Solving Process Issues With a Detail-Oriented Approach

Reader Question:

We’ve got great machinists here. Most of the time, we find issues quick, solve them quick and get our parts out on time. Where we’ve struggled is when an issue doesn’t seem to react immediately to a more traditional fix or lingers throughout long runs. What’s a better way to solve issues when the solution is less obvious?

Miller’s Answer:

Troubleshooting for a machinist is an essential skill. Whether it be program errors, machine alarms or broken tools. It sounds like your team is handling these with ease. However, those issues typically follow a simpler cause-and-effect model. For example, I cut too deep; therefore, I broke a tool. Once you fix the cause and reduce the depth of cut, the tool no longer breaks. The root cause was obvious, and the corrective action was clear.

Many dimensional issues are solved in a similar manner. If a bore is too small, adjust the boring bar. If a depth is wrong, correct the tool height. If a position is shifted, check your work offsets. In fact, given the consistency and accuracy of our machines and the tools we put in them, our process problems are often limited to simple adjustments to the machine’s offset table.

However, what is a machinist supposed to do when their process issue lingers or their first attempts at a cause-and-effect solution prove ineffective? We often encounter issues that happen occasionally, or we are working with a process that is on the ragged edge of in spec. The key to solving issues like this is a thorough review of the process, and a commitment to working the details one by one.

To begin, identifying when you’re stuck can often be the hardest part of solving an issue like this. As machinists, it is often very easy for us to throw metal at an issue. If I’ve done this twice without significant improvement from either of my corrections, I start to question my perceived root cause a little more. It is essential for a machinist and a shop to have the discipline to pause a process and reassess, even when schedules are tight. Time spent making bad parts is the same as time spent making no parts. Therefore, making a game plan and then executing a robust, lasting solution will often require the same amount of time, but will result in much less material consumed.

In addition to identifying an issue before it festers, we need to gather some general information about the problem. Most importantly, how often is it occurring, and when? Can the issue be related to a shift or an individual who may need some additional training? Is it related to surrounding shop events or a new lot of raw material? Does it happen when the machine is cold or towards the end of a long run? Answering all of these questions helps drive the next step of root cause analysis.

After identifying the issue, and its potential for long-term headaches, we must identify all the potential causes for that issue. We’ve already recognized there is no single root cause, so make a list of every potential culprit. This should include tools, fixtures, tool paths, programs, order of operations, machine conditions, coolant and so on. Be very thorough here! The purpose of this exercise is twofold: First, it is to identify the items we should address, and second, it is to get a shop thinking about the machining process in its entirety rather than laser focus on a short list.

The next step is to rank these items relative to their impact on the issue at hand. For example, culprits related to workholding may be ranked higher for a flatness issue than our cutting tool of choice. It’s not that the tool has no impact, but we want to address the causes with the potential for the highest impact. Next to each item, identify what you plan to do about them related to the issue. For workholding, it could be to modify the soft jaw shape or reduce clamping force. For tooling, it might be speed and feed changes or freer cutting tool geometry. You also need to identify what solutions are simpler to implement today and for free, and what improvements may need time and money to work out. With a big list of suspects, a ranking of probability and an idea of difficulty, lead time and price, we can now begin addressing each one.

The only thing left to do is to execute the list. Look for free, simple and high impact and do those first and keep working down. If you see an improvement to your lingering issue, then keep this change as part of the machining process. It may not be the massive win you were hoping for, but it’s a win we want to keep. If you see a neutral result, but you know it’s a better process for other reasons (consistency and repeatability may be a side benefit not related to the current issue), consider keeping it in place as well. Most important is that with each one, we properly document the changes and level of improvement. If something makes a problem worse, then we want to be able to quickly revert and change course.

Our goal with each process improvement is to keep stacking up small wins until we are in tolerance, then perhaps stack a few more to give us a cushion for long term stability. For example, correcting runout in our tools may be worth a few microns, while adding a fixture support may be worth a few more. Maybe a tool path or speed and feed improvement is also worth a few more still. Throughout all these changes, we will eventually get to consistent, in-spec parts. The final resolution to any lingering issue is going to be a sum of each mini-improvement we implement, so be sure to take the wins as they come.

A machining process should be understood like a system with multiple inputs. All inputs within that system have a part to play in final part quality and how robust that system is throughout a long part run.

To reiterate, we are working in the foggy areas of the process, and the toughest issues in a system are rarely obvious to fix. Focus on the details, work them diligently and keep stacking those small wins. Eventually you’ll be stacking good parts as well!