europa-list
[Top] [All Lists]

electronic scales [was: NEW BUILDER - SCOTLAND]

Subject: electronic scales [was: NEW BUILDER - SCOTLAND]
From: Rowland & Wilma Carson <rowil@clara.net>
Date: Wed, 17 Nov 1999 19:26:22

>  digital postal scales, battery powered.
>Resolution is 1 gm which is plenty fine enough for Europa epoxy work.

That may be an over-optimistic assumption.

First, resolution is not the same as accuracy. To determine the 
suitability of any electronic scales for resin proportioning you need 
to know the amount of uncertainty associated with any displayed 
value. In the manufacturer's specification, this may be stated in 
terms of absolute or percentage accuracy, or by reference to 
linearity and repeatability.

Earlier this year an SP Systems representative (supplier of the 
Ampreg 20 resin used in the UK-sourced kits) told us that for Ampreg 
20, acceptable strength is obtained with mix ratios between 101:24.5 
(4.122:1) and 99:25.5 (3.882:1) by weight. Thus for optimum strength 
the ratio should be within 3.06% (resin-heavy) or 2.94% 
(hardener-heavy) of the theoretical value.

Assume 1-gram resolution scales are used to mix a nominal 125gram 
batch of epoxy. If the scales are absolutely accurate (most unlikely) 
the hardener displayed as 25g could be anything between 24.5g and 
25.5g, and the "100g" resin 99.5g to 100.5g. This gives worst-case 
mix ratios of 3.902:1 and 4.102:1 - both within the acceptable range.

Then we look at the scale manufacturer's specification sheet and find 
that the accuracy is quoted as (say) +/- 1% or +/- 1 count, whichever 
is greater. So any weighing up to 100g can be in error by up to 1g. 
Cheap postal scales may have an even slacker spec than this.

So in reality the hardener portion displayed as 25g could be anything 
---From 23.5g to 26.5g, and the "100g" resin could be 98.5g to 101.5g, 
giving worst-case mix ratios of 3.717:1 to 4.319:1 - no longer in the 
optimum range.

Of course, if you use the same dispensing technique for both 
components (eg stop adding drops immediately the last digit changes 
to the desired value), you are much less likely to get ratios at the 
extreme ends of the range. And the linearity of your balance may be 
such that errors are all in the same sense and therefore can be 
neglected. In fact, if the scale is truly linear, and the 
repeatability is good, then the absolute accuracy hardly matters! But 
these are areas not usually well-documented by manufacturers at this 
end of the market, so we are back to the accuracy specification.

My own choice was the Ohaus LS200 (200g by 0.1g for under 100 pounds 
UK). The accuracy specification (+/- .75% or +/- 3 counts, whichever 
greater) allows me to mix epoxy batches using as little as 20g 
hardener / 80g resin and be confident they are within 2.5% of the 
theoretical ratio. If I went down to 10g hardener / 40g resin then 
the ratio could be in error by up to 4%, which I'd prefer not to risk 
- and I may not need such small batches anyway. On the other hand, 
larger batches of 40g hardener / 160g resin should be within 1.5% of 
the theoretical ratio.

Some of you may now have no doubts about why, on a team-building 
course in my former existence, the assessments of my personal 
characteristics included the term "pernickety". But I console myself 
that I've never heard anyone complain about me being too pernickety 
as I pre-flight the rented aeroplane they are about to accompany me 
in as a passenger ....

Having said all that about balances, I may finish up getting a resin 
pump as well, after having enjoyed the convenience of one on the 
Europa factory Weekend Workshop course.

Hope this helps.

Rowland

           ... that's Rowland with a 'w' ...


<Prev in Thread] Current Thread [Next in Thread>