iSixSigma

Six Sigma for Surface Roughness

Six Sigma – iSixSigma Forums General Forums Tools & Templates Six Sigma for Surface Roughness

Viewing 12 posts - 1 through 12 (of 12 total)
  • Author
    Posts
  • #54188

    Kanchan Joshi
    Guest

    Hi,

    For surface roughness measurement:

    If the roughness limits are say 100-200 micro”, and if the same surface can have different roughness (within 100-200) at different locations,do Gage R&R and six sigma method apply in this case?

    0
    #194104

    Mike Carnell
    Participant

    Yes

    0
    #194106

    Joel Smith
    Participant

    @Mike-Carnell Can you quit boring us with your lengthy explanations and just provide a simple answer?

    0
    #194111

    Mike Carnell
    Participant

    @joelatminitab My apologies but it was a closed question.

    Let’s see. Let’s restate the question. I have a lot of variation in my surface finish process so do process improvement methodologies apply? That is one of those questions where you are wondering if they are just messing with you.

    0
    #194128

    MBBinWI
    Participant

    @Mike-Carnell, @joelatminitab – perhaps “!” would have been more succinct?

    0
    #194223

    Christian Nelson
    Guest

    Can you provide more detail on how the Gage R&R would be set up for surface roughness? Say the surface has variation in the roughness. The different operators (and the replicates) would need to be in the exact same locations with the same measurement direction. Can coupons with variations be used for the study?

    Or should you get several standards with known roughness values and use those?

    0
    #194227

    Chris Seider
    Participant

    You need to be able to measure known variation within the product sample. Don’t do down the route of trying to measure the exact point. If the measurement system doesn’t look at some average smoothness across some standardized area, then you will fail all MSA checks.
    It doesn’t have to be an “average” but don’t expect two machines or operators to get the same exact point if you have known variation.

    0
    #194228

    MBBinWI
    Participant

    Christian: The intent of a GR&R study is to demonstrate that the measurement tool and the measurement methods are adequately precise to make decisions. You need to have samples that are representative of the range of variation that are observable in the real population. And finally, you need a measurement approach that gets sufficient measurements that allow you to see the difference between good parts and not good parts. In this case, the fact that you have parts with different variation across the surface, means that you will likely need a sampling approach to get the reproducibility to a level that is satisfactory.

    0
    #194269

    Kanchan Joshi
    Guest

    Thanks in advance for your patience.

    Which method between the two mehods given below would be more appropriate here?:

    1. Use say ten different coupons with a single location designated with a template, three operators will measure roughness on each coupon at the designated location three times each- the location for three operators will vary within one to 5 mm.

    2. Use a long coupon with ten locations designated with a template, three operators will measure roughness at each location three times each- the location for three operators will vary within 1 to 5 mm

    0
    #194270

    Chris Seider
    Participant

    I suggest you see what range of readings you have across the web or width. This should give you an idea of whether you are “cheating” the MSA evaluation by having everyone measure the exact same spot.

    0
    #194272

    Kanchan Joshi
    Guest

    Chris,

    Thanks for providing clarity in this area.

    The purpose of this query was to figure out the best possible genuine Gage R&R method applicable to the process.

    0
    #194273

    Chris Seider
    Participant

    I’m trying to hint that unless you measure across some standard area and get some average, weighted mean, median or whatever technique, you won’t have good confidence that actions will be taken appropriately if you get a reading that’s different from one operator to another. Taking the sample from the process is PART of the MSA that is often neglected.

    Think of sandpaper (especially course stuff). The granularity ensures you won’t get the same toughness or roughness reading within a mm of another because of the known process variation across the sheet. Does that mean the sand paper might not be in spec or process variation should be acted upon if using SPC?

    Hope the hints are coming more clear. Good luck.

    0
Viewing 12 posts - 1 through 12 (of 12 total)

You must be logged in to reply to this topic.