Printer Friendly

SPC for small to medium foundry's sand system: Part 1.

SPC for Small to Medium Foundry's Sand System: Part 1

With its emphasis on testing and record keeping, statistical process control is proving to be a valuable tool in producing better castings, reducing costs and preventing problems.

The purpose of this discussion is to introduce statistical process control (SPC) to the small to medium foundry and describe the steps needed to implement such a program for your sand system. There's a lot more to SPC than is covered here, but this will get you started.

Although SPC is not a new idea, it is a powerful tool that is just beginning to be appreciated by foundry personnel. Foundries using SPC are seeing significant improvements in their bottom lines.

If you have a good operation, you probably are doing much of what is required for SPC. All that's necessary is to use your data a little differently.

Where do I start?

You start by understanding what you want from SPC and what it takes to achieve it. Then you put together an implementation plan, with objectives and tasks for you and your staff, and reasonable schedules for completion. Be sure that everyone (including you) knows exactly what he is responsible for.

Introducing SPC won't be painless and it will cost you some money, at first. But it does work to reduce costs and improve casting quality. That's its goal: continued process and cost improvement.

What do I look for first?

As you get started keeping records of the results of tests on your sand system, see if your scrap increases when your sand tests give results that are different from normal. This means you have to determine what "normal" results are. It also means that you must have a way of relating the sand test results to the quality of the castings that are made with that sand. You have to keep good production records.

You will have to decide what tests to make, how often to make them, where to make them (and who in your foundry makes them) and how to be sure that they are accurate.

Sounds like a lot of testing and record keeping. What do I do with this data?

That's what this discussion is all about. Frequently, the interpretation of the data will be obvious. Other times you will have to dig to figure out what's going on. The point here is that SPC is a tool to help you produce better castings. It's an ongoing process with a goal of preventing the problems that could occur.

One last point before we get started. SPC takes a total commitment from management or it won't work. But once your management gets behind the program and sells it to the rest of the employees (including the sweeper), it will pay off in better castings, lower costs and more satisfied customers.

What is SPC?

Every metalcaster knows that in order to prosper, especially in today's very competitive economic climate, total dedication to continual, never-ending improvement in quality and productivity is required. This means that there has to be a commitment from everyone in the organization, including management. Your customers demand it and your competitors will supply it if you don't.

So how does SPC help?

In two ways. First, SPC is a tool and a method that will give you hard numbers about what your problems are and where solutions to them lie. Second, because everyone in the foundry is involved in the program, SPC is an effective way to keep your employees committed to an ongoing program of continual improvement.

The traditional approach to producing castings is to depend on production to make the product and on quality control to inspect it and screen out items that do not meet specifications. These are then repaired, or scrapped out, and the good ones are shipped.

We've been operating like that for years. Why change now?

To begin with, it costs a lot of money to inspect castings and it costs a lot to fix them when they're not right. It's a lot less expensive to make them right the first time.

That sounds sensible, even obvious. You probably already have told your employees to "do it right the first time." But this kind of cheerleading isn't enough. There has to be a program that makes "doing it right" easier than doing it wrong. And that's where statistical process control comes in.

You mean that statistical process control makes it easy to do things right?

That's not quite what we said. Statistical process control is the use of statistics and statistical techniques (such as control charts) to analyze your processes and their products so that you will know what actions to take to correct your process and ensure that the product will be right.

You see, if the processes (like your sand system or your gating systems) are correctly designed and working properly, your product just has to be right. So, to get good castings, you need a way to be sure that your processes are always running right. That's what SPC will help you do.

Just what is involved in an SPC system?

A good way to look at SPC is to think of it as a feedback system that has four elements: * the process; * information about the process; * action on the process; * action on the output of the process.

This is the whole combination of people, equipment, materials, methods and environment that work together to produce a product--in short, your foundry operation.

The total performance of the process--the quality of its output or its productive efficiency--depends on the way the process has been designed and built, and on the way it is operated. The process control system is useful only if it contributes to improving the performance of the process.

Information about the Process

This comes from two sources: the products that are produced and the process itself. You are familiar with the information that the products give: scrap, rework and good products. But at the same time, the process itself is generating information on how it's doing: pouring temperatures, pouring times, sand test results, machine cycle times, etc.

If this information is gathered and interpreted correctly, it can show whether action is necessary to correct the process so that scrap and rework are avoided. But you must gather the information and interpret it in a timely fashion for this information to be helpful to you.

Action on the Process

Think of this as action you take to affect the future--the future quality of your product. You look at the information on the process, decide that the process is beginning to get a little out of control, and so you correct the process before the process produces scrap castings.

This action can vary all the way from changing the temperature on a hotbox core machine or increasing the amount of western bentonite in your sand mix, to retraining an operator or even buying new equipment because the present equipment can't quite do the job. It can even mean changing the entire process. As soon as you make the changes, you start right in again, monitoring the changes that you have made to be sure that you are keeping the process under control.

Action on the Output

Think of this as action to affect (or correct) what you did in the past, like reworking castings. This action is what you're trying to avoid when you use SPC. So when you're taking a lot of action on the output of your process, you are getting a clear signal that your process isn't under control. And you're going to continue to rework and scrap castings until your process is under control.

It's obvious that inspection followed by correcting the product is a lot more expensive than controlling the process so that it produces a good product to begin with. When using SPC, it's important to focus on gathering process information and analyzing it so that action can be taken to correct the process itself.

How do I get this information that the process is supposed to be generating all the time?

Before you start to get information, it's important to understand what it's going to look like. The first concept you must understand is variation.

You know that there is variation in every process all the time. What you're trying to accomplish here is to minimize the variation so that the product is acceptable. To do this, you must be able to trace the source of the variation. You want to know whether the variation that you see is common or whether it's caused by something else, something which is unusual or special.

What's the difference between common and special variation?

Let's start with common variation. This refers to the many sources of variation within a process that is under control--statistical fluctuations arising from minor chance causes. These tend to form a distribution of values. The distribution has three characteristics: * typical or average value (think of this as the location of the distribution on a graph); * spread or range (the amount by which the smaller values differ from the larger ones); * shape (the pattern of the variation--whether it is symmetrical, peaked, etc).

Special variations are those which disturb the common variation distribution. They are in addition to the common variations. They indicate that the process is not under control and their cause (or causes) must be identified and corrected.

Are common and special variations related?

No, not usually. Special causes of variations can be detected by reasonably simple statistical techniques. The discovery of a special variation and its cause, and correction or removal usually is the responsibility of someone who is directly connected with the operation, not management. The resolution of a special cause of variation, then, usually requires local action.

On the other hand, while the extent of the causes of common variation can be indicated by simple statistical techniques, more detailed analyses are usually needed to isolate them. Because these common causes are generally part of the manufacturing system that is used to produce the product, fixing them (decreasing the variation) usually means changing the system--a management responsibility.

It appears that only about 15% of process problems are correctable locally by the people directly connected with the operation. The rest (85%) require management action, such as new machines, supplies or processes.

Remember, your goal in implementing a process control system is to make economically sound decisions about actions affecting the process. You want to avoid both overcontrol (taking action when none is necessary) and undercontrol (taking no action when action is necessary). Isolating causes of special variation from causes of common variation is essential in deciding when to take action and what kind to take.

Well, that all sounds good but how do statistics help?

A process is in statistical control when the only source of variation is from common causes. W. Edwards Deming (the father of SPC) has said that a state of statistical control is not a natural state for a manufacturing process. Rather, it is an achievement arrived at by elimination, one by one, with determined effort, of the causes of special variation.

The initial function of an SPC system, then, is to provide a signal--which statistics will show--when there is a special variation in the process. You know from statistics that there is a problem and you can look for it. It's better than guessing that there might be a problem, and spending your time and money for something that may not be there.

But to do this, don't I have to know what my common variations are?

Yes, you do. There's a term for common variations in SPC. It's called process capability. Process capability is the minimum variation that can be achieved after all special causes have been eliminated. It represents the performance of the process itself, as demonstrated when the process is being operated in control.

Process capability can be represented in terms of the proportion of output that is within product specification tolerances. When your process is in statistical control, there will be a predictable distribution of parts, with a predictable proportion of them meeting specifications. As long as the process remains in control, it will continue to produce the same proportion of satisfactory parts.

Of course, you want all of the parts you produce to meet customer specifications. To achieve this (assuming that you have removed the causes of special variations), you must improve your process capability. This will require management action to decrease the common variation.

So, the point is that all I have to do is find out what my process capability is, then look for exceptions to it, and then find the causes for those exceptions and eliminate them.

Yes. But then, when you've eliminated all the causes of special variations, you still have a job to do. You have to decrease the amount of common variations until you're producing no scrap or rework.

SPC Requires Planning

To start an SPC program, you must preplan your campaign. You must determine your process capability so that you can establish your control limits, specifications and physical requirements. You should form a project team with your own people, using outside help only when necessary.

Sounds like a job an outside consultant could do better than we could.

Sorry, but hiring an outside consultant and expecting a turnkey SPC system isn't going to work. You see, you and your people have to believe in what you're doing if SPC is going to be a way of life in your foundry. Outside consultants can help, but only you and your people can make the system work.

Remember, a sand system includes a molding line, cooling line, shakeout, return belts, elevators, screens, storage hoppers, prepared sand delivery and the molding machine. Control must be exercised in each part of this system to give optimum performance. Your system should be designed to maintain consistency in the process.

Doesn't it come down to my designing my sand control from my own experience?

Yes, very much so, although you may find that your experience needs to be improved after your SPC program is in place. Your control system will be based on: * selection and control of base sands; * castings produced, their size and mix of sizes; * sand additive selection and control; * formulation of molding sand mixtures; * mulling procedures; * sand conditioning after mulling; * sand testing procedures.

Before you start an SPC system, remember that it is not a substitute for common sense. You still have to keep in mind that the ratio of sand to metal determines your burnout rate and therefore, your bond addition. The shape of the castings and the amount of coring they require also affect the bond addition. The general rule of thumb for ferrous castings, which suggests the addition of 100 lb of bond and 300 lb of new sand for each ton of iron poured, is only approximate. You already know what rules work in your foundry. It is unlikely that SPC will cause you to change them if they are correct.

Because controlling your sand means adding just the right amount of bond and new sand, you should keep good records of how much metal you are pouring each hour and how much bond you use in each muller cycle. So starting an SPC program means keeping good records.

What else does it mean?

Starting an SPC system means that you have to measure your process. You must decide what to measure, where to measure it and how often to measure it. You also should give consideration to how to measure it--what equipment and procedures to use, and how to be sure that the equipment is giving accurate readings.

PHOTO : Information gathering is a critical step in implementing an effective SPC program.

PHOTO : To be effective, SPC requires that the data collected be compared with actual results.

Committee (4-M) members who contributed to this article are Van Champion, Talladega Foundry & Machine Co, Inc; Bill Edison, George Fischer Foundry Systems, Inc; Steve Neltner, The Hill & Griffith Co; Pat O'Meara, Intermet Corp/Lynchburg Foundry Co; Tom Piwonka, Univ of Alabama, Tuscaloosa; and Jack Vincent, Mueller Co.
COPYRIGHT 1989 American Foundry Society, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1989, Gale Group. All rights reserved. Gale Group is a Thomson Corporation Company.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:statistical process control
Author:Piwonka, Tom
Publication:Modern Casting
Date:Sep 1, 1989
Previous Article:Heat treating, aging system also permits core sand removal.
Next Article:Granite City foundry springs back to life.

Related Articles
Combine training and SPC for quality improvement.
SPC leads to the right mix for investment casters.
SPC for small to medium foundry's sand system: part II.
Product showcase.
New force are revolutionizing the foundry industry.
Using SPC in the coreroom.
Gain a competitive edge with TQM.
'Shop-floor' process controls for lost foam.
Portable sand tester provides accurate, repeatable results. (Case history).
Casting answers & advice.

Terms of use | Copyright © 2016 Farlex, Inc. | Feedback | For webmasters