mike miller,
I ran the engineering department at place that did a ton of sheet metal. Though this is not explicitly answering your particular questions, it does address how we resolved the same issues.
We did a little homework and made some measurements and calculations. This allowed us to improve our part and flat pattern design to the extent we dramatically reduced our scrap and WIP, eliminated our bottleneck (Yes! Everyone in engineering/manufacturing should read The Goal at least every other year!), and were even able to design-in more features - all because we had good data and knowledge. Since we had our own sheet metal fabrication we were responsible for the flat patterns as well as the design of the finished parts.
Here is what we did:
- Measured the actual thickness of our metals. Believe it or not, they were not in the middle of the published ranges for the gauges. We updated this every six months or if we got a different supplier. It is astounding how many people just use the nominal gauge value and expect accurate results. That might be okay for the part design for some (we weren’t satisfied with that), but that will not work for generating accurate flat patterns.
- We invested the time to do some bend studies. We took scraps of all the gauges and measured their thickness (again) and their lengths. Then we gave them 90 degree bends and measured the resulting legs.
- We did step 2 for all gauges and for all press brake punch/die combinations.
We used a spreadsheet to calculate our K-factors for our combinations of tooling and thickness. You can go as detailed as you want with this, but we made separate tabs for each material and were careful to indicate the tooling. (We even looked at grain direction, but for us the typical difference in the K-factor was on the order of 0.02 so we determined that was not significant enough for us to worry about.) - We had about 15 people in our department so I made them all participate in the data collection and calculations. Some were understandably resistant to this until they saw with their own eyes that we could get accurate results! Then they became diligent in properly updating and applying the technique.
We were making multi-bend parts (typically 3-4 parallel bends in one direction and then more bends in other directions) in 12-gauge with only .002-.004" deviation from target. Most of our bends were 90 degree, but we did not encounter any errors with more open bends. This absolutely worked!
See the attached files for more explanation and an example spreadsheet.
K-Factor Development.xls (43 KB)
K-Factor Development.PDF (68.5 KB)