plastic injection machine?

Hi guys, just like to find out how to make the interaction between a plastic injection molding machine and a servo motor driven robot arm as fast as possible. Heard that many new robot arms today is having very short interval time. How do they do that? by using solid state relay?

We have several Sepro three axis robots interfaced to Nissei injection presses. The standard they use is SPI(Society of Plastics Industry). Our presses came with a 32 pin I/O interface connector which is part of the SPI standard. Other than that its I/O board to I/O board. No relays involved. Management did want some other "bells and whistles" added to the process to which I used a PLC to interface between the robot and the press.

Two of my clients have large Engel mouling moulding machines with Engel robots, suitable for making wheely (rubish) bins. When the die starts opening the robot starts heading down towards the 3D image that I defined as a no-go area. It also "chases" the opening die, which you can imagine has to open a long way to get the bin out. When opening is finished, robot only has to drop 10mm to compress spring loaded suckers and grab the bin. The robot is in charge of giving the OK for all special movements by the moulder such as permit cores out(for pins in axles etc) and ejectors forward. The chasing is the real time saver. Time taken for relays to operate is not worth considering. Most robots have i/o boards with volt free contacts on relays just waiting for your interface wires.

CAUTION: Severe robot/mold damage is possible with the following method!

That said, I'll continue. The Hekuma high-speed (REALLY high-speed!) side-entry robots we had on our Netstal presses used an ingeneous technique they called "Dynamics".

Two proximity switches mounted a 'set distance' apart actuated near the fully open position of the press. Connected to high speed inputs, they measured the speed of the platen as it passed. The speed was used to 'anticipate' the position of the mold. If it was moving fast enough, the robot was allowed to start it's move BEFORE the mold was physically open enough. Since the clamp was moving fast enough, you could assume it would be open far enough by the time the robot got there.

The same method was used to anticipate that the robot was clear of the mold on the way out. The press was instructed to close BEFORE the robot was physically clear, with the assumption that it would be clear by the time the mold reached it.

We're talking only milliseconds of time savings here, but those milliseconds add up to a LOT of additional press cycles (read: $$$) over a year's time!... :nodi:

Things can and will go wrong though... We did have some spectacular crashes over the years!...

beerchug

-Eric

you don't happen to have any pictures? ;-)

I no longer work there, so sorry panic, no pictures...

Hekuma's website has nothing, but if you can get a hold of the Jan '02 issue of British Plastics & Rubber, it seems they have an article about it entitled "Robot over-rides machine control to initiate opening and closing". See THIS LINK

beerchug

-Eric

Oh, wait a minute... The pictures you were hoping for were of the 'spectacular crashes' rather than the "dynamics" system. Sorry, the answer is still no.

beerchug

-Eric

Yes we sure had some crashes, but only the spring loaded suckers. The rest of the robot head was forbidden to enter the 3D zone. Some pretty complicated maths here I imagine with this system of "chasing" on the Indramat controllers and die position transducers, but I just put in the numbers and we got there in the end - really zippy and impressed me. Tourist customers are always shown these machines.

Thanks

Hi guys, thanks for the info. Learnt something in this discussion. Anybody has the jan 2002 article? thinking of making my own side entry take out robot that has the shortest interval time ever. any suggestions?

How fast do you want to go?

The trick is keeping the tooling weight to a minimum, so 2 words... Carbon Fiber... :nodi:

Sorry to say, someone has already beaten you to the punch...

Back in the early 90's, we had a few of the first-generation ultra high-speed robots from CBW on 4-cavity video box molds. Impressive in and out times on these early units, and they've probably improved over the years. Looks like they're concentrating on side-entry now. HERE'S a link.

Panic likes pictures, and I actually HAVE a picture this time. Too bad it was taken AFTER we shut down that line. Therefore, the only picture I have is of what USED to be a T-192 CBW robot...
[attachment]
No, this was not after a 'spectacular crash'. They were dismantled on purpose...
beerchug

-Eric

Thanks Eric

Remember "Bad day" topic? Now if only every 10th post
had a picture

SPI serial is not the answer.

The mold control and robot must be controlled by the same controller so they can be tightly coupled/monitored.

Its not difficult to imagine a mechanism attatched to moving platten and something stationary, that when die opens, robot arm is forced into the middle, and on closing, the arm is pushed out again. Saved some time there, huh?. All you have to do is grip (or suck). Robot speed is determined by moving platten speed, not that it matters. Just a thought
Jon

We tried one of these. I think it was from Conair-Sepro (I remember it was Sepro orange). A giant bar mounted on the back of the press with a cam profile cut into it. As the mold opened, the tooling was cammed into the mold. In theory, it's a great idea, but clamp speed was limited because of the huge mass of the arm. The arm had to rotate a full 90° over only about 6" of platen movement. Open or close the mold a little too fast, and the cam followers would snap...

Nothing beats a servo!...

beerchug

-Eric

Yes Eric, size is the limiting factor on the these mechanical devices. Generally they're just sprue pickers on smaller machines.
Ken, can you give us more detail on what you are doing. We made a small side entry picker once. Just pneumatic and simple. Still, why not get a bit racey if you have the tools.