2 wires, a ball bearing and a buzzer

[Eric Salop]

About a year ago I made myself a bigger welding table and one of the things that got me scratching my head was how to ensure the top was flat. YouTube was a good source of inspiration at the time and as usual, ever since it keeps throwing up suggestions for things it thinks I should watch. Most of the time I ignore them, although this one caught my interest when I came in for a warm up and a coffee this afternoon and after watching both parts, I thought it worthy of a mention here as I know many of you appreciate a bit of good engineering that is well presented and may not have seen this.

The first part has a lot of intro (I must confess I mostly skipped through the section where he was just measuring the flatness of a commercial welding table), then shows a high quality welding table build How to build a flat welding table. Part 1 , Plans Available Fireball tool - YouTube

The second part is what really wowed me. It shows a simple technique for checking/adjusting for flatness that could be applied to any metal surface, using just 2 lengths of taught music wire, a ball bearing and a buzzer - something I have never come across before. There is also a good explanation at the end about the thought he put into reducing human error in marking out a flat plane How to build a flat welding table. Part 2 , Plans Available Fireball tool - YouTube

 

[tomplum]

He does some good videos. I've seen his stuff before. And it isn't like he's way out there- just creative in a different way.

 

[PJSprog]

Interesting, Eric. Of course, it presumes the flatness of the bench he's moving the clamped bar along to be accurate. It is an interesting method to establish a plane, to be sure.

 

[RedNeckGeek]

Lots of good fab tips (and plugs for his various squares), but I'm not sure how practical the method is in anything but a well equipped industrial shop. A saw horse will move as soon as you try to apply tension to the music wire, as will a ladder. And unless the floor is very level, there's going to be errors because the fixed string will reflect any un-levelness in the floor. Unless you can figure out a way to use a level on the wire, and a very good level at that. But then as soon as you touch the wire with tle level, you will change the lay of the wire.

If all you're trying to do is get the plates on the table into the same plane, I'd start by leveling a center plate, then use a straight edge and shims to adjust the surrounding plates to that first plate. If you have three 1-2-3 blocks (two under the straight edge on the center plate, the third on the plate being leveled), they would be easier to use than shims. Yeah, it would take some time, but it would eliminate all the monkey motion and Rube Goldberg stuff with the wire, ball bearing, and beeper.

Here's a link to a 50" aluminum straight edge, but it's only straight to 0.005" over that 50" span:
5" Anodized Aluminum Straight Edge Guaranteed Straight to Within .5" Over Full 5" Length SE5: Amazon.com: Industrial & Scientific

 

[Eric Salop]

Sag on the wires was my first concern when working to a few thou, however looking at some other sites suggests gravity effects on a thin taught wire are insignificant over a short distance. I agree his setup looks a little awkward, but the principle is there to be applied to whatever anyone's imagination can come up with.

With my limited welding skills, a table that flat would be wasted on me. Once I had straight edges on my table, I got my slatted top near enough into a single plane using thin fishing line pulled tight across each diagonal, raising and lowering a corner until they only just touched one another.

 

[Kyle_in_Tex]

And, after making all your welds on your fab, when you unclamp it from the table, it just "twangs" to it's new natural equilibrium.

But yes, starting with a reasonably flat table is better than nothing.

 

[RedNeckGeek]

And, after making all your welds on your fab, when you unclamp it from the table, it just "twangs" to it's new natural equilibrium.

If the weldment is restrained by being clamped to the table, the hot material around the weld and the weld itself will yield as it cools to room temperature and there will be less distortion in the resulting assembly. Piening between passes can reduce distortion even more if the assembly is restrained.

 

[Eric Salop]

Piening between passes can reduce distortion even more if the assembly is restrained.

Thanks RNG, I used to feel real bad when slag didn't peel away as it cooled and I hack at it with a hammer, now I can tell myself I'm just minimising distortion. :laughing:

 

[RedNeckGeek]

:laughing::laughing::laughing:

 

[Dadnatron]

Interesting, Eric. Of course, it presumes the flatness of the bench he's moving the clamped bar along to be accurate. It is an interesting method to establish a plane, to be sure.

Actually, it doesn't. He is using the second transverse WIRE as his point. The other end can be off any amount, he brings the wire back to plane against the other wire. The wire will be 'flat' given enough tension between its points. And if the wire is essentially 'level' or a close approximation, the table will also be level.

The most important thing is that the table is flat compared to itself. Not compared to the poured concrete floor.