Warping / Flatness SPC - X bar-R chart - Control is Max 0.15

[wanamir]

I need a help to plot a good X bar-R chart for warping/flatness specification.
From our customer drawing, it's stated the overall warping/flatness control is Max 0.15.
From the given control, I took the 0.15 as USL, zero as LSL and 0.075 as nominal.
I'm not sure it's correct or not to use those above for my calculation.

From actual part reading, mostly our reading more to LSL where we get 0~0.05 and yes all understand the best warping/flatness result is zero. Thus the plotted chart not to the center but to the lower. It's not a good chart I think.

From this kind of specification, I also need a help to get Cp, Cpk, Pp and Ppk.
I do have the formula but don't know which one I should have Cp or Cpk, Pp or Ppk.
Maybe because I don't really understand the meaning of those things.
To all experts please helps me with this.

WAN

 

[Clausterphobic]

I need a help to plot a good X bar-R chart for warping/flatness specification.
From our customer drawing, it's stated the overall warping/flatness control is Max 0.15.
From the given control, I took the 0.15 as USL, zero as LSL and 0.075 as nominal.
I'm not sure it's correct or not to use those above for my calculation.
WAN

IMHO it is NOT wise to make the customer's specification as your control limit. What process were you monitoring anyway? plating? stamping?

From actual part reading, mostly our reading more to LSL where we get 0~0.05 and yes all understand the best warping/flatness result is zero. Thus the plotted chart not to the center but to the lower. It's not a good chart I think.
WAN

If you don't mind I would like to ask, have you received any complaint from your customer so far regarding the flatness/warping? It seems that you have a pretty much stable process (reading between 0~0.05) so would expect there should be zero. If there is an issue, you need to verify your process if you are seeing real data on you Xbar-R chart.

From this kind of specification, I also need a help to get Cp, Cpk, Pp and Ppk.
I do have the formula but don't know which one I should have Cp or Cpk, Pp or Ppk.
Maybe because I don't really understand the meaning of those things.
To all experts please helps me with this.
WAN

Cpk is defined as process capability index.

Cpk is computed by the following steps:

CpL = (xbar - LSL)/3*standard deviation
CpU = (USL - xbar)/3*standard deviation

Cpk = min(CpU, CpL)

Cp is defined as process capability.

Cp is computed by:

Cp = (USL-LSL)/ 6*standard deviation

I can go on but it would be a little hard to explain here if you are saying that you do not have a background on statistics and SPC. I would suggest you get yourself a training on SPC or search the web. You may search using the term "statistical process control".

you may also examine the related topics found at the bottom of this page.

 

[apestate]

Hello wanamir

I don't have time to pull together anything that would be of help, but here's a great thread on the subject:

http://elsmar.com/Forums/showthread.php?t=1087

I also recommend AIAG's SPC manual.

 

[pmwong]

our max warping control given by our customer is 0.3 max. Our customer did not ask for any control charts. what we practice is that we state in our inspection plan that we carry out some test every 2 hrs using the feelers gauge fitted onto the car audio set given by one of our customer as a test/fitting sample.

 

[wanamir]

our max warping control given by our customer is 0.3 max. Our customer did not ask for any control charts. what we practice is that we state in our inspection plan that we carry out some test every 2 hrs using the feelers gauge fitted onto the car audio set given by one of our customer as a test/fitting sample.

Yes correct, when we deal with Sanyo, they also doesn't ask for control chart and process capability result but different with our main customer, Blaupunkt Germany & Portugal. To them the flatness is very important thus recently they request for control chart and Cpk.I treat this requirements as CSR.

 

[Caster]

I need a help to plot a good X bar-R chart for warping/flatness specification.

From our customer drawing, it's stated the overall warping/flatness control is Max 0.15.
From the given control, I took the 0.15 as USL, zero as LSL and 0.075 as nominal.
I'm not sure it's correct or not to use those above for my calculation.

From actual part reading, mostly our reading more to LSL where we get 0~0.05 and yes all understand the best warping/flatness result is zero. Thus the plotted chart not to the center but to the lower. It's not a good chart I think.

From this kind of specification, I also need a help to get Cp, Cpk, Pp and Ppk.
I do have the formula but don't know which one I should have Cp or Cpk, Pp or Ppk.
Maybe because I don't really understand the meaning of those things.
To all experts please helps me with this.

WAN

WAN

The control limits for Xbar and R are calculated from the process variation and not the spec.

You have set limits based on the spec, so the chart you dscribe is more like a Pre Control chart (Google) than an Xbar and R chart. Still a good chart, and a nice place to start.

It is good that you are hugging the lower physical limit, it means you have a capable process.

Cp is tolerance/process spread so you have 0.15/0.05 or 3.0 ....pretty good. A Cp of 1.33 used to be OK, recently increased to 1.67, and you are at 3.

SPC books by Wheeler (Google) are excellent places to start (and finish) on this topic, good practical advice from basic to advanced.

I was taught capability as truck and tunnel. The width of the truck is what your process can make (0.05). The width of the tunnel is the customer spec (0.15).

In your case no problem, the truck can move around a bit and still get through the tunnel.

If the tunnel gets smaller (tighter specs) or the truck gets wider (process gets worse) you have more trouble. Cp of 1 is truck = tunnel. Not a good situation.