Dual dimensioning rounding ?s / is it legal to say "dims in brackets are for reference only&quo 6

[ingallspw]

Is it "legal" to call out your secondary dual dimension as "REFERENCE ONLY" ?

If not is there a standard that forces you to use a set amount of decimal places in your dual dimension?

We have metric dimensions that are technically tighter than the English and vice versa that may cause someone to reject a part due to rounding of the two.

I realize dual dimensioning is not everyone's favorite so thanks for bearing with me.

 

[drawoh]

dgallup,

I've heard that a lot but some times it does not work out that way.

Take
+/-0.06in
[+/-1.5mm]

I issue sheet metal drawings with metric dimensions and tolerances. I know for a fact that some of our vendors request the SolidWorks files, then they convert them to inches. This is fine as far as I ma concerned, as long as the vendor understands that we will inspect to our metric drawings. They must account for round-off by tightening their inch tolerances.

You asked above about parts sized to inch dimensions. If you use inch threads, gears and pulleys, you have to specify the inches sizes so that the vendor can select the correct tooling. If a hole is Ø2", I call up Ø50.8mm. I wish SolidWorks would display an alternate units reference dimension without the tolerances.

--
JHG

 

[Keith1029]

"The spec. is the spec., that's why it is the spec."

As a customer, I provide a drawing with the specific dimensions and tolerances that I feel I need; I inspect it (or have it inspected) to that print. If it is out of tolerance it gets an NCR and goes back. If the vendor wants to work in another unit system they can feel free; they can work in furlongs for all the difference it makes to me; at the end of the day it falls within the tolerance of my print, or it gets rejected.

As a vendor, I take a customer print, convert all the dimensions to a consistent system (if required) and produce an internal print used for manufacture. Since dimensions sometimes don't convert exactly, where rounding changes are made they are made in the direction to keep the part within the customer specification, if the rounding is too tight I go to more decimal places to reduce the rounding error. For internal inspection I inspect to the manufacturing print because it matches my process/tooling, if the customer wants inspection documentation we inspect in their units system. Either way, parts that are accepted should always be within the tolerance specified by the customer drawing.

Frankly, I don't see the need for dual dimensions, if someone needs me to do a conversion for them, I question if they have the skill to produce my part at all; BUT if a dual dimension must exit, by definition, it should be converted EXACTLY (with tolerance exact), or it should be reference, or the drawing has an error. Something is what it is; the units we use to report the value are irrelevant and do not change the fundamental size, the point of the drawing is to establish the single maximum and minimum fundamental size and shape.

To use your example, Ingallspw, in my opinion I don't think it would be anal at all to reject a part based on the .001" non-conformance, to do otherwise says "oops, I toleranced that incorrectly" at which point the drawing should be changed and it is not out of spec at all or it says "feel free to ignore my specifications, I won't enforce them anyway" which establishes a bad precedent.

-Keith

 

[juergenwt]

What a mess! I am working for a company selling internationally. All new product is designed in metric. However these products are to be attached to existing machines designed in imperial. So I can understand the question about the pulley.
Add to that the fact that all the material (steel, brass, aluminum etc) is purchased in imperial and the fact that the company has it's own fasteners with threads used by this manufacturer only and you can imagine the problems.
I am sure many other US companies face the same problems.
Now, since all new prints are to be in metric but the existing machine is all imperial most designers and engineers will design the part in inches since this will make the part fit the existing machine. Add CAD for the conversion and you end up with the metric .xxxx. Now this stuff goes out to the shop floor and with set up and inspection having very little feel and understanding for metric you are now flirting with disaster. Common sense will tell somebody familiar with both systems where to draw the line. The biggest problem is created when we have fits on new parts (DIN ISO 286) and old parts (shaft)are machined to imperial tol. of let's say .500 -.0002". First you have to establish what type of fit you desire. Press, sliding, running etc.. I see no other way but to take the .500 -.0002 and convert it to metric 12.7 -.005 = 12.7 h4 and use that to establish the required fit for the new part i.e 12.7 H7 for a sliding fit and put that on the print.
At least you are now working in one system only. Any vendor would have to work to this and that is what inspection would use.
I do not know if listing the imperial equivalent to 12.7 H7 would be helpful.

 

[CheckerHater]

I can feel your pain.

Unfortunately nothing will change until somebody actually makes an effort. And here is deep underlying problem: metric conversion is seen as long-term investment and many years of bad economy resulted in fear of long-term commitment. Today’s business model is more of a “hit-and-run”.
Nevertheless, many things can be done and it is Engineers job to educate the management.

Material: metric material is available and big manufacturers are buying them. In fact, American cars are stamped from metric sheet. Eventually it will trickle down to small guys. Material that requires machining is interchangeable - you leave allowance anyway. Pre-machined material like precision rails and shafting is available in both inch and metric.

Fasteners, bearings and other purchased components are increasingly metric. Traditionally inch components like tapered tool shanks don’t need to be “metricated” – you buy your tools as they are.

Machines: if you are talking about old manually operated lathe with its gearbox specifically designed to produce inch threads, then yes, there isn’t much you can do. The rest doesn’t look so bad. Mills can be retrofitted with digital readouts. CNCs can move between inch and metric with the flip of a switch. More complicated machines like CNC sheet metal punch presses may require memory re-flushing, but it still do-able.

Tooling and instruments: Drills, reamers, taps and dies (and the gages to check the results) are all available in both inch and metric. If necessary, there is nothing wrong about producing inch thread in otherwise metric part, there is nothing wrong about specifying inch thread on otherwise metric drawing. But make it metric! Digital calipers switch between inch and metric as well. Not to mention that in today’s world of $10 micrometers it’s not a huge problem to supply your shop floor personnel with proper measuring instruments.

Naturally all of the above is just a short overview. Everybody has to research what’s available in his industry, or even in his area. But someone has to start somewhere.

 

[3DDave]

The preferable units are hands and stone, for length and weight. Rods or furlongs if you are going a little farther, angstroms if you are not going as far.

The inertia against changing measurement basis is because there is no particular benefit in one over the other in application in a sufficiently large ecosystem. Past that, they are both arbitrary.

In English linear measurement the basis is the inch, which is divided into as fine a scale in both factors of two and ten as one would like vs the meter which is typically divided only in factors of ten.

In temperature measurement, the degree Fahrenheit is a finer base division than the degree Centigrade. While the measurement techniques for either are subject to the same physical limitations (conversions of temp deltas to voltage or expansion of liquids) degrees F requires less display precision for similar temperatures than degrees C.

The main advantage to the metric system is that the units are coordinated. Watts are used for electrical and mechanical power as opposed to converting them to Horsepower. If this was universal across human experience, it might be noticed, but the benefit to this coordination is seen as small; so small that in engineering, the 'slug' is often ignored as the unit of mass (it's comparable to the gram, though more massive), leading to calculations where the pound-force and pound-mass drag in the local acceleration of gravity (G-sub-c) to balance the equation.

One might wonder why, with 1/3 of the world's population speaking Chinese (some version) the rest don't convert. Supposedly, there will be/are more English speakers in China than the rest of the world, but that's because the English speaking ecosystem offers a clear, rather unrelated, economic advantage. Until there is a clear economic reason, the division will remain.

 

[ewh]

Mixing units on a drawing should not cause issues (as far as part definition). Mixing units in a dimension does.

“Know the rules well, so you can break them effectively.”
-Dalai Lama XIV

 

[CheckerHater]

One might wonder why, with 1/3 of the world's population speaking Chinese (some version) the rest don't convert.

Hieroglyph-based languages are not very effective way to convey information when compared to alphabet-based, but this is mildly off-topic.

Yes indeed in metric system watts are used for both electrical and mechanical power. It is little less known, that watts are used for hydraulic power as well.
Just like in electrical power watts are calculated by multiplying voltage by electric current, in hydraulics power is calculated by multiplying pressure by the flow of hydraulic fluid.

Let say we have 200 l/min flow at pressure of 30 MPa. As metric units are usually divisible by 10 rather than 3 or 4, it is not difficult to remember that liter is cubic meter multiplied by 10^-3, minute is 60 seconds, and megapascal is 10^6 newton per square meter.
Therefore the power is equal to (200*10^-3)*(30*10^6)/60 m³*N/m²*s or 100000 N*m/s or 100 kW

Now, with the flow of 50 gallons per minute and pressure 400 pounds-of-force per square inch good luck to calculate number of mechanical horsepowers without using your local library or the internet.

This is why any REAL designer / engineer / scientist whose job requires calculations will prefer using metrics hands down. This is why all engineering / scientific software is actually programmed in metric and customary units are converted. Ever wonder why the largest object you can create in SolidWorks is exactly 39370.07874016 inches?

Naturally, for “CAD Engineers” and “Senior Certified Spell-Checkers” there is no particular benefit in one over the other.

 

[3DDave]

I clearly wrote - "in a sufficiently large ecosystem." Also Chinese writing is not hieroglyphs (old Egyptian); I referred to the spoken language.

50 gpm * (231 in^3/gal)*(1 min/60 sec)*(400lbf/in^2) *(1ft/12 inches)*(1hp/(550ft-lbf/sec)) = 11.66 hp and didn't have to look anything up. It's like remembering Avagdro's number 6.022*10^23 atoms/mole or the Gc is 9.81 m/sec^2 at the surface of the earth (approximately) or that standard air density is .00238 slug/ft^3. Just numbers that come up that need to be remembered.

If I had a copy of MathCAD software it would have done the conversions for me. If Microsoft would get off their butts and make their software unit aware, that would make life better all around, preventing miscalculations on, for example, sheets that have man-hours and dollars/pound, which under current circumstances can be added or multiplied without triggering any error.

Or just be stunned and amazed:

http://www.wolframalpha.com/input/?i=50+gpm+times+400+psi+in+hp

Where did you get the idea that CAD/CAE software is always metric on the inside? PTC software is not programmed in metric; it is unit agnostic, sometimes unit antagonistic. I guess Solidworks is unsuitable for solving problems over 10,000 meters.

It was around the time of the Jefferson administration the US formally adopted the metric system, because we were unhappy with Britain at the time and happy with France. This before the whole Freedom Fry debacle. The American inch is legally defined as 2.54 centimeters, since 1959. See

http://en.wikipedia.org/wiki/Inch#Modern_standardisation

for more and uglier details.

I prefer the units my customers prefer, for which I take into account the tools, materials, and training available in the customer ecosystem.

 

[CheckerHater]

You have to agree that while you used 3 conversion coefficients (gallon to inch, feet to inch, and pounds to horsepower), I used zero (not counting minutes to seconds in both cases).
So, while both systems are arbitrary, one is far more effective.

Also, all modern Chinese characters are evolved from ancient hieroglyphs, so the system is indeed hieroglyph-based.
Not to mention that hieroglyphs themselves still in use in some places:

http://www.theworldofchinese.com/2010/06/hieroglyphics-in-china/

 

[ingallspw]

Me like dem dar fancy numbers! Gollll-ee dair purrdy!

And yup, it's true we merican's be stuck in miles of our inches and horse powers but who came up with integrated circuits, Lasers, computers, laptops, Kevlar, Liquid Crystal Displays, Refrigerators, MRI's, Laser Eye Surgery, Video Games, air conditioning, motion pictures, Assembly lines, e-mail, internet, light bulbs, LED's, sky scrapers, copiers, pace makers, the electric motor, microwave ovens, air bags, satellite in space, Nylon and most importantly Bra's and Toilet Paper. And probably most of it was designed in feet & inches.

Honestly I don't care what system is the norm. I really want to move to metric. It would make my life easier.

But to say using Metric makes you a real engineer is like saying having a master's degree makes you smart. I know people who's knowledge of engineering would blow your mind that hate metric with a passion. (Why? I don't know.) And I swear you have to be a CEO with a master's from Harvard to make the stupidest decisions.

Thank you all for your input! I think I have all that I need!

Have a good day!

 

[3DDave]

I can't find the right XKCD comic for this. Sigh. Still, I would not have recalled the Wolfram Alpha site otherwise, nor had the interest in ancient writing and transition to modern day. All in all, not so bad.

"Hieroglyph" is limited to old Egyptian. Being Egyptian is the key characteristic.

That one can find an erroneous page on the Internet written by someone outside their field of expertise is not surprising. Since it is a website maintained within China**, it is likely the page is a mistranslation for a term like ideograph, pictograph, or logograph, which are classifications hieroglyphs can fit in. It would be just as odd to suggest that Chinese is Mayan, who also used non-alphabetic writing, but are also picture-based.***

I did find that modern Chinese writing is apparently getting expanded the way modern French is. While the French are forced into le Big Mac and other injections of English almost unchanged into their language (using the same alphabet, the fools, that'll teach them) the Chinese apparently massage the words from other languages to meet up with phonetics from their existing words, even when they have to squeeze a bit on the pronunciation to get the word to fit. So it isn't written or always pronounced in a way recognizable to the original-language speaker. And it is written in essential gibberish to the Chinese reader.****

Oh wait - found it.

https://xkcd.com/386/

**

http://www.theworldofchinese.com/about-us/

***There's a great documentary on how the meanings of the Mayan symbols was deciphered. Unlike hieroglyphics, there was no equivalent to the Rosetta stone to help the translation.
****Apparently in Chinese writing the order of the strokes that make up the characters is also important. Your handwriting can be picture perfect, but if your handwriting instructor (or parent) watches you do it in the wrong order, you are still a failure. Way to add the guilt.

 

[ingallspw]

Oh and what I said bout us merican's is not to diminish anyone or any other country's accomplishments.

 

[CheckerHater]

Should I point out that statements "real professionals use metric" and "using metric makes you real professional" are not the same?

And since you asked "who came up with", here you go:

Integrated circuits - Germans
Lasers – Americans
Computers - British
Laptops - Americans
Kevlar - Americans
Liquid Crystal Displays - British
Refrigerators - British
MRI - Russians
Video Games - Americans
Air conditioning - Americans
Motion pictures - French
Assembly lines - British
E-mail - Americans
Internet – Al Gore
Light bulbs – 22 other guys before Edison and Swan
LED - Russians
Sky scrapers - Americans
Copiers - Americans
Pace makers - British
The electric motor - Hungarians
Microwave ovens - Americans
Air bags -Germans
Satellite in space - Russians
Nylon - Americans

 

[MintJulep]

If you don't control the calculation of unit conversions and the associated precision on your drawing then who does control it?

Who is making the conversion? The guy who took the order at the machine shop? The guy turning the handles (or are there multiple guys?)?,
When are the making the conversion? Every time a new blank gets set on the tool? Is there one marked up copy floating around in the shop?
What did they use to make the conversion? An iPhone, a calculator, a slide rule, a lookup table?
How many significant digits did they bother to write down? Do they even know what a significant digit is? Did they mistype, misread, mistranscribe...

 

[CheckerHater]

3DDave, if you bothered to read the article, you would see that the whole idea of it is that Naxi hieroglyphs actually precede ideographs, pictographs, or logographs. Not to mention that the article itself was written by the guy named Nicolas Richards (sounds like someone who doesn't speak English, right?)

And MintJulep, thank you for bringing the thread back to its roots.

 

[3DDave]

I did bother to read that and many other articles, articles you didn't bother with. What is essentially a travel guide for, you know, tourists, was the -only- one to refer to Chinese writing that way.

Checkerhater, I do admire your ability to stand by an unsupported conclusion. Look up Monty Python's Argument Clinic. You'd be perfect.

I went to college with a guy named Kojak. He was from Indonesia and his given name was too difficult for English speakers to pronounce, but I should have guessed he was a bald, New York detective based on the name.

While the Russians put a satellite in space, an American first proposed the utility of doing so and an American first demonstrated the practical means to do so. Thank you Arthur C. Clarke and Robert Goddard, wherever you are. Jules Verne would be in the running, but his destination was not an orbit and the means was not practical.

 

[3DDave]

"If you don't control the calculation of unit conversions and the associated precision on your drawing then who does control it?" is an interesting question.

It's a risk like any other risk.

Consider that the CNC machine with it's ability to operate with a switch in metric or English is likely depending on a microprocessor maintaining a count of turns of a drive motor that is carried through a gearbox and converted into a distance based on the rate of advance of a screw. Those counts are probably not going to be in any even division of the primary units of the machine interface; they certainly can't be both metric and English at the same time. The advance on the screw is not uniform - each turn of the screw may advance a little more or less depending on where on the screw the mating follower is, so the conversion is imperfect.

The results are subject to temperature changes of the machine and the workpiece. Perhaps the machine moves 10 inches with as much precision as the NIST can bring to bear, but the part is warm at the time and later is cooled to inspection temp.

There are more analog to digital errors (converting the part size to an analog of part size by inspection equipment to a numerical readout) that continue this process.

When a part is specified in one unit system but will be accepted in another, these and many other factors are already in play, beyond drawing control. It is up to the buyer to ensure the acceptance process will meet their requirements.


I disagree** with the idea that inspection equipment must be better than the last digit. It needs to be much better than the tolerance zone; the tradition seems to be 10%, but it still needs to be accounted for just like any other error contributor.

Why penalize someone with better measuring equipment by truncating numerical conversions on the drawing? For example, in converting 1.0m to inches, would the measurement accuracy only be +/-1 cm and the matching conversion is to 40. inches +/- .1 accuracy based on maintaining significant digits while avoiding putting 39.37 inch on the drawing because it would require +/-.001 resolution at inspection?

Does .3125 +/- .1250 require .00001 to be sure? Higher precision makes for a better discriminator, but it isn't a simple requirement. I've mentioned before that if a supplier is consistently hitting so closely to the zone edge that only the most sophisticated methods can tell good parts from bad, you probably need a new supplier.

**Disagree in the sense that it is a workable concept, but it may be more expensive than necessary. This is unlike simply being wrong. If this was part of a supplier plan I would not reject it for that.

 

[juergenwt]

Manufacturing in imperial or metric will most likely come out the same as long as you design and process in the same system. One is not more accurate than the other one . What every manufacturer needs to do is decide which way he will go. If he just wants to build garage door than there is no need to change to metric - that is until the first doors "made in China" show up. If he wants to sell to the rest of the world than he has the choice of going imperial and go under or build in metric.
If you deliver to the automotive industry you have no choice -it's hard metric. Building industry - imperial. Both systems will work. Your choice. One thing you can not do is mix the systems. There is no such a thing as a 6.35mm x 20 thread.
As far as calculating and engineering goes - with today's computers you can work with inches or millimeters equal fast.
That was not the case when the US first decided to go to the moon. Many of the first calculations had to be done using pencil and paper. The team that built those machine did all the calculations in metric. It would have been impossible to do this in imperial because of the urgency and not because the team consisted mostly of European engineers. After everything was designed it was than converted to imperial so that US industry could produce the necessary hardware. NASA of course had the manpower and money needed to do all the conversions.

 

[Keith1029]

Calling a thing metric or imperial changes the form no more than calling it happy or sad. The drawing describes the thing in itself; it will have a unit structure chosen that will give a numerical, standardized depiction of that thing, with only one interpretation. The conversion of units, and precision of the conversion IS controlled by they drawing (or maybe more appropriately "the required outcome of the act of conversion is controlled by the drawing"). How the conversion is done, if it is done, when it is done, how often it is done, who makes a mistake doing it is really none of my business as the customer/designer, any more than what process to use, what tools to use, what fixturing or jigs to use.

What IS my business, and I think what you are getting at Mint, is my exposure to risk and the cost of the product. The more conversions that are required the more chance for an error to be made (increasing risk) and the more people and time are required to do that (money). But everything comes down to delivery time, money, and rejection rate; people make vendor selections based on that (which is essentially a critique of how efficiently those vendors accomplish all the tasks that you mention).

If you want to go a step further in controlling your risk, like if you doubt your vendors ability to do the conversion correctly but trust them to do everything else, then by all means make the conversion on the drawing if that is the most efficient course of action - although it should still have only one interpretation (most likely one of the drawings has to be reference or the conversion is made to the base system with no dual dimensions). As a general rule though, this practice, in my opinion, falls under ASME Y14.5 - 2009, 1.4.e "The drawing should define a part without specifying manufacturing methods". To me, deciding if, how, when and who is going to do a conversion falls under the same umbrella as deciding how you are going to make that hole, which machine you are going to use and which operator you are going to trust to operate it. If a manufacturer decides to do the conversion by hand with pen and pencil every time he chucks it into the machine (without even writing it down on the print), that's his decision to make, but it can't think it would be good for business.

Now if I am responsible for both the engineering/drafting and the manufacturing side; I tell my engineers to design in the units that we manufacture and inspect to so that they are the only ones doing any conversions and they do it in the design and check phase.

-Keith

 

[drawoh]

... There is no such a thing as a 6.35mm x 20 thread.
...

The problem with screw threads and gears is that you are telling the fabricator what tool to use. You could say M6.35X1.27, except that the ISO screw standard is not otherwise identical to the Unified National.

I have just designed a differential screw adjuster that consists of a 0.8 pitch metric thread working against a 32TPI English thread. Do the math!

--
JHG