Process Stability why not?

Programming vs. Process

We need to make part A. It is an overly tolerated product smothered in GDT. (For my design friends out there!). We can program this product a few different ways.

1.      We do it as we have always done. The quickest way to get it running. This might be 2 -3 operations. Whatever works.

2.      Depending on demand, we might invest in equipment to machine complete. Great idea. But we apply our same principles, get it to the floor now.

In both instances, we use tooling, etc. they we have used in the past. We need to get this moving, there is no time to reinvent the wheel so to speak. Though we will be able to make the product, do we have a robust process? Do we have an efficient process? And most importantly, do we have a stable process? Possibly. This is what I call programming. Get Er done!! It can be expected that any manufacturing facility can make a few pieces successfully. Any decent “Job shop” can achieve this.

To develop a stable process means we need to look at part A differently. Having process stability means not only do we make part A successfully, but we make part A successfully every time, every set up, every change over. This requires process.

   Since we are developing a new process we should first check our normal distribution. Pp.

What is our target size, and without making any adjustments how much does this size change part to part? How close to the correct size? Or centered. Are we? Ppk. 

Most companies surprisingly do not take the time to do this.

We measure the finished product, and see how it changes. This is a good start. We have gauged our process but have we?

1.      What steps are taken to create the feature being created?

2.      What is the interrelationship between these steps? How does step 1 effect step 2? Step 1 and 2 effect 3?

This is the level of process control that can set this process apart.

We create a feature using multiple tools. The first tool leaves “finish stock” for the second. When the feature goes OOT or close to OOT we adjust tool 2. We then might replace tool 2 the finisher. But what has happened to tool 1?

Since we have taken the time to check the stability of the finished product why not go a step further? At what point does the wear of tool 1 affect tool 2? If we take the time to study tool 1 we might learn that after 2 pieces the amount of finish material has changed form .003” to .005”.

We also might learn that this trend continues until we reach 6 pieces, then the wear might stabilize for the next 4. The point being is that unless the capability or stability of the roughing process is known, you cannot truly have a stable finishing process. Or you might but at what tool cost? If you change the finish tool every part, or change both the rough and finish every part, you might be stable but are you efficient? Cost effective?

If time is taken to know the capability of roughing, finishing becomes much more stable.

For example: If tool 1 wears .002” per 2 parts, thus creating a shift in tool 2 by .0005” we can allow for this. Using logical program control, tool 1 can be automatically adjusted every 2 pieces. This then allows tool 2 to be cutting the same every time. If it supposed to remove .006” it is doing so every cut.

The importance of knowing correct tool wear, and correct tool life cannot be understated. Not if you wish to operate efficiently. Knowing the capability of each step of the process, determining the cost per cut of all tools, is how to create a Lean, Stable Process.

The reason most companies never quite make it this far is time, or it is a lack of resources, or both.

I personally enjoy “getting the numbers” as I call it. I like to know what each tool is doing, what is its capability. What is the cost per cut of any given tool? All these principles should be applied whether you are doing process improvements or new product development. Maybe I am just a numbers geek, or a control freak, who knows? But when I am tasked with “Programming” a product, I myself prefer to create a Process.