Conducting a MSA - Measurement of plaster by use of a scoop - I need Help

[PWithey]

I need some help in conducting a MSA-

The process involves the measurement of plaster by use of a scoop- the spec is set at 260 +/- 20 and I dont think the process will be accurate enough - any suggestions on how i should proceed

 

[wmarhel]

R&R

I need some help in conducting a MSA-

The process involves the measurement of plaster by use of a scoop- the spec is set at 260 +/- 20 and I dont think the process will be accurate enough - any suggestions on how i should proceed

First things first, be sure to clearly identify, document and train the operators on how the scoop is used. It is a heaping scoop, a level scoop, etc.

Have a device (a scale for weight, a container volume) which will provide enough resolution and which is accurate and repeatable.

Do a search in this forum, or use googles advanced feature and do a search for gage r&r (you can always try R&R, GR&R, etc.) selecting ".xls" as the filetype.

Granted, this is a really simplified explanation, but the study doesn't have to be a mammoth undertaking. Don't make it any more complicated than needed.

Just out of curiosity, if you don't mind. What is the application that it would require a GR&R on plaster?

Wayne

 

[PWithey]

We add an extra amount of plaster to a mould when certain jobs are produced, the stated spec is 260g +/- 20grms but this is not achieveable with the scoop- i want to demonstrate this by using a MSA. I think i have to do the following
3 operators
10 parts
measure the parts twice

But how do i measure the same scoop twice if the variation is in the scoop size- do you understand what I mean?

Phil

 

[Jim Wynne]

We add an extra amount of plaster to a mould when certain jobs are produced, the stated spec is 260g +/- 20grms but this is not achieveable with the scoop- i want to demonstrate this by using a MSA. I think i have to do the following
3 operators
10 parts
measure the parts twice

But how do i measure the same scoop twice if the variation is in the scoop size- do you understand what I mean?

Phil

You could confirm your hypothesis without using MSA, per se, if all you want to do is demonstrate that the scoop method won't work by having operators do the task x times and recording the results. If having the +/- 20 gram accuracy is critical, you might consider premeasured portions of plaster in order to establish control.

 

[Wes Bucey]

I have an idea that a small change in the design of the scoop could bring the results closer to the optimum you seek.

Simply stated, the scoop requires a design which will allow it to be filled to the brim and leveled off prior to adding the contents to the batch. Thus said, the typical "flour scoop" design offers too much opportunity for operator variation.

Once the optimum scoop design is introduced, a process to fill and bang the scoop to ensure no voids have been introduced before leveling should be taught to all operators. (optional - use the flour scoop to fill the measuring scoop.)

If even further accuracy is required, a small scale can be introduced into the process to weigh the entire new design scoop and plaster contents (scoop tared out) before adding the plaster to the batch. My impression is the scoop and plaster are only added to some batches, not all, so the equipment can be reserved strictly for the addition of plaster.

How do you measure the original amount of plaster if the subject plaster is an "extra amount of plaster?"

 

[Hershal]

wmarhel has the right idea, I think.

Use an analytical balance, calibrated AT LOCATION OF USE by an accredited cal lab.....have a specified beaker that you can tare out, and use it to measure the plaster. The scoop is only a transport means from the mixture bag to the beaker, nothing more.

Oh, never move the balance. Also, brush or blow away any dust that drops onto the pan of the balance.

Hope this helps.

Hershal

 

[PWithey]

We add the extra to certain jobs which are predetermined due to their size, the other moulds which we run as normal are measured at source so there is no chance of variation but when this process was added to our (against my better judgement) system I could see the opportunity for variation which is what I am trying to prove.

I have now conducted a test to prove the hypothesis that the process is incapable- 3 people measured out a level scoop which is supposed to measure 260g +/- 20g and came up with results varying from 234g up to 350g and that was in a controlled environment not out on the busy shop floor.

I will tonight (I am in the UK working a night shift ) be testing a method of measuring the extra plaster correctly and consistently.