Is it normal for a straight edge to be doing this?

UKworkshop.co.uk

Help Support UKworkshop.co.uk:

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

Tetsuaiga

Established Member
Joined
4 May 2012
Messages
573
Reaction score
1
Location
UK
This is a piece of douglas fir I planed, my straight edge is the type that has a tapered edge. As you can see as I tilt it back or forward it lets light through either in the middle or the edges, does that mean my straight edge is bent?

When buying one I though one the narrow edge type would be more accurate than flat edge but i'm confused why its telling me different things when its tilted at angled.

The last picture is with it at very close being at 90 to the surface sitting by itself. The change seems to be a little less when its pushed towards its flat side. Maybe the edge is straight but has bow in it, if so is it still reliable?

Thanks
 

Attachments

  • 20151021_132658.jpg
    20151021_132658.jpg
    91.3 KB
  • 20151021_132715.jpg
    20151021_132715.jpg
    99.1 KB
  • 20151021_132730.jpg
    20151021_132730.jpg
    103.5 KB
Straight edges / steel rules can be unreliable! An engineering quality straight edge (for a example a B grade workshop SE) will be far more accurate than most cheap steel rules. But you will pay quite a bit more than the sort of thing you typically get from woodworking tool suppliers or Amazon. A 1 metre Starret for example is in the £60 area (plus VAT) for bench grade. You can get a pretty decent carbon one for maybe half that.

The point I am making is you need a decent reference point to know whether rule, wood or both are off ;-)
 
A straight edge isn't really a woodworkers tool at all. Better to judge straightness by eye, and twist with boning rods. Though I sometimes use a spirit level - it's a rigid box construction so won't bend like your edge. A thin edge is only useful laid flat e.g. for cutting card with a knife. On edge it may bend.
 
I think that means it's only flat only along one side (the working side, in this case quite narrow) but not along the tapered sides. I don't if they are made so that only one side is flat. They should follow a standard, ie BS or DIN, but I've only read eg "within 2 microns per metre" etc, but not whether a particular straightedge has which surfaces that are flat. I would presume only one side on a straightedge is flat.

I've noticed this with engineer's squares, if I don't hold it square on, I get a false reading.
 
The straightedge may have a slight bow in it's length. It is probably at it's most accurate stood upright on it's edge, not tipped over.

There is a way to generate your own workshop straightedges, by making three of them. Then compare A to B, then A to C, then C to B. It's possible for two of them to have a slight dip in the middle, and one to have a slight hump. The hump might match the dip in the others, but two dips compared will show up immediately. By gentle planing or scraping the reference edges until all three show no deviations from one another, you can get all three straight enough for all woodworking needs.

That's the method originally used to generate reference straightedges and surface plates. No longer necessary in normal commercial manufacture, but handy for those not having expensive manufacturing and metrology tooling available!
 
Cheshirechappie":3gpi063v said:
The straightedge may have a slight bow in it's length. It is probably at it's most accurate stood upright on it's edge, not tipped over.....
Yes but how would you know? A flat bendy straight edge is no use except when laid down flat - for a straight line across a surface, not to check if the surface itself is flat. For that you need a rigid box construction like a spirit level, though for short lengths say 12" or less a ruler (combi square etc) may be good enough on edge
 
Thank you, I might try the three part method or buy myself buy a new one that doesnt hae the thin edge.

I have laid it on my bandsaw table which is largish bs400 table, doing so tells me there is quite a significant bow. That's presuming my table is flat but there's no way it's bent as much as i'm able to push the straight edge down the remove the bow on one side.

This straight edge is a maun one which was around £50 quite old now and I have actually dropped it once or twice. I'm thinking if I buy a new one i'll make a wooden straight edge and just use the metal one for checking and setting planer blades.
 
Could I also ask peoples opinion of beveled vs flat edge? Beveled would seem more useful but flat can balance on its own of course. Are there any other differences to consider?

Presuly you could just tilt a flat edge onto its edge and get the same result as the beveled edge?
 
Tetsuaiga":piwisecw said:
Could I also ask peoples opinion of beveled vs flat edge? Beveled would seem more useful but flat can balance on its own of course. Are there any other differences to consider?

Presuly you could just tilt a flat edge onto its edge and get the same result as the beveled edge?
For marking out and cutting a bevel edge helps. For checking for flatness you need a box section - a thin edge is no good
 
Thanks Jabob. What makes a beveled edge not good for flatness? I thought they were recommended for checking plane soles?
 
Tetsuaiga":36cfq428 said:
Could I also ask peoples opinion of beveled vs flat edge? Beveled would seem more useful but flat can balance on its own of course. Are there any other differences to consider?

Presuly you could just tilt a flat edge onto its edge and get the same result as the beveled edge?

My Rabone straight edge has both and find I use the beveled edge the most. It's just more precise and light is more easily seen under the thin edge. It's vital you you hold it at 90 degrees to the work or it will just sag to the shape of what you are testing.
 
[/quote]
For marking out and cutting a bevel edge helps. For checking for flatness you need a box section - a thin edge is no good[/quote]

Used a bevel edge for years to check for flatness with no problems. Just have to hold it it at 90 degrees to the work.
 
Tetsuaiga":10aw6jeb said:
Thanks Jabob. What makes a beveled edge not good for flatness? I thought they were recommended for checking plane soles?
Bevel edge usually means a thin one which might get bent. If it's at all flexible then it may be no good for checking surfaces but perfectly OK for cutting/marking lines across surfaces. OK if it's thick. A box shape is good cos you can tilt it to sight just the edge.
 
Tetsuaiga":2ohly58x said:
This is a piece of douglas fir I planed, my straight edge is the type that has a tapered edge. As you can see as I tilt it back or forward it lets light through either in the middle or the edges, does that mean my straight edge is bent?

Yes! But only in a plane perpendicular to the specified straight edge.

In the way you're meant to use it, it's fine.

Straight edges with fine edges are easier to read - some high precision tool makers' straight edges are actually knife edged for this reason. Straight edges where the edges have actually width/thickness can mislead.

Here's more info:

http://www.luthraprecision.com/straight-edges.html

BugBear
 
bugbear":3bo9qsf2 said:
Tetsuaiga":3bo9qsf2 said:
This is a piece of douglas fir I planed, my straight edge is the type that has a tapered edge. As you can see as I tilt it back or forward it lets light through either in the middle or the edges, does that mean my straight edge is bent?

Yes! But only in a plane perpendicular to the specified straight edge.

In the way you're meant to use it, it's fine.
You are meant to use a flexible straight edge flat, not on edge. On edge - even if no light shows there no way of knowing whether it's sitting on a curved or a flat surface.
.... Straight edges where the edges have actually width/thickness can mislead.....
You can tilt a rigid box section like a spirit level to make a 90º "knife" edge and be sure that it's straight.
 
Jacob":2s4ccj4z said:
You can tilt a rigid box section like a spirit level to make a 90º "knife" edge and be sure that it's straight.

Only if both faces leading to the edge are both machined to tolerance.

Spirit levels have flat edge faces (obviously), but the main (large) face isn't worked to the same precision, if at all.

Perfect knife edge being the intersection of two perfect planes and all that. :D

BugBear
 
bugbear":m2zhvrm8 said:
Jacob":m2zhvrm8 said:
You can tilt a rigid box section like a spirit level to make a 90º "knife" edge and be sure that it's straight.

Only if both faces leading to the edge are both machined to tolerance.

Spirit levels have flat edge faces (obviously), but the main (large) face isn't worked to the same precision, if at all.
They are on better quality ones with forged or cast bodies. Precise enough for woodwork purposes anyway
Perfect knife edge being the intersection of two perfect planes and all that. :D

BugBear
Until you put a bend in it
 
I would put the straight edge to one side for now, a sure route to madness.

Try planing three pieces of timber say 3" x 1" on your machine, flip them round and check the edges and faces against each other. This is a reliable way to make a wooden straight edge. If you do not have any serious gaps and the wood looks straight by eye that is good enough for woodwork. It is also a reliable way to find inaccuracies in your machine. For example my Wadkin RZ has very accurate tables, so the true face of the 3" x "2 is dead flat, but the fence is worn at the bottom on the infeed side so the true edge comes out ever so slightly twisted. Don't forget that timber can move as you plane when you release tension from the board in the same way as it does through the rip saw, just not as obviously. If you're confident that your machine and technique is accurate, yet your timber is not being straightened when machining, it it sometimes good practice to flip the board and try and break the tension by machining the other side. Technique also comes into play, too much pressure on the timber simply means it will spring back to where it was rather than taking out the high spots.

What machine are you using? A modern overhand planing machine with decent capacity of comparable quality to a British maker will set you back around £10,000. Those small fractions can be very very hard to find on a machine, most of the time impossible. A jointer plane is probably the easiest and most cost effective way of getting it, a lie nielsen no8 will set you back £360. The machine has done all the laborious work for you so a few passes of a hand plane to get it accurate is no big deal. I'm saving up for a Martin over/under planer at the moment, a Record No8 is keeping pretty much keeping me in business :lol:
 
Thanks james, i do use my no7 quite a bit but like to check things with the straight edge even then. Ill have to do the test you suggested.


Im currently sorting out some wooden parts for a 4 way panel clamp for a workbench im making. Hence why im trying to get the most flatness. When putting an edge on the laminated top strips i used the no7 after running through bandsaw.

My planer is an old startright inca planer without thicknesser says no 2 on its base and its only 75cm across both tablds. Cant remember what the proper mode name is.
 
A straight edge is just that, and the bevel is to show the "straight" side or edge
I still have my original "werkrite" and 3"x1-1/2" box from my plastering and drylining days, as Jacob says perfect on the Knife edge.
Regards Rodders
 
Back
Top