The question of deadlines has reared its head again.

Last year I did a series of 3 articles questioning the overall effectiveness of deadlines.

These issues have recently come back to bite me on a project.

The project was to modify an injection mold tool by converting it from manual operation to hydraulic operation and thus make it automated. The reason behind this is sound, as the manual operation was slow and inconsistent which caused problems with the plastic overheating in the machine.

What was less sound was the methodology used to design these mods. Their customer was first quoted for the mods (as they own the tool) which they approved of. They then waited until the day of the deadline to start the project which is where the problems began.

The tool comprises of a fixed and moving bolster, one main moving core and two small side cores. The two small core plates are issue in question here. They are of approximately 9sq inches and are located solidly within the bolster. The locating geometry of the bolsters determines the position of these cores and the final dimensions of the part. As the tool originally was these cores were manually removed by two operators as the two halves of the tool were split. The purpose of these mods was to replace the two operators with a pair of two tonne hydraulic rams.

This was a management led project who were proceeding on the basis that it was going to work because they believed in it. When I became involved I started to question the viability of the project as it looked a bit iffy, however they were not willing for me to spend a day carrying out a detailed feasibility study. The project proceeded and the tapered locating faces were machines away to allow clearance to the side cores to slide to and from the main core.

On the first test the side cores were pushed away by the molding pressures. For the second test, with the customer screaming as the deadline is something like 6 weeks overdue, a temporary m12 screw was added as an ad hoc heel plate to hold the cores in place during molding. Under the molding pressures these screws mushroomed and the threads also began to deform.

At this point I insisted on doing the calculations. The machine was achieving and injection pressure of around 32,000 psi, a typical car tyre is inflated to around 30psi. A pressure of 32,000 psi over a surface area of 9 sq inches equates to a force of around 130 tonnes. The surface area of an M12 threadform to a depth of 25mm will begin to yield at 15 tonnes and will ultimately fail at around 30 tonnes.

During initial tests, the injection pressures were started at a lower figure before gradually being ramped up. The next issue was the by positioning the rams on the fixed half of the tool, the sequence of operation that the tool required to work simply didn't match what the machine could offer. This issue was known, understood and explained clearly by the machine setter, but his concerns were not passed on.

So whats happening now is the portions of the bolster that were machined away for clearance are now being reinstated and the rams are being mounted on the moving half of the tool.

This tool mod is now in its 8th week. Had it been properly thought through with all of the available data this would have been no more than a two week project.

As of 2017, coffee and paper are still cheaper than workshop time and labour.