posted November 20, 2002 04:44 PM               

Are the following possibilities? How likely? How much of a problem?
1) Bridged extruder hopper (happened to us already - definitely top of the list bad)
2) Unstable screw speed
3) Air entrapment in melt that makes it through the extruder?
4) Degraded material or gels?
5) ?

posted November 20, 2002 10:20 PM               

You have just proved that a single screw extruder is very sensitive to changes. Bulk density and viscosity changes are very likely to result in unstable extrusion. It is very hard to maintian constant viscosity when reprocessing polyesters. I would bet your economics would be better if you sold the densified scrap, and calculated the savings in good product yield.

posted November 20, 2002 10:31 PM               

Welcome to the extrusion forum- looks like you've got most of the potential problems nailed. Inconsistant drying could be added to the list.

I like the Erema process for recycling PET- so if reusing scrap is a must do for your company, I'd take a look at their equipment.

posted November 21, 2002 09:58 AM               

I wonder if your problem is caused by the pellets being obliterated into powder in the feed section of the screw.

If you could feed solid pellets (i.e. remelted material) instead of densified pellets, so they don't fall apart, I suspect you'll get a more consistent output from your process.

When the pellets break into powder, the powder acts like a lubricant between the virgin pellets and you'll get inconsistent feeding. The pellets can "tumble" instead of get sheared against the barrel wall.

If you don't have the option to remelt the material, is there a way you can make the densified material less likely to crumble in the feed section of the extruder (i.e. maybe run your pelletizer hotter)?

posted November 21, 2002 11:38 PM               

The machines we have are called 'pellet mills' around here, but I'm unsure of thier exact name or manufacturer. The shredded film is forced through a die to make the pellet; the die is shaped like a large, wide wheel with holes. The friction and pressure of the material getting pushed through the die is what creates the heat to form the pellet. The die is not actually heated externally.

In the samples we have taken from the extruder hopper the pellet is already broken back down into shreds and granules. I know we need to improve the pellet condition. The trick is, I need ammunition to convince others the pellet quality needs improvement....

What I'm really hoping to gain from this discussion is theories on precise mechanisms on how this low density, granulated pellet mixed with high density chip could create something like a gel, degraded material, or force an air bubble through the extruder. Having a convincing, provable theory on why low quality pellet causes web breaks will help me.

I should explain our extrusion process a bit more: we run two single screw extruders in tandem, and meter output through a positive displacement pump. The melted polyester is filtered, then extruded through a flat die and biaxially oriented. The process is normally very stable and can run for days, sometimes well over a week without a web break.

When a batch with broken down pellet material is being extruded we experience a higher than normal rate of web breaks. Let's also assume that high quality raw material does not cause web breaks, keeping the rest of the process including polymer line pressures as a constant. I know this through many samples of raw material from the extruder hopper during good and bad process stability. There is no evidence that a macro-bubble (that you could see with the unaided eye) gets through the die. I suspect there is something, some small nodule creating a weak spot or stress concentration that causes the web to break in the transverse stretching oven.

Neither extruder is vented, and I think I need a lecture to explain how air is prevented from getting forced through the extruders. An air bubble is the only thing I can think that can get through the filters. The filters do a very good job of catching gels and degraded material; but, if the production of gels/degraded material is very high then of course some gets through the filter.

I need something more concrete to explore than just my vague idea 'there's some nodule/air/gel thing getting through'.

posted November 21, 2002 11:50 PM               

Thanks for your explanation of how the powder can act as a lubricant in the feed section of the screw.

Is it possible (how possible??) that the powder sticks to the screw or barrel walls and degrades?

Or that the inconsistent feed creates pressure pulses that promote bubble formation in the polymer line?

How can I detect an inconsistent feed condition in the first of two tandem extruders?

posted November 22, 2002 09:17 AM               

A potential source of your problem could come from resin that sticks to the screw and then is disturbed when the process becomes unstable. That resin which sticks to the screw for a long time can transform into a number of different organic compounds which can get through the screens, but do not stretch.

Again, your company should give strong consideration to the fact that you are working in a sensitive resin system, in a sensitive process. I'm sure the margins are good when using prime material. Take it from someone that has been reclaiming materials for decades, your rewards are low or non-existent, while the risks are high.

posted November 22, 2002 09:36 AM               

I do not like screen changers for gel elimination. I have seen gels come through an extrusion line that are much bigger than the mesh of the screen - the process engineer that was running the line explained that gels can get through screens. Apparently the come up to the screens, kinda ooze through it, and join into a single gel again on the other side. I'm not enough of an expert to confirm that. Tighter screens will increase the backpressure which will help decrease gels, but the screens themselves shouldn't be use to eliminate gels.

Another possibility to consider is that gels are being formed at some hangup point in the downstream piping- an elbow, a piece of the die that isn't aligned, etc. - eventually these break off and a gel will come out of the die. The more degraded the material entering your process, the higher the propensity for the material to degrade inside.

Anyhow, on to the crux of your question - if you are feeding virgin pellets only, these get melted primarily by getting smushed (sheared) against the barrel wall. If you put a lubricating powder between these pellets, instead of getting sheared they will just slide around. Therefore the powder interferes with the melting action and you will have a harder time melting the whole pellet.

It would be worthwhile to see if your gels are unmelted particles, or if they are degraded material. Your resin supplier may be willing to offer a test if you don't have the equipment to do it.

Personally, I have never seen air get through an extruder, although it certainly is possible. Most of the time the pressure developed in the downstream portion of the extruder is enough to push the air backwards.

When the web breaks, you should collect the material right at the break and examine it. Is there a gel? Or did it break because of a weak spot in the film (in which case I'd look at the IV of the recycled material)?

I have to admit that I would be very surprised for an unmelted particle to make it through two successive extruders. I would think that anything not melted in the first extruder would get melted in the second. Unless they are very short (12:1) extruders. Why do you run two tandem extruders?

Also, do your web breaks occur at the die or in the orientation process?

The easiest way to detect extruder surging is with a pressure gauge for the extruder. The melt pump will keep the product coming out at a constant pressure, but what is going in to the pump?

I'm no expert, but hopefully something I say will help ask the right questions to get to the right conclusion...

-Bob Cunningham

posted November 22, 2002 11:10 AM               

Some types of reclaim present many problems (i.e. the relationship between fluff and extruder surging is well documented) but nowadays margins on the end product are way too tight to overlook every possible cost savings.

There are several ways of doing reclaim, some better than others, but most processors don't really have any choice nowadays. They have to do it.

I have not heard of a company in quite some time that sends all of it's scrap to a landfill - just about everyone reclaims somehow, even if it's sending their material out to a local recycling house to be repelletized.

Reclaim is one of those necessary evils that has to be dealt with in today's economic climate.

I don't know about Doug's process, but it's not uncommon for 15 - 20% of a cast film web to be unusable because of neck-in, and you can't overlook the cost of increaseing your raw material usage accordingly (especially with PET!).

posted November 22, 2002 12:01 PM               

The IV of the recycled material is slightly lower than new resin, we have tested it, and IV testing is part of daily quality checks.
Since this is a continuous process I'm sure that some amount of the polyster has been processed several times over, but I would think that if severe IV loss occurred after being processed there would be a cumulative effect, which is not the case.

The mixed new/recycle material is a part of the reason for tandem extruders; any pressure fluctuations from the first extruder (we do get some pressure instability for whatever reasons - periods of raw material change due to low density pellet, or over/under dried material) is dampened in the second extruder. The second extruder is also responsible for assuring homogenous temperature and thorough mixing. I'm sure the material is completely melted by the end of the first extruder because there is a filter before the second extruder. We have pressure sensors after both extruders, I will re-investigate pressure changes vs. batch and material quality (this could take a few days).

The web breaks are actually _after_ the orientation process, in the hottest part of the oven where heat shrinkage properties of the film are established.

I will make a concerted effort to collect and examine film samples from these types of web breaks.

We do find gels in the finished film that are much larger than the filter screens; since air bubbles are not a likely candidate I will concentrate my search for gels. It is possible too that dead spots in the melt flow through the pipes are accumulating degraded material that makes its way out through the die.

Just a note on the selling vs. recycling scrap: there is no way we would be profitable without using recycle in our process, we even buy certain forms of scrap polyester to use in our process. This is very good for our unit consumption numbers. However, we do sell some forms of our scrap that we don't have the equipment to recycle in house.

Also, the volume is such that we have to use pellet as opposed to a re-extruded cut chip type recycle (like the Erema machines?). We have a re-extruder machine but the output is lower than demand.

posted November 22, 2002 01:52 PM               

Filtering out gels is basically imposible, detective work and eliminating their cause is the only cure. Air entrapment is posible, but a very rare beast.

On Apet sheet, I've noticed increasing brittleness after two or three passes through the cycle of drying, extrusion, forming and granulation. How much of an IV loss have your tests picked up?

posted November 23, 2002 09:54 AM               

I'd suggest you do a DSC-OIT test to see what your material is stabilized for.

posted November 23, 2002 10:42 AM               

Suggestion: More consistent mixing of your virgin/reclaim batches. This'll take some doing but it worked at an injection plant and also at a BF extrusion plant I worked at.

Get a brand new cement-mixer, the kind you use for small contracter/home-improvement jobs. Also get several LARGE clean barrels/gaylord boxes

Determine which precentage of reclaim/virgin works best, write it down. Then mix that in the cement mixer, tumble for 5-10min then dump into barrles/gaylord boxes. Directly feed this into the machine either by sucker-hose or by bucket bragade. (Temps work great for that! )

This worked real well for ISOPLAST 302EZ at an injection plant and it worked for our UHMWPE lines at Tyco.

Oh, and also check your heat history/regrind history. Ya can only reuse that stuff a few times before it turns into snail snot... And Marshal Williams told me it isn't possible to biax snail-snot.

posted January 05, 2005 01:20 AM               

Overall I believe you are doing everything correctly, in-as-much that is within your power.

The dual inline extruders would work well as a buffer and add to stability and consistency
of the process. So . . .why the occasional problems?