specific gravity

Quite some time ago I stumbled onto an article on specific gravity on the internet that I had to leave before I got to read it all the way through, and, of course, I didn't bookmark it. The article covered the usual methods and contained info on building your own gear. I believe the article was written by Alysson Rowans. I have not been able to relocate the article. If anyone can point me in the right direction I would certainly appreciate it.
Thanks
Gord

[www.nexus08.clara.co.uk]

John Sinkansas's "Mineralogy for Amateurs" has a section on building your own SG balance which I built when I was a kid, worked great. You might want to try that.

John Bett's gave me a tip about two years ago that I've found incredibly useful. It's so simple and so easy to do and actually fairly accurate - at least for my purposes. You don't need all the fancy equipment that's sold unless you're doing detailed lab work. It's so easy in fact that you can do it while they're in the field. All you need is an electronic scale and a container to hold water. I usually set my scale to read "carats" rather than "grams" or "ounces". Since the S.G. is a ratio to generally two significant digits, setting it to "grams" works out fine. Also, since it's only two significant digits - just assume the the S.G for the water is "1" without factoring in any temperature difference from 20C.

First measure the weight (mass) of the sample and write the number down.

Second, place the container for water on the scale and zero it out. At this point you're going to dip your sample - I've used dental floss in the past to hold the specimen - but have since worked up a little thin aluminum plate with string (dental floss again). I usually dip the plate in the container as well and re-zero. I then place the the sample on plate and dip it in the container and take my second reading and write the number down.

At this point it is simple math. Take your first value and divide it by the second. This will give you a S.G. value that is close enough for most of what you need to for identification...along with other characteristics like hardness, cleavage, crystal system, etc. This is a simple and quick way to get at a S.G. for most identifications.

If you need to get more than two significant digits, then go ahead and knock yourself out with a more elaborate and expensive system.

Thanks to all of you.
David, that is the article I've been hunting for. Reiner, I will try and find this book, not sure if its still in print. Good info Steve, I know I'll be refering back to it once I get started.
Thanks again
Gord

Yes, I use John Bett's method as well. It's very easy and pretty accurate. I use a digital postal scale that's available for $10 - $15.

The method was posted on his (John Betts Fine Minerals) website homepage on 6/9/09, and is still there in his archive. (I'm taking the liberty of pasting it below):

"I have developed a simple technique for determining specific gravity using a digital kitchen scale:

1. Turn on power and the scale should "zero" itself.
2. Weigh the specimen and record the weight.
3. Find a container large enough for the specimen and fill with water with room enough to submerge the specimen.
4. Place the water-filled container on the scale and "zero" it out.
5. Suspend the specimen from a wire or paperclip in the water, but not touching the bottom or sides.
6. Record the weight reading.
7. Divide the weight in #6 into the weight in #2 and you get specific gravity.

Do not ask me how or why it works. The explanation will only confuse you. All you need to know is that it works, it is fast, it is easy and does not require any lab. equipment. This technique is accurate to the first decimal place. And that will narrow down the possible mineral species of an unknown specimen making it fairly easy to get an accurate identification in your own home without spending money for testing."

Accurate or precise ?

John Sinkansas's "Mineralogy for Amateurs" is not in print, but I keep hearing that there will be a new edition of the book, For people who are serious about mineral id, Don Peck's book is excellent (published by Mineralogical Record)

You should find not only the how to's of using a range of methods for determining specific gravity (relative density) in my article, but how to construct the instrumentation.


The Jolly Balance, for example, if the spring is carefully matched to the sample size, is capable of a precision of the order of better than 0.1% - and that is just using scrap materials and common workshop tools. At that level of precision, the difference between tap water and distilled water will be measurable.

Please remember that the article is a work in progress, and is still not actually complete (yet).

When you invited me to see your etchings, I didn't expect to see so much degraded quartz. Really.

I believe accurate is the right term for the shortcut (John Betts) method, because it is fairly reproducible, and my measurements of known minerals compare well with published data. It is not precise because it only weighs to 0.1 g and so most determinations are only good to two significant digits. I don't doubt at all that a Jolly Balance is more precise, but really, if your tap water introduces a 1% error it probably has a darn high dissolved solids content.

Speaking of "reproducible", you throw in another term with yet another meaning. In fact, there is no way for you to determine the reproducibility of the technique, but you might well have determined its repeatability.

The difference between tap water and distilled water will produce a difference in SG of up to 0.1%, depending on the levels of dissolved carbonate. With an accuracy to about 0.1%, a 0.1% difference in working fluid SG is quite a significant detractor from the precision of the technique.

Of course, one must always state the worst-case figures for the uncertainty of a reading. By making your corrections and by using a standard sample (usually a glass bead of known SG) you can reduce the errors introduced by using impure water by a simple correction.

When you invited me to see your etchings, I didn't expect to see so much degraded quartz. Really.

Sorry, I stand corrected on my misuse of the term "reproducible", I did mean repeatable.

To be even more "specific", the temperature of the water is a factor. Over a moderate range of room temperatures, say +20C to +30C, the density of water will vary a couple tenths of a percent.

Gene

Alysson, it was the Jolly that caught my eye the first time around and I remember thinking how easy it would be to build. I'm looking forward to the rest of the article when your done! Kelly, I have used digital scales before along with a 50 ml grad with OK results, but I gotta wonder about how close I really am when I do that. I will try your method next time I get the chance. The specimens I have to use with the grad have to be the right size-too small and I can't see the difference in the grad reading and too big and they won't fit. I'm sure there are sample purity issues with larger samples, as well. With the Jolly all that goes away. Should be able to use whatever size I have providing the spring is the right sizing. Can't wait to start.
Thanks Again
Gord

Hypothetically, I have completed doing the same experiment as Kelly. We have compared data and it agrees to, say, a standard deviation. Is it OK for Kelly to now use the term reproducibility?

Gene

As Gene correctly states, the temperature of the working fluid is also important.

Other factors that can affect results to a greater or lesser degree include the amount of air entrained in the fluid ans how well you ensure that there is no air trapped on the surface of the sample. Size, shape and surface texture of sample fragments can have an effect - and if air is trapped on the suface of the fragments (poor wetting) it is possible for the experimental error (correctly known as experimental uncertainty) to exceed the results obtained - rendering your effort useless.

I have seen the SG of lead (dust shot) appear to be only marginally above that of water.

Good technique and consistant methodology counts for an awful lot when performing any scientific measurements.

When you invited me to see your etchings, I didn't expect to see so much degraded quartz. Really.

I have done SG tests using the basic method outlined in John Sinkankas' book "Mineralogy for Amateurs" and have used tap water which contains a lot of calcium, and it was at about the temperature of the ground around the bottom of the well. The way I tested its accuracy, was to simply test its calibration: I first tried a couple of samples (of fairly pure substance, and of roughly the same size; a good clear quartz crystal of about the same size is handy - also use the same type of thin thread that is planned for the unknown) that I knew the SG of, then compared them to the results I got. I was therefore able to figure out the amount of error, and whether that was acceptable for my purposes or not.

The method for determining density that we all learned in school (determine mass in air, suspend object in water in a graduated cylinder) is not very accurate. The major problem, as was cited by Gord, is the uncertainty of the change in level of the water in the graduated cylinder. John Bett's solution is much better. And I like the idea of using dental floss to suspend the object (I use either silk or polyester thread) . Steve's ise pf a small aluminum plate gets rid of the hassle of tying knots - - I like it.

Alysonn, I haven't seen your paper yet, I need to do that. Jolly balances are great, but the springs are so dog-gonned delicate; and I am fumble fingered.

Peter, I learned that "accurate" is the real value (not entirely measurable as errors get included) and precise is how close do repeated measurements group around a point. A measurement can be wildy inaccurate but be very precise. Of course accurate and precise is to be desired. Am I mistaken?

On the subject of springs for Jolly balances - for small specimens, you could always use a nylon spring:

Coil some 5 or 10 pound monofilament fishing line around a small resin or wooden former, allowing a small gap between turns. Holding the coil taut (I use a wooden former with holes drilled to maintain the tension, heat the monofilament carefully in the live steam close to the spout of a boiling kettle (don't use a whistling kettle unless you really enjoy the noise!).

Once the monofilament cools and has been run under cold water for a few seconds, you can remove the former and you should find yourself with a robust spring.

Using different lengths and weights of monofilament will give you a good range of springs. These springs will return to size even after being almost stretched straight.

2¹/₂ pound fishing line is probably the lightest monofilament you will find, heavy duty strimmer (weed whacker) line the heaviest. Outside of this you will need to use your ingenuity!

For the sand-grain sized specimens, a curled horse-hair will give good results (but don't let it get wet!)
For multi-tonne specimens, steel hawser would work. (I have seen one of these used with 4mm steel cable for obtaining the net density of rocks and soils for soil-mechanics studies - a very serious piece of laboratory equipment!)

... ... ... and yes, both accuracy and reproducability are extremely desirable. An accuracy to +/- 0.005 g/g_(w) relative to water is an ideal target as most mineral S.Gs are quoted at X.XX, i.e. 0.01 g/g_(w).

Precision (as in precision engineered), by the way, relates specifically to the systematic reproducability of results when consistent experimental technique is used. (i.e. The equipment produces consistent results which are only affected by differences in the way the equipment is handled?) Reproducability relates to the consistency of your results when actually repeating the measurements.

When you invited me to see your etchings, I didn't expect to see so much degraded quartz. Really.