Do Americans use metric system?

Ad: This forum contains affiliate links to products on Amazon and eBay. More information in Terms and rules

Thanks Wuzak for the aid in explanation.
I didn't reply to some post because, to explain, I had to start from scratch.
So, let's begin...

ERRORS

Errors are normally classified in three categories: systematic errors, random errors, and blunders.
Systematic Errors
Systematic errors are due to identified causes and can, in principle, be eliminated. Errors of this type result in measured values that are consistently too high or consistently too low. Systematic errors may be of four kinds:
1. Instrumental. For example, a poorly calibrated instrument such as a thermometer that reads 102° C when immersed in boiling water and 2° C when immersed in ice water at atmospheric pressure. Such a thermometer would result in measured values that are consistently too high.
2. Observational. For example, parallax in reading a meter scale.
3. Environmental. For example, an electrical power ìbrown outî that causes measured currents to be consistently too low.
4. Theoretical. Due to simplification of the model system or approximations in the equations describing it. For example, if your theory says that the temperature of the surrounding will not affect the readings taken when it actually does, then this factor will introduce a source of error.

Random Errors
Random errors are positive and negative fluctuations that cause about one-half of the measurements to be too high and one-half to be too low. Sources of random errors cannot always be identified. Possible sources of random errors are as follows:
1. Observational. For example, errors in judgment of an observer when reading the scale of a measuring device to the smallest division.
2. Environmental. For example, unpredictable fluctuations in line voltage, temperature, or mechanical vibrations of equipment.
Random errors, unlike systematic errors, can often be quantified by statistical analysis, therefore, the effects of random errors on the quantity or physical law under investigation can often be determined.
Example to distinguish between systematic and random errors is suppose that you use a stop watch to measure the time required for ten oscillations of a pendulum. One source of error will be your reaction time in starting and stopping the watch. During one measurement you may start early and stop late; on the next you may reverse these errors. These are random errors if both situations are equally likely. Repeated measurements produce a series of times that are all slightly different. They vary in random vary about an average value.
If a systematic error is also included for example, your stop watch is not starting from zero, then your measurements will vary, not about the average value, but about a displaced value.

Blunders
A final source of error, called a blunder, is an outright mistake. A person may record a wrong value, misread a scale, forget a digit when reading a scale or recording a measurement, or make a similar blunder. These blunder should stick out like sore thumbs if we make multiple measurements or if one person checks the work of another. Blunders should not be included in the analysis of data.

the rest is here
New Mexico State University - Department of Physics
There are a good few errors and reasons for them missed out there. Everyone understands what the diameter of a circle is. When it comes to the diameter of a pipe end there are many different ways to measure it for many different reasons. You have major problems when a client specifies a criteria and system of measurement without knowing why or the ramifications then doubles down on "the client is always right" when the people making their product try to give them advice. I spent far more of my life than I like to admit discussing ways to measure the diameter of a pipe end.
 
Since people are not perfect and neither are their machines it goes without saying that all measurements have UNCERTAINTY and in a correctly made measurement that uncertainty is in the terminal digit.
Next the edge (or whatever you are measuring) will always end up between two marked (on the instrument doing the measuring) marked lines. When you read the measurement from the instrument you start from the largest marked point and work your way down one marked line at a time until you come to the two lines that the actual object edge lies between. Here is where you ESTIMATE the value. There are no marked lines here so you have to estimate. So all correctly made measurement contain all digits known with certainty and ONE uncertain digit. These digits 1-9 are SIGNIFICANT DIGITS. ZERO however has TWO functions. ONE of those functions is its use as a PLACEHOLDER. Placeholder zeros are NEVER Significant as they were never read from an instrument. In 93,000,000 miles only the 9 and the 3 are significant the six zeros are place holders and the 3 digit was estimated and is a uncertain digit.
IF a terminal zero is to be significant you have to specifically indicate it so in 3450 cm the zero is a placeholder and the 5 is uncertain and there are 3 significant figures but in 3450. cm the zero is significant and is the uncertain digit so there are 4 significant figures

You are arguing from a theoretical point of view, not a practical point of view.

I've never seen a decimal point without a digit following it. But maybe that's just the engineering world where clarity is required.

if I draw something that is to be 3450mm then I expect it to be 3450mm +/- 0.5mm. The person making that thing will use a measuring device that has markings at 1mm increments, or better.

When I check the thing I would do the same. If the measurement falls between two lines, I determine whether it is closer to the upper of lower measurement. That is, if it is between 3450 and 3451 I don't estimate it to be 3450.3. If it is closer to 3450 it is OK and approved. If it is 3451 it will sometimes be OK, but other times it will be unacceptable - it depends what the thing is for and what, if anything, it fits to.

If the person who built the thing is like you and worked to 3 significant figures and the measurement is not an acceptable measurement, it will be rejected and sent back to be fixed or remade.
 
I cannot speak to the practicalities of the manufacturing or engineering end of things. Perhaps because in the sciences so many different types/kinds of instruments measuring so many different parameters were in use that it became incumbent that the experimenter state in specific terms the degree of uncertainty in each and every measurement. The sole problem is ZERO because of its dual usage both as a simple place-holder and in some cases an actual measurement digit.
We also used a BAR written over the last significant zero so 93,000,Ō00 miles would indicate that the three zeros after the 3 were actually measured and are therefore significant digits. The fourth zero (with the bar over it) is an estimated digit and therefore an uncertain digit though still significant. The last two zeros are placeholders and not significant.
It was also necessary to not have uncertainty wander all over the place so, for example, you made certain to use the same instrument all the time. So one would always use the same balance to measure mass so that its uncertainty was constant and in the same direction rather than one balance that read high followed by using a different one that read low.

I determine whether it is closer to the upper of lower measurement. That is, if it is between 3450 and 3451 I don't estimate it to be 3450.3. If it is closer to 3450 it is OK and approved
That's because you have that 0.5mm tolerance to play with. Your part is acceptable as long as the actual length falls between 3449.5 and 3450.5. So you might not actually write down the estimated .3 but you are cognizant of it none the less. And if that part is 3450.9 it has exceed tolerance and will need to be shaved down a bit.

The difference between us is that you were making something that had a practical direct use. If I'm measuring the speed of light through quartz then I don't have a fixed standard to compare my results against. All I can do is make several measurements and compare their precision. I can't find accuracy unless I have an accepted standard. So I measure to my instruments limits, perhaps 4 significant figures. A better instrumentality may later measure to 5 significant figures or better. As the significant figures increase and reading begin to cluster we can get an accepted value to a stated degree of error.
So the new Kibble Balances give us a Planck's Constant of: 6.62607015 × 10-34 kg⋅m2/s Many years ago I consistently used a value of 6.6262 for common calculations. Today that would be in error
 
1569777791238.png
 
I used to run a CNC lathe many, many years ago( one of the many jobs I tried out but didn't particularly care for).
I used to just measure the parts the way they told me with the calipers they gave me. If the parts were within tolerance they went into the good bucket if not into the reject bin they'd go and id adjust the X or Y axis as needed and try again until good.
Seemed like a fairly simple job at the time.
Never realized making parts could get so complicated.:)
 
I used to run a CNC lathe many, many years ago( one of the many jobs I tried out but didn't particularly care for).
I used to just measure the parts the way they told me with the calipers they gave me. If the parts were within tolerance they went into the good bucket if not into the reject bin they'd go and id adjust the X or Y axis as needed and try again until good.
Seemed like a fairly simple job at the time.
Never realized making parts could get so complicated.:)
You would not believe the hours I have spent in rooms full of engineers discussing go-no go gauges (which is what I presume those calipers were). Also drifts, tapes calipers, lasers and all things used to measure pipe ends. I frequently noted that about a quarter of the people there didn't know what they were measuring and or why. It actually is complicated. What you were doing is quality control, adjusting x and y axis before you had to put stuff in the reject bin is quality assurance.
 
I was told many years ago that the reason some cars ran forever and never had a significant mechanical problem while others were lemons from day one was due to the way in which part tolerances came together randomly in the assembly process. By chance a car would come along where all over tolerances were matched with all under tolerances. Or the reverse where everything was at the limit of too big or too small.
When we rebuilt an engine we'd buy 40 pistons, for example, and weigh and mic each and every one until we had a set of eight that were identical. The rest were returned. The same with valves, push rods, piston rods, rockers. etc. The running engine had just about zero vibration
 
You would not believe the hours I have spent in rooms full of engineers discussing go-no go gauges (which is what I presume those calipers were). Also drifts, tapes calipers, lasers and all things used to measure pipe ends. I frequently noted that about a quarter of the people there didn't know what they were measuring and or why. It actually is complicated. What you were doing is quality control, adjusting x and y axis before you had to put stuff in the reject bin is quality assurance.
Yes, I was just running parts and making minor adjustments to the program as needed to keep them within tolerance.
This was back when I was about 19 or 20 and had just tacken some machining in general and Cnc in particular at the local junior college. It never did occur to me just how much might have gone into getting it to that point.
 
I was told many years ago that the reason some cars ran forever and never had a significant mechanical problem while others were lemons from day one was due to the way in which part tolerances came together randomly in the assembly process. By chance a car would come along where all over tolerances were matched with all under tolerances. Or the reverse where everything was at the limit of too big or too small.
When we rebuilt an engine we'd buy 40 pistons, for example, and weigh and mic each and every one until we had a set of eight that were identical. The rest were returned. The same with valves, push rods, piston rods, rockers. etc. The running engine had just about zero vibration
That is a difference the Japanese (and others) made in production engineering, narrowing down the variance in machined tolerances.
 
I was told many years ago that the reason some cars ran forever and never had a significant mechanical problem while others were lemons from day one was due to the way in which part tolerances came together randomly in the assembly process. By chance a car would come along where all over tolerances were matched with all under tolerances. Or the reverse where everything was at the limit of too big or too small.
When we rebuilt an engine we'd buy 40 pistons, for example, and weigh and mic each and every one until we had a set of eight that were identical. The rest were returned. The same with valves, push rods, piston rods, rockers. etc. The running engine had just about zero vibration

I've been circle track racing for 30 years, built many engines myself, and been in on building a bunch more.
The approach of sending back more parts than you buy would make you a unwanted customer/client with any parts suppliers I've dealt with, and would quickly get you blackballed with any other supplier. These guys do talk to each other you know.

The way me and my friends balance a engine is balance a set of rods, small and big ends, by machining the pads on the ends intended for that, until we get the rods as close to equal as possible, then do the same with the pistons.
Then put the lightest rod with the heaviest piston, etc until we assemble the engine, if there's any difference left, it's corrected by balancing the crank dynamically.

Big NASCAR race shops do buy pistons, etc. in big bunches, and try to sort them out by equal weights as close as possible. Then use the same approach we do.
They only return defective parts.
 
In 1988 I bought in the same day from the same car dealer two absolutely identical Fiat Uno

Fiat_Uno_2_p_S_Chiara.jpg


one for me and one for my Wife, as they were very handy to use in the crowded and narrow streets of my City.

b-b-l-antica-torre.jpg


After three years my Wife told me "..bring my car to the garage, the crank for lifting the driver's glass is broken…"
Exactly three days after the crank of my Uno car broke also.
Has anybody heard of " programmated obsolescence"?
 
In 1988 I bought in the same day from the same car dealer two absolutely identical Fiat Uno

View attachment 554704

one for me and one for my Wife, as they were very handy to use in the crowded and narrow streets of my City.

View attachment 554702

After three years my Wife told me "..bring my car to the garage, the crank for lifting the driver's glass is broken…"
Exactly three days after the crank of my Uno car broke also.
Has anybody heard of " programmated obsolescence"?
It is caused by the genetically engineered crank weevil that takes three years to munch through a crank.
 
I was told many years ago that the reason some cars ran forever and never had a significant mechanical problem while others were lemons from day one was due to the way in which part tolerances came together randomly in the assembly process. By chance a car would come along where all over tolerances were matched with all under tolerances. Or the reverse where everything was at the limit of too big or too small.
When we rebuilt an engine we'd buy 40 pistons, for example, and weigh and mic each and every one until we had a set of eight that were identical. The rest were returned. The same with valves, push rods, piston rods, rockers. etc. The running engine had just about zero vibration
I think there's alot of truth that with some bad luck you can get a hold of a car that just by chance ended up with a whole lotta parts that are at the edge of tolerance( and maybe a few that are over) and therefore a real lemon.
I've never heard anything but good things about Subarus so we bought one. Mistake.
We got it brand new and within a few months it had engine troubles, brakes trouble, transmission trouble, and died occasionally when pulling up to a light.
Took it back to the dealer a couple times and even they couldn't fix it. We just traded it in on a Toyota truck a couple weeks ago.
 
The approach of sending back more parts than you buy would make you a unwanted customer/client with any parts suppliers
Well its been close to 60 years so I can 'fess up'. In Chicago one of the biggest auto parts retailers was Warshawsky or J C Whitney. One of my best friends was a Parts Manager. Everything I bought was through him. Generally all parts were graded as Good - Better - Best. So I'd call him and order Good and he'd put Best in the box. So while I took home 30 of their Best pistons I was only charged for the 8 I kept the rest went back on Monday and were never charged.
Had another friend in a tire retailer shop. I'd order the smallest cheapest and he'd bring out the biggest and best tires they had in the store.
Was the only way I could afford that beast
 
It shouldn't be, from a pilot's perspective. All you're looking for is the number on approach. I fly a Tiger Moth in mph, a Nanchang in km/h and any GA aircraft in knots, never had any problem as e.g. approach speed is 60 in the Tiger - 150 in the 'chang.
The Spanish Air Force flew both T-6's and SNJ's. They kept mph in the T-6's and knots in the SNJ's. I wonder if they kept a single type in a particular squadron but I don't know. The training school used T-6G's so they would have been consistent in mph.
 

Users who are viewing this thread

Back