Calibrating a Sand Hourglass

Add or remove sand...

Presumably one ond of the hourglass could be left with a small acess hole which is then sealed with a bung or somesuch...

Del

If it does not have an opening, then the only thing I can think of is changing the gravitational constant of the universe.

No,No,No, you have to determine the accuracy of hourglass within 1 second uncertainty, and usually that type of devices are one-piece (no holes). The glassmith hot seals the holes after the sand is inside.

A bit more imagination!

Use an NIST-traceable stopwatch and thermometer.

How? Your hands are NIST-traceable?

How can you press the start button on the stop watch and tilt the hourglass at the same time? What's your bandwidth? and your uncertainty?

Human reaction time to unexpected events is 1/5 sec; when the event is expected, it is near instantaneous. Well within the 1 sec prescribed.

Human reaction time to unexpected events is 1/5 sec; when the event is expected, it is near instantaneous.?

Are you sure?

Mmmmm.... Our prime minister just yesterday was still saying, as from months ago, that we have no crisis about oil rising prices, unemployment growing every month, inflation higher than in the last 15 years...

Do you mean our prime minister isn't human?

Do you want to start III world war?

George W. Bush will be available here in about seven months. Think things will be better with him?

Help! Please stop the world I'm gonna get off!!!

Thank you anyway for the offer, but I prefer politics that lie in my mother language. I can understand them best and that allows to defend myself accordingly.

If we are still refering to no watches, but we do have timers that are said to be "on the money" I would start out with a specified amount of sand in specific sized hourglass and time it with the "standard" hourglass. The opening in the hourglass would be started slightly larger than prescribed. Once the timing is finished you could spot heat the center portion to close it off more to adjust the flow of sand. The tough part would be if it closed too much and you had to expand the hole. You could break the glass to retrieve the sand and start over or you could cheat and adjust the Standard to match the new one. However after doing this over the years, the standard 3 minute timer may be somewhere around 13.76 minutes give or take a few grains.

As a glass blower, you can't spot heat the center. You will crack it to pieces even if you use borosilicate glass. You would have to preheat the whole piece, work it, then stress relieve it for 24 hours at around 1050F, allowing it to cool slowly in the furnace. I truely respect a gaffer that can got that hole the same size everytime.

I've not seen hourglasses hand blown, but I'm sure it takes much skill.

Hey! were talking about a world that has no watches. I'm sure we could figure out how to spot heat a tiny piece of glass. Of course the sand would probably start to turn into more glass and really screw things up. How about we make them out of plastic? Does this stuff exist in this world. Then we could just leave the top open and drill through the center to adjust..... then oh well just hire some guy to count to 540 seconds and then remove the eggs. A good second counter could probably make at least 30 cents an hour!

I'd say that the repeatability of the period would depend on the consistency of the sand. You'd need to ensure a uniform grain size (and maybe shape?) as I reckon a few 'big' or 'long' grains together (or even one 'extra-big' grain) going through the orifice would have different effects depending on whether they were at the bottom or top of the heap. How much difference it would make, I haven't a clue.

If you drop it from a known height you can easily predict when time runs out for the hourglass well within one second.

The downside is this only works once.

It's a smashing suggestion, though!

Tilt the thing at a predetermined angle.

O yes for calibrating without a watch one can use a pendulum of known length.

Happy counting. Or use ® Viagra the only medicine that counts.

I don't live in a country that does not allow clocks so i can use one!!!

Hourglasses are calibrated by the maker then sealed. If left unsealed so that some one could change or correct the calibration the sand would draw moisture which would vary the time.

The thread says 'How would yourself calibrate one of these devices' well I've said how I would do it...off for a nap now...

Del

Purchase your hourglass from Rolex.

Well, it seems I'm having my usual problems with English language! (By the way, why don't you write in Spanish, it's easy... I was able to read it just 3 years old!!!)

Del has pointed one of the main constraints: the calibration must be done by yourself. This was intentionally written to avoid answers like "send it to a certified laboratory".

To buy it from any recognized supplier isn't a valid answer. Even most sophisticated measuring instruments have to be calibrated at regular intervals.

So, adjusting sand before closing the glass could be valid for the manufacturing, but not for repetitive calibration.

And finally, Where the heck my initial post forbids the use of watches or any other clock type? I'm a freedom lover and I would never would do such stupid things.

A bit more brain straining!

Esta es porque nosotros utaliza inglase! no puedmos escribe en español! Yo viva circa Almeria para la ultima ocho años y tengo problemos grande con tu idioma!!! Los cientos! (Duloroso para te yo si!) Tu inglaise es cien vezes mejor de mi español intonsis, para mi, puede sige en inglase? PFV

I won't even attempt to spell check that!

Anyway, hope all is well in Madrid

El Niño de Sheffield

Where the heck my initial post forbids the use of watches or any other clock type?

Refresh memory here.

Specifically (with emphasis added): "You're a cook in a restaurant in a quaint country where clocks are outlawed. You have a four minute hourglass, a seven minute hourglass, and a pot of boiling water. A regular customer orders a nine-minute egg, and you know this person to be extremely picky and will not like it if you overcook or undercook the egg, even by a few seconds. What is the least amount of time it will take to prepare the egg, and how will you do it?

I'm a freedom lover and I would never would do such stupid things[?]

So why would a freedom lover with watches or any other clock type choose to use a sand glass, or give up free time and liberty to calibrate the un-calibrate-able, a sand glass? Unless seized in an episode of stupidity?

Other...

To buy it from any recognized supplier isn't a valid answer. Even most sophisticated measuring instruments have to be calibrated at regular intervals.

But it is a valid joke; except, maybe, for someone who knows were to find a "recognized" "supplier" of "sophisticated" hour (or minute) glasses. Or services that will calibrate these at regular intervals. Or someone who would hope to seek such services but not be laughed at.

adjusting sand before closing the glass could be valid for the manufacturing, but not for repetitive calibration

Adjusting a timer glass, other than as part of a joke, could easily be done, if doing so ever made any sense, which it wouldn't. You could simply manufacture in a filler hole, sealed with plug like a piggy bank. But even this wouldn't make good sense because you could still not attain accurate or reliable precision you were seeking. In their time, sand glasses were adequate because (1) their precision (or lack thereof) was known (or didn't matter) and (2) they were adequate for almost all needed timing tasks--measurements of seconds or fractional seconds within minutes, hours, or days,...were not a pressing need, if at all. Since then, until today, timing glasses have been abandon and relegated as novelty items as the need, and the capability, for more precise timing has arisen and become widespread. They do retain some utility for the timing of processes where only rough precision is needed, as with the timing of egg boiling. In such instances, they can even be preferred due to (offsetting) advantages enjoyed over more advance, more precise timing instruments: (a) They are easiest to operate; (b) They are easy to see (and no design enhancement is required to make them more viewable); (c) they are very simple and durable with very long wear-out times compared to precision instruments; (d) they need neither repair or calibration--because their cheapness makes it more feasible and cost-effective--thus there is no alternative--to repair or calibrate by disposal and replacement (they could, in fact, be thought of as an epitome as regards reliability-availability-maintainability); they are completely, totally, and directly recyclable, with no possible harm to the environment.

It is good...the English learning thing; with continued effort towards mastery you will come to appreciate that language's benefits for abstract expression and technical/logical conversation. It is also good that you have a site such as this at which to hone not only your English proficiency, but also you logic and reasoning proficiency.

Guest, your "Refresh memory here" link was to another thread.

Granted, but was not that other thread starter invoked here, in Kwetz' thread starter? And, given no instruction to the contrary, may we not—nay, should we not—take the other thread's premise to have also been invoked...here? Not that the omission made any difference...

"And, given no instruction to the contrary, may we not—nay, should we not—take the other thread's premise to have also been invoked...here?"

NO!

A) Without an explicit reference by the OP cross-linking the threads, each should be considered as a separate discussion. I, for one, had not read the 'other thread' until stumbling on your red herring link. The other thread was, indeed, mentioned, but not invoked as a set of preconditions for this thread.

B) Would you apply that logic to an exam paper? (... but question 3 gave boundary conditions, so I applied them in my answer to question 4). To problems at work?

[Edit - see also #16]

A persistant difficulty on these fora seems to be the inability to know when to let an innocuous point made remain just that: a point made.

It appears--I don't say this to prove anyone wrong, only to encourage your dispassionate second look--that the meaning of "red herring" was not clearly understood; and that the example posited about an examination does not prove the point attempted to be made, but is, in fact, an actual red herring (though apparently not intentionally so).

An approprate analogy might posit an exam question making allusion to the study material (perhaps an explanation with example in the text book), not to another question in the text (which, in effect, would be deemed to be a separate [uninvoked] matter). There (with such a non-fallacious analogy) a correctly reasoning examinee would be justified to hold, that any premise in the text book material being alluded to, applied also in the exam question, unless it was stated otherwise.

Moreover, in your quest to disparage or rebut, rather than simply seek to arrive at common consensus, you overlooked the one argument which would have validly supported your dissent without resort to fallacy or group think--the valid argument which would have been debatable (i.e., which I would have been justifiably willing to concede in part).

Then again, your throwing in of the exam red herring might have been a knowingly considered debate tactic meant to bring (what you thought to be approprate) ridicule to what you "understood" to have been a red herring, in which case you cannot not be faulted for failing to realize that Kwetz would not be well served by the attempt.

Um, er, snort, what?

snort

....Famous last word.

Sorry, you miss the real post. Go to the top of page and read it again carefully.

You're talking on another thread which give me the idea for this one...

I think I'm not who has to refresh memory

Best regards

Good afternoon "Guest":

Question:

So why would a freedom lover with watches or any other clock type choose to use a sand glass, or give up free time and liberty to calibrate the un-calibrate-able, a sand glass? Unless seized in an episode of stupidity?

Answer: Just because he is FREE. Did I need to watch a dictionary to see if FREEDOM has a different meaning than the Spanish word LIBERTAD?

Could you perhaps help me with my poor English? Perhaps then I would understand the relationship between freedom to use an hourglass and episodes of stupidity?

It is good...the English learning thing; with continued effort towards mastery you will come to appreciate that language's benefits for abstract expression and technical/logical conversation. It is also good that you have a site such as this at which to hone not only your English proficiency, but also you logic and reasoning proficiency.

I absolutely agree with you, not only the English learning, but any other language that allows people try to understand each other. And of course, I'm registered on this site because I'm sure I'm in no way omniscient and I think one should learn as much as possible: There is no freedom without knowledge.

Best regards.

I share your sentiments.

Best regards back. Hasta luego

OK, how about:

1) An electromechanical "turner" to invert the hourglass.

2) A light beam focused through the waist, illuminating a phototransistor.

3) Output of phototransistor (taken to an ADC), & plotted against time.

4) (Obvious) - timing mark on plot to indicate start point. Presumably when turning move finishes. The move would have to be swift and repeatable.

Initially, visual inspection used to determine the 'end point' on the plot. Once the general form of the output has been observed, an algorithm could be generated to find the end point automatically, and thus automate the measurement process.

May need to experiment to see whether temperature and/or barometric pressure significantly affect the timing period - in which case, calibration would have to be done at standard conditions (or a correction factor applied).

Should also repeat the measurement with sand flow in the other direction.

Well, one answer that really try to solve the question!!

Of course, temperature and barometric pressure should be recorded. All certified laboratories indicate in their certificates the ambient temperature, barometric pressure and relative humidity. And have you noted that paragraph usually included which says more or less "the results herein expressed are related to this device in the time and conditions...." that assures you that your device was OK at the calibration moment but don't assures the same one nanosecond after?

Noted!

I didn't include the RH on the basis that the envelope should be sufficiently impermeable for RH not to affect the result, but it should, indeed, be recorded - if for no other reason than that it may (in some obscure way) affect the calibration system.

Hour glasses?! (as in cremated eggs?) Or minute (+/-) glasses?

As the timer is being calibrated, one end would be open to atmosphere until tests are completed

Time the sand runout with three different amounts of sand. Weigh each amount or total weights. Extrapolate the curve created to reach the time you require and read off the amount required. Proof check and extrapolate again if necessary closer to your required point. Don't breathe or you will change the humidity level.

Unfortunately if you take over 4 minutes to do the test, you won't be conscious to see the final result.

I assume you mean adjusting it. If you vary the temperature, you change the viscosity of the air inside. The viscosity of air varies as the 3/2 power of temperature (Sutherland's formula). A higher viscosity of the air makes the sand flow slower and increases the hourglass time.

Calibration: The act of checking or adjusting (by comparison with a standard) the accuracy of a measuring instrument (Webster)

I mean just the first (checking) because of periodicity of such action and due that hourglasses (at least al I've seen) aren't adjustable.

If the "hourglass" was placed on a variable heat warming surface, the changes in temperature of the sand, glass, and air inside the glass should change the time for complete reversal of the sand by small repeatable amounts. That should accomplish the OP question.

Now for another question. If the hourglass is inverted every three minuets for ten hours a day for ten years, how much wear will take place in the restricted area of the glass? How will that affect the timing?

Well, some folks are getting bored so I'm going to tell you in what I was thinking.

One of the tasks I've had was to keep the measuring devices calibration system of the company. I had to suffer some "certification audits" conducted by auditors that were rather public notaries than engineers. When I saw the post cited, I thought: What could I do if one of the devices under my responsibility were an hourglass?

My own response was: Take a certified calibrated digital stopwatch with a resolution of 0,01 sec (just normal one), the hourglass, a camcorder and a laptop with some video software.

Put the hourglass just near the stopwatch and the camcorder recording both. After start recording, start the stopwatch and then tilt the hourglass, when all the sand had fallen, tilt again the hourglass. This operation should be performed about 10 times to perform an uncertainty analysis Type A according to GUM.

Then look at the video recorded, use the low speed playing (even one picture a time) and zoom capabilities of most video software to detect the moment in which the first sand grain falls. Pause the image and look the time showed by the watch. Make the same for the last sand grain and get the time difference: it's exactly the time of the hourglass. Resolution of camcorder when viewing video just picture by picture should be between 0,04 and 0,05 sec. Of course we can use a high speed recorder to improve that but for a simple hourglass I think it's enough.

Sorry but I'm too lazy to make now the calculation of uncertainties.

Best regards

Only big problem I can see with that is "... detect the moment in which the first sand grain falls ..."

This will happen while the hourglass is being turned over. Your problem is going to be getting repeatability on the turnover speed & pinpointing the start time (which is why I went for a mechanical "turner-over").

You're right, but my real problem is quite different. I've got all devices named but no money to buy a mechanical or electrical "turn over" device....

Anyway, even when turning it over by hand, the only caution is to keep the fingers far from the hourglass neck and not to hide it from the camcorder.

I've made hundreds of data recordings when PCs had just 640k RAM and HD just 20 MB and I had to save memory space because the software recorded data on the memory and just save it to HD on finishing after asking "Save data (Y/N)". My "maximum repeatability error" (time between I started recording and I started measurements) was less than 0,5 sec. Note I recorded data at 1kHz/channel sampling rate.

The ten times I proposed to repeat the measurement accounts for repeatability too.

Let me see if I understand. You have two glasses. You want to calibrate one (either one, or both) so that between the two you can be certain of required egg cook time with error of one second or less?

Nooooo. I haven't glasses. I won't calibrate anything. I'm just a poor stupid man asking stupid questions.

I'm just punishing myself writing one thousand times "I'll never post stupid questions"

From now on I will ask just things as "How does affect the datum position to determine pump NPSH?"

Thank you very much to all you folks.

Chill, Kwetz, there are idiots everywhere .

I reckon you could knock up a mechanical "turner-over" device with a few bits of junk, if you look in your local junk shops, skips, (?dumpsters) or whatever's the Spanish equivalent of a local dump. Just something to act as a chuck to hold it, & two detents to stop it at ±180º, plus a barrel that you could roll (with the flat of your hand) would do the trick.

What trick? Turning the glass over? This might work if the glass could be turned exactly one per second for an even number of seconds--one end would be empty if calibrated, not empty if not...but how do you calibrate the turner if the glass is not calibrated? Kwetz is not idiotic--just chagrined.

Maybe sedate the customer with sodium pentathol. Add/subtact sand grains as you cook eggs, one by one, respectively. When the customer smiles and/or doesn't snarl or puke, the hour glasses are calibrated.

"Idiot" was not in reference to Kwetz.

The trick in question is a method of rapidly inverting the hourglass without obstructing the line of sight from the camera to the hourglass and/or reference timer. The camera records the time. It is not necessary to control the timing of the turning device, just to ensure that it turns rapidly so that the start point can be established repeatably.

Aha! But don't you think that, saying, "rapidly," you are introducing a variable to replace an impediment? One that might not be able to be "trimmed down" to satisfy the one-sec-max posited tolerance? I realize you have clipped and attached a different thread to this one--in this one it was reminded that the other forum had prescribed "no clocks" other than the minute glasses--that, accordingly, the glass or glasses would need to be calibrated by their own properties--but let's go with the rapid flip idea anyway.

First we take for granted (by assumption) that the assisted-flip duration will be exactly equal from flip to flip, but also that it cannot be instantaneous, which, as we shall see is probably not the worst thing. Within the margin of error specified by Kwetz, we have, with any one flip, two (emergent/incipient) situations, even allowing for use of a "clock" such as the digital camera. If while calibrating the speed of flip by testing incrementally from the low side of an ideal flip speed (I say ideal because, before the fact [that may or may not reveal itself], we can't know the "genuine" (the one and only) flip speed exactly...), it could be that we never arrive at an ideal time that corresponds to exact time because the "experiment" requires multiple flips during which any error will compound over successive flips...and one second is a very fine tolerance, even for no, or just one, flip, leave alone multiple flips.

Now let's consider the high side, as calibration flips are made more and more rapidly. As we approach instantaneous flip time duration (even while still at a point far from it), can we expect that the ever increasing speed of a flip will have no effect on the sand being hurled up and around as the glass rotates and then stops suddenly? And can we expect that the collection of sand particles will behave exactly identically, individually and collectively, with each flip? Would it not be possible--nay, likely--that the increasing rapidity with which the sand particle are flipped could introduce variation in (say) the elapsed time before the first grain, and succeeding grains after that, makes it to and through the waist? And what of variations in the arrangement and massing of particles that were just hurled rapidly around? Can these be made to settle and begin their descent exactly the same with each flip, no matter how precisely controlled? And, wouldn't there also be the chance of full or partial bottle necking as the sand (much of it all "bunched up and inpinging on one side of the glass) begins to change direction and fall? (I have on more than one occasion, in fact, cooked using a 2-minute glass...and watched the food begin to overcook as I wondered why the sand was taking so long...until realizing that sand grains had blocked the way for each other and for the remaining, like Abbot and Costello going through a door. I believe this real chance of blockage, or sufficient blockage to impede getting the flow going steadily, is the reason one might see the instrument typically set down with a bit of a thump--to get the flow going smoothly thoughout the sand "column." It--this variablity in initial sand column descent speed--is probably also a reason that timer glasses are imprecise at best, with probably little chance of measuring within a second reliably and consistently...even assuming you could visually (and instantaneously, with no reaction time) mark the last instant that the last grain departed the top chamber...for every flip.

So I believe that Kwetz's winsome self recriminations were simply his belated way of conceding the "infeasiblity" of his original quest, given the inherent inaccuracy and uncertainties with timer glasses.

And yes, I got it about your "idiot" talk levity; my rejoinder in kind was not intended to be taken literally. (But I was counting on Kwetz, not others, not to catch on.)

Please, don't get angry.

I never recriminated myself, just was a sort of joke when after more than 30 posts somebody (joking too or not reading previous posts, I don't mind) says things like " send it to a laboratory" or " so, you have 2 hourglasses and want to calibrate one using the other?"

I know the inaccuracy of sand timers. The thread I posted was just to make a funny (at least I thought so) exercise on an absurd premise. Nobody who really needs measuring time with a minimum accuracy would use that devices.

I don't think somebody insulted me in this thread, but even if someone has tried to, he failed: as we say in Spain "Only can insult those who can, not those who want"

And finally a technicality: One sand glass timer can have (after calibrating) an accuracy of... let's say 9 ± 2,3 seconds but determined with a method whose uncertainty is less than a second.

Best regards

P.D Please when you participate with several posts in the same thread I recommend you to register (if you aren't yet) or not to post anonymously. In the night all "Guests" seems equal.

You would have to rotate the hourglass at the same speed as it would be normally turned! If you turned it rapidly, the sand would be held momentarily in the bowl at the top due to the centrifugal force (like the bucket of water span upside down trick! ) then all of it would drop at once and the cycle would start!

You're on to something John! If the whole system was placed on top of a weight sensor attached to a computer (for example), when the thing was spun round, you would get a definite trigger point when the sand hits the bottom of the top glass, coupled with some optical trickery to detect the last grain etc. Your a genius John!

I'm getting horribly close to wasting a weekend building a lash-up of this thing.

Two three current problems, tho': 1) I don't have the factory keys this week-end, and don't have enough workshop gubbins at home to do it in a reasonable time; 2) I've got to spend a considerable chunk of the weekend taxiing Small to various classes, rehearsals and competitions; 3) I've got a mountain of work to get through between taxi duties.

BTW - if my light interruption method was arranged along the axis of rotation, it could detect both the first and last sand grains passing through the orifice. One could experiment with the inversion time (going back to an electro-mechanical system¹, (which Kwetz has already ruled out!), or just trying handraulic 'slow' and 'fast'). I'd say about ¼ second for the 180° feels about right. As has been said, too fast would be unrealistic, and probably give false results (but probably not outwith the normal range of variation due to all the other factors mentioned in the thread).

¹ This could also be achieved mechanically, with all manner of contrivances (weights & pulleys, escapements, dampers, spring power etc.) - it's just that electro-mech is easier to control.

Send the hour glass to a calibration lab and let them figure out how to calibrate it.

Well, there was a time when you could just dial POP-CORN and get the time to synchronize anything!

Hello Kwetz

Hourglass Variations over time:

The biggest problem with using sand in an hourglass is that over time there are alterations to the geometry of what you thought was going to be constant.

Apart from room temperature variations, which do alter the hourglass timing, there are three other factors which should be immediately evident:

1. As the sand grains jostle and rub against each other, over time the sand grains become smaller, thus the "hourglass timing" becomes shorter, until over many millions of uses of the hourglass, the sand becomes so abraded it becomes a powder, and cakes together, refusing to run through the neck of the hourglass.
2. Likewise, as the sand grains run through the hourglass neck, they abrade the neck inner diameter, enlarging it, also causing a shortened time as measured by the hourglass. This time alteration will eventually decrease to a minimum, when the sand particles are so small they are a fine powder, and cake in the upper bulb.
3. As the neck inner diameter is abraded, the glass surface becomes not so smooth, slowing down the movement of falling sand grains, lengthening the time measured by the hourglass, which counteracts to some degree (2) above.

Kind Regards....

Is the bottom if that glass about to suffer a ruptured aneurism? Or does it just have gas?

Hello Guest,
Glad you noticed that, and it is easily seen that you have a keen pair of

The movement in that animated picture, of the black frame, independently of the brown frame, indicates the instantaneous (or nearly so), 180 degree "flip" of the hourglass, so the unit is inverted and re-timing commences once again.

Kind Regards....

Run for your lives! A beam of enlightenment from a lighthouse has changed me into giant, disembodied Clippy.

Before you go, thanks. I should have been wearing my infinite-speed-or-nearly-so headband strobe light.

Hello Sparkstation

You're completely right. All sand in this planet have been forming by this process from thousands years ago. I thank you to put this technical answer that is a good example of failure analysis mode.

Anyway even I've said it's a strange thing to calibrate an hourglass, have in mind that all measuring instruments degrade in some extent according to use and time. That's the reason by which all standards relating to measuring devices ask for periodic verification of their accuracy (calibration). And one more thing that usually forget those notary-like auditors, the calibration that really validates a measurement is that made before. Calibration of a device indicates you its accuracy and uncertainty at the moment and conditions of calibration, but not a nanosecond later! So the next calibration indicates that in the meanwhile, there is a high degree of confidence on the device correct function.

Best regards

Hey, didn't Fred Flintstone have an hourglass wristwatch? Is that the reason why Mr. Stone was always yelling at him for being late for work?

Calibrate it against your sun dial.

Even better, integrate the glass into the sun clock's stylus. Then calibrate them both, reciprocally.

Well, this brings a whole new meaning to the saying, "Pounding sand down a rat hole!"

If I use stopwatch and video camera to calibrate hourglass then use recording to confirm sand stop and start times do I have to have the video camera calibrated to confirm exact time span ?

NO! The camera is recording the hourglass and the time shown on an independent timer (which should be calibrated).