Injection Molding Control

[ndzied1]

In the vein of pre-packaged Injection modling machine solutions, Bosch Rexroth has 2 specialized cards for use in injection molding machines. With these pre-packaged cards, a very simple PLC could be used to initiate and coordinate the moves.

The HACD-DPQ is for Injector control
The HACD-DPC is for Clamp control

Some information can be found at:

http://www.boschrexroth.com/hacd

You can also download the software for card setup (called BODAC) there.

Here is a screen shot of the injector setup:

Here is one for the Open Mold Setup screen:

Disclaimer: I work for a Bosch Rexroth distributor.

If you want to persue this option, send me an e-mail or PM and we can discuss further.

 

[Unregistered]

Archie,
I use PLC's all the time to do IMM control retrofits. I find cost is determined by type of control needed. If this is an early 70's machine, with all bang-bang valves, cost should be fairly inexpensive.

There are many specialized control companies out there for IMM's. I've used SCI, Barber-Colman, Hunkar, Van-Dorn, ect. Just depends on what you want for control.

Most OEM's have there own controller upgrade available.

Pete....when did HPM switch to Seimens? Last I knew they were still using Barber-Colman.

Could'nt remember my password....that's why unregistered.


Puck

 

[Jiri Toman]

IMM retrofit is a very complicated undertaking unless you use a pre-packaged solution such as AB's PROSET. Most people here are underestimating the complexity of this project by an order of magnitude.

Open-close-squeeze and squirt you think that's all there is to it?
Hardly. I have retrofitted a molding machine years ago using the AB's QDB module.
I did not use the Proset.
To begin with, if you are going to do your own controls you will need to have a deep understanding of the molding process.
Judging by the posts, most people here don't.
They are trivializing it.

AB's QDC module was specifically developed to control Inject as well as clamp sides of an IMM. It is undoubtadly the most complex module that AB has ever developed.
There are literally hundreds of parameters that need to be intialized. I think that with QDC I needed about 27 block tranfers just to start the module up.
You think that if a general purpose motion controller could do the job that AB would go through the huge expense of developing this module?
Incidentally I was involved with this right at the bleeding edge of it. That's when late Frank Dyke and Dean Giancola were still in charge.
Actually some of the scaling values are based on my feedback.

On the inject side you need to do RAM pressure vs position, RAM pressure vs time and you need the capability to switch over
to pack/hold based on RAM pressure OR position OR time OR cavity pressure. You will need to control the shot size, pre-suck back, post suck back and screw rotation.

General purpose motion module can't do all that. In addition you have to control the clamp side.

we have worked long and hard at make the transistion from position to pressure control very smooth

I have no idea what this means.
There is no transition from position to pressure control.
The RAM pressure is profiled vs position or time. The switch over to pack/hold is based on position, time, RAM pressure or cavity pressure.
How can generic motion controller deal with cavity pressure?

AB has a built in ERC which is a tuning process automatically
adjusting PID parameters based on last few parts produced.
Generic motion controller will not do that.

How will generic motion controller deal with RAM pressure vs time profile?
How will it deal with screw rotation, suck back and shot size?

This off loads the PLC so it can to the temperature controls, HMI and safety interlocks.

You also have to control the clamp, core pull sequences, slide core interfaces, knock outs and robotic interfaces.

Finally, the HMI screens always seem to take the most time to write and get right.

I agree with that I had to develop over 100 sreens to have a state of the art IMM control. That is a major undertaking.


Finally there is a question of hydraulics. Having reliable
P/Q type valve for your inject is a must. I suggest that you contact Norm about this. Bosch- Rexroth are the leaders
in this field and you will need hydraulics expert to
help you select and size the valves.

 

[Unregistered]

Jiri,
I was in no way trying to trivialize it. I've been doing this over 30 years, and fully understand all aspects of injection molding. Over 50 machines and still counting. I took Archie's post as meaning "can a plc be used".

I think you fail to understand the concept of cheap. Injection molders as a whole do not like to spend a lot of money on old, used machines.

I believe the comment about not understanding the transfer from position to pressure quote, should have been "velocity/speed to pressure".

You can absolutely have all the bells and whistles your speaking about for a price. That price is usually cost prohibitive to the molder.

I agree that a retrofit is complicated, even with a canned controller. If you don't understand the lingo, and how the process works, it makes specifying a control way more difficult. I've corrected a bunch of messed up systems that "generic guys" have installed.

Puck

 

[Jiri Toman]

You can absolutely have all the bells and whistles your speaking about for a price. That price is usually cost prohibitive to the molder.

I agree it's not cheap. You will not find too many (if any at all)
OEM's who would be using a PLC based solution as a standard
control for an IMM. Some manufacturers will do it as an option.
OEMs usually sell thousands of machines and so it becomes more cost effective to develop their own controls. Spare parts business and support will go back to them.
Every time you want to make a change to the IMM controls you need to call them, they like that.
PLC control will make end user much more self sufficient.
I know that AB has tried for years to get an OEM to use their
controls but I don't think that they have succeeded.
(Apart from an optional control with high price tag).

 

[Peter Nachtwey]

I have no idea what this means.

It means that you want to transition from the velocity profile to pressure control or limit as the pressure builds up in the mold. You want to approach the pressure set point quickly with out over shooting and settle quickly. Normally the pressure that must be applied at the beginning is very low and the injector executes a velocity profile as a function of position along the barrel. At the same time the pressure is monitored. When the injector gets close to the end of the barrel the pressure in the mold starts to rise rapidly. At some time the controller must control the rise in pressure so the pressure does not blow the mold apart. We don't wait for the pressure to get to the set point before trying to control it. We actually calculate a pressure trajectory and control the pressure as it approaches the set point.

There is no transition from position to pressure control.

Who said that? You start out with a velocity profile. The pressures are low. Eventually the pressure will increase. Somewhere you must make the transition from velocity control to pressure control or from being velocity limited to being pressure limited. You can control position/velocity or pressure or you can limit both. You can't control both. We have an option to velocity or pressure control, or do velocity and pressure limit. Just because you have done ONE injection molding system doesn't mean that all the rest are like the one you have seen.

The RAM pressure is profiled vs position or time.

Define profiled. Do you mean controlled or limited? You can limit the pressure; you can’t control the pressure as a function of position. Be careful before you take exception to that.

The switch over to pack/hold is based on position, time, RAM pressure or cavity pressure.
How can generic motion controller deal with cavity pressure?

What is with this generic controller stuff?
It is obvious, the controller must have analog pressure sensors and have the ability to run a PID control for both the position AND pressure feedbacks. Again, there must be different modes of using the two PIDs. Also, one needs to map the pressure sensor of interest to the PIDs. Who said our controllers are generic?

How will generic motion controller deal with RAM pressure vs. time profile?

A 'generic' controller can’t because it wasn't designed to do injection-molding machines, but the QDC is not the only module that was built for injection molding. We can handle all the changing set points with a built in state machine. If actually has commands for changing pressure set points and it as a separate PID for the pressure feedback.

How will it deal with screw rotation, suck back and shot size?

Easy, with a quadrature or tach feedback. Most systems I have seen rotate the screw open loop from the PLC. It is not a speed critical issue. Those that need precise control can use the motion controller. The suck back and plasticizing positions use a MDT feedback. The plasticizing also requires pressure control. The screw turns and is forced back as more plastic is forced to the front of the screw. The MDT feedback tells the controller when the screw has been force back far enough. The backpressure is reduced to about 100 to 200 psi depending on how much and the plastic is forced into the barrel. This keeps air from getting into the system. The screw turns to plasticize/melt the plastic beads. The screw position is monitored until the shot size is achieved. Then the screw is and the hopper are stopped. These distances may be modified a bit from shot to shot.

You also have to control the clamp, core pull sequences, slide core interfaces, knock outs and robotic interfaces.

The motion controller is usually only controlling those items that need closed loop or very fast control. So this means only the clamp. Digital output can control the other items you mentioned but normally that is left to the PLC.

I think a GENERAL motion controller can do the job very well.

Finally there is a question of hydraulics. Having reliable
P/Q type valve for your inject is a must.

You don't use a P/Q valve for injection molding. This leaves too much of the control in the hands of the valve. I like the Rexroth valves with the Q2 spool, which was designed specifically for injection molding. This is an odd valve and it works completely different in each direction. This valve is not intended for position control. It can't be used to hold a position because the B port is normally ported to the tank. The Q2 spool works very well for controlling velocity profiles while extending and it makes it very easy for the controller to control pressure as the pressure gains seems to be a little lower with this spool.

I suggest that you contact Norm about this. Bosch- Rexroth are the leaders
in this field and you will need hydraulics expert to
help you select and size the valves.

They do know their valves and I like them and recommend many of them. Now what do you want them to tell me about hydraulics?

I have found that each manufacturer does injection molding in a slightly different way. My goal is to provide the means by which they can control the velocity or pressure or limit both as accurately and simply as possible. I can set up injection, pack and hold, backpressure and clamp sequences in just a few hours after the machine is ready. Simple machines can be done in much less time. I don't claim to be an expert at injection molding. I have seen enough to know there are too many differences. There certainly is an art to determining the speeds and/or pressures that must be applied to fill the molds rapidly and evenly without over pressurizing the cavity. However, if the customer knows what he wants then it is usually very easy to provide it. The tough ones are those that think they can control velocity AND pressure at the same time. They need to be reeducated.

 

[Jiri Toman]

You don't use a P/Q valve for injection molding

Of course you do that is what Arburg uses. Talk to Bosch.

You start out with a velocity profile. The pressures are low.

You did not mention velocity , you mentioned position!

Define profiled. Do you mean controlled or limited? You can limit the pressure; you can’t control the pressure as a function of position. Be careful before you take exception to that.

Profiling in IMM terminology is pretty much the same across all the machines, HPM's, Arburgs, Sodicks, Engels, New Britains, Huskys, Demags etc.
Profiling is pretty much shown by Norm in his pictures.
You set up 10 step profiles Pressure vs Time or Pressure vs positon etc.
If you are profiling velocity you will have pressure limits
and you will eventually switch over to pack/hold based
on one of 4 different triggers (if you have them enabled).
The cavity pressure is something you are obviously not familiar with in addition to many other aspects of an IMM.

 

[ELake20]

I've looked at doing something very similar to what you guys are talking about, and was overwhelmed by the amount of things that would have to be achieved...we have 9 Van Dorn Pathfinders (for you Eric with the crappy controls) we also have an early van dorn with relay logic controls...talk about a mess and a hugantic cabinet. I need a digital camera.
For our larger machines we have a 2 1100s and 4 750s...those are cincinatti milicron. They love us calling them for repairs and program changes. I've even discovered a bug in there program and they did come out and fix it free of charge...the newer milicrons are 486sx based..haha imagine all the processing power!!

Cool topic..its right up my alley but I was easily discouraged when I saw what would be required to use a PLC...you are correct when you say injection molders dont like to spend money...I know all about that!!

 

[Eric Nelson]

...we also have an early van dorn with relay logic controls...talk about a mess and a hugantic cabinet.

We also had a few little 150 ton machines with that 'giant clicking cabinet'... Most of the contactor had broken free of their mooring and were just hanging by their wires...

I'm actually glad I didn't go through the trouble of upgrading them with PLCs, as these machines continued to run for many years just fine. As far as I know, they're still running...



-Eric

 

[Unregistered]

Guys,
Lots of good stuff covered here. Those old Van Dorn relay machines due very well just by eliminating those AB relays with a PLC and adding some positioning sensors. Done lots of those. I find that the PLC hardware is less costly than changing out these relays.

When speaking about pressure limiting during injection, prior to tranfer, I have to disagree with you guys. You do not want to pressure limit during inject, on velocity control. You want to control velocity, not pressure.

Arburg and I believe Boy use the P/Q valve. Works well.

Van Dorn tried using Seimen PLC's for several years as part of their standard machine. They developed their own system as they were having problems with the Seimens hardware and software.

You bet the OEM's want their own control system. Lots of big money in field service calls, and software upgrades/changes.

Cinci was using Vickers hardware on the 486 systems, and is now using Seimens on the Extreem controller. Software development may be their's, but hardware is not.

The old Pathfinders, with the Seimens control had lots of problems, until resolved, which I agree made them crappy. However their new Pathfinder controls seem very robust.

That's my nickel's worth.

Puck

 

[ELake20]

Cinci was using Vickers hardware on the 486 systems, and is now using Seimens on the Extreem controller. Software development may be their's, but hardware is not.
Puck

Neat. We just got a new extreme cinci, I'll have to check that out..
I've got the vickers software, its very cumbersome and runs in dos, but gets the job done.

 

[Jiri Toman]

When speaking about pressure limiting during injection, prior to tranfer, I have to disagree with you guys. You do not want to pressure limit during inject, on velocity control. You want to control velocity, not pressure.

Puck,
I agree with you 100%. If you want to profile inject based on velocity and control velocity, but as Peter points out and Norm's picture shows you always need to have a pressure limit.
I have seen many molding set-ups that don't use velocity profiling at all, they just profile pressures. But I do agree that if you choose to use velocity profiling as your inject mode of operation,
you need to tweak it so that you don't hit those pressure limits.

AB's QDC module has all these goodies built in.

 

[Neal Gigliotti]

I work as an applications engineer in the Plastics Market Group for Bosch Rexroth, responsible for many OEM's as well as support for the rebuild market. I'll try to be as neutral as possible for purposes of this forum.



As a rule; don't do the PID loops for closed loop control of a hydraulic axis in a PLC. Hydraulic systems are very non-linear by nature. We use software and hardware techniques that were developed over decades with a very broad range of motion control experience. These are necessary to properly handle hydraulic system and valve characteristics. While it is possible to program a PLC and control the hydraulic axis, the results will always be less than we would desire. Processing speed is only a small part of the requirements for a quality hydraulic controller. AB gave up on it several years ago and asked us to work with them on the DPQ/DPC solution for hydraulic injection machine control. I can supply you with lots of information on the subject; if you would like it send me an email please. Even the injection OEM's that still use their own "black box" PC and PLC solutions for control spend an inordinate amount of time on the subject. Those solutions usually don't work very well either.



Use either a Control or Compact Logix PLC (you'll save a lot of money with the Compact Logix) for the basic machine sequence control and the temperature loops. Use stand alone dedicated controllers for the hydraulic axis. There are options for those; we manufacture the DPQ (injection) and DPC (clamp) axis controllers for plastic machinery. You'll find them relatively easy to implement and at the top end in control and data acquisition capabilities. There is documentation from AB available comparing this solution to the ProSet 200 that is quite enlightening. If it's implemented well the cost of such as system is less than you might think. There are several plastic machinery integrators that we use who have this type of solution programmed and ready to go. It's possible to implement it yourself if you have the time and experience with PLC programming and injection molding machine requirements.

 

[Jiri Toman]

AB gave up on it several years ago and asked us to work with them on the DPQ/DPC solution for hydraulic injection machine control. I can supply you with lots of information on the subject; if you would like it send me an email please. Even the injection OEM's that still use their own "black box" PC and PLC solutions for control spend an inordinate amount of time on the subject. Those solutions usually don't work very well either.

Proset 200 is alive and well, AB did not give up on it by a long shot.

This is off the AB's web site:

Pro-Set 200 features pre-programmed and tested machine ladder logic and PanelView screens which enable full operation of injection molding machines. With other control systems an engineer must invest 3-6 months time to write the control programs and operator interface screens necessary to control the machine. Pro-Set 200 logic programs are already written and only 2-4 weeks of time is required to perform system configuration and machine setup.


Expert Response Compensation (ERC2) is a model predictive control algorithm for control of pressure and flow loops (injection, clamp, and ejectors) that:


eliminates the need to manually tune the pressure and flow loops for the clamp and injection units:

minimizes overshoot and steady-state error

provides comprehensive control over pressure and velocity profiles

provides machine performance metrics

The ERC2 algorithms are embedded in the SLC 5/04P processor and are part of operating system. ERC2 algorithms recieve data in the form of pressure and position information from the ladder logic included in Pro-Set 200. This information is used to calculate new output values (pressure and flow).

In terms of DPQ/ DPC solution and cooperation with AB I found the following App note on AB's web site:

http://domino.automation.rockwell.com/applications/gs/region/gtswebst.nsf/files/VP+JVH+Engineering+FINAL+5-14-04.pdf/$file/VP+JVH+Engineering+FINAL+5-14-04.pdf#xml=http://127.0.0.1/texis/search/pdfhi.txt?query=injection+molding&pr=Domino+Servers&prox=page&rorder=500&rprox=750&rdfreq=0&rwfreq=0&rlead=250&sufs=2&order=r&cq=&id=4272bce51


All they are saying is that on one project JVH has used DPQ/DPC tied to AB via D-net.
It does mention that AB has worked with Bosch on the DPC/DPQ.

What it means is that AB has broadened it's product offering by being compatible with
DPC/DPQ. It does not mean that AB has declared it's own controller dead!

Even the injection OEM's that still use their own "black box" PC and PLC solutions for control spend an inordinate amount of time on the subject. Those solutions usually don't work very well either.

Most OEM's don't use PLCs, they use their own controls built from ground up.

Tell me, is Arburg or Engel using DPC or DPQ in their controls?

Before I retrofitted Arburg with AB based controls I was told
by Arburg engineers that their controls are the best and that I will never be able to make AB perform as well as their controls. Guess what. When I got done I ended up with better controls than what Arburg had. Machines I retrofitted have been now running for 13 years. Total of 52 machines.

If you have a definitive proof that AB is abandoning Proset
based controls in favour of DPQ/DPC please publish it here !!

 

[Neal Gigliotti]

Jiri,
Sorry for the confusion, AB hasn't given up on the Proset 200; they gave up on development of any new dedicated modules for hydraulic applications like injection molding. We were also told by AB that the SLC product would soon (whatever "soon" means) be silver series. Logix will obviously be their future.
AB has several other publications on the DPQ/DPC including application notes and a very interesting head-to-head test against the Proset 200. I also can't find any of the other links, but I was given the files by AB and would be happy to email them to you. It's possible that those links are only for the AB integrators.
Also, sorry but I can't discuss specifics regarding OEM's due to Bosch Rexroth relationships with machine builders. I hope that you understand my position. However it's not a surprise that a resourceful controls person could exceed the capabilities of OEM controls.