When the surface plate is just too small ..

[graham-xrf]

The granite 18 x 18 in inches. The diagonal is about 25.4 inches. The need is for a straightedge 36 inches long. A good granite that size costs a bundle, and takes up lots of room in a small shop. Also, I already have the 18 x 18.

I get it that spotting 2/3 of the length is likely to get messed up by the overhang, and it gets mad confusing if one then spots it by shifting the work along to get at the 2/3 from the other end. What if there was a high spot near the overhang end? That would produce a whole new spotting pattern over the 2/3 already done.

I thought there may be a strategy, but also perhaps that this problem already has a well proven strategy that I just don't know about yet.



1. Null the overhang.
I am thinking to eliminate the overhang bow effect using something like in the sketch. The jack is set to carry half the weight of the overhang. This could at least allow a spotting pattern on the tested 2/3, but not any hinging moves.

2. Getting at the overhang end
The next move is to turn the work around to have the overhang now included in the 2/3, but resolutely ignore all results other than those from the previous overhang, plus a very little extra. Work only on what used to be the overhang. If there was a high spot, it will be seen. If there was a local low, then that too will be seen. This can inform a strategy on what to do about the initial 2/3, and how to bring it all into a common line.

3. Getting a single spotting result over the whole work
Suppose the condition of the overhang was (extreme case) all a bit low, needing the entire initial 2/3 to be scraped down to meet it. This too might be done by looking for a spotting pattern that has a middle third that repeats, even though it now has the initial overhang shared over the plate. I even thought one may be able to check with a 2-colour spotting, to see if one overlays the other in the middle third.

All of the above is just my first attempt at a plan. If experienced folk here say it is doomed, then plan B may be to start saving up for a bigger granite. Of course, it may be that a load cell and all that stuff is just not needed, and there is already a well know work-around. I sure hope so !

 

[C-Bag]

Plan B.

Or Plan C, save up for a nice 36" B&S camel back like I did. But I have a 18x24 granite.

 

[graham-xrf]

Plan B.

Or Plan C, save up for a nice 36" B&S camel back like I did. But I have a 18x24 granite.

It had crossed my mind, but the only one I saw was not only £££, it also looked like it needed to be put onto a surface plate, and given some TLC. I do keep an eye out! You got me looking for a B&S outlet in UK

Regardless, it would be nice to know if one can spot a straight edge in overlapping sections, and still get a sensible result

 

[C-Bag]

Regardless, it would be nice to know if one can spot a straight edge in overlapping sections, and still get a sensible result

Connolly specifically warns against it many times, over and over as it's a waste of time and a rooky mistake.

 

[graham-xrf]

Connolly specifically warns against it many times, over and over as it's a waste of time and a rooky mistake.

OK - I hear it. Connelly maybe did not envisage going the way of load cells, etc. I am in a field where I just don't know, and I mean to get to know some before I expend on discovering what others have already exhausted themselves in proving. When I read the accumulated wisdom, I do soak it up, but I am not a slave to it! There is a logic to observing that a reference flat surface has a line in it's plane that aims through space to continue it's reference some distance beyond it. We use the same principle to enable making gunsights, and theodolites. Seeking a (possibly new) practical strategy is surely OK.

With all respect to a fellow member, I quote @Richard King 2 from 2019 ..
"Remember Connelly was not a scraper, he was a organizer of thoughts from several craftsmen, much of that book is wrong in my opinion".

It occurs to me that where one should scrape - or not, might yield to something like illuminating the work in monochromatic light, or laser raster, in an interferometer comparison to an optical flat, to display the deviations. Perhaps some third technique. I keep having all these sideways thoughts, but mostly I dismiss them on good practical or cost grounds. I think the present best route is keep looking for a bigger granite in good condition, and get either a blank, or a used SE with potential, and gain some of the skills that @Richard King 2 teaches right here. This, I guess, might be sometimes contrary to Connelly, but that's OK. I may be rookie, but you are witnessing me not blindly blundering into a mistake.

Edit: [ Looking about, I find £340 + £30 P&P gets a somewhat used scratched and dinged 36" SE with some rust pitting. A little 12" raw unmachined castiing is £150. That last one is maybe reasonable, but the present need is for 36" or 48". I pass over anything bigger because they are just too heavy and unwieldy]

 

[rwm]

I understand why large surfaces and surfaces on machines get scraped. For a straight edge, why would you not just have it surface ground it? Or are you specifically trying to do this yourself and learn scraping?

 

[C-Bag]

With all respect to a fellow member, I quote @Richard King 2 from 2019 ..
"Remember Connelly was not a scraper, he was a organizer of thoughts from several craftsmen, much of that book is wrong in my opinion".

And with all due respect to you too, RK2 has referred to Connelly a lot and is still his selling book. I know from my time as a mechanic you ask 10 of them how to do something, and you’ll get 11 answers.

So I refer to what makes sense not “historical engineering“(we’ve always done it this way). The most used and crucial check for me is hinging the work. And Connelly said in the passages I remember that having more than a couple of inches sticking out on both ends of your reference is going to make it really sketchy to hinge anything properly.

I get it’s really easy to come up with an alt solution and go with it. But I understand this one parameter and without it I see no way to properly do a very basic check that will keep from down the wrong track. Which once you get off just goes further into the weeds.

But I’m certainly nobody and not an expert so shouldn’t have responded in the first place.

 

[rabler]

With load fee etc, I was able to acquire a 3'x4' surface plate for less than US $200. I'm sure I'll spend more than twice that having it calibrated. I think there is an aspect of intellectual pursuit to your method, rather than just purely pragmatic progress.

 

[Weldingrod1]

Cast iron levels are pretty cheap to buy. You could do the three plate method with them and originate some straights...

Sent from my SM-G892A using Tapatalk

 

[graham-xrf]

I understand why large surfaces and surfaces on machines get scraped. For a straight edge, why would you not just have it surface ground it? Or are you specifically trying to do this yourself and learn scraping?

I want to fix up the ways of a lathe, perhaps eventually on both, and I have nothing to lose in attempting some scraping. Of course, I do not intend to have the lathe ways be the very first thing I scrape on, and up to now, I have only scraped on some small things. A straight edge capable of being useful for the later scraping seemed logical, but I am open to a short cut if I come across a reasonably affordable one.

I do have a "straight edge", which is too short, but is of the type used with feeler gauges across engine blocks. About 5mm thick, 600mm long, with a beveled edge down to about 2mm. So far, it has only been used to rule lines with a pencil for cuts on woodwork!