posted February 23, 2003 06:04 PM               

We are two engineering students from Universtiy of Waterloo and we are trying to do some research for Extrusion Control Software.

We are trying to determine if there is a demand for software that can better predict software process requirements for a desired output, for our 4th year project. If you could please look over the following questions, we would appreciate any input anyone can give regarding any of these questions. Thank you for your time.

1) Roughly how often does an extruder experience downtime?

2)How much of this time is due to troubleshooting during start-up of runs?

3)What are some rough costs loss due to this time? How often do companies typically restart a extruder processing run?

4) Are there any single or twin screw extruder control features that industry would be interested in seeing implemented?

5) How much would companies be willing to pay for this software?

6) Do companies typically hire extrusion consultants of handle processing problems in-house?

7) How much are companies willing to pay for consulting services?

8)What are some current extrusion processing issues of interest in industry?

9) We would appreciate any further comments or input regarding extrusion processing software.

posted February 24, 2003 05:24 PM               

quote:

---

Originally posted by WaterlooStudent:
Hello,

We are two engineering students from Universtiy of Waterloo and we are trying to do some research for Extrusion Control Software.

We are trying to determine if there is a demand for software that can better predict software process requirements for a desired output, for our 4th year project. If you could please look over the following questions, we would appreciate any input anyone can give regarding any of these questions. Thank you for your time.

1) Roughly how often does an extruder experience downtime?

That would depend on a lot of things. If a resin or color is changed over frequently, the down time would be higher.
If there is a lack of preventative maintenance, it would be higher as well.

For the most part, the less the better.

2)How much of this time is due to troubleshooting during start-up of runs?

A few factors.....how well is the process been defined and characterized, the skill level of the technicians, and the preventative maintenance of the equipment.

3)What are some rough costs loss due to this time? How often do companies typically restart a extruder processing run?

Costs are dependent on labor and overhead, labor absorption, resin costs, etc.

Restarting a run? Depends on if the extrusion is captive or not, and if there are any programs in place from logistics personnel to increase inventory turns.

4) Are there any single or twin screw extruder control features that industry would be interested in seeing implemented?

Many already are. There is recipe controls already available, main panels with multiple extruder inputs, pressure control loops, gauging feedback loops, lines operating as complete systems.
Depending on the design of the controls, they can make it easy or difficult to operate the equipment.

5) How much would companies be willing to pay for this software?

Depends on what you are saving them. If there is a good, well documented processes in place, the bang for the buck would probably not warrant high cost software. Depends on the pricing structure.
One could always write their own with something like labview and RS485 or industrial buss controls.

6) Do companies typically hire extrusion consultants of handle processing problems in-house?

Depends on if they have the expertise or not, and the systems in place to identify problems and probable solutions, as well as a good PM program (do you sense and underlying theme here?).

7) How much are companies willing to pay for consulting services?

They usually pay a premium. Some people offer their services for as much as a few $K per day, though if you do, you had better have a track record of getting results.

8)What are some current extrusion processing issues of interest in industry?

Medical industry processing can be quite difficult.
Using what many use for lab equipment as standard production lines can be difficult. We can measure our output in grams/minute, not pounds per hour.
Bulk density and pellet geometry can make or break you.
I have used as low as 1/2" extruders for production equipment over the years. With a 1/8" flight width and a root diameter of 3/8", the channel depth is usually what many use for a metering depth and the channel width is quite narrow. We have many runs with more resin left in the machine when we are done than what is saved for product. Many of our compounds are easily $15-$50 a pound.
I'm sure the high volume guys have quite a few of their own problems as well (I wouldn't know much more then theory, I've spent my entire 24+ years in medical
extrusion).

9) We would appreciate any further comments or input regarding extrusion processing software.

Good luck. You might want to include validation services with the software (CFR 21 part 11), it will get you in my door to at least consider it.

Thank you very much for you time and input.
Amanda & Greg

---

posted February 25, 2003 09:54 PM               

1) Roughly how often does an extruder experience downtime?

Generally, most companies figure 85% uptime and shoot for 90%. Some factors can be maintenance, frequency of screen changes (unless a continuous screen changer is used), etc.

2)How much of this time is due to troubleshooting during start-up of runs?

Not very much on an established product. If it's a product that has never been run, then a set of conditions is usually "guessed" and then fine tuned after it is running, over the next shift or week etc.

3) How often do companies typically restart a extruder processing run?

Some companies shut down every day. Most try for 24/7, with a weekly shutdown (for an hour?) for screen changes, etc. But "It depends".

4) Are there any single or twin screw extruder control features that industry would be interested in seeing implemented?

The control systems have been continually revamped for many years by industry professionals, and are faily mature products. Most of the quibbling amongst the control gurus comes down to control schemes (i.e. is it better to control the line by rate, thickness, or line speed). Also things like the HMI (or MMI or whatever) are touted, but really that's just artsy-fartsy stuff to make it easiest for the operators. The control schemes behind the interface is fairly well established.

5) How much would companies be willing to pay for this software?

Generally they want it for free, included with the control system of the line. A customer doesn't mind paying for the control hardware (boards, drives, cabinets) but will expect the software to be thrown in for free.

6) Do companies typically hire extrusion consultants of handle processing problems in-house?

Depends. Large companies generally have their own experts. Small companies usually have to fall back on expertise from the machinery or resin manufacturer when they get confused.

7) How much are companies willing to pay for consulting services?

Usually $5k - $7k per week is fairly common, and could be more depending on the nature of the consultant.

8)What are some current extrusion processing issues of interest in industry?

More rate, less waste, and lower operational costs. How to implement that into a control scheme? Well I dunno...

posted February 26, 2003 03:05 AM               

Larry and Bob seem to have your questions "nailed".

From a slightly more volume perspective than Larry's- lot to lot changes in resin, undocumented changes in additive packages, varying gel counts all add "interest" to the job of making something usefull.

Most raw materials suppliers have their own "in house" techos to assist long suffering processors. The rates quoted for "hired guns" seem fairly accurate.

Recycling and waste minimisation are number one and two on my list.

posted February 26, 2003 05:06 PM               

quote:

---

Originally posted by Steve H:
Hi Guys

Larry and Bob seem to have your questions "nailed".

From a slightly more volume perspective than Larry's- lot to lot changes in resin, undocumented changes in additive packages, varying gel counts all add "interest" to the job of making something usefull.

Most raw materials suppliers have their own "in house" techos to assist long suffering processors. The rates quoted for "hired guns" seem fairly accurate.

Recycling and waste minimisation are number one and two on my list.

Steve H

---

Well, I don't know about "nailed", but I think the perspective is pretty good.

posted February 26, 2003 10:15 PM               

Ultimately processors want to make money.

That means running a process at least as efficiently as the competition.

When that does not happen, then processors want to control their process up to the point where they make money.

“Control” is a big word here. It might involve materials, equipment, people and other influences.

In the world of extrusion how much does the electrical or logical control systems, that are currently available, influence the processor’s profitability? Not much.

What will make a difference are control systems integrated with aids like simulation, fuzzy logic, expert systems, statistical process control, and neural networking. Extrusion is a very non-linear process, and will require these non-linear control techniques if any progress is to be made. A few companies use these tools effectively, but I doubt they will be accepted widely.

On the bright side recent attempts to simulate twin screw reactive processes have been successful. This is truly amazing!

Tom Cunningham
PlastxAlliance

posted February 27, 2003 04:57 PM               

quote:

---

Hello,

1) Roughly how often does an extruder experience downtime?

2)How much of this time is due to troubleshooting during start-up of runs?

4) Are there any single or twin screw extruder control features that industry would be interested in seeing implemented?

---

Well, from an operator POV:

1) Depends on how often we need to change screens, weather or not the line has a problem, or when the techs can come out to fix the machine.

2) Quite a bit. When you got a blend of 7 resins, 3 additives and process-aid going into an old Glouster you get quite a bit of troubleshooting time.

4) I would like to see better control-layouts. I dislike flipping through pages of data to adjust setpoints, and then flipping back to adjust another. Also, the control package needs to be WISIWYG and very simple to use. When I go to change my heats I don't need three pages of PID tuning, trending and datalogging.

I would also like to see an "On Line" database of troubleshooting info. Not internet linked, but sort of an "expert" system built in that could help guide me through a tricky spot.