape. A third one will reveal all; i.e. it can't fit both if any of them are off.


Hello,

This is true, but only necessary if we have absolutely no flat reference surfaces available in the workshop. Unless we want true engineering tolerances then our best long plane and a well ground machine table will give us enough truth for wood work and saves us making three straightedges!

Mike.
 
Thanks Custard for this thread I've got a few off cuts of quarter sawn oak I'll make myself a set over the Christmas break.
 
Thanks Custard.

Love the "bow" curve template which has gone straight to the top of the "to make" list - I'm kicking myself for all the time I've spent fiddling with trammels !
 
SammyQ":17i0r7e7 said:
Jacob's right. Three to achieve trueness in straight edges.

He is indeed right. I wonder where he learned it?

*chuckle*

It's also quite true what Custard says about not obsessing about 'absolute' truth. For woodworking purposes, Custard's way is plenty good enough - but the Rule of Threes could be very handy in a communal workshop, especially as three or more straightedges can be checked against each other, even if regularly used by different craftsmen (or students).
 
Nice thread. Just for interest, there is a very good section in The Anarchist's Tool Chest book (highly recommended by the way) which may be where Custard gets his reference to putting a curved top on so as to expose more end grain and achieve better stability. Some people also cut fancy ends so that that they do not mistake their straight edge for a piece of usable wood. I put a hole in mind and it hangs on the wall next to my steel rules and levels.
 
AJB Temple":glm9a83r said:
Nice thread. Just for interest, there is a very good section in The Anarchist's Tool Chest book (highly recommended by the way) which may be where Custard gets his reference to putting a curved top on so as to expose more end grain and achieve better stability.

I think this notion may have originated with the curved sides of wooden coffin bodied smoother planes.

I've seen the claim a few times, but never any evidence.

BugBear
 
I think the 3-straight-edges method should not be dismissed. If you've already got a straight edge, then making a straight edge is trivial and it'll never be more accurate than the reference straight edge.

The curved edge to me seems unnecessary. I think it would more useful to make it bevel/knife edged.

As for making it different so you don't mistake it for a normal piece of wood, there lots of other ways!
 
JohnPW":2v566xg9 said:
I think the 3-straight-edges method should not be dismissed. If you've already got a straight edge, then making a straight edge is trivial and it'll never be more accurate than the reference straight edge......
Nobody dismisses it - it's the foundation of precision engineering!
https://en.wikipedia.org/wiki/Surface_plate
The 3 method can actually be more accurate than a particular reference plate or straightedge.
And is a practical way of making yourself three accurate straightedges, knowing that they are accurate, and having an easy means of checking them at any time.

PS i.e. you follow Custards instructions but make three, Then check them against each other. If all combinations (1 + 2, 2 + 3, 1 +3) fit with no gaps then all are accurate. if any one match doesn't fit then one, two or all of them are not accurate. Not that I'd bother myself, I just thought it was interesting.
I'd do as Custard does - just make one and also eyeball it for straightness
 
I have read elsewhere on the interwhatsit that exposing as much end grain as possible helps with stability - yadda yadda...

I think a bevel was planed on 3 edges to assist with this.
 
Back
Top