All opinions welcome

[Qanalyist]

OK, so a hypothetical (or not) question.

Say we have an injection molding machine we use for the same mold, that machine has the program for the settings run every time, no changes made ever, we have lots of machines but only this particular one is ever used on this mold. A set of trials are run and the parameters and setting agreed on to make the best possible quality product. After a while the parts are no longer in specification, but the program has not been touched and the material has not changed. Again it is the same machine with the same mould and settings.

What happened?

[hobda1ad]

In my experience if the process didn't change the machine or auxiliary equipment variables could have.

-Is the machine still holding a proper cushion?, is the Hydraulic pressure still correct?
-If you use a Thermolator is it working properly?
-Do you dry your material? and is the dryer working properly at the correct dewpoint?

Our tools seem to run different (Typically better) as they warm up. It could ultimately be your process needs to be tweaked as the tool warms up and during the tools lifespan.

A well tuned process can only be as capable as well tuned equipment.

[joeprocess]

How long before the parts fall out of spec.? Does this happen every time the job starts up? Is it a hydraulic press?

[Qanalyist]

How long before the parts fall out of spec.? Does this happen every time the job starts up? Is it a hydraulic press?
Good question, yes a hydraulic press, it seems that we will have months of good parts then a bad run. They are made in batches of 100,000 (small parts). The entire batch will be bad. These are small parts and you need a microscope to see the cold flow on some inside surfaces so usually a bad batch is made before the problem is noticed. The team that run this machine say that as the settings are the same as the good runs and they have changed nothing, it is not their fault?
I take their word that settings are not changed but I can't help but think that Injection molding is more than just the settings. As mentioned by hobda1ad there are many other variables. For instance the thermocouples were changed just before the run, was a calibration needed after sensor change? Regarding material drying, again they claim same procedures followed as usual, no changes made to process.
Are settings the be all and end all of injection molding?

[Qanalyist]

Since you all have taken the time to reply I will give you all I know about this problem.

The part is a small component that has some critical dimensions that mate to other parts, so long as these critical mating surfaces are good then the part is good. We have been making these parts for 15 years or so, maybe a batch per month 50 or 100 thousand parts per batch. About once a year a batch will be made that has cold flow and the mating surfaces will be out of spec. It seems that the bad batches are made in the winter months, that may be a coincidence, maybe not. When talking to the setters and techs they allways have the same response, 'the settings are the same'. This is not a usefull response, something has changed, obviously and they seem unwilling to investigate further. As I am not an expert I cannot argue with this, but as an engineer it does not sit well that the molding process seems to be in the lap of the gods as to whether it is good or bad.

My question is this, is it as simple as settings, that once the settings are in a machine then the product will be 100% good every time and if not then there must have been a supernatural intervention? Do I need a Voodoo doctor?

I am the QA that has to investigate the issue every time it comes up, it is me that gets it in the neck from the managers that use these parts because it is the same problem evey time and it looks like I am not doing my job. Repeat issues should not happen, yet this one does.

[JayDub]

Good question, yes a hydraulic press, it seems that we will have months of good parts then a bad run. They are made in batches of 100,000 (small parts). The entire batch will be bad. These are small parts and you need a microscope to see the cold flow on some inside surfaces so usually a bad batch is made before the problem is noticed. The team that run this machine say that as the settings are the same as the good runs and they have changed nothing, it is not their fault?


These runs must cover multiple shifts - but you don't do a first piece inspection? No in-process inspection? If you caught the process making bad parts you would have a much better chance of identifying the problem. And if you can get management approval to shut the press down as soon as you find bad parts, then you would force the techs to troubleshoot the problem.

In general, presses should be calibrated periodically (pressure and position sensors as well as t'couples) - but this isn't your problem: bad sensors drift, but they don't drift in and out, they just keep getting worse.
Do you have lot traceability on the resin? If you know that a lot of resin has produced both good and bad lots of parts, you can eliminate the resin, otherwise it is suspect. A wider than normal molecular weight distribution can make a resin a bear to process and will not show up in the MFI value.
First thing you can do is have a QA tech or intern copy down all the press and thermalator settings for every run, so next time you get a bad lot you can see if everything is consistent or not.

Good luck!

[joeprocess]

Injection Molding has one of the highest sets of variables of any manufacturing process. It starts from the inherent variability of the resin and can compound from there. May it be lot to lot variation in the material>

[rickbatey]

So prior to you stating that bad parts are made every winter, I could see the resin as the root cause. Minor changes in viscosity can have huge impacts on actual cavity fill rates and pressures. But since you've noticed that this occurs in the winter, I think you need to place a temp sensor in area adjacent to the machine and start recording the ambient temp. If the press has internal mold heats, a t/c tied into the press and then have the press collect process data every shot. Now look at the bad lots and see what the temp was around the press. Standard thermocouples read your temps as an amount above ambient temperature. So cool the area around the press rapidly and guess what happens to your melt temps. Barrel blankets help tremendously with this issue and I used this method to get the price of the barrel blankets approved. The only issue with the blankets is if you change resins/barrel temps then you'll need the newer style that allows for the top of the front section to be removed.
Rick.

[iautry1973]

I know everyone is talking about the settings being the same, but with scientific molding they tend to be more concerned about the outputs than the inputs. Do you keep track of the peak pressure, material cushion, inject time, part weight, etc.? Might be a good way to track down the variance in the process.

As rickbatey said temperature around the machine can make a huge difference. One of the shops I worked at in Montana if you opened the backdoor up in the middle of winter it would cause the machines closest to it to alarm out for bad parts. Also how is your cooling water chilled? Is it a tower or a chiller. The only tower I worked with caused varying processes depending on the season as it wasn't able to keep a stable temperature.

[Qanalyist]

Thanks guys, I now stand on a better footing with this information. I suspected there was more to it than 'the settings' and quite obviously there is. I will ask the questions you have asked me and see what reply I get, they will probably accuse me of reading a book on molding!!! This forum is better than a book, you can't ask a book a question.

We are in Ireland, so there can be quite a difference in summer/winter temperatures, I know the guys use external water heaters to heat molds and that keeping temps up can be an issue for them. So unless one of the settings is 'the weather' and 'the season' then they will ahve to deal with this some other way.

Cheers.

[3NGLENN1889]

Google "John Bozzelli" and his SIM articles (and associates). The man is a wizard. When you get a "good" run all your "outputs" need recording. Read everything the man has written.

[iautry1973]

Some other places for scientific molding articles and webinars: fimmtech.com, ides.com, and Andy Routus at www.traininteractive.com.

[brentb]

I know everyone is talking about the settings being the same, but with scientific molding they tend to be more concerned about the outputs than the inputs. Do you keep track of the peak pressure, material cushion, inject time, part weight, etc.? Might be a good way to track down the variance in the process.

As rickbatey said temperature around the machine can make a huge difference. One of the shops I worked at in Montana if you opened the backdoor up in the middle of winter it would cause the machines closest to it to alarm out for bad parts. Also how is your cooling water chilled? Is it a tower or a chiller. The only tower I worked with caused varying processes depending on the season as it wasn't able to keep a stable temperature.

Where in Montana were you located?
KOM
brent

[iautry1973]

Brent, Bozeman, MT. with Blackhawk/ATK. I am in Vancouver, WA now with US Digital.