Pd/M Magazine Article

Title: Tooling: The Last Maintenance Concern

Author: Tom Ulrich, Warren Stamping Plant, Chrysler Corporation

22800 Mound Road, Warren, MI 48091

(810) 497-3821 • FAX (810) 497-1750

The purpose of this article is to relate a story about a journey from a hopeless, subsistence existence to a fulfilling, profitable present that envisions a rewarding future. To change a mission statement from:

"Get Through Today So We Can Try Again Tomorrow" to:

The difference between the two statements is four years and several millions of dollars - four years of work and planning and several millions of saved downtime dollars.

This story is also about improving service to our in-house customers by eliminating most rework and providing consistent product to sub-assembly operations which enable us to provide world-class parts to the final assembly operations of Chrysler Corporation.

Prior to 1992, Warren Stamping Plant’s die maintenance was concerned with putting out maintenance fires and patching broken dies together on average of one per week. Every troubleshooting die maker knew the welder better than their own family members. Several dies required two and three visits from the welder during each running shift. In short, the dies were out of control and completely unreliable; rework areas were large and growing; and costs were spiraling upward at an alarming rate.

Documenting die failures was deemed a necessary first step. For two years (prior to 1992), a database program (DataEase) was used to store die specific information, track engineering changes and make text-based entries relative to repair work performed by each die repair area. Before long, any printout of repair information about any specific die resulted in a very lengthy document. One problem with text is that it is extremely difficult to quantify and qualify. Trend analysis is impossible. We needed a better system.

To address this common problem, Chrysler’s three major stamping facilities formed the Die Maintenance Committee and asked Harbour & Associates to help them formulate an effective die maintenance procedure for all stamping operations. Monthly meetings provided the opportunity to share experiences - good and bad. This committee decided that the most basic need was an effective die preventive maintenance (PM) procedure developing into a fully predictive maintenance (PdM) process. To this end, PM areas were established at each plant and a rudimentary, time-based PM process was enacted. Critical dies were identified and first time capability, run to run and die related downtime were all tracked to provide a basis for evaluating the benefits of what was being done. In the final analysis, die related downtime was chosen as the hallmark of tooling performance.

During 1992, while auditing and evaluating the dies and trying various methods of maintaining them, the average die-related downtime, per month, was 1379 hours. The average cost per hour of downtime was determined to be $1128.00. There certainly was an incentive to improve tool performance. About $1.55 million per month was the basis for re-engineering the entire maintenance function. This process was, however, approached very cautiously

One of the main problems with implementing something dramatically different from business as usual is the resistance from those who make the plant run on a day to day basis. Hourly employees and first line management provide the foremost challenge to any new process. You had better deliver. We did. Incremental, measured improvements pushed the program forward enough to justify its continued existence.

The PM Area die makers performed an audit procedure on a select number of dies and passed off any dies that needed repair to a die repair area. Obviously, this caused some friction. We pressed on. Crane availability is always a problem and the PM procedures exacerbated the problem because every die was taken completely apart during the audit - all the while, the same number of die were being repaired as before.Careful note was taken of the particular needs of the PM area and attendant die makers. A "wish-list" of tools and accessories was compiled for the area. Manpower usage was studied and attrition loss was considered.

In Detroit, the skilled workforce is aging and incredible numbers of tradespeople are able to retire - some through "30 and out" contract provisions and many more are at or near 62 years of age. When senior management was made aware of the possible attritional losses, they reacted with a hiring program that resulted in the addition of 50+ die makers to a total die maker force of 228. That was in 1994. By year end, we achieved a net increase of only 30. The numbers are falling every month but, the improvements to the die maint-enance program have lessened the initial need for more die makers because the "fire-alarms" are not ringing, the welders rarely visit the production line anymore and downtime has been drastically reduced.

Since there were more dies than resources available to control them, we instituted an initialization procedure wherein every die was refurbished to like new condition before it was allowed into the PM program. During the remake, components and sizes were recorded into a so-called "Blueprint Book". Every perishable part was recorded along with the company order-code number. The Sterling Plant PM crew even appended a drawing of the part with every hole identified by size, button number and punch number. These booklets are extremely popular. The sizes are referenced during the PM procedures to make note of excessive wear and order replacement parts. Major alterations are also recorded in this document as are weights and overall measurements (overall measurements are used to allocate floor space for die storage adjacent to the die repair areas).

Late in 1993, management was sufficiently convinced of the value of an on-going PM process that they asked for our plans to establish a PM program which would encompass the entire stamping operation. A list was prepared that included the purchase of new tools such as hydraulic die separators, work benches, tool grinders, steel horses, chains, carts and other ordinary tools that are used every day. We needed more because we wanted to break up our three repair areas into eight PM Repair Areas. Every die maker was to become a PM die maker as well as a repair person.

More that anything else, the re-organization of the die shop was critical to the success of the PM program. Psychologically, most everyone bought into the idea that they were personally responsible for the quality of the tools that produced body panels for Chrysler’s truck lines. I have dwelt on the culture issues quite a bit I know, but, culture issues are the most important to overcome. Having checklists and hand tools and common procedures doesn’t work unless the work crew and first-line management buy into the program and provide enthusiastic support for the articulated objectives.

Let us now address the process. And now we treat this process as a PdM program. Every new die is automatically included in the program. Older tools must be "keystoned" (repaired to peak performance and quality) before inclusion. Blueprint Booklets are prepared for each tool entering the program. These are filed and made available for subsequent cycle-based checks. The counter is set to zero and every panel produced is counted. All new entrants are put on a 50,000 cycle procedure. Checklists with specific activities are used to ensure a common, measurable procedure. All checklists are retained since they prove that the activity has, indeed, taken place and because die makers are encouraged to write pertinent comments at the bottom of the list.

At the 200,000 cycle procedure, all dies are subject to review by first and second level management personnel. For this review, all checklists, all incidental fixes that have been performed since birth and empirical information is used to evaluate the die and the cycle-based frequency. This step allows a tool to be placed on a different cycle count. The most common movement is to 100,000 cycles. Some are kept at 50,000 cycles.All decisions are based on available data and best guess analysis is removed from the process.

When the tool is subject to the 400,000 cycle procedure, care is taken to ensure that all wear areas have maintained size and surface. Plates are replaced; pins and bushings are replaced if necessary; nitro cyclinders are thoroughly checked and recharged. In short, the tools are returned to "keystone" condition. Again, the tools are re-evaluated and a switch to a longer - or shorter - cycle count is considered for each and every tool. In a nutshell, that is the current PdM process at Warren Stamping.

The next step that the PM Staff has proposed and that is currently under consideration by plant management is to employ SPC data and bring the program to a totally predictive mode - or as close to predictive as we feel is possible. By using old, cheap computers and an array of sensors that mount to a simple fixture and connected to the terminal to provide a continuing SPC display. When this objective is achieved, each die maker will be permanently assigned to specific tools. By viewing SPC data, they can observe whether the panels produced, though still within control limits, are edging toward a limit in any specific sensor-controlled area. If that is the case, that particular area will receive attention after die set to bring the panel back into the mean of the specification.