I was wondering. On my designs I end up using “full” measurements. It is easier to cut a piece that is 56 cm long rather than, let’s say, 22 inches, that would equal to 55.88 cm. I am pretty sure this would interfere with some proportions as well - if I make a rectangle measuring 56 x 78 cm it would be way easier to cut it measuring 22 by 30 inches.
So… Will the measurement system used on the design affect on the final result of the project?
I would guess only slightly.
There is indeed a principle of preferred sizes: if your design calculations call for a dimension that’s not a nice even number, you tend to round to the nearest whole number that makes the design still work. This makes it easy to use off-the-shelf components (which are typically offered in sizes that are nice round numbers, or at least a few commonly used sizes).
Look in a bearing catalog, and they’re all listed with ODs, IDs, and thicknesses that are nice even numbers like 1 inch or two inches (or common decimal fractions, e.g. 0.25 inches, 0.625 inches). If you’re working with SI units, you might look in the catalog for a bearing with a 25mm OD; if you’re using imperial units, you’d pick one with a 1-inch OD.
Look in a metals catalog, and you’ll find their stock is also offered in similar common sizes. If your strength calculations dictate a shaft 0.98 inches in diameter, you’ll look through the catalog for 1-inch shaftstock (unless your design has unusually strict weight requirements), or 25-mm stock.
So you would really only expect minor effects on your design by using one measurement system or the other.
Wasn’t there a space probe that failed because part of the team was using Imperial measures and the other part was using metric? Or was that just navigation so it doesn’t count?
Depends on the application. The Tupolev Tu-4 was a virtually exact replica of the Boeing B-29 bomber, reverse-engineered from B-29’s that had made emergency landings in the Soviet Union. The skin of the B-29 was 1/16" aluminum, which comes to 1.587 mm. Obviously this particular thickness wasn’t available in the USSR. I can’t tell from what I’ve read what was available, but supposedly they switched to a varying thickness that ranged from 0.8 mm to 1.8 mm. Anyway, even a few percent difference could be utterly critical in an aircraft application (which are pretty much always weight-limited).
Yes, the Mars Climate Orbiter.
Analysis I’ve read about this program (the Tu-4 NATO codename Bull) say that in hindsight it was a huge waste of manpower & resources, partially because every component had to be converted to metric before being duplicated. Add to this the fact that the B-29 was obsolete by the time the Soviets finished their copies. It was a purely political propaganda program that actually wound up setting them back nearly a decade (if they had instead put those resources into designing & building their own inevitably more modern intercontinental bomber).
Wouldn’t surprise me. The skin was always the obvious thing that gets brought up with regards to that project, but there much have been zillions of tiny problems to be solved. Stuff that you wouldn’t necessarily think about, like screws that intersect important bits if they’re 5% too long.
I think that 1.6 mm aluminum must not have been available. It seems like it would have been close enough. I’d guess there was 1.5 mm, and 2.0 mm, but in between there was only one more thickness–the 1.8 mm that I saw mentioned (I’d think that 1.75 would make more sense but it seems that metric types prefer to add more decimal digits only under duress). Between them there’s nothing that really works well.
Worked at a company that was transitioning hand drawings to CAD after going from real measurements to metric, or celsius, or whatever they call it. 
It was easy for the guys modifying the hand drawings–all they needed was a pencil, an eraser, a calculator, and some ungentle rounding. It got hairier putting it into CAD–I would convert the metric (to very few decimal places, as Dr Strangelove noted) to decimal imperial, then to the nearest fractional imperial because that was how it had been drawn 50 years before, back to decimal, then to metric with lots of decimal places because computerized drawings need to be precise. It was a half-assed form of reverse engineering.
If you have the capability to measure in inches, it is equally easy.
What sort of design are you talking about, using what sort of measuring/cutting technique?
Also, what degree of precision is “22 inches” meant to represent (+/- how much)? Calculating the cm equivalent to 4 figures (down to hundredths of a cm) may well be unnecessary.
In my career as a Chemical engineer, I believe I have made fair bit of money converting units (judging by hours charged )
Most intuitive units - length / mass / temperature are easy and most design disciplines convert is easily - wait I take that back. I have seen many a times mechanical engineers confused over the Rankine temperature measure - most thermodynamic properties need to be converted using the rankine temperature like kelvin in the metric temperature measurement.
Another area is the volume/weight / heating value of gases. Take for example : Normal m3, Standard ft3, Standard Tons, Metric tons, MM BTU, k Joules etc are all units for natural gas quantities - most non chemical engineers will not like to do the conversion but it is essential part of trading/markets.
The units of energy are such that most people have very bad comprehension of the ratio of power consumed by their cars versus their A/C versus their grill. But dimension wise they have very good idea of the ratios of the sizes of these equipment .
The most challenging units (this has nothing to do with conversion but more to do with definitions) are those of radiation exposure. You need to spend at least an hour to understand the units.
Some of the other areas affected by units are :
Simply not true.
When expressed as BED (banana equivalent dose), it is intuitively obvious what any particular exposure is.
You simply imagine what it would be like to eat a pile of the corresponding number of required bananas.
And of course it is trivially easy to convert between metric bananas and imperial bananas - so no problems there WRT to the OP.
I’ll bet it makes a significant design difference some of the time.
e.g. in architecture. If you’re picking whole numbers of units, eventually the unit choice could make the difference between whether there’s room for you to add, I dunno, an extra restroom or a fountain or whatever.
In the extreme you might pick a whole number of kilometres for a path, whereas someone working in imperial might have picked a whole number of miles, and all the knock on effects of that.
The big problem is the common sizing of available materials.
You can make up any plan measurements you like. To any precision.
But. What are the usual sizes of the materials that you want to use? If you are going to go shopping in a place that has common metric sizes listed, then it would be easier to start planning in metric. You will be dealing with it, start to finish.
Adding and subtracting the material thicknesses etc in the same system.
Precision level can be sloppy when presented in multiple systems dimensions in advertisements or even supplier specs. Could be enough to cause problems.
Plan it out in the most common material measurement system in your area. The one that is actually listed, not the one that might be the supposed measure of the law and land. Makes for the least amount of conversions.
Sometimes the problems don’t even involve different measure systems, but different assumptions. Pressure: absolute or relative? (The difference is 1 atmosphere). Concentration in %: chemists normally work w/w, pharmacists do it v/v when not w/v or v/w.
People look at you funny when you ask “is this % w/w or v/v?” but it turns out that very often they don’t know it themselves.
Certainly it affects your ability to sell your end design in certain parts (i.e. most) of the world.
Back when I was sourcing and purchasing pharmaceutical equipment for the UK I could only use equipment that was constructed using metric parts. Knowing lots of similar industry people around the globe this was common practice and must have restricted the market of those companies unwilling to design and build in metric.
Isn’t this why SI units were introduced in the first place?
If the OP is talking about aesthetics rather than science, then I do not believe that it makes any difference. Chartres Cathedral did not suffer because a dozen masons used different measuring systems - the important thing was proportion. A golden rectangle is golden in any measuring system.
Yes, having different unit systems around does interfere with design, here in the US. The rest of the world exists in SI (the international “metric” units based on meters, kilograms and seconds) but the US spends a huge amount of extra money on the luxury of nobody having to change their units right now. Those of us caught in the middle see this the most.
I’m a scientist who’s expert in a particular element of machinery and I work with a few dozen design engineers helping improve designs. This problem is a mess.
First, there are conversions between different units. Lengths are pretty easy because an inch is exactly 25.4 millimeters. But the units for heat transfer get extremely messy and there are very many possible units (over 200 I think) for heat transfer coefficient, to give one example. It is a great deal of extra work to convert all of these things when comparing, say, insulation materials available from various manufacturers. Each of these steps carries the possibility of error, and this is true also for the various manufacturers, so their own specifications sometimes have errors (in fact in some cases they will give their specification in more than one unit, and these will contradict each other).
Second, there are all kinds of things we need to buy that come in specific dimensions, such as bolts and rivets and dowel pins and bearings and pipe fittings. Because we may have to use subassemblies based on different units, we need adapters between all these different systems. When we try to test different subsets of a system or when fixing things, we may wind up stopped in our tracks because we need some obscure adapter, even though we have hundreds of things here whose only purpose is to adapt between different unit systems, because the variety of things you could need is mind boggling.
Third, the need to behave globally makes it pretty hard for a company to stick with Imperial dimensions, but the availability in the US of construction materials in regular SI dimensions is more limited. We wind up working in something called “soft metric”, which is a system where you buy Imperial sized raw material and cut SI dimensions into it. Of course this design can’t be reproduced in the rest of the world, where they don’t have Imperial sized materials.
Fourth, those of us that do a lot of science deal with a wide variety of equations that make very simple sense if they are done in what’s called a “coherent unit system”. SI is coherent, and Imperial is not. The equations work in a straightforward way in any coherent system (not just one). However they require all sorts of extra constants, which are basically correction factors, to be used in a non coherent system.
I could go on and on. But, yes, this is an absolute mess, and it’s woven into the foundation of all the technical work that happens in the United States. People who work in any other country typically don’t mess with any of this at all, but here we carry this extra burden – and it’s no goddam fun at all.
One of my clients was the biggest company in its business for most of the world - all countries but one, in fact. When they said this during the initial company presentation, one of my coworkers joked “c’mon, in Somalia too?” “We DO have clients in Somalia, actually.”
The country they didn’t do business in at all was the US, and they gave two reasons for this:
I know part of the problem was units, but not how much. And in some cases the problem was that where for other locations they could identify certain standard parts by metric, for the US they could not (according to my source, I can’t vouch for his accuracy on this specific subject).
Of course, there are variant unit systems based on the American system which are coherent… but the problem there is the plural. There are multiple such systems, which means you have to figure out which one is being used, or if it’s the bad old incoherent one.
For example: In a coherent unit system, the unit of force is equal to the unit of mass times the unit of length divided by the square of the unit of time. For instance, in SI, the unit of force is the newton, the unit of mass is the kilogram, the unit of length is the meter, and the unit of time is the second, and 1 N = 1 kgm/s^2. In the most commonly used version of the American system, however, the unit of force is the pound, and the unit of mass is… also the pound (technically a different pound, and a pound of mass only weighs a pound of force in a standard gravitational field, but most people ignore or are unaware of the distinction). So you can either use the common form, and have to stick factors of g into all of your formulas for no good reason, or you can take the pound to be the unit of force and define a new unit called the “slug” as your unit of mass, defined such that 1 lb = 1 slugfoot/s^2, or you can take the pound to be the unit of mass and define a new unit called the poundal to be your unit of force, defined such that 1 poundal = 1 lbm*foot/s^2. Now figure out just what someone means when they say “pound”. Worse yet, try to figure out what someone means when they’re lazy and sloppy and don’t mention the units at all.
It’s a shame the OP has disappeared because I think a few people might have misunderinterpreted the question.
I don’t think he’s asking about problems mixing units.
While there are a lot of units used for quantities of natural gas in different markets and contexts, the idea that it would be especially difficult for non-chemical engineers seems absurd. In the course of my work I convert all the time between therms, MMBtu, kWh, GJ, and normal cubic metres, as well as between LHV and HHV. I don’t think this would pose a problem for an engineer of any discipline (or for that matter, any first-year engineering student).