How to know if wooden beams are completely straight?
- This topic has 15 replies, 6 voices, and was last updated 6 months, 2 weeks ago by
.
-
Posts
-
Hi,
I have a project that involves building some beams with a section of about 50mmx50mm, with lengths varying between 2m and 2.40m.
My main question is how to know that the faces of the beams are completely straight, as I only use hand tools.
I don’t have a surface planer, and I didn’t feel like buying one (even a small one) just for this project…
How can I know that the 2.4m beams are completely straight, using only hand tools, without any belly or hollow?
The only straight edge I have is from Veritas, but has only 60cm length.
I will need to drill the beams to fit some bars between the beams, so it is very important that the beams are really straight, otherwise the bars will not be straight.
Any thoughts?Thanks
Hi Antonio, you could try this technique that Paul teaches:
https://www.youtube.com/watch?v=A_y0PJxPA_U&ab_channel=PaulSellers
P.S. – First time poster, but I’ve been lurking for many years. 🙂 The input of this community has been invaluable!
I was going to point to the same video.
But with a comment.
For really long boards, one should use the method laterally (the string in front of the edge to be tested and not above it {turn first the board a quarter turn of course} ) in such a way that the string sagging (if any) doesn’t give false indication.
Thanks, but I don’t think I understand the reasoning.
What is the intention of putting the string in front, and not above?
What’s the real intention?
How can I tell if the beam is straight by putting a string in front?
I’m not understanding, I’m sorry.And is this method really effective for testing whether a board is straight, putting the string in front or above?
Because it only takes a little bit of sagging in the string (even a little) for it not to work.
I think it is difficult to stretch the string well so that there is no sagging, no?Make the same thing as shown on the video and then turn the whole board/string assembly a quarter of a turn in such a way that the string is in front of the board.
That way if the string is sagging a little bit, it will not influence the measurement of the space between the board and the string.Instead of ordinary string, use dental floss for Paul’s method. Floss is remarkably strong and can be pulled piano-string tight with no measurable sag across tens of feet.
Tradesmen who hang suspended ceilings use an orange-colored version to get room sized ceilings as flat as with a laser. ( laser beams are much thicker than the floss). There are about 50 meters of floss in a small dispenser, so you won’t run out any time soon.Or make your own straightedge of any length you want.
Pick out a board that’s quartersawn or even riffsawn and plane one edge as striaight as you can by sighting down it as a starting point. Then use your prepped edge to draw a line with a fine tipped pencil. Spin the straightedge 180° and compare the edge to the line you drew. That will show twice the error in your board. Tune the edge and repeat and test until line and edge agree. This will get you within a couple thousandths of an inch pretty easily as long as you are even a little fussy.For accuracy as good as a machinists edge, use three boards to compare them to each other. If they all agree when you test them against each other, all three edges must be straight. . ( with just two edges to compare, you might have complementary curvedi edges)
When you are done you will have three perfectly straight edges, or one straightedge and two boards with perfect edges on them, or at least as straight as you are fussy. It might take a while. This method is how machinists proof their metal straight edges.Then make your new proofed board look like a tool.
You can dress the board up with a curved back edge that tells you it’s not the straight edge you proofed, and cut fancy ends so the board doesn’t look like a piece of scrap and get cut up for shims or something. A coat of finish will slow wood movement, but you can always repeat the proofing process if you suspect your new tool has gone out of true. Finally, drill a hole in one end so you can hang your new tool vertically on a nail in your shop off the floor. It will likely stay true for years.I have a 7’ straightedge that’s maybe 30 years old and still true when I test it.
Finally, for a quick and dirty test for straightness, I’ve found that the finished edge on melamine shelving from the box store is remarkably true in most cases. If you have a piece in your closet, use that as a test edge to get you started. If I was in the field I’d test my edge against a customer’s shelf or a granite counter edge to see if I needed to tune my edges. Granite is often cut with CNC water jet cutters to very high tolerances.
Or just try the string method and observe what the string shows. Sometimes you just have to do something to understand it.
Thanks, but I don’t think I understand the reasoning.
What is the intention of putting the string in front, and not above?
What’s the real intention?
How can I tell if the beam is straight by putting a string in front?
I’m not understanding, I’m sorry.And is this method really effective for testing whether a board is straight, putting the string in front or above?
Because it only takes a little bit of sagging in the string (even a little) for it not to work.
I think it is difficult to stretch the string well so that there is no sagging, no?Now I get it!
Smart!
Instead of ordinary string, use dental floss for Paul’s method. Floss is remarkably strong and can be pulled piano-string tight with no measurable sag across tens of feet.
Tradesmen who hang suspended ceilings use an orange-colored version to get room sized ceilings as flat as with a laser. ( laser beams are much thicker than the floss). There are about 50 meters of floss in a small dispenser, so you won’t run out any time soon.Or make your own straightedge of any length you want.
Pick out a board that’s quartersawn or even riffsawn and plane one edge as striaight as you can by sighting down it as a starting point. Then use your prepped edge to draw a line with a fine tipped pencil. Spin the straightedge 180° and compare the edge to the line you drew. That will show twice the error in your board. Tune the edge and repeat and test until line and edge agree. This will get you within a couple thousandths of an inch pretty easily as long as you are even a little fussy.For accuracy as good as a machinists edge, use three boards to compare them to each other. If they all agree when you test them against each other, all three edges must be straight. . ( with just two edges to compare, you might have complementary curvedi edges)
When you are done you will have three perfectly straight edges, or one straightedge and two boards with perfect edges on them, or at least as straight as you are fussy. It might take a while. This method is how machinists proof their metal straight edges.Then make your new proofed board look like a tool.
You can dress the board up with a curved back edge that tells you it’s not the straight edge you proofed, and cut fancy ends so the board doesn’t look like a piece of scrap and get cut up for shims or something. A coat of finish will slow wood movement, but you can always repeat the proofing process if you suspect your new tool has gone out of true. Finally, drill a hole in one end so you can hang your new tool vertically on a nail in your shop off the floor. It will likely stay true for years.I have a 7’ straightedge that’s maybe 30 years old and still true when I test it.
Finally, for a quick and dirty test for straightness, I’ve found that the finished edge on melamine shelving from the box store is remarkably true in most cases. If you have a piece in your closet, use that as a test edge to get you started. If I was in the field I’d test my edge against a customer’s shelf or a granite counter edge to see if I needed to tune my edges. Granite is often cut with CNC water jet cutters to very high tolerances.
Wow!
Thank you for your inputs, extremely valuable!
I was checking the Dining Table project, episodes #2 and #3.
He used the string method only for some beams, but not for the faces of the boards.
Why didn’t Paul use the string method to check the straightness of board faces?
How does he know that the board faces are completely straight, since the winding sticks are only for checking the twist?One thing to think about is, does it matter? If we are laminating two boards together, edge to edge, then it matters that they are straight along their entire lengths. If we are making a box with dovetails at the corners and nothing in the middle, then the only thing we need is for the ends of the boards to be out of twist relative to each other. It doesn’t matter what happens in the middle as long as it looks good. Those are two examples of “does it matter.” Also, the hand-tool approach of using a face side and face edge can help to reduce the need for perfectly true material.
That said, remember that planes tend to make things uniform. So, if the ends are out of twist relative to each other and a plane is taking uniform shavings, then everything in between has a good chance (but no guarantee) of being in plane with the two twist-free ends.
It has been a long time since I saw the table projects, but things to think about: If you are laminating boards, you may make the edges perfect, get the faces close, and then glue up knowing that you will need to level the entire glued-up assembly. So, it is a waste of time to make things perfect before hand. Also, with some tables, you will be attaching the completed top to a frame (aprons), turnbuckles will pull the top down onto the apron to keep it from cupping, and then you will plane the top surface to look good. This goes back to the “does it matter” part of this reply. Sometimes, perfecting things just doesn’t matter. Sometimes, a step later in the project will take care of things. When using hand tools, it is important to not use machine methods executed with hand tools. It can waste a lot of time.
Thanks for the input, Ed.
That makes a lot of sense.One thing I don’t quite understand very well is if I have to use string method to check all 4 faces of the beam, or if it’s necessary to check only the 1st face.
Can you please tell me a time point and video name (project, episode number) that is related to what you are asking about?
In the recent Dining Table project, episode 2, Paul uses the string method to check a long board for straightness.
He also demonstrates how to use it in the youtube vídeo that @LiquidWood mentioned in the post up above.
Imagine that it was a board with all 4 faces with the same width, say, 2″x2″.
Is it necessary to check all the faces of the beam for straightness, using the string method?
Or just the first face, and then square the other 3?I believe Paul is preparing stock for the table top at that point in that video, just a few minutes at the end. The edge that he tests with the string must be straight to get a good, gap free lamination, so he is using a lot of care with it. Even if he wanted a slight curve (some people do that when laminating), he’d probably want to start from perfectly straight. So, this is why he is taking such care. The wide faces don’t matter too much. After gluing the stock together, the whole assembly will be planed. He’s using a fair amount of care with those surfaces, but he really can’t get in trouble with them. You might ask, what about the other edge of the board? Do you need a string there? It depends upon the worker. Paul can probably pull a gauge using the trued edge as a reference and plane to the gauge line without checking. Others may need / choose to check.
You asked about a 2×2. The answer is, it depends and it’s the job of the worker to decide. If the wood is for legs and all of the joinery is at the top for the apron, then you just need to square two faces and work relative to them for the joinery. Everything else is just “looks good, is good.” If the stock was for something that would receive a through-tenon, then it might pay to take more care with all of the faces so that you can transfer the mortise from one side to the opposite.
As you get more experience and think about the “why” of things, you’ll get a better feeling for what is necessary. Often, once you get one face true, it can be a reference for truing / testing the others. I’ve never used the string method and rely on the action of the plane plus a straight edge, but if I were doing Paul’s table, I might use the string.
Hope this helped.
- You must be logged in to reply to this topic.