What Data do we need for Welding and Repair Process Capability?

E

Elvin

I am new here :bigwave: so i may not know some of the stuff or rules here.
Anyways i am doing a project in my school, is for a certain company that specialise in welding and repairing. We want to do a process capability test on it, which involve Cpk or Cp, however we do not know what data we need.

What we have:
1) They have irregular size product, they are split into a few categories (Collar/ Rotor/etc)

2) They use machine to do pre-machining and use labor to do welding.

3) They have quite a number of different process depending on what is the customer requirement

4) We are there to observe the company for only 2-3 months

5) Our motive is to identify the process that took the longest or process that cause most defects.

6) We were told that there were no USL and LSL. (not certain about it too)

*We hope if someone could enlighten us on what data should we be collecting? and how we should be collecting them? what to take note.

Machine time? Labor time? Setup time? Process time?

This really mean a lot to us, hope someone could help us.:)
Thank You :bigwave:
 

John Broomfield

Leader
Super Moderator
I am new here :bigwave: so i may not know some of the stuff or rules here.
Anyways i am doing a project in my school, is for a certain company that specialise in welding and repairing. We want to do a process capability test on it, which involve Cpk or Cp, however we do not know what data we need.

What we have:
1) They have irregular size product, they are split into a few categories (Collar/ Rotor/etc)

2) They use machine to do pre-machining and use labor to do welding.

3) They have quite a number of different process depending on what is the customer requirement

4) We are there to observe the company for only 2-3 months

5) Our motive is to identify the process that took the longest or process that cause most defects.

6) We were told that there were no USL and LSL. (not certain about it too)

*We hope if someone could enlighten us on what data should we be collecting? and how we should be collecting them? what to take note.

Machine time? Labor time? Setup time? Process time?

This really mean a lot to us, hope someone could help us.:)
Thank You :bigwave:

Elvin,

Welcome to the Cove!

Before applying statistical techniques, I suggest that you first focus on process design and mistake-proofing to the extent feasible.

Your primary concern (bottom-line) is the speed at which the system enables the process to add value to its inputs.

After process design and mistake-proofing (for right first time products) you may then see if it is worth investing in the use of statistical techniques.

John
 
Last edited:
Top Bottom