Welcome to our 340-2 Injection Molding Design
Prepare to be A-MAZE-D...
-------------------------------------------------------------------------------------------------------------------------------------
Our project group utilized the ProEngineer software package to create a part to be plastic injection molded. When designing our part, we considered three charactersitics:
§ Practicality (would our part have a purpose?)
§ Usability (could we actually use our part?)
§ Complexity (was our part challenging to make?)
After considerable thought, we decided to produce a part for our younger friends. A maze game that required guiding a small ball through a maze and into a hole would be challenging for both the designers that created it and the young children that played with it. The main challenge for us (the designers) was creating such an intricate design out of one mold.
Once we settled on a design, we had to tackle several tough issues:
- Using one mold would require a complex "floating" core to create a cavity between the two part halves where the ball would be deposited.
- With the size of the mold only 3.5" x 3.5" x 1.4", the maze was going to be pretty small.
- Given the height of the walls, the design would require considerable draft to remove.
Here's
a view of the top half of our maze...
...and
a view of the bottom half (which fits inside the top, like a shoebox)
MACHINING THE MOLD
Once the design was completed in ProE, it was time to actually machine our mold and core. The mold was machined using NC programming that were sent directly from ProE to the machining center. Using different size mills allowed us to get into tight spaces using a local milling process in addition to the major face milling. We had to have a runner with two gates because of the size of our core and the thinness of our walls (you can see the runner and gates below).
Check out the our
NC Programming results…Not too shabby….
MACHINING THE CORE
Waiting for the automated mill to work on the mold took a lot of time and gave us a chance to work on the core. This was the most difficult part of our project and by far took the most time. Our idea to have four different layers, each machined separately by hand, proved to be a bad idea and if we had to do it again with floating core we would definitely have machined it all from one piece of metal. Anyway, we started with 4 pieces (for each of our 4 layers) of aluminum stock and cut them to the appropriate outer dimensions (3.4" x 3.4"). Then, using both manual milling on the Bridgeport machine and automated milling on the CNC apparatus, we had all four of our layers cut out. Unfortunately, our final core did not fit comfortably into the mold, forcing us to grind and mill it down, layer by layer, piece by piece, until it fit.
Here’s a picture of our complex core, with all four
layers showing…
**Note how the connecting .
rods
bent.** .
TIME TO INJECT....
Finally, WE WERE FINISHED, and it was time to shoot the prototype in the injection molding machine. Well, we shot, and....
Oops, not a great first run. It turns out that we didn't have enough draft and our part got stuck, but we eventually got it out. Then we saw our other problems.
WHAT HAPPENED TO OUR GREAT DESIGN??
Here were our main problems:
1. Our walls were too thin.
2. We allowed to much room for clearance between the inside walls of the top and bottom parts.
(1) Although we had the dimensions of the core correct in our heads and on paper, when it came to the automated milling we inputted the wrong numbers into the computer. In order to create a 2.5" x 2.5" top part with a wall thickness of 0.15", we should have made the top part of the core 2.2" x 2.2". Instead, we mistakenly programmed 2.5" x 2.5" so that it would fit snugly into the mold (forgetting about the walls) and eventually had to manually mill the dimensions down to 2.4" x 2.4". In the end, even this proved to not be enough.
Similarly, for the bottom part of the core, we needed a 1.8" x 1.8" core size to produce a 2.1" x 2.1" final part (to include 2 walls and a 0.05" clearance). Instead, we gave dimensions of 2.1" x 2.1" to the core and had to mill down to a final dimension of 1.9" x 2.1".
As a result of the above errors, our none of our walls were thick enough, and the final bottom part was pretty much swimming inside the final top part.
(2) Instead of adding to this sizing problem by leaving even MORE clearance, we should have allowed for shrinkage and not even added extra clearance. After all was said and done, we ended up with a clearance of 0.15" all the way around our bottom part.
GIVE US A SECOND CHANCE...
If we could do it all over again, we would make some major design modifications. First, we would use two molds, one for each half, instead of trying to squeeze the whole part into one. This would eliminate the need for a core entirely. Although this might get rid of some of the design complexity (and FUN!), it would probably produce a better overall part. If we chose this option, the parting line for the top half would be between the maze and the walls. There wouldn't even BE a parting line for the bottom half.
Why is this so much better?
1. All walls would be machined mechanically and their thickness would be uniform
2. Machining time would be significantly reduced (only the walls would have to be cut out of the molds)
3. NO CORE!
This new strategy would cut production time down by hours! Then we could make all the kids happier quicker.
If you
have any questions, feel free to email us at j-edmonds@northwestern.edu