Working in Metric 8

[DaveAtkins]

For the first time in my career, I will be working on a project which requires the use of metric (hard metric at that!) units.

I am curious about the experience of those who normally work with imperial units like I do. My hunch is I should not try to convert units back and forth, but instead dive right in and do calculations and drawings in metric. Train myself to think in metric, so to speak.

DaveAtkins

 

[BridgeSmith]

Of course, the number one thing is to attach units labels to all the values. Don't like NASA.

 

[IRstuff]

" Don't like NASA."

That was not a units problem, per se, but a management problem. Someone(s) failed to check the interface control documents (ICDs) and more someone(s) failed to test the interfaces. Additionally, upon investigation, it was discovered that NASA actually saw the problem several times in actual flight, manifested as position errors of the spacecraft on its way to Mars, but no one bothered to check why. Had ANYONE been paying attention over the course of a couple years, the disaster could have be averted.

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529 Entire Forum list

http://www.eng-tips.com/forumlist.cfm

 

[BridgeSmith]

I expected you would have a more intimate knowledge of that screwup than I, IRstuff. My point was simply that the error began with a failure to label or clarify the units of the numbers conveyed. Although, I just realized I didn't do so well at conveying my thoughts, either - "Don't like NASA.", what the heck was that?

 

[IRstuff]

Probably some Freudian thing...

The units were likely labeled, since the ICDs are critical deliverables in any aerospace design. There's typically a massive table of data being transferred across an interface, and there would be a item name, numerical range, scaling factor (if needed), and unit. Both sides of the interface would have had reviews and agreed upon all the data types and formats, etc. Usually, the Interface Description Document (IDD) would be transferred a few times across the subcontractor interface for the two parties to agree on the exactly byte counts, etc. It's likely that both sides probably even agreed on the SI units.

While it's never been reported, I suspect that like a couple of engineers in my company, their propulsion engineers ALWAYS design in US units and then convert to SI when documenting, and this one time they forgot to do the conversion in the deliverable code. Even still, something should have been detected during the software unit testing, so someone screwed that up, as well.

However, while that was an egregious error, several times during the flyout, the navigation engineers noted that the course corrections were not resulting in the correct flight path corrections, and they blew it off as possibly too much radiation pressure, but they never checked or ran calculations to verify

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529 Entire Forum list

http://www.eng-tips.com/forumlist.cfm

 

[JStephen]

I spent quite a few years in school with metric, and it was always the SI system. But surprise, the real world doesn't always use SI. So there was a post the other day referencing kgf. We get foreign mill certs on occasion that will show kgf/cm^2 as stress units. And the metric drawings I've seen just showed every dimension in millimeters without any indication of units. So instead of "31.823 meters", it would just show "31823" on the dimension arrow. The point is, check local usage to confirm what they want, if you haven't seen similar drawings from there.

 

[drawoh]

JStephen,

There is a very important advantage to treating kgf and lbm as The Work Of The Devil. You always know if, when and where to insert the gravitational constant in your equations. This is not important in static structural analysis, but the moment things start moving...

--
JHG

 

[Sparweb]

I think I've tried everything that's been suggested so far at least once.
Combined units in Excel is the WORST and I will never NEVER do that again.

Here's another way:
Using either Mathcad or SMath, do your calculations in the default units system, putting all the "x=#" values in a column on the right. Leave a spare 2 inches of blank space on the right side of the page. Copy-Paste the right-hand column of values into a second column in the blank space. Correct the units in the new list so now your page tells the reader the value in BOTH SYSTEMS. You can leave one column in metric, the other column in Imperial. This allows the yankee curmudgeons and the high-strung europeans alike to read all of your calculations.

No one believes the theory except the one who developed it. Everyone believes the experiment except the one who ran it.
STF

 

[IDS]

Or in Excel:

Units4Excel

Doug Jenkins
Interactive Design Services

http://newtonexcelbach.wordpress.com/

 

[JStephen]

I was just over watching the latest video from The Hydraulic Press channel on YouTube. The guy's got a load cell attached to his press, and in this case, he's measuring the force required to push nails into wood. And the unit of force is: Kg.

I don't remember for sure, but I think his press hydraulic pressure is in bars, if so, there's another non-SI unit cropping up.

 

[ThomasH]

Hi

If you want to be strickt you have to differ between SI base units and derived units. There are also units that are related to the SI system in a very simple manner, like bars.

The unit for force, Newton, is not a SI base unit but derived from kg, m and s. I have more than once noticed that force and mass are treated as being the same. Especially some software's don't make a difference between the two properties.

Thomas

 

[robyengIT]

see ISO 1000:1992 revised by ISO 80000-1 :2009

 

[BridgeSmith]

It seems several people have strong opinions one way or the other about MathCad, SMath, and Excel. Personally, I much prefer Excel. I'm fairly sure it's a matter of familiarity, for myself, at least (haven't tried Smath, and MathCad only a couple of times). I suspect it's the same for others.

For those who aren't already 'locked in' to one of them, I suggest trying out whatever ones you have access to, so that you can decide for yourself what works for you. There's a ton of stuff that I can do easily in Excel that I don't know if the other programs will do, and I sure wouldn't know how make them do (like rotating a table of values, columns to rows).

Obviously, they will all work for converting units. Excel does it fairly easily, either by reference, lookup, for by the preprogrammed function. It looks fairly easy in the other programs as well.

 

[IRstuff]

It's much more than just ease of use; units are inherently part of the math systems of SMath and Mathcad, to the extent that unit conversions are completely transparent:

since pascal is the default pressure unit, the conversion is automatic.

Additionally, because the units are integral to the calculations, they provide inherent dimensionality checking:

is messed up because I used a velocity, so result is not a simple pressure.

I defined nmi as nautical mile elsewhere, so this works:

Nevertheless, there are obvious things that Excel excels at, like any table-formatted data and the graphing thereof; graphing data is something Mathcad is still struggling with.

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529 Entire Forum list

http://www.eng-tips.com/forumlist.cfm

 

[BridgeSmith]

If I was picking a program to start out with, especially if I was doing alot of unit conversions, I would probably take a long look at one of those other programs. Having units inherent in the program would definitely be easier, and would catch some potential errors.

 

[IDS]

IRstuff examples in Excel, using Units4Excel:

Currently the Evalu function does not recognise units embedded in the function to be evaluated, so you have to define them as named parameters with a value of 1, and enter them as a factor, such as 5*dyn.

You can name the units whatever you want, so for instance I have named nmi (nautical miles), which is a defined unit in the spreadsheet, as nmile.

You can also redefine existing units, such as kt (knot), which I have defined as 1.852 km/h, which is the exact official definition, agreeing with the SMath result. Note that the current spreadsheet uses an approximation which is only accurate to about 5 SF. I will upload a corrected version later today.

You can also define new units, such as the ksmoot, equal to 1000 smoots.

Units4Excel can be downloaded from Units4Excel.

There is also a Python version at Units and solvers with Pint and Sympy. Note that this is a work in progress, and requires installed versions of Python and xlwings.


Doug Jenkins
Interactive Design Services

http://newtonexcelbach.wordpress.com/

 

[TehMightyEngineer]

Star for using smoots!

Ian Riley, PE, SE
Professional Engineer (ME, NH, VT, CT, MA, FL) Structural Engineer (IL, HI)

 

[IRstuff]

I get that if someone has nothing else, then Units4Excel is a great workaround, but I'm not convinced it's as easily usable as Mathcad; obviously, Mathcad requires considerable coin, so it's not for everyone, but SMath is free.

For example, you can't put in ft2 in to K13 and get a valid answer in the example spreadsheet because you need to input "sq ft," which means that you need to remember that for whatever reason, the English units were defined with sq (unit), while the metric was defined with (unit)2, neither of which are natural notations. In Mathcad, ft^2 is simply the square of the basic "ft" unit, and that means that ft^n, where n is any number is automatically usable without additional work or thought from the user.

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529 Entire Forum list

http://www.eng-tips.com/forumlist.cfm

 

[IDS]

IRstuff - I'm not saying my spreadsheet is better than SMath, or even as good as SMath, just for people who don't know SMath, or who have an application where they prefer to use a spreadsheet, it provides the units functionality to get the job done.

Regarding derived imperial units, you are right that it won't do them, but you can add any unit you like on the Ext Unit List sheet, and it will then be available after saving and restarting.

Doug Jenkins
Interactive Design Services

http://newtonexcelbach.wordpress.com/

 

[IRstuff]

No worries; like I said, I use Excel quite a bit, in conjunction with Mathcad.

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529 Entire Forum list

http://www.eng-tips.com/forumlist.cfm