MSA Studies on a scale that is used in the calculation of Specific Gravity (SG)

X

Xman

MSA Studies

I am trying to make sure that I have all of my MSA studies covered off before my next QS9k surveillance audit next month. This includes linearity, bias, stability, etc, etc.

One are I am having trouble is with a scale that is used in the calculation of Specific Gravity (SG). I have the scale set-up with a side attachment that allows me to weigh in water. I weigh a piece of material in air, then in water, then calculate the SG. The SG must then be within a specified range.

For example let's say SG must be between 2.15 to 2.43
Lets say the part measured 356 gm in air and 209 gm in water
That is an SG of 2.42

Now, given that I have 7 different SG specs for 7 different materials, each of which have hundreds of part number variations, can I calculate Linearity for the scale being used???

K

Ken K

This could be a complicated situation but I'll give it a shot. Hopefully, others will offer their thoughts also in also.

The various parts numbers should have no impact. Same material, different configurations. You need to concentrate on the seven materials. My suggestion would be either use a mid point material for your MSA or better still, a low end, high end and mid range material. This would give you coverage throughout the range of materials you produce and tell you how stable the scale is over the entire range. I would also suggest you use separate measurements and charts for each material.

Yes, you will need to do three studies, but it sure beats doing seven. Now, if your auditor would accept the mid range only, that would save you even more work.

We use the three range method on our Mettler scale and have not had any problems.

Any other thoughts?

A

Atul Khandekar

IMHO, Ken K is right. Material and number of parts have no bearing on linearity of scale. Linearity study is to be done over the entire range of the equipment - in your case this translates to the operating range - minimum SG spec and maximum SG spec. It is upto you to decide the number of 'levels' between these two ends. Choose the levels at values that occur most freqeuntly. However, since you have only 7 different specs (or levels), 3 studies as suggested by Ken K should suffice. You must have at least 3 levels to draw a regression line on the linearity plot and to compute the goodness of fit.

X

Xman

Ken / Atul - thanks for the input. That's actually the way I was thinking. The only thing is, what do I use as a tolerance? I have "ranges" for each SG, I could use the midpoint of each range as a "spec" and use +/- half the range as the tolerance?!? But what about the scale? There is also measurement uncertainty there as well? Arggg! This MSA stuff can be SO frustrating!

A

Al Dyer

Yes I have a specific comment!

I have no idea about specific gravity measurements, but sounds like if you 20 parts from the same lot of steel, they should have appx. the same S.G,

Is it possible to test different types of materials to prove the linearity of the gage?

Aluminum, cast, steel, foam, liquid etc...

Please realize this is from a novice!!!!!!!!!!!!!!!!

A

Atul Khandekar

Xman,
My take is you use the minimum of the tolerances from the 7 specs that you have. Midpoint +/- half range at every spec is okay if you are doing 7 R&R studies.

Al,
I believe this measurement has 2 steps - Weigh a part in air (Steel, Al , what have you...) and then weigh it again in water. Use a formula and Eureka!

X

Xman

Let me clarify...

I guess I should clarify this a little...

We manufacture disc brake pads. We have seven different friction formulations. Each friction formulation has its own SG specification range (eg: material "A" is 1.50-2.30; material "B" is 1.32-1.89; material "C" is 1.75-2.05, etc, etc...)

During testing, we "shear" the friction material from the steel backing plate. Then we take this sheared friction pad and weight it in air and then water, on the same scale, to calculate the SG for that particular material. This is one of the tests involved to qualify that specific mix batch of material prior to release for production.

What I need to work out is exactly how to properly conduct a gauge linearity study on the scale which is used to weight these test pieces when calculating SG.

I hope this makes my problem a little clearer.

K

Ken K

If I understand your question correctly, your trying to focus on the scale and not the material?

Is that correct?

Page 36 of the MSA manual has a very good example of a linearity study.

S
Replies
2
Views
6K
Replies
11
Views
16K
Replies
3
Views
10K
Mr.Happy
M