Heat Loss from a pipe and not sure how to tackle this problem

[Rvanpelt]

I have tried a couple different ideas of how to approach this problem and would like some advice as to a method to use.

The Problem:

The pipe in question is a liquid feed tube for a spray nozzle. The feed tube has an ID of 0.067 inches, an OD of 0.125 inches, and is made of Titanium. The length of the feed tube is 2.582 inches.

The liquid being transported is a mixture of 66% water and 34% polyvinyl alcohol by weight. When the liquid is being sprayed from the nozzle and is being transported through the feed tube, it is at 140 F temperature which is maintained because the nozzle is being fed from a heated main header pipe.

Preceeding the nozzle liquid feed tube is a 16 inch length of 3/16" ID PTFE tubing with a tube heater and insulation around it...so the 140 F temperature is maintained. Between the PTFE tubing and the nozzle is a 3-Way solenoid valve that diverts flow back to the header or to the nozzle. The exit end of the feed tube is to the ambient air, about 80 F.

My problem is not really with when the liquid is flowing through the nozzle but after the flow to the nozzle is stopped and diverted back to the header. At that point, there is some residual liquid left in the nozzle. The entire feed tube holds about 0.15 ml of liquid when liquid is flowing and I would approximate that about 0.1 ml is retained after the flow is stopped. As this retained liquid sits in the nozzle feed tube, it cools and the polyvinyl alcohol bgins to solidify which presents problems when restarting the flow to the nozzle.

I am looking at installing a miniature heating element into the inside of the feed tube of the nozzle to maintain the 140 F temperature while the system is idle. it would be installed through the swagelok fitting that connects the 3-way valve and the feed tube and it would protrude a short distance down the feed tube toward the nozzle exit.

I am looking to determine what wattage must be provided by this heating element to maintain 140 F in 0.1 ml of a 66/34 mixture of water/polyvinyl alcohol. How would one approach this problem?

 

[IRstuff]

Seems to me that the accuracy of the temperature might be a bigger problem than the actual wattage. How are you going to maintain the temperature setting?

What is the airflow around the feed tube? What's the ambient temperature? Can you insulate it to minimize the heat loss? The heat loss for a fairly high air flow is still only about 0.5W, just as a reference point.

You haven't indicated what happens if the temperature gets too low or gets too high. To minimize the temperature excursions, you'd need some sort of PID temperature controller.

TTFN

FAQ731-376

 

[Rvanpelt]

I haven't determined an appropriate method of temperature control yet. At this point it seems like it will be a challenge to just design a method for installing the heating element. A thermocouple could work but I would then also need a way to insert that through the fitting at the entrance to the feed tube.

The airflow around the majority of the feed tube is non-existant. The feed tube is enclosed in a 2 piece shroud made of stainless steel that is screwed together like a bottle cap with viton elastomer o-rings providing a fairly strong seal. Without a heated liquid, the temperature inside the shroud is not much different than that outside the shroud (about 80 F in this case) but obviously that would not be the case here with the heated liquid. The shroud is barrel shaped, 1-1/8" ID and 1.437" OD and length of 1.271". 0.831" of the feed tube protrudes out the front of the shroud, ending in the nozzle tip, and 0.509" protrudes from the rear of the shroud which is where the swagelok fitting is installed. These two end portions would be exposed to the ambient outside of the shroud but the airflow is negligible here anyway.

I cannot insulate the feed tube inside of the shroud due to the electrical components installed around it. The nozzle uses ultrasonic standing waves to atomize and these are produced by electrodes and transducer crystals around the feed tube. On the front end, any kind of insulation would dampen the standing waves, affecting performance. The rear end could be insulated but the fitting connection gets in the way.

The main concern is to keep the temperature from getting too low as it causes the polyvonyl alcohol to solidify which stalls the nozzle upon start up. Too high of a temperature could start to cause damage to the electrical components I mentioned. A PID loop would be necessary based on a signal from a thermocouple and that is something I plan on doing, assuming I can find a way to rout both that and the heating element into the feed tube through the fitting or some other point without compromising the liquid seal.

For reference, below is a sketch of the device. The dashed lines running lengthwise are the ID of the feed tube.

 

[IRstuff]

From what you're describing, the issue seems to be primarily at the nozzle end? Which would seem to imply that applying heat at the nether end might of minimal utility?

Have you thought about a radiant heater ring at the nozzle end? If you keep the separation down to something like 0.1 inch, you'd get some amount of thermal conduction directly through the air in the gap and you could keep the temperature well above 120ºF at the nozzle end.

The back end might be wrapped with a narrow strip heater on the outside of the tube. If you hold that end at 140ºF, then there might be sufficient thermal conductivity in the tube itseld to keep the portion of the tube inside the shroud at a reasonable temperature.

TTFN

FAQ731-376

 

[sreid]

Is it possible that you could arrange to purge the nozzle with air through another valve?

 

[Rvanpelt]

Using an equivalent thermal circuit, I was able to calculate the thermal resitances across the different boundaries, i.e. the wall of the titanium feed tube, the convection and radiation from the OD of the feed tube, wall of the stainless steel shroud, and convection and radiation from the OD of the shroud. I have never gotten data on the temperature inside the shroud so I guessed at that being roughly 120F and came up with an overall loss of 0.13 Watts approximately which seems reasonable. I think a 0.25 W heating element would be more than sufficient and using it with a thermocouple would prevent over temperature. The trick will be routing the wires for the heating element and thermocouple into the feed tube and placement of each.

 

[Flareman]

I agree with sreid.
Clearing the material from the pipe and nozzle would seem to be a better long term solution. A short burst of pressurized flow (air/N2 or CO2) input immediately after the three way valve when it switches over might not be too difficult to arrange. Perhaps a second 3 way valve, close-coupled?

David

 

[Rvanpelt]

I like that idea and it is definitely a more simple idea. I will pursue that line also. Down the road, there may still be a need to heat the nozzle for other applications so I will look into the feasability of doing it but for this application, an air purge seems to be a good bet.