|
Building a Better Mold, Part 2 By Steve Johnson Our Guest Speaker column this month is a continuation from the November issue’s column by Steve Johnson of Progressive Components. Asking someone to expend an effort to collect and maintain a maintenance database is normally received with the same enthusiasm as asking someone to start tracking how many pizzas they eat in a year, how much money was spent doing so and what were the consequences (calorie, fat and intake, heartburn frequency, blood pressure changes, off-gassing, etc…) Who cares? I need a pizza I pick up the phone and order one---damn the consequences. I’ll worry about those later. Too many American plastics manufacturers spend money on molds and the subsequent repairs (tooling and labor) with the same mind-set. We don’t track and analyze mold performance and repair issues, so too often we clone old problems into new molds. Then when the company CEO calls an emergency meeting to announce that the business is going elsewhere or the plant is being shut down to “save costs” we wonder what happened. Which costs? Whose costs? What products? But without a photographic memory, track we must to improve. As discussed in the last issue of this magazine, grey data is in the mind of the beholder. Seldom collected, hardly accurate and never utilized, “data recall” continues to be in short supply, yet it costs companies thousands of dollars each year in excess tooling and wasted labor. One piece – and one hour – at a time. How to track is another article. For now let’s get right to the pay-off. Let’s look at mold performance and maintenance issues through a few simple reports vs. a mind probe or shop poll. The right reports in the right sequence will not only point to the bad apple, but it will allow you to drill down into what the issues are really costing you. Find the Bad Apple Instead of randomly picking a mold from a database to review its performance we will back up a couple of steps to show how specific reports can guide the user to molds that are quietly eating away company profits. Using the actual database of a plastics molding company, let’s first examine a report (chart #1) that was used to quickly tell us which molds run the most (top 20) and their repair costs divided into production hours to get a CPH (Costs Per Hour). Maintenance costs must be divided into cycles or run time hours to set an accurate baseline in which to reference. Total costs won’t tell you much by themselves. They have to be related to mold or part /production defects.
Chart #1 shows the Top 20 Running Molds sorted first by Run Time, then by CPH (cost per hour). This shows how much it costs (total of labor and tooling) to maintain the top 20 runners per production hour. The molds highlighted in red run the same part from identical molds built by the same company and only 1 year apart. But for some reason the maintenance costs of the 6870 far exceeds (by $58,442.00) the costs to run the 6816 even though the run time column shows only a 32 day difference. What is going on here? Why is this newer mold using much more tooling compared to its older sister? Also notice the (green) 6136 Valve Screw “B” costs about 2 dollars per hour less (or $13,052.00) to run than its twin sister the 5895 Valve screw “A” even though “B” ran 277 days longer! So right off the bat, 2 big flags appear that warrant these molds’ histories being further reviewed by examining and comparing their defects and subsequent corrective actions to see where the dollar difference is. It is logical to expect that if the molds were built by the same vendor at the same time, the molds should perform the same within consistent process parameters and standardized maintenance practices. Any variation in these three criteria, then it could be a whole new ball game concerning performance and maintenance characteristics. Imagine, then, if a call came through requiring a couple more molds just like these be built, one could create a brand new money-guzzling-lemon if this data is ignored. Look Closer This report also raises a few interesting questions about the three identical molds highlighted in blue, even though they are the most efficient running molds in the top 20. The molds are again clones, but the 6147 used almost $6,000 worth of tooling, while the other two used hardly any. Of the three, 6147 also ran the fewest cycles. So again, what tooling was replaced and why was it replaced? Was the tooling worn out or broken? Is the same tooling requiring replacement? When was it replaced and who replaced it? These and other questions need to be answered accurately, without hours of searching through work orders to analyze entries. Would anybody in a shop “remember” the details surrounding these molds through several production runs and repairs? Hardly – at least not with enough accuracy to make a well in-formed decision. But if this type data is not analyzed it is just standard cloning practice to say “build it” to the vender.
For now we will drill into the most expensive (to run) mold of the top 20 and see what we can find out before these issues are built into another mold. Chart #2 shows a summary of both Clear Cups Top Three Defects. Although each mold had over 200 mold or part defects during this three year time period (not seen on this report to save space), these three defects surfaced as the most frequent and costly for both molds and are a good place to start a comparative analysis. The report clearly shows there are three different defects (long gate, drag and flash on the frosted skirt) that are suffered by both molds, but not to the same degree (198 counts to 103). This needs to be drilled into further since the molds have about the same amount of cycles on them and the tooling being replaced to correct the three issues is extremely expensive. The top issue for both molds was easily the Long Gate problem, so let’s look at corrective action data for that defect.
Chart #3 (Corrective Action Analysis) displays what tooling was replaced to correct the Long Gate defect in both molds. The total cost of the tooling and associated labor is seen on the far right. $52,000.00 is quite a lot of money spent to correct only one defect. The next step was to sort and filter on corresponding criteria that surrounded the defects looking for patterns or trends such as start times, start technician, press number, repair dates and repair technician. Also inspected were spare tooling (valve pin and gate inserts) components to verify design specifications from the vendor to validate interchangeability and stack dimensions between the two Clear Cup molds. Here was the outcome with this case. When nothing out of the ordinary was discovered in the shop, a call was placed to the hot runner manufacturer to discuss the problem. When dealing with the mold builder or hot runner manufacturer, it pays to do a little investigative work before questioning their design or work. The hot runner manufacturer immediately wanted to know how many cycles the valve pins and gate inserts had on them, and if the tooling being replaced was coming from any particular locations within the 32 cavities. Here is where the conversation could have ended, but we were ready with even more data. Having accurate reports just a mouse-click away to show cycle counts, defect position, number and type of PM’s given to the mold at specified frequencies, and then providing documentation to demonstrate that standardized procedures exist – and are followed – for removing, cleaning, inspecting, repairing and installing the tooling is indispensable. Although this exercise ended with no fault determination, this company was considering purchasing – or cloning – more of these molds using the exact same technology and design. Through accurate data collection, not only was a design issue (tapered shut-off valve pin to a thin land area at the gate) uncovered in the hot runner system, but R.O.I. justification was established through correction action costs, allowing them to redesign/rebuild the old molds and improve upcoming clones. The ability to track and sort critical data allowed an accurate resolution thus a happy ending without the typical conflicting information or memory lapses. As maintenance awareness increases in companies looking to reduce costs, some companies design in-house electronic programs to track mold performance and maintenance costs. If you do, be sure to associate all costs with specific events or issues such as cleanings and mold or part defects to understand exactly the past problems and to aid in troubleshooting and establishing corrective action costs, before you clone these problems into other molds. Track the past…secure the future is more than just a catch phrase.
Steve Johnson is Maintenance Systems Manager at Progressive Components and also the creator of MoldTrax, the industry’s most comprehensive performance and maintenance documentation system for molds and dies.
Be a TPN Guest Speaker! If you'd like to submit an article for a future issue of the TPN, please contact Editor Cyndi Kustush at editor@tooling-product-news.com for editorial guidelines. Be sure to provide complete contact information and any proposed topics or ideas. |
