Cooling a oven belt

Hello, I got a customer that has a shrink wrap machine to bundle up beer cans to make beer packs. They have a problem with the oven's all metal belt conveyor. When the belt gets too hot, the shrink wrap stinks on the belt and if the belt is too cold, the wrap doesn't shrink under the pack.


The machine has a big fan to cool the belt but it's manually operated with a potentiometer. If the operator notices that the wrap sticks, he turns the fan to a higher speed and if the wrap doesn't shrink under the pack, he turns the fan to a lower speed.



They would like to regulate/automate the belt temperature. A VFD will be installed.



I was thinking a simple PID temperature loop with the fan. A temperature probe would be installed after the fan. We would do some tests to figure out the optimal belt temperature in the oven versus the belt temperature after the fan.


Anwybody has a better idea how to achieve this?


Thanks!

You really need to test the temperature of the actual belt, this could be done with an infrared sensor, using a probe just above the belt is prone to problems, even if you allow for offset any slight air movement will probably change the reading and as direct contact is difficult with most sensors I would certainly look if a non contact method i.e. infrared with an analog output (if one exists) or the possibility of coms.

Many years ago, I used one of these (before they were bought by Fluke):

https://www.flukeprocessinstruments...perature-solutions/spot-pyrometers/compact-ci

There was also a little box of electronics that went with it which supplied a 4 to 20mA current output that was scalable to the measured temperature.

I would spray it with water instead cooling it with fan. Also you need to select good probe with fast response. Adding a photoelectric switch is also good choice, because you don't want to measure temperature on gaps. Infrared temperature sensors don't like reflective surfaces.

IR camera

What I don't understand is that if the plastic should shrink and not stick inside the oven, why the temperature of the belt is regulated outside the oven?

Maybe the temperature of the lower part of the oven should be regulated separately from the upper part.

I would use Infrared thermocouples on the belt. I have used them to test panels with shiny Mylar laminated on them and there were accurate and reliable.

I would put one right after the fan area, if you want to tune the PID for it maybe add one right before the fan also in case the conveyor is warmer or cooler than normal going in .

lfe said:

What I don't understand is that if the plastic should shrink and not stick inside the oven, why the temperature of the belt is regulated outside the oven?

Maybe the temperature of the lower part of the oven should be regulated separately from the upper part.

Click to expand...

The belt is not currently temperature regulated, it passed through the oven heats up overtime.

I will check for infrared probes

matwon said:

Hello, I got a customer...

They would like to regulate/automate the belt temperature. A VFD will be installed.

Click to expand...

I realize the customer wants to automate temperature control, but the description sounds more like a measurement problem in that product quality is the feedback on whether the belt temperature is in an acceptable range.

Depending on circumstances and your relationship with the customer, you may consider a two-phased approach: (1) install belt temperature measurement with an operator feedback mechanism, and (2) add automatic control if human-in-the-loop is still not good enough.

As others pointed out, non-contact temperature sensing is probably the best option for measurement.

Operator feedback could be a visual indicator such as a three-color stack light (cold/good/hot) and/or a digital temperature readout. The operator would continue to use the existing potentiometer, but with advance information on temperature going out of range, they can make the manual adjustment before wrap quality deteriorates.

If this is good enough, you will avoid a lot of baggage associated with an automatic control that might be overkill for this situation.

But if the customer has already made up their mind, so be it. They are always right, you know.

Mispeld said:

I realize the customer wants to automate temperature control, but the description sounds more like a measurement problem in that product quality is the feedback on whether the belt temperature is in an acceptable range.

Depending on circumstances and your relationship with the customer, you may consider a two-phased approach: (1) install belt temperature measurement with an operator feedback mechanism, and (2) add automatic control if human-in-the-loop is still not good enough.

As others pointed out, non-contact temperature sensing is probably the best option for measurement.

Operator feedback could be a visual indicator such as a three-color stack light (cold/good/hot) and/or a digital temperature readout. The operator would continue to use the existing potentiometer, but with advance information on temperature going out of range, they can make the manual adjustment before wrap quality deteriorates.

If this is good enough, you will avoid a lot of baggage associated with an automatic control that might be overkill for this situation.

But if the customer has already made up their mind, so be it. They are always right, you know.

Click to expand...

I think this is the best approach. At the moment you have no data to work with, and any attempt to control the process will just be speculative.

Install the measurement system first, along with visual feedback to the operator (like a temperature readout), then let them run it the same way they always do for a month or two. Any operator more advanced than a trained monkey is good at at least one thing: working out ways to make their job easier. I don't imagine that peeling stuck plastic off a hot belt, or re-wrapping pallets manually are fun tasks for them, so if you give them any sort of way to see those issues coming and circumvent them, I guarantee they'll get you the data you want. After a couple of months, go back and chat to them, and you'll likely get some really valuable information like "yeah, the plastic seems to start sticking as soon as the temperature goes above x°C so I start to wind the speed up as soon as I see it get to about y°. And then when it comes back down to z°C I back it off again before it stops shrinking the plastic." Basically, you're using the world's most powerful organic self-learning computer (the human brain) to gather the data you need, and then you've got a halfway decent shot of controlling the process, now that you actually know how you need to control it.

Or, as Mispeld says, you might find that the customer is happy with the manual control, now that they have the ability to foresee and prevent the issues before they occur.

From a sensor point of view, I've used the Raytek MI3 for non-contact temperature measurement with good results. My application was quite different, in that it was conducting product samples on surface temperatures as low as -40°C (or °F, that's the magic point where it doesn't matter which you use), and the sensing distance was only 20-30mm. But I'm sure there are other ranges. It's a good reliable sensor, had no problems with it after 7 or 8 years, and it's a nice easy 24VDC 4-20mA instrument.