Measurement tools?

[CPDDET]

Im wondering what tools and techniques others are using when taking measurement from ship plans.

 

I'm slowly, very slowly, moving toward scratch building and finding taking accurate measurements from ships plans a bit clumsy. For instance: I will be making my own deck pumps for Bluenose, turing them from brass stock on a Sherline lathe. This is a small part and the lines on the plans can be quite thick. Do I measure from the outside of the line? The inside on the line? The center of the line? What tools are others using to make the very fine measurements accurately? 

 

Thanks in advance for any advice passed along.

 

Dave

[BETAQDAVE]

I think the center of the line would be the way to go.  I use a pair of drafting dividers with a center wheel so it won't loose the setting.

[CPDDET]

Thanks betaqdave. That will work when transferring directly to the workpiece.

 

But how about when that divider distance needs to be translated to a decimal measurement? Using the lathe I need to dial in the measurement so I need the measurement from the plan in decimal form.

 

Dave

[druxey]

Do you have decimal reading calipers? That would be the most direct way.

[CPDDET]

Druxey, that's what I have been using but find it a bit clumsy in some situations. Thought there might be a better method I wasn't aware of.

 

Too bad they don't make these in a smaller size.

 

D

[BANYAN]

D, what I do is use the calipers or 'calibrated' ruler (one that has been checked and certified accurate to a 'poomteenth' - sorry for the techo jibber jabber)   and mark the closest full centimiter on the plan, etc then I have this viewing/magnifying glass that has a little calibrated pointer and scale on the bottom that allows me to wind the needle/pointer to exactly where I need it.  That gives me a very accurate measurement overall.  I don't have an issue transfering this to wood etc with a good caliper (I use Mitutoyo -Japanese) calipers which are very accurate).

I'll try an remember to take a photo and get more info on that viewer/magnifier for you.

 

cheers

 

Pat

[CPDDET]

Thanks Pat, I will in interested in seeing it.

D

[BANYAN]

No problem, I also forgot to mention, I don't mind working from any part of the lines, as long as I remember to use the same point consistently.  I'll try an dget that photo later today.

 

cheers

 

Pat

[BANYAN]

Hi again D.  As promised here are the photos which may give you an idea what to look for.  Unfortunately, there is no brand or even makers name on it at all.  I got this from a mate who bought them in a deal (he buys end of line lots, seconds etc).  The twin LEDs in this one don't work unfortunately, but I really don't need them. 

 

The magnifier (yellow arrow) is very clear, bright and about 4x so I get a great magnification (may be even better as I may be a little underestimating the zoom/mag factor) of the viewing area (green arrow).  I think it is set up in the factory to the scale you want.  Mine is for mm, but there are three other scales around the sides  but I have not tried to change it.  As you can see you twist the knob to wind the pointer/needle (red arrow) with the screw being a very fine thread so lots of control. 

 

The whole 'kaboodle' folds and is stored in a pouch to protect it.  When I next see my mate I will ask if he has any more.

 

cheers

 

Pat

[Bob Cleek]

Hence the once common machinist's magnifiers:

 

[bruce d]

That folding gizmo is called a 'linen tester'. Wonderful tools.

[CPDDET]

Ah, linen tester! Readily available and economically priced. Thanks!

[Nunnehi (Don)]

My favorites.........

 

 

[KeithAug]

16 hours ago, BANYAN said:

Unfortunately, there is no brand or even makers name on it at all

Edited November 24, 2020 by KeithAug

[wefalck]

Paper drawings, as commonly used by shipmodelers, are not very accurate, so taking precise readings may not help you in the end, because errors of several measurements tend to end up. There are couple of strategies to overcome this:

 

- always take measurements working from the largest outside dimensions, working inward and make sure that subsequent measurment add up to the total of the outside measurement first taken

 

- redraw the parts in a 2D CAD or similar program from the readings taken from the paper copy; the result should look like the one on the paper and you can take the precise dimensions off the computer without having to worry about line thickness; you can add dimensions to your drawing as in a normal technical drawing, which is helpful for machining.

 

- a variant of the above is to scan (parts of) the paper drawing and copy this image into your drawing program; in another layer you can trace the outlines of the part you want to make and correct dimensions so that all parts fit together; this is today my usual method.

 

If I am working from paper drawings the old way, I am using an analog vernier caliper. I have very small one that is only 10 cm long and comes very handy for working on small parts.

[Dr PR]

Any measurement taken from a printed drawing is subject to many errors. As mentioned above, line widths make finding dimensions problematic. Some printers do not actually print to scale and add some error to the measurement.

 

However, this isn't always a big problem. Things are not built to random dimensions. Engineers usually work in some units of measurement (inches, feet, millimeters, meters, etc.). For WWII US ships things were designed in feet, inches and fractions of inches (1/2, 1/4, 1/8 and occasionally 1/16 and 1/32). It is a lot simpler if the original units are metric!

 

Engineers usually don't design things in odd fractions, like 1.297 inches. If your measurement from the drawing scales to 1.297 inches (1:1 scale) it's a good bet that it really should be 1.25 inches. If you know the units of measure (inches, mm, etc.) the part was designed in you can correct some of the measurement errors by rounding to the nearest common fraction.

[Bob Cleek]

Adding to Eberhard and Dr.PR's comments, this discussion reminds me of something my late boatbuilding mentor, a older fellow who'd been one of the last to have gone through a traditional trade apprenticeship and had run a boatyard of his own for perhaps fifty years, said to me on the subject: "A house framing carpenter cuts to the nearest quarter inch, a finish carpenter cuts to the nearest thirty-second of an inch, and a boat builder cuts to the nearest boat." His point was that it isn't the measurements that matter, but rather the fit of the piece to the ones next to it, so forget about the dimensions on the plans and pay attention to what you are fitting together.

 

Any sort of ship or boat plans, at least until the advent of CAD, are never absolutely accurate. What they are, really, is simply "scaled plans for drawing full scale plans." You can't draw a scale line fine enough, even at 1:48. Back in the day, they'd draw the lines of a 150' ship on a six or seven foot long piece of drafting vellum and the scale lines drawn would still be so wide if blown up to full scale that you couldn't take accurate measurements from the plans. While at modeling scales, the problem isn't as great, how often do we see plans drawn to 1:48, even? The rule in full size engineering is always that measurements are never to be taken from the drawings, but rather must be taken from the notation of the distance on the drawing. In modeling, we can cheat somewhat, but only if we "build to the boat" and not to the plans. What the pre-CAD draftsmen did was to take up a "table of offsets" from the drawings with dividers and read the distances from scales, knowing that no matter how carefully they placed their divider points on the center of the line, the table of offsets would never be perfectly accurate. Indeed, if the offsets for a 150' ship were accurate to within an inch or two, they were quite good. 

 

The purpose of the table of offsets was to enable the loftsman to loft the patterns for the ship full size. The loftsman takes the table of offsets and the lines drawings and uses these to draw the vessel full size on the lofting floor. When doing so, the loftsman uses battens to spring fair curves, using the offsets as a guide, but the offset points are rarely all on the fair curve sprung with the batten. (There are many tricks to the loftsman's trade. In "fairing the lines" from the draftsman's offsets, the loftsman uses the "diagonals" to test the accuracy of the lofting, for example. Further discussion of this is beyond the scope of this post, but for those interested, Lofting, by Alan Vaitses is highly recommended.) The loftsman trusts the batten, not the draftsman's offset measurements to develop the full size patterns for the shape-defining parts of the ship. Only once in a while, when there are a number of identical vessels to be built, will you get lucky and find that a loftsman has generated a corrected table of offsets from the full size lofting that are "tighter than a gnat's ***." In this case, there will usually be a notation on the table of offsets like "Corrected offsets." or "Offsets as lofted." Otherwise, the offsets will have to be "faired" on the loft floor.

 

The loftsman's full size patterns were usually only those essential to get the vessel "in frame." From there, the "wood butchers" "built to the ship," not to the plans. They'd set up a few basic frames, sometimes as few as as a midship frame at the widest beam and a couple forward and aft of that, plus a stem and transom. Then they'd tack battens sprung across the faces of these frames and the resulting "basket" defined the shape of all the frames in between. In such fashion, a fair hull would be constructed. This is sort of the way planked models used to be built, although once in a while, an author would draw up a full set of frames and publish them for modelers to use, as we see in the old modeling books by Davis and his contemporaries.

 

Today, CAD makes it possible, in theory, at least, to generate far more accurate drawings and it seems modelers are seduced by CAD and then find themselves sucked into believing they have to become micro-machinists using extremely accurate (and expensive) machines with DRO, or even CNC, to turn out parts accurate to .0005 if they want to build a good model, even from a kit, but this isn't so. "If it looks right, it is right." was the old time ship builder's maxim and it serves the modeler as well in miniature as it did the old timers working in full size. The old timers didn't have to worry about cutting each side perfectly square and to exact size when making a box. They just cut half of the sides a bit large and when the box was built, they planed the overhangs on the edges to fit, yielding a perfectly jointed cube. I'm not knocking CNC, for it certainly has it's place. (We wouldn't have IKEA knock-down furniture without it!) For building one-off models, though, the old fashioned measuring tools are more than sufficient and often much less expensive, not to mention a joy to own and even collect. Our goal is to create a compelling impression of reality in miniature. That doesn't always mean NASA-level tolerances in our measurements. (Even at that, John Glenn orbited the earth in a rocket ship designed with slide rules!) Sometimes, even slight deviations from exact scaling, such a a smidgen smaller rigging lines, can actually produce a more compelling impression of reality than perfectly sized ones, and that's when modeling becomes an art and not just a craft. So as the man says, "Don't sweat the small stuff."

 

Edited November 26, 2020 by Bob Cleek

[shipman]

An erudite and practical view of this topic, Bob. Watch YouTube vids of boatbuilders at work. Craftsmen share a lifetime's experience, confirming your comments.

[wefalck]

I have a natural science background and grew up in a natural science-househould. So, the natural thing for me was to make measured drawings, cut pieces according to the dimensions on the drawings and then assemble ... but often this did not work, perhaps due to poor manufacturing tolerances. I learned from watching craftsmen, who tend to cut pieces somewhat oversize and then fit them. Not the most efficient way in a production process, but very effective in an artisanal context - and shipmodel building is an artisanal process.

[Danstream]

@Bob Cleek thanks for your clear and realistic explanation. Things are (or were) exactly as you said also in aircraft construction where I have some background. I smile when I read about confrontational discussions based on measurements taken directly from assembly drawings. As you said, the only reliable measures in drawings are the ones reported in writing.

Thanks and best regards,

Dan.

[Roger Pellett]

Well said Bob!

[CPDDET]

Thanks to all for very informative information. There are several ideas that I will add to my building process.

 

Dave

[lehmann]

There are digital dividers, such as IGaging Digital Divider/Compass  Not cheap, but not exorbitant.  

 

However, before becoming obsessed with small measurements, remember that it is also not possible to manually draw small objects accurately.  CAD can do better, but there is still always a line width that is typically several pixels wide.

 

If the part is small, the dimensions should correspond, in scale, to something a person can pick up or grab by hand.  So, adjust your numbers accordingly.  Actually, for small parts it's the relative dimensions (proportion) that matter the most.   For instance, a ship's wheel. The diameter is easy to measure, but the spokes and handles have be in proportion otherwise it will look like a gear.  The eye and brain can be fooled by scale, but they pick up errors in proportion easily.  Making small adjustments to ensure proportion, or the appearance of proportion, is one of the skills of modelling.

 

If the part is bigger, then scaling with the points of calipers should be close enough.  At 1/8 to the foot scale, 0.005 inches, which is obtainable with calipers on a drawing, is only 1/2 inch in full size.

[wefalck]

The line-width in a (real) CAD system is zero. It is different, however, when we are talking about pixel-based drawing programs, where indeed a line (particularly if it does not run parallel with the axes of the system) can be several pixels wide. It would be bad practice to draw something on a CAD system, print it out and then take measurements from the print-out. Even in the old analogue days I would draw an item in 10:1 or 20:1 scale to reduce the effect of drawing inaccuracies and the problem of line-width. In the CAD-age the drawing itself is only something to orient yourself, the measurements are taken from the information stored within the system. Adding measurement bars in my CAD (EazyDraw) is a little bit awkward and makes the drawing rather 'busy', so I tend to make a print-out and note the relevant distances by hand. This then is the drawing from which I work on the lathe, the mill, or by hand.

 

Lehmann is right in saying that proportions are what matters. That's why one should work 'outside-in' from the overall dimension, as I noted earlier. For many details there are handbooks (now also on the Internet) that give you typical dimensions. The human being hasn't changed that much over the past few thousand years or so, hence the 'handy' diameter of handles, rails etc. hasn't really changed, because hand-sizes haven't really changed. In fact ergonomics as such haven't really changed (though behaviour and use of people may have changed), so that modern dimensioning more or less gives you an indication.

 

In most cases, for our kind of models the 'look' is what counts. If you are into working steam-engines and the likes, of course, exact dimensions to specifications are most important, otherwise the engine may not work. So, in our case we probably don't need measurements to 1/100 of milimetre ensuring the right kind of fit. Therefore, digital calipers are usually an overkill (though you can buy them from 15€ upward nowadays). I find them also too bulky most of the time and it is another item to maintain, i.e. making sure it has a working battery in it. My preferred caliper is a small vernier of only 15 cm length that can be easily used to control the work on the lathe or mill, but also on the bench. It has a 1/20 mm vernier, which is probably better then the working tolerances of most of us. You can estimate 1/60 mm easily, as thirds between two engraved lines can be guestimated. This, however, is better than the mechanical precision of the measurement, because of varying pressure you may apply on the part.