Search the Elsmar Cove!
**Search ALL of Elsmar.com** with DuckDuckGo Especially for content not in the forum
Such as files in the Cove "Members" Directory
Social Distancing - It's not just YOUR life - It's ALL of OUR lives!
Me <——————— 6 Feet ———————-> You

How to start Implement SPC for Injection Molding Process?

KCIPOH

Quite Involved in Discussions
#1
Hello Cove Members,

I would like to perform an analysis for the injection molding parts running in my company.

What is the first step i should do to start a SPC?

Appreciate inputs :confused:
 
Last edited:
A

adamsjm

#2
I would like to perform an analysis for the injection molding parts running in my company.
Thanks for asking us. My first question for you is "Why do you want to perform an analysis of the injection molding parts running in your company?" Who wants "what" problem to improve and "why"? Can you clearly state and qualify the problem? (what, when, where, how many.)

This will help you and us to understand the next steps we should take. I do not know if SPC will work until we define the problem.
 

Stijloor

Staff member
Super Moderator
#3
Thanks for asking us. My first question for you is "Why do you want to perform an analysis of the injection molding parts running in your company?" Who wants "what" problem to improve and "why"? Can you clearly state and qualify the problem? (what, when, where, how many.)

This will help you and us to understand the next steps we should take. I do not know if SPC will work until we define the problem.

SPC is not necessarily initiated because there is a problem.

Stijloor.
 

bobdoering

Stop X-bar/R Madness!!
Trusted
#4
SPC is not necessarily initiated because there is a problem.
Correct, primarily it is a tool to...well, control a process. It should give you feedback on how a process is performing, and whether there are any unusual conditions to address.

Think of it like a gas gage. You don't need a gas gage, but just think of the misery you can save yourself by having one that works? Same thing with SPC. Only thing is some minimalists would rather do without the gage and live with the misery. Pity.

I recommend looking at this as a starting point.
 
P

pearsonow

#6
Have a look at the thread I have started on introducing SPC into the supply chain, It will gradually have more details on how we are going about it for our injection molding sub contractors.

Olly
 

Bev D

Heretical Statistician
Staff member
Super Moderator
#7
Adamsjm has a response that is on point: why does the OP want to analyze the process? Is there an existing Problem or is it just a desire to understand process capability, etc. The answer to this clarifying question enables us to answer the question the OP has.
 
Last edited:
T

TShepherd

#8
Just for kicks - and the lack of a definable problem - there can be many reasons why you might want to apply some SPC techniques to your process.

- You might want to apply SPC to test your own capability.
- You might want to ensure that the process is stable, in control and capable.
- You might want to show your customer that you you willing to go beyond the normal checks to protect them.
- You may want to educate co-workers to SPC.
- and the list goes on.

If you do not have a definable problem or issue - here are a few basic steps.

Review the print for your part and look for critical dimensions (sometimes identified on the print through various symbols or letters - ask your customer to assist if you can.

Select one, preferably an easy one to gage (Thickness, Length, O.D, I.D)

Determine what type of chart you want to use (Average & Range, Median and Range and so on).

Develop the chart to meet your needs. (measurement ranges ect.)

Determine the sample size and frequency.

Collect the Data overtime - sufficiant time to ensure that you have captured all the variation introduced into the process (Man, Material, Methods, Set-ups, Shift, and so on)

Analyze the Data to determine if the process is STABLE, IN STATISTICAL CONTROL & CAPABLE. - Only inherent variation exists.

If the process is STABLE, STATISTICALLY IN CONTROL AND CAPABLE - Drastically reduce or eliminate SPC, your wasting your time - UNLESS IT IS MANDATED BY YOUR CUSTOMER and the added cost should have been in the RFQ.

If the process is not STABLE or NOT STATISTICALLY IN CONTROL OR IS FOUND TO NOT TO BE CAPABLE and those plot points that indicate that they are out have been noted (Shift, Time, Material Changed, New Operator and so on) - you will already have some clues as to what Special cause may be being introduced into the process.

Continuing to apply SPC to a proces that is not stable, in control & not capable is normally a waste of time.

Other quality methodologies specifically may be more appropriate.

The details of the above basic steps can all be found in the AIAG SPC book including blank charts, formulas to calculate control limits etc.

Good Luck,

Tom:agree1:
 

bobdoering

Stop X-bar/R Madness!!
Trusted
#9
If the process is STABLE, STATISTICALLY IN CONTROL AND CAPABLE - Drastically reduce or eliminate SPC, your wasting your time - UNLESS IT IS MANDATED BY YOUR CUSTOMER and the added cost should have been in the RFQ.
The sampling rate should be a function of the variation found in the process, not how content you are with a short term study. You may be stable, statistically in control and capable today, but ongoing charting will allow you detect when a special cause occurs and requires your attention. Some charting actually tells you when to make adjustments. So the "cost" of ongoing charting should actually be cheap insurance so that when a bad thing happens, you may actually notice it is happening before you create a disaster...which just might not be a waste of time.

SPC is just as much about detecting special causes and dealing with them when they happen as it is getting a comfort level that the process is in control. Just because it is in control during the initial study does not mean it will stay that way over time when the rest of the variances in the total variance equation start to influence the process, such as more material lot variation, new operators, etc.

As far as RFQ, SPC should be in the overhead, as it should be a part of your regular process control. If it is so special the customer has to "order" it, you just might be missing the point of the exercise. But, on the other hand, you won't be alone.
 
T

TShepherd

#10
I don't recall saying anything about a short term study - I said sufficient time to ensure that all variation is captured (I think the book says something like 125 sub samples of 3 - 5 for a short term study - that could be a month or a year. Then drastically reduce the frequency or eliminate the SPC on that characteristic and apply resources somewhere needed.


Time and time again companies are getting jobs (Automotive is good for this) only to find out later that there are KPC's (Safety, Fit & Function) characteristics that require variable data that may require special gages ect. They expect the supplier to know this and sometimes can be resolved after RFQ - but they will push the supplier as far as they can to eat the cost themselves.

Tom
 
Top Bottom