“The difference between screwing around and science is writing it down.”
― Adam Savage
Dear Friends,
This week, like many this past month, has been filled with yet more mill troubles. Yesterday was a low point indeed. This week we are once again without medium bars and I have been working diligently on getting the new production fixture finished. It has taken the better part of a week to make, including two late nights.
Each day that passed was full of complex decision making, not the least of which was due to the fact that I was partially redesigning the fixture as I went along. This was partially necessary as my original plans hadn't properly considered the fact that the fixture itself also needed to be able to be manufactured (within the limitations of the machine) and wouldn't just spring into existence.
I'll give you an example...
With the fixture, it not only needs to hold the material properly, but it needs to allow for the chips of metal being created by the cutting process and the coolant that is being sprayed on the cut to bleed off into somewhere out of the way. Realistically, the coolant is a liquid so it will find its way out, but the chips are not so simple. A stainless steel chip can get hot during the cutting (in fact that is a part of what cools the system, the shavings taking the heat away), but that means it work hardens. If it is not properly evacuated it could be pulled back into where the tool is cutting and introduce a really hard, oddly positioned, random obstacle for the cutter. Like a piece of gravel between two metal plates, it grinds.
This random chip can (and often does) shatter a small piece off the very hard and very sharp edge of the cutting tool. The next time that cutting tool spins around and that broken edge is engaged, it can and usually does, snap. Under the best of circumstances the other teeth compensate for its failure, but that causes additional wear to them.
Thus, the plate needs to have methods for the chips to be evacuated. The way to do that is to have your bottom plate open all the way to the table below (for example, a slot would be cut all the way through both fixture plates). The table has channels in it that are designed to allow you to attach fixtures to the table, but also creates an easy way for chips to flow out of the way.
But the table is extremely hard and ground incredibly flat (for my table its within one thousandths of an inch across its entire 11x18" surface). They are also a foundational part of the machine. So you functionally can't replace them, and if you do hit them with a tool, it will likely shatter the tool. This means that before I began on the fixture I needed to make some type of riser that could lift the raw aluminum plate that would become the fixture plate off the table to be cut through.
Well, the only real way to make these things is perfectly (as well as you can manage) parallel and flat. That allows you to identify how far you need to go in 3 dimensional space to make the holes, slots, and channels. To make the fixture flat without it being on the table, the risers also needed to be flat, parallel, and identical in height to create the same flatness above the table.
Days of careful, precise work, lead to the risers being finished and then more days lead to the fixture plates being made. The plates had a unique challenge as well. Since the machine isn't perfect at knowing precisely how far its traveled in space (although its very good) things are not exactly where they should be in a perfect world. But you need the two plates to be identically paired. So you have to make them at the same time, and use the same passes to cut through the both to match.
This process was more days of careful, precise work. In the end I had a fixture that was accurate enough to the original specs to begin production cutting.
And that brings us up to yesterday...
An enormous amount of time had been spent setting this up. Far more than I had originally estimated. Since I had to be doing this all day, my other duties either were left undone, or were being picked up by my team. This took a toll on all of them as well, and I could really feel that mounting. But, yesterday was the day it was all finally going to come together. Months of work, tens of thousands of dollars, and a crash course in a very complex subject were going to lead to us having a simple, reliable way to make high quality bars in shop. There was an exhausted hope in the air, like a runner at the 25th mile of a marathon, known the end was near and things could return to their usual level of insanity. I went into the day feeling pretty good.
By noon, that feeling was gone...
Time and time again the program that I had written failed in one way or another. Taps snapped, drills broke, and endmills exploded over and over again. I kept tweaking the code, trying to find the sweet spot. All materials have a window of success when it comes to cutting them. Some, like aluminum, have a wide tolerance for error. Run the spindle too fast, or the feed too fast, or the cut depth too deep, and your carbide endmill will still do what it needs to without snapping. 304 Stainless Steel seems to have one of the narrowest windows for success of the common metals. This window is shrunk further by the limitations of the machine. So experiment after experiment failed. The snap of an endmill sounds like the snap of a bone when it happens enough times, and you wince the same for both.
After a specific group of failures I noticed that there was a bit of a pattern. The endmills were snapping at the end of the slots. With some close inspection I realized that the ends didn't have any gap right at the very end for the chips to evacuate, and furthermore, the lack of gap meant that coolant wasn't being sprayed on it there either. In an effort to solve this, I took a small bit out of the ends of the slot in the fixture to open it up. The next time I ran it there was this horrible harmonic screaming that came from the fixture. That is a clear indicator that the material is vibrating in the fixture.
The next time I ran it, the 8th endmill snapped. I had to stop. I believed that by taking that tiny bit out of the end of the slots I had rendered the fixture unusable. I went and sat in my office and wondered aloud if this was going to beat my record for my most expensive mistake yet. I was crushed. Crushed under the weight of my own expectations, and the feeling that I had lead my team on an exhausting hike only to find we had gone in an elaborate circle.
Despair is not only a sin, its bad strategy...
I often say that line to myself when I feel this way. I find it helps reframe my mindset. Winston Churchill famously said "When you are going through hell, keep going." It's frankly a very important point. To stop and wallow in a moment of failure cements it. You get stuck in a dark place, and no solution just arrives on its own. It must be sought out.
So I reached out to Skiddy and explained the elements of the failed experiments. Thankfully I had a good set of notes I had taken along the way (written down in my Murdy No. 7 Pocket Cut). As we reviewed the results of the various experiments he proposed some new ideas, specifically around how I was cutting the slots. Instead of "plunge cutting" which takes the endmill straight down through the material he suggested "ramp cutting" which cuts down the length of the slot in a downward ramp and then cuts back up the slot at a constant depth. That essentially splits the slot into two triangles which is easier on the endmill.
There were some other changes in the program that were discussed and I left the day feeling that perhaps all was not lost. That leads us into today.
The sun was shining and I was hopeful that we would see some improvement in our bar production with the new changes to the programming, and indeed we did. This was stymied by the drills breaking unfortunately. I to this very moment, do not know what the correct feeds and speeds need to be to properly drill through our stainless steel without work hardening the holes or snapping the drills. If I can find out that ideal speed and feed that will also help reduce the number of taps we snap in the holes.
There is still work to do to fix the design, but hopefully tomorrow I will have all the parts I need in shop to make those adjustments. Today however is another day without a reliable method of producing bars.
In other news this week, We are now able to ship internationally!
We have been battling the issues getting the website to properly translate still, so it's currently only available in English, but you can shop in your local currency and check out in it as well with all the import duties and taxes already included.
I'm not sure how this will go, but I'm excited for the prospect.
Stay tuned for more letters on Thursdays in the coming weeks and be sure to go subscribe to our YouTube channel. If you like and watch the videos it helps us get promoted more by the algorithm to people who may never have heard of us.
Ever your servant,
Colin Murdy
CEO/Owner
Murdy Creative Co.
Cell: 414-434-9001
MurdyCreative.Co
