High Accuracy Weight Cals - A Dumb Question

[Hawaii596]

This is  not my forte, so this may be common sense to some...

Using Class 0 or other high accuracy weights to calibrate Sartorius, Ohaus, etc. high accuracy analytical balances...

One of the procedural requirements is to let the weights soak in the environment where the cal will be done for around 24 hours.  We are a remote lab and bring weights to customer locations, and don't have that luxury. 

For dimensional items it makes sense for thermal mass fluctuations to properly stabilize.  And from my tech, I understand there are some physical considerations regarding buoyancy and apparent mass.  But I'm still not getting how the weight destabilizes on the platter of the analytical balance unless it's properly acquiesced in the cal location. 

Can one of you well versed folks (sincerely) explain to me what characteristic it is that's stabilizing?  I wouldn't want to actually crack a book on this, as I probably have it in my vast library.  But I thought maybe the right person could quickly and easily explain.

Thanks for any help.

[OlDave]

OK, first off, lets clear up one misconception.

Sartorius, Ohaus, etc. high accuracy analytical balances...

Ohaus does not belong in any sentence that includes "high accuracy".

While it is true that the weight will actually vary in size slightly due to the temperature and the amount of air it displaces will change, that isn't the primary source of error that you will see.

A weight that is not properly equilibrated will set up convection currents in the weighing chamber. A cold object will pull warm air downwards towards it and the viscous drag of the air on the weight will make the weight appear heavier than it is. Warm weight, opposite effect.

Now how long is long enough? Good question. How much of a delta is there between your weight and the balance chamber? How many decimal place is the balance display? What is the nominal weight of the calibration mass?

If you are smart, you will keep your weights in a somewhat controlled environment during transport. (Don't leave them in the truck overnight in February) As soon as you get to the work site, open the lid of the box while you are doing the initial paperwork so the weights are already equilibrating to the room temperature.

If you were able to maintain the weights to a somewhat normal temperature during transport, it will probably only take 5 or 10 minutes at the most to put the 200 gram weight on the pan and check the span accuracy of a Mettler XP205 balance (5 place, 220 gram capacity) to its mfg. spec of ±0.00040g at 200g.

The problem becomes a lot more complicated when you are testing the repeatability of the balance though. It has a repeatability spec. of 0.03mg using a 100 gram weight. There, if the weight isn't stable you will see drift that will adversely affect your standard deviation of the 10 replicate weighings. But even this can be (for the most part) compensated by properly zeroing the balance between each weighing.

Does this give you something to work with? How does your SOP actually calibrate the balance? If done correctly, you only even need 1 traceable weight to calibrate a modern electronic balance. The span weight. All the rest are reference indications only.

If you want to really get into the dirt on this one, PM me and we can go into even deeper detail.

[MIRCS]

Ol Dave................The Metrologist

Hawaii....................The Cal tech

And don't forget that these are what you look at in dead weight testing for pressure..........as stated by Ol Dave

[Hawaii596]

My apologies, as I am an electrical/electronic type.  Ask me about the subatomic structures of thermocouples and how it impacts uncertainties and I can provide some useful inputs.  But as a Navy Electronic Tech, precision weight calibrations isn't one of my fortes.  To make me feel better, though, when I showed the posting to the tech (who has much more dimensional background than I) he already understood it - and (by the way) agreed completely with Ol Dave's response.

We don't do much in the way of precision weight cals at this lab.  Mostly large shipping scales using the 50 lb Troemner 'bricks' (as I call them).  Just a few analytical balances; and none are highly process critical.

[MIRCS]

My apologies, as I am an electrical/electronic type.  Ask me about the subatomic structures of thermocouples and how it impacts uncertainties and I can provide some useful inputs.  But as a Navy Electronic Tech, precision weight calibrations isn't one of my fortes.  To make me feel better, though, when I showed the posting to the tech (who has much more dimensional background than I) he already understood it - and (by the way) agreed completely with Ol Dave's response.

We don't do much in the way of precision weight cals at this lab.  Mostly large shipping scales using the 50 lb Troemner 'bricks' (as I call them).  Just a few analytical balances; and none are highly process critical.

It wasn't a knock on ya..........you do the right thing and ask instead of pulling things out of your ass. Bravo Zulu

[OlDave]

No, no knock intended for you Hawaii. I just ragged on you for putting Ohaus in the top tier for balances. In my opinion, only Mettler and Sartorius deserve that ranking. All the others are a distant third.

Mass is one of the few measurement areas that I classify myself slightly above average in so if you have any more questions feel free to ask.

[Hawaii596]

My misnomer description of the best brand names (which my tech also harrassed me about) should be a description of my background in that area.  No offense taken.  I know about a lot of disciplines, and there are a whole lot more that I am not so expert in.  That is, in my opinion, one of the great values of such a forum as this.  Thanks for the assistance.