R
rdanis
Hello experts,
is it good to do a destructive and non reproducibility MSA (repeated with single operator only)?
is it good to do a destructive and non reproducibility MSA (repeated with single operator only)?
R
rdanis
Re: Destructive non reproducibility MSA
It's a breaking strength inspection of extruded tiles (final sampling inspection). We take 5 pcs tiles and use breaking strength machine to know the breaking strength of the tiles. The spec is 1500 N minimum.
I've read Miner's blog and planned to take 30 pcs consecutive produced tiles (expect from 1 batch material mixing) as testing specimen. So what is the next to do? What are the points to be considered in conducting this type MSA?
It's a breaking strength inspection of extruded tiles (final sampling inspection). We take 5 pcs tiles and use breaking strength machine to know the breaking strength of the tiles. The spec is 1500 N minimum.
I've read Miner's blog and planned to take 30 pcs consecutive produced tiles (expect from 1 batch material mixing) as testing specimen. So what is the next to do? What are the points to be considered in conducting this type MSA?
not sure 30 consecutive is good. from previous data do you get the full range of process variation within 30 tiles? if not consider 2 pairs of consecutive tiles taken from across a lot or multiple lots. if the largest range of variation is seen within a lot you only need a single lot. if the largest range of variation is across multiple lots you will need multiple lots. (there is a way around this but it is better to have your samples come from a representative range of product.
the 2 consecutive pairs become your 'repeated' measures. (you may also need to consider what constitutes 'consecutive' if you use a grid fixture...understanding the next level of process details will help with defining what consecutive really means.
on the other hand if you feel comfortable that the 30 consecutive pieces are representative full range of variation or for some reason you can't take other samples, you could consider breaking the tiles in half and using the halves as your repeated measures.
the 2 consecutive pairs become your 'repeated' measures. (you may also need to consider what constitutes 'consecutive' if you use a grid fixture...understanding the next level of process details will help with defining what consecutive really means.
on the other hand if you feel comfortable that the 30 consecutive pieces are representative full range of variation or for some reason you can't take other samples, you could consider breaking the tiles in half and using the halves as your repeated measures.
R
rdanis
Bev,
I can say 30 pcs consecutive products will not represent the range of the product.
We batch around 70-78 times a day. When we do the next batching, we don't empty the bin from the previous batching.
About selecting multiple lots, We should consider material and process environment characteristic, such as sand moisture, temperature during the curing process and temperature on the yard stocking. The variation may become larger and larger.
Splitting the tiles seems to be not good idea since there is a standard dimension for tile. Cutting the tiles may lead to different broken strength value since (may be) the cutting process it self may affected the strength characteristic and also the tile is shorter.
At first, I want to take the first 30 pcs at the beginning of production process in the morning, but now I'm seeking the better method how to choose the sample, need advice on that.
PS:
I don't fully understand about this statement:
"(you may also need to consider what constitutes 'consecutive' if you use a grid fixture...understanding the next level of process details will help with defining what consecutive really means"
I can say 30 pcs consecutive products will not represent the range of the product.
We batch around 70-78 times a day. When we do the next batching, we don't empty the bin from the previous batching.
About selecting multiple lots, We should consider material and process environment characteristic, such as sand moisture, temperature during the curing process and temperature on the yard stocking. The variation may become larger and larger.
Splitting the tiles seems to be not good idea since there is a standard dimension for tile. Cutting the tiles may lead to different broken strength value since (may be) the cutting process it self may affected the strength characteristic and also the tile is shorter.
At first, I want to take the first 30 pcs at the beginning of production process in the morning, but now I'm seeking the better method how to choose the sample, need advice on that.
PS:
I don't fully understand about this statement:
"(you may also need to consider what constitutes 'consecutive' if you use a grid fixture...understanding the next level of process details will help with defining what consecutive really means"
what constitutes 'consecutive' for you?
are the tiles extruded through only one 'cavity' or are they from multiple cavities? the idea of consecutive for destructive testing is that the two consecutive tiles are as alike as possible. same material, same process conditions, etc. so two tiles extruded consecutively form the same 'cavity' are consecutive, but two tiles extruded at the same time from side-by-side cavities probably are not as similar...
are the tiles extruded through only one 'cavity' or are they from multiple cavities? the idea of consecutive for destructive testing is that the two consecutive tiles are as alike as possible. same material, same process conditions, etc. so two tiles extruded consecutively form the same 'cavity' are consecutive, but two tiles extruded at the same time from side-by-side cavities probably are not as similar...
R
rdanis
Bev,
Yes, the tiles are extruded through only one 'cavity'. It's a continuous extrusion which then being cut into pieces of tiles. The extrusion materials are continually batched 70-78 times per shift during the production process.
Yes, the tiles are extruded through only one 'cavity'. It's a continuous extrusion which then being cut into pieces of tiles. The extrusion materials are continually batched 70-78 times per shift during the production process.
R
rdanis
Hello Bev,
Where I can find a good nested MSA template for 10X1X3 (parts=10, appraiser=1 and trials=3) design?
Where I can find a good nested MSA template for 10X1X3 (parts=10, appraiser=1 and trials=3) design?
Similar threads
