Metal forming - Brake presses - Need error-proofing advice

[Matt Swartwood]

Good afternoon all. Unfortunately it seems that lately I am only on here when I have a question or need help. For that, I apologize; business has had me on the road or in the air the last few months.

I do; however, have a question to those who have dealt with Metal stamping, primarily Brake Presses. We are a job shop on a large scale and at the moment produce about 10,000 different part numbers. Our brake press department currently and consistently contributes the largest amount of product to the PPM. My question is how do you error-proof a process that is 80% operator driven and 20% machine/product variation without hindering production demands? It is very difficult to build guides or locating pins due to the amount of movement and different positions required to manufacture the part. In addition, brake press tooling is considered perishable tooling and is not generally purchased to run a single part; more like a multitude of parts.

Our company is very prevention oriented and I am highly opposed to resorting to detection methods (check fixtures or 100% inspection) just because we have hit a stumbling block. Although the check fixtures would prevent product from getting out to the customer, the amount of investment it would take to build fixtures for every part and the productivity losses incurred with the additional inspection is just not feasible.

I believe that we have a much broader problem than just the lack of error proofing in the department, but still think it should be one of the items addressed. I am starting a fairly detailed benchmarking campain, which includes the comparison of the following (but not limited to):

--Orientation of new hires and previous experience
--Training program and the verification of effectiveness
--Length of service at their current position
--Pay rate and the class of technicians they are able to acquire (we are not in an industrial area).
--Turnover rates
--Mistake-proofing methodologies
--Technological advancements available for equipment
--Types of product produced
--Types and quality of material used
--Customer specific requirements


Having said all of that, I hope to identify what is causing the higher PPM rate (may be a multitude of things). Anyway, back to the subject; any suggestions on what we could do to improve the mistake proofing methodologies for our brake presses? Any advice you could offer would greatly be appreciated.

[Ron Rompen]

If you are running 10,000 different parts, how can you NOT invest the necessary money in dedicated fixturing?

I have worked in a brake-press shop before, and every part that we made had the necessary fixturing.

80% operator-driven is (IMHO) an incorrect interpretation of the problem; the operators are unable to consistently produce a quality part because they are not given the tools they need to do it. This is NOT an operator failure, it is a management failure.

Fixturing can (I am sure) be designed which can be easily and quickly modified from one part to another.....all it takes is some effort and commitment of resources.

[Marc]

I would think that at the contract stage this issue would be addressed.

I remember a client that wanted 100% conforming product at one place I worked. The contract was accepted where it was known there was no way, considering the volumes, to provide 100% unless an optical system was used (which of course was out of the question). Needless to say, nonconforming product got out from time to time.

Reality stares you in the face sometimes.

[Wes Bucey]

Ron and Marc probably reflect what 90% or more Covers would respond - Deming and Red Beads demonstrate that the bulk of nonconforming parts are due to the process, not the operator.

Mistake proofing is a favorite topic of mine. It works best when you have a complete root cause investigation of the nonconformities you produce. It may very well be some sort of mistake proof fixturing is necessary to eliminate the common causes of error.

Please do not lull yourself into believing "mistake proofing" is the same as "brainstorming." The very best mistake proofing comes from engineers who know how a process works and should work before suggesting "fixes." This is not a time for a bunch of know nothings to sit in a room with lots of caffeinated beverages and shout out stuff like "Why don't we . . . .?

Brake press operations depend upon the order in which multiple bends are made to assure the most error-free finished product. The order in which the bends are made is contained in the Control Plan. Just like most businesses, the control plan should include a list of "evaluation flags" (some call them "traffic lights" or something similar) which are readily apparent to signal a stop to refixture a part if the process begins making nonconforming material. Sometimes, it can be as simple as keeping one "Gold Standard" part at the machine so it can be held up for a visual comparison, especially on those characteristics which experience says may cause the most problems in delivering an error free product to a buyer.

[Matt Swartwood]

Thank you for all of the responses. I did not thoroughly explain the magnitude (or lack thereof) of the situation, but probably should have. The PPM's aren't catastrophic; we are just looking to improve the process. As stated before we have 10K different parts; each run could be 10-500 pieces, dependant on customer requirements. We don't produce entire lots of discrepant parts (normally). Generally, when issuing RMA's it is for 1-5% of the lot run (or just a couple of pieces). We see issues like:
--part wasn't against the back stop
--part crooked
--part backwards
--part formed on the wrong end

We currently have a very thorough APQP process that identifies potential problem areas. Also, our FMEA's and control plans are geared to detect potential PPM hazards. However, without an inspection frequency of 100% or a steadfast error-proofing process, the likelihood of detecting those one or two pieces is slim.

That is why we want to look into other avenues, if any are available. I am in hopes that during the benchmarking process, I can find new machine/part gaging technologies available such as moving guides or pinned plate holders. I am sure that our company could design something of the sort..... Maybe that's my answer.

[Jim Wynne]

Originally Posted by Wes Bucey

The order in which the bends are made is contained in the Control Plan.

I think it might be more accurate to say "work instructions" than "Control Plan" especially with capital letters. An automotive-type control plan (and the phrase is usually associated with the AIAG form) wouldn't normally include detail down to the order of bends. There might be a process step that is called "bending" and associated controls might be listed, but in general the down-and-dirty details are found in a router or work instructions.

That said, I'll add a word about "conformity." Sometimes the goal of 100% conformance leads us away from where we need to be, imo. Actually, it might be more accurate to say that the definition of "conformance" is the problem. There might be instances where making "bad" parts isn't necessarily a bad thing in the final analysis, so long as everyone agrees on the goal. The impetus for a job shop should be to supply customers with what they need in order to keep their customers happy, and to make a reasonable amount of money doing it. This involves not perfection of processes (an unattainable goal), but optimization of them--making them as good as can reasonably be expected, given known constraints. So if the best, most economical way to get the customer 500 "good" parts on time is to make 510 of them and tell them to throw bad ones away and keep the extra good ones, if there are any, then that's what should be done. It's not a difficult thing to convince a customer that he might come out ahead, and save all of the soft costs associated with pointless corrective action processing. The point is that there should be no universal definition of "nonconformance"--it should be defined for each job.

[CarolX]

Hi Matt,

Sorry I didn’t see your post until just now. And I can give you my input.

I work for a company that is pretty much on parr with yours. We, too, are large job shop, with runs between 10-500 pieces, with similar return rates for the same issues…..

First, we have to admit the really obvious…..we have humans placing parts against the backstop. In order to form a perfect part each and every time, the operator must place that part against the stops, perpendicular to the stop, with the same force, the correct bend order, in the correct burr direction. Add in things like operator attention, fatigue, distractions…I could go on all day listing the potentials for defect in this area.

I bet your don’t have the same problem with your punch presses. With your punches, you load your tools, load the program, and punch your parts. Most of the work is done with the punch press. Yes, there are some operator dependant steps, but they are miniscule compared to the press brakes.

From my experience, you can 300% inspect the part, and still have returns. It is the human factor.

Our solution….automate. In the past 2 years we have purchased 7 robotic press brakes. Are they perfect, no, but they are reducing our returns for forming errors.

[Jim Howe]

So if the best, most economical way to get the customer 500 "good" parts on time is to make 510 of them and tell them to throw bad ones away and keep the extra good ones, if there are any, then that's what should be done.

This reminds me of the Pershing II radar guidence system that I worked on back in the late 80's. We had a delivery problem of finished units because our magnetron supplier could not deliver enough magnetrons per month that met US Army criteria. After several meetings with our supplier we found out that the US Navy used the same magnetron in one of its missiles but was on the low end of the power curve. The supplier would make the magnetron, test them, deliver the high power curves to us (obviously a lower amount) for Army missiles and send the lower power curves ( higher amounts) to the Navy. Solution was to make the pie bigger. Some how the Prime convinced the Pentagon that if they would order more missiles for the Navy the Army's delivery of missiles would improve dramatically! It did and the P-II's were deployed on time.

As to the subject matter our brake and roll presses are layed out by hand (large parts) and fixturing is on hand to check radii and degrees of bend. In our sister plants that do a higher volume CNC's are used but the fixturing is there to check on a sample basis.