Do you REALLY know your melt temp?

I understand the 30-30-30 rule and it all makes sense but what are the thoughts on checking melt temp as soon as you build the shot compared to building the shot and then let the material set approximately the same time it would be in cycle? For instance, I was taking melt temp on a process with a 61 second cool time (I know this needs to be fixed!!!) and saw when the shot was injected quickly quickly after building it there was about 10-15 F hotter melt than the check done when the material set in the barrel for the approx. cycle time. The process is running 125 psi BP (not my process) with a unfilled POM, so I'm thinking this is high shear heating with the temperature being normalized by the cycle time in barrel??

Hi Guys
I have done quite a similar collection method and found to be accurate enough to make decisions for the last 4 years. I used the Chinese tea cup (porcelain?) and made a 1 mm hole at the back for a fast response thermocouple. Use a nylon bracket to hold the cup and fit a handle to the bracket. Purge directly to the cup, ensure the thermocouple is 20 ~ 25 mm protruding from the bottom of the cup, keep the purge to fully cover the thermocouple. Observations are up to 4 collections may need the reading to stabilize. The last 3 to 4 readings from 8 reading collected showed 3 ~ 4 degree Celsius range. 30 seconds readings(from the 30/30 method) showed a lower value by 4 degree C than the maximum readout. (So I use maximum readout, not the 30 second readout). The maximum readout was at around 26 seconds and then the temperature started to fall. Shot size and speed do have 4 to 6 degree C influence on LCP material.
But at the end of the day, you have to get the melt temperature at the gate, not at the machine nozzle for a good molding. So its only half the way...

Richi

Great article. Hope moulders out there are reading - not only, but especially for semi-crystalline resins! (From experience largest proportion of part failures for POM, PA and PBT traceable to inhomogeneous melt).

I found an adequately sized Teflon(R) cup machined from rod stock a useful reservoir to use. The cup needs a handle fitted of course. The stabilised melt is then injected into the cup and measured with a quick response temperature probe. The PTFE insulates melt to prevent cooling.

Gentlemen,
A Patented "TeMP" sensor #7,585,166 Sensor reads the Thermal elastic Melt Pressure.

The sensor TeMP profile indicates each mold cavity Fill to Final Peak Pack Point time for:
1. Machine shot size VOLUME
2. Mold cavity volume for each cavity, with a sensor

The sensor can be placed on the mold part line, in the vent groove and define fill to peak pack.

System is proven and ready for the market.

A power point presentation can be obtained by e-mail to fbuja@rochester.rr.com

This sensor define melt density "K" of the injected melt volume into a mold cavity that changes with melt FORCE, TEMPERATURE and PRESSURE

I used a sophisticated thermal imaging camera on molded products, the results were good - it showed the temperature distribution on the entire product. As to its sensitivity to emissivity, angle, color, shininess and air contaminants, I heard, these cameras are based on IR sensors like IR guns but have much more of them to be able to build an image. As to its usage; if its high price is affordable v. the total apx. USD250 of fast response dig. thermometer and apx. USD30 thermocouple, it can be considered.

George M./APD

IR Guns

All this equipment is COLOR SENSITIVE. If you calibrate it to a color you're good. BUT if you have the same material in two different colors an IR gun will read two different temperatures while a needle probe will not. While it's not a bad thing, it is something to keep in mind.

How about thermal imaging cameras? Are they more accurate than the IR gun, or is it the same problem?

Hi Richi,

I have forwarded your e-mail to Kistler(R) - they should contact you soon. As to RJG/e Dart I have no hands-on experience.
George M./APD

Thank you for the EDL probe info - I'll try it and compare it to my 0.062 inch Dia. probe performance.

G. Markus/APD

Hello Markus
Good to hear your experience on 32 cavity mold with Kistler. my email is richi@lng-res.com The e Dart RJG systems are more popular here, compared to Kistler, what is your experience with RJG? I would like to have a conversation with the Kistler's technical side for better understanding of their product & service. Thanks :-Richi

Melt Probes

I have found a .040" reinforced needle probe from EDL Labs to give a very fast response. I would highly recommend for accurate melt temps.

Hi Richi,

You are planning to take a giant leap toward sensor-based robust process development and near 100% quality monitoring. I worked with electronic connectors for decades preventing/solving numerous problems, e.g. warpage, dimensional issues, etc., the former with special part design modifications the latter with precision packing. As to Kistler(R) I have used their products with great success, e.g. a molder could not produce a single acceptable part from a 32-cavity mold for nine months, Kistler data collection/analysis and our transient thermal analysis of the hot valve gate and gated section of the part pinpointed the problems; the mold was in production in six days. I am not a rep. for Kistler(R), please send me your e-mail and I'll get you in contact with their tech. manager for proper sensors, charge amplifier/signal conditioner and their price.

G. Markus, APD

Hi Markus
Thanks, It makes more sense to go till the cavity, if you really want to parts right, not just at the nozzle level. I agree with you. I did check out the Kistler's site. Have you used it? Both temperature sensing & Pressure? I see they have Multi sensor set (P_T). Does Multi sensor perform as good as having individual pressure & temperature sensors? We are in electronic connector business, and parts are pretty small, but cavitation has to grow as the product matures. I have tried melt flippers, but does not give enough satisfaction. Some products can not be accurately weighed to recognize the difference in weight.
If I were to use 8 probes just before the gate (on the runner), for both temperature & pressure, together with rest of the amplifying, sensing and monitoring (all hardware & software)how much would it cost(a ball park figure) me? (It may not be economical to install the equipment for each mold here, but my thinking is to have mold debugging / qualification using this, and later dismantle and use for next mold debugging. What do you think?)

Richi MR

Hi Markus
Thanks, It makes more sense to go till the cavity, if you really want to parts right, not just at the nozzle level. I agree with you. I did check out the Kistler's site. Have you used it? Both temperature sensing & Pressure? I see they have Multi sensor set (P_T). Does Multi sensor perform as good as having individual pressure & temperature sensors? We are in electronic connector business, and parts are pretty small, but cavitation has to grow as the product matures. I have tried melt flippers, but does not give enough satisfaction. Some products can not be accurately weighed to recognize the difference in weight.
If I were to use 8 probes just before the gate (on the runner), for both temperature & pressure, together with rest of the amplifying, sensing and monitoring (all hardware & software)how much would it cost(a ball park figure) me? (It may not be economical to install the equipment for each mold here, but my thinking is to have mold debugging / qualification using this, and later dismantle and use for next mold debugging. What do you think?)

Richi MR

Hi Richi,

In case of hot runner system it is relatively easy by setting the same temp across the board, provided the system is equipped with high-end, properly calibrated sensors/controls. In case of cold runner system one way is a quick, expert Filling analysis. At a given flow rate the melt temp. neither increases nor decreases in the cold runner/gate; HOWEVER, this injection flow rate may not be optimal one for the part so it is possible that we have to maintain higher or lower melt temp at the nozzle to keep the cavity happy. Another possibility is, as mentioned earlier, installing an e.g. Kistler(R) temp. sensor in the cavity, close to, but not at the gate to monitor melt temp. shot-to-shot, 24/7/365, free after the initial investment.

G. Markus, APD/Advanced Plastics Design, Inc.

Thanks George
A lot of the time we see topics on scientific molding or systematic molding but on melt not so many, good to see more thought on this topic. Is there any simple logic thumb rule for making the melt at the gate is same temperature as the melt out of nozzle?
Richi MR

Hi Richi,

Herein is my nickel's worth:
1. The purging and the injection shots (sizes) are normally different, will it not influence the melt temperature?
Not markedly - I do not expect more than a few F. If that makes the difference in part quality the project is in trouble. That polymer temperature is important, which enters into the cavity. In cold runner systems we encounter heat losses or shear heating depending upon the injection flow rate. We computed as high as 25F increase and 50F decrease of temp. in cold runners. Interestingly, various gates heated up the polymer only by 6 – 10 F due to the short time the polymer travels through it. In case of hot runner systems the manifold and hot drop setting most of the time has significant influence on the melt temp entering the cavity, especially if the process tech sets up the temps all over the map. I immediately correct the latter scenario to: last barrel zone=nozzle=hot manifold sprue=hot manifold=hot drops. I change the latter setting in case of improper gate functioning (drooling or freeze off). It is also important to calculate where the material of the next shot sits; often for small parts the entire injected volume is in the hot drop - in this case their temp. control is of utmost importance.
2. The speed of the injection and the speed of purging normally not the same, how about its influence?
A few F – shear heating in the faster flow scenario can be computed upon request. The project should be robust enough to tolerate this with ease. Again, that polymer temperature is important, which enters into the cavity and it can be significantly different from the purge temperature even if the latter measured rapidly and accurately.
3. It takes may be 45 to 60 seconds to do (quickly do) the checking of melt temperature, but my molding cycle is less than 15 seconds, will this give me different input than the real?
It will up to a few F if the thermocouple dia. is small and responds fast. Otherwise significant, e.g. 20-60 F deviations are expected between measured (improperly) and actual melt temps. The difference between cycle time and measurement time should not make notable difference - it is the accuracy of measurement that counts.
George M. - APD

Hi Guys
Thank you all for the different opinions. Few questions more on the same subject.
1. The purging and the injection shots (sizes) are normally different, will it not influence the melt temperature?
2. The speed of the injection and the speed of purging normally not the same, how about its influence?
3. It takes may be 45 to 60 seconds to do (quickly do) the checking of melt temperature, but my molding cycle is less than 15 seconds, will this give me different input than the real?

Appreciate someone can put more thought on this.
Richi MR
Golden City Plastic. Malaysia.

Because a thermocouple probe has both low heat capacity and high thermal conductivity and the opposite is true for the melt sample, it is actually better to preheat the probe to a temperature higher than that expected for the resin melt temperature. It needs to be high enough to not drop below the melt temperature during rapid movement from a preheater to the melt sample. Once placed in the center of the melt sample the probe temperature will drop relatively quickly, then plateau at the melt temperature of the sample. As the cooling of the melt reaches the center of the sample the probe resumes a faster rate of temperature decline. This is easiest to demonstrate to machine operators graphically by recording the temperature reading versus time, but ultimately the plateau is easy enough for operators to record routinely from a non-recording display. The 30 second wait time suggested by Bill is OK for targeting precision across operators, but the above method will improve both precision and accuracy.

I considered the IR gun technique a good few years ago; however, my research showed the measurements with these are sensitive to the angle of measurement and the emissivity of surfaces, e.g. color and surface finish. So I stayed with the accurate, fast-response thermocouple/digital thermometer-based contact measurements of both mold surface and melt temps. It was interesting to find that our computed mold surface temperatures were within a few Fahrenheit from actual measurements both at the beginning of injection and right after part ejection.

G. Markus, Advanced Plastics Design, Inc.

Bill:

Your comments are well made and yes the differences are quite amazing. My experience has shown as you mentioned that the differences can be as much as 25F or more depending in the technic. You have to run your machine for at least 30 minutes to just start to stabilize. If you have any doubts regarding melt temperature it is sometime valuable to check it again after you have been into the process for a good hour or more if you have heat sensitive materails. Prehaeting the probe with 100 F of the expected melt temperature reduce lag time. Good Article.

Bob Alvarez

I'm surprised there was no mention of melt temps taken with IR guns.
I would be very interesting in seeing any information regarding the accuracy of that method. With the low cost an ease of use I'm sure a lot of people are using that method.

I would like to add a few thoughts to this issue. We have been measuring melt temperature accurately for the past twenty-five years. The key to it is the preheating of the thermocouple (as you well said) and employing not a heavy, e.g. 0.12 inch dia. needle having significant thermal inertia, but rather a much faster responding one from Omega Corp., possessing only 0.062 dia. We gave the part number for the latter to numerous molders. Kistler offers in-line temperature sensors for the nozzle region for continuous monitoring of the melt temperature. Better yet, the temperature of the polymer can be (should be) monitored in the cavity itself. Combined temperature/pressure sensors give molders near 100% QC running free after the initial investment. G. Markus, Advanced Plastics Design, Inc.

Nice report again Bill, as always good to see on how you put this in perspective.