what to do with the straight edges

leblondboy · Jan 13, 2014

I made 3 straight edge, 24"x1 1/4" since I was fascinated about how it's made. So, the next question is what do I do with it? Should I use it to make a flat surface? I don't know. What would a hobbyist use this thing for? I am thinking of use 1 of it to make a DIY precision level.

lowCountryCamo · Jan 13, 2014

Is it april 1st already?

allloutmx · Jan 13, 2014

How straight is it? How did you measure?

Mark Rand · Jan 13, 2014

If you made three by comparison against each other, what precautions have you made to measure/eliminate twist? (a precision level will work).

Regards
Mark

leblondboy · Jan 13, 2014

It's hard to know how straight it is, but but comparing the 3, I suppose they should be very straight, and I can always make it "straighter" by putting more time into it.
Yes, I compared 3 against each other, and also rotate each to compare. That's a lot of comparison. They should produce a flat surface, which means also not twist (twist is not flat).
A precision level would work, but I think comparing them would provide a good indication of amount of ink transfer. The more ink dots per inch, the higher precision. If I keep doing it, I suppose it can be super precised.

leblondboy · Jan 13, 2014

lowCountryCamo said:
Is it april 1st already?

hahha, why it's so hard to believe such a thing that most people in this forum would make for themselves? About what to do with them? Well, you'd be surprised that most tools are not used, and some people just happen to work in a field that they use them often, and need to buy/make a new edge often.

Mark Rand · Jan 13, 2014

leblondboy said:
It's hard to know how straight it is, but but comparing the 3, I suppose they should be very straight, and I can always make it "straighter" by putting more time into it.
Yes, I compared 3 against each other, and also rotate each to compare. That's a lot of comparison. They should produce a flat surface, which means also not twist (twist is not flat).
A precision level would work, but I think comparing them would provide a good indication of amount of ink transfer. The more ink dots per inch, the higher precision. If I keep doing it, I suppose it can be super precised.

You can only verify a lack of twist by comparing the surfaces at an angle (90° 60° etc. not 180°) to each other. or using an external reference, like a level or a larger surface plate. If they're straight edges and are all twisted the same amount, they'll still match each other when tested in-line :eek:

. Been there, done that, had to re-do the work...

For the straightness, working the three against each other will result in perfection, but with straight edges you also need the 'third party' check for twist. If you were producing three square or round surface plates, you could check for twist by rotating them, but that's a bit difficult with a straight edge.

leblondboy · Jan 14, 2014

Mark Rand said:
You can only verify a lack of twist by comparing the surfaces at an angle (90° 60° etc. not 180°) to each other. or using an external reference, like a level or a larger surface plate. If they're straight edges and are all twisted the same amount, they'll still match each other when tested in-line . Been there, done that, had to re-do the work...

For the straightness, working the three against each other will result in perfection, but with straight edges you also need the 'third party' check for twist. If you were producing three square or round surface plates, you could check for twist by rotating them, but that's a bit difficult with a straight edge.

With all respect to your experience, by no mean to question that.

However, I just don't get the part that 3 flat (even narrow) pieces would have "twist" issue. There are two things to this. 1st, theoretical, I don't think it's possible to have 3 twisted pieces matching each other. There 2nd, on the practical issue, is it possible that due to the method, it can be twisted, but not detected due to the narrow surface?

I read your post several times, and it seems to be the case of the 2nd scenario. However, even just 0° (no rotating), matching 3 pieces together would indicate non-flat areas.

Here's a reference of why that is the case:

Making Accurate Straight-Edges from Scratch

Mtndew · Jan 14, 2014

leblondboy said:
1st, theoretical, I don't think it's possible to have 3 twisted pieces matching each other.

First rule of thumb in this line of work... Murphy's Law.
Second rule is.. never assume.
After that it's all downhill

leblondboy · Jan 14, 2014

Actually, you did mention of straight, versus flat surface. It is correct that a long pieces can have all horizontal lines on a surface straight, but that surface doesn't have those lines in a same flat surface. In this scenario, by rotating 180 degree, the short lines (spanning the narrow edge) would be the one to be tested, and they would match against both sides of the other pieces. This forces them to be straighten up. However, if it's very narrow, it can easily cause the matching process to fail, due to minor tilt that is not detectable by hands. So, this is probably what you were referring to.

The surfaces I made has 1 and 5/8" width, so that's probably very wide to have the issue you noted. I think if it's less than 3/8", then it would be a problem.

Screwmachine · Jan 14, 2014

What is the form of your straightedges? Are they just straight bars (if so how thick?), or do they have the curved back for stiffness? If they are straight bars, even if they are an inch (or two) thick you will be surprised at how much they can deflect under even very moderate pressure- certainly including gravity.

leblondboy · Jan 14, 2014

leblondboy said:
Actually, you did mention of straight, versus flat surface. It is correct that a long pieces can have all horizontal lines on a surface straight, but that surface doesn't have those lines in a same flat surface. In this scenario, by rotating 180 degree, the short lines (spanning the narrow edge) would be the one to be tested, and they would match against both sides of the other pieces. This forces them to be straighten up. However, if it's very narrow, it can easily cause the matching process to fail, due to minor tilt that is not detectable by hands. So, this is probably what you were referring to.

The surfaces I made has 1 and 5/8" width, so that's probably very wide to have the issue you noted. I think if it's less than 3/8", then it would be a problem.

One more scenario that maches exactly what you said. Think of each piece looks like two perfect straight edges put together side by side. However, only the middle lines up, while the other ends are off (twisted?). So all 3 are like that would match by putting on on top of the other, or turn 180 degree. However, if you slide them against each other along the length, they will not be straight, and the bumps will show up.

leblondboy · Jan 14, 2014

Screwmachine said:
What is the form of your straightedges? Are they just straight bars (if so how thick?), or do they have the curved back for stiffness? If they are straight bars, even if they are an inch (or two) thick you will be surprised at how much they can deflect under even very moderate pressure- certainly including gravity.

It is 3 inches height (thick). It's a C shape beam, where the straight edge is at the bottom of the C. The vertical is 3 inches.

leblondboy · Jan 14, 2014

Mtndew said:
First rule of thumb in this line of work... Murphy's Law.
Second rule is.. never assume.
After that it's all downhill

You don't assume, you check your work against each other. Even with Murphy's law, air planes drop often, but people do successfully travel from one place to another using air planes. To know if they successful or not, ask some people who flights before, or do it yourself. Things fail once awhile doesn't me you should give up. Only the successful that counts

Screwmachine · Jan 14, 2014

If you develop a twist during scraping, it will be on all 3. You can slide the two you are marking up all you want, it will print up like a dream and you think you have perfect straight edges. Get a cheap granite plate from Enco and check out your work.

Sounds like they are well formed, and glad to see someone interested enough in scraping to fabricate their own straight edges- please post some pics!

leblondboy · Jan 14, 2014

Screwmachine said:
If you develop a twist during scraping, it will be on all 3. You can slide the two you are marking up all you want, it will print up like a dream and you think you have perfect straight edges. Get a cheap granite plate from Enco and check out your work.

Sounds like they are well formed, and glad to see someone interested enough in scraping to fabricate their own straight edges- please post some pics!

View attachment 96671

Yes, sliding will color all areas that touches. However, the way I do it is to put lapping compound and then actually grinding against each other to remove the high point. I understand lapping compound does have its limit to the precision. So if I need better precision, I will need a correct "reference" to scrap it better. Also, do you think matching at half width would show up the twist?

leblondboy · Jan 14, 2014

Just to make it easy to imagine, take an extreme case of twist of a bolt and nut, where they slide effortlessly against each other, the nut's inner surface actually is the same of the bolts, but it's also a compliment versus exact, meaning taking 3 pieces to match, you will either have 2 bolts and 1 nut or 2 nuts and 1 bolt, and you know two of them would not slide against each other.

For the two concave surfaces (bolts) will roll on each other, which appears to match, but holding 1 end together, the other side will be apart (hence rolling).
For the two convex surfaces (nuts), they will not mach in the middle.

I think for the kind of twist you're talking about is more like a DNA twist. I think you're right.

Yes, you guys are correct!

Without a true reference, I probably angle it also, to have diagonal match. I think that will show up any twist. Probably not doable if it's 3/8" or less.

hickstick_10 · Jan 14, 2014

Screwmachine · Jan 14, 2014

Trying to mark up something that narrow along half its width would be a futile exercise in my opinion.

Lapping is not the way precision metal straight edges are made. I won't say it can't work, but you will honestly have an easier time with a better outcome scraping. Lapping is great for very hard materials like 60RC steel or granite, not good for soft stuff as you will inevitably charge your piece and it becomes a lap for whatever it touches afterwards. And with scraping, you have an extreme amount of control regarding where and how much material you remove.

David Utidjian · Jan 14, 2014

I think that comparing long rectangular surfaces has this problem because one is only switching them end-to-end (180*) instead of also comparing them at 90*. Seems the three-surface-method would work best with square, nearly square, or round surfaces.

-DU-

what to do with the straight edges

Plastic

Stainless

Titanium

Diamond

Plastic

Plastic

Diamond

Plastic

Diamond

Plastic

Titanium

Plastic

Plastic

Plastic

Titanium

Plastic

Plastic

Stainless

Titanium

Titanium

Similar threads