Use of nanoparticles to enhance heat transfer

[jwseifed]

Hey guys. I have a counterflow heat exchanger which passes liquid nitrogen through tubes and hot exhaust gas over them externally. At the same time, we are injecting nanoparticles into our inlet pipe so that it mixes with the exhaust gas prior to entering the heat exchanger. The purpose of this is to enhance heat transfer, which has been shown to occur on some lower scale experiments in the past (albeit for cross-flow configurations). However, it is having the opposite affect in our case. The exiting temperature of the LN2 is decreasing and the exiting air temperature does not change compared to the baseline test.

There is most likely some frost enveloping the LN2 pipes, which is reducing heat transfer, but it still doesn't explain why the nanoparticles would be REMOVING heat from the coolant. I've attached a screenshot of the plot to help illustrate better.

Green - inlet temp of hot air
Black - outlet temp of hot air
Blue - outlet temp of cold LN2
Red - inlet temp of cold LN2

Nanoparticles were injected during the period between the vertical black dots shown. Steady state starts at around 300. Note the rapid drop in LN2 outlet temp (blue) and rising air outlet temp (black), these should theoretically be trending in the opposite directions. I also don't know why the temperatures continued to change even after the injection was completed. If anyone has some insight on this I would most definitely appreciate it. Thank you

 

[IRstuff]

Seems to me that your premise is not necessarily valid. The nanoparticles would only be effective IF, AND ONLY IF, they were able to get better heat transfer to the nitrogen and be more efficient in doing so. So a bunch of particles are floating around in the air stream; the majority of them will never contact the HX walls, so they'll never transfer their heat to the nitrogen, AND they simply transfer their latent heat back to the air at some later time. The only thing that would make sense is if the particles could somehow make more impacts against the HX, otherwise, no benefits will accrue.

The nitrogen temperature is dropping because you've actually made it harder to remove heat from the air, so less heat is getting to the nitrogen, not to mention the fact that nanoparticles are also adding Joule mass to the heat transfer, i.e., Q_air+Q_particle all have to be removed, since you appear to be wanting to get the air temperature below 200K. If your nanoparticles were pre-cooled to, say, 150K, that might make them useful.

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529 Entire Forum list

http://www.eng-tips.com/forumlist.cfm

 

[Compositepro]

I concur with IR. Your hypothesis seems blatantly wrong. I would expect exactly what you observe.

 

[LittleInch]

I've always tended to find anything with the word "nano" in to be the modern equivalent of snake oil.

If you start to ask questions about what exactly these mysterious "nano particles" are, you normally get zero technical information you can actually do anything with other JUST BELIEVE ME IT WORKS...

I don't know what timescale this represents, but it doesn't look long enough to show any trends to me.

Given you've told us not a lot about the size, flow, type of nano particle or why you think you're getting frosting when the air temp is 200C I'm not sure how we're supposed to guess at what is happening here.

My guess for what it's worth is that your nano particles are fouling your HX.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[jwseifed]

Thanks for replying IRstuff.

Our hypothesis was based on previous experiments which showed an increase in heat transfer, based on increased delta_T, of air passing over a hot wire after nanoparticles were injected. However that was for a CROSS FLOW configuration, and ours is essentially counterflow. I think your assumption is bang on in assuming the np's aren't contacting the HX tubes, more likely they are flowing adjacently along them. It's a unique design that I wasn't involved in, so studying it now I'm not all that surprised at our results.

What I am surprised at however is how significant the drop in LN2 temp is given how little nanofluid we are loading (about 2% mass fraction). Do you think this lower heat transfer would be due to lower thermal conductivity, thermal capacity, or to other properties of the air/nano mixture? The nanoparticles are mixed with acetone and the solution is injected at ambient temp into the hot air, dropping the temperature going into the inlet (green), which could also decrease HT I suppose. I'm still perplexed as to why the LN2 temp kept dropping after injections were completed as well.

(btw the temperatures on the y-axis are in Celsius)

 

[jwseifed]

Yea I'm guessing it is LittleInch. I agree there aren't really any confirmed findings with nanos. We are going based off of previous experiments, albeit much smaller scale ones, which showed some heat transfer enhancement of fluid after nanoparticles were introduced compared to before.

The theory with nanoparticles is that they have superior heat capacity due to greater surface area to volume ratios, lower particle relaxation time, higher Cv, etc. Our particles were 127nm in diameter.

Frosting occurs b/c of the cold LN2 going in (-180C). We've placed our petals in ovens before and still found it enveloped in frost even at oven temperatures around 430C, albeit the air was still and not moving like in this case.

 

[jwseifed]

The np material is boron carbide. In the previous successful test referenced, the material was aluminum oxide.

 

[chicopee]

Just a thought, research the use of inserted retarders within the tubes; perhaps that'll increase the contact between the nano particles and tube walls. The retarders are twisted flat strips which you'll find in your residential gas fired hot water storage tanks.

 

[jwseifed]

Thanks I'll look into that.

I think LittleInch is right in that it is mainly just extra resistance due to fouling. These particles stick pretty good, have to open up our HX and see.

 

[Compositepro]

The main effect will be the heat capacity of the particles cooling your hot gas and therefore lowering the heat transferred to to your nitrogen. If the problem were fouling, the heat transfer would not return to normal when you stop feeding your particles.

 

[enginesrus]

Maybe try Hydrogen dioxide nano particles. They seem to have worked good for heat transfer for years.

 

[btrueblood]

Wait, you are injecting particles into the air stream, not the LN2, correct? And from temperatures, it would appear the air side has iced up long before any nano particles get injected. Adding nano-particles to humid air would seem counterproductive, as they just add nucleation sites for more frost to form.

 

[btrueblood]

Before doing much else, I would look into putting a closed-loop control on the HX air-side surface temperature, or possibly on the leaving air temperature, to hold temperatures above 0 deg. C and hopefully control the icing. Control would need to throttle or pulse-modulate the flow of liquid nitrogen...

 

[jwseifed]

A closed-loop control of the surface temperature is very interesting. Decreasing LN2 flow rates would increase the surface temp but it would also decrease heat transfer rates, and thus our outlet air temperature would increase, and we want to decrease it as much as possible.

Trying to decrease the humidity seems like the simpler way to go. You are right that the nanos and acetone (which is added along with the nanos to better facilitate flow) are just creating more moisture and thus more frost. Perhaps passing everything through some sort of dehumidifier prior to entering the heat exchanger would help.

My theory isn't great on this, but is it possible to add a substance to the air which would remove some of the moisture? I mean, since the acetone/nanos are adding moisture, can we add a substance that would do the opposite? I think it comes down to the partial pressure of the water vapor correct?

I'm also curious how you deduced that our system was already iced up beforehand. Very intuitive

 

[IRstuff]

You're going down the same silly road again; adding "stuff" to your air stream just adds cooling load to your nitrogen. You could try methods described here:

https://content2.smcetech.com/pdf/BP12_Drying_Air_by_Lowering_Dew_Point.pdf

Another way to dehumidify is to freeze moisture out of the air, which is also a method of vacuum (cryo) pumping, but that is essentially what's happening starting at t > 230, since the N2 outlet temperature is just below freezing. You need to get the air's dew point to much less than freezing to do any good.

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529 Entire Forum list

http://www.eng-tips.com/forumlist.cfm

 

[Caloooomi]

There's really not a lot to go on here to be honest.

What sort of surface area, flowrate, pressure etc. are we dealing with?
Are the tubes enchanced, externally was fins, or internally with twisted tape?
Is it vertical or horizontal?
Are you monitoring surface temperatures on the tubing?

Could be that the nano-particles are creating more nucleation sites promoting freezing on the tubing. This is why I asked about the orientation - at least if it is vertical any condensate formed in the warmer sections can at least run back down the tubing and remove excess ice. Although I'm surprised by the amount of moisture in exhaust gas!

If there is no enhancement internally, you will be film boiling through the tubing as well.

For me, it is essentially an ambient air vaporiser inside an oven. You will always be air-side limited for heat transfer.

 

[btrueblood]

"I'm also curious how you deduced that our system was already iced up beforehand. Very intuitive "

It's hard to prevent (see below) so I automatically suspected it. A look at your temperature plots confirmed it, although if I was your lab TA I'd take points off for not labeling the graph axes.

Shortly after graduating college, I went back to visit my old profs and I was introduced to a group of students/PhDs/professors who were building a LN2-vaporization powered mail jeep and driving it around campus at the U of Washington. They also did some studies that showed it would be more economical than gasoline at the then current pricing of the two liquids in bulk. But, in our humid coastal environment, the heat exchanger that vaporized the LN2 before expanding the gas in the engine would always frost up eventually, which reduced the heat transfer drastically (rime frost is a pretty good insulator). Tricky problem, and icing becomes the limiting factor in that design.

Why did you use acetone? Could you use ethanol instead? Ethanol or any of the glycols would act as antifreeze, and lower the ice point on the surface, if only temporarily until it drips/blows off. Do you have a way to preheat the air going thru the HX? Can you put electrical heat tape along the tubes? Can you switch the airflow through two different heat exchangers, thawing an iced-up one while running the second until it ices up?

De-humidification of the air could be done in a upstream dehumidifier, you might even be able to run that dehumidifier using the chilly nitrogen coming off the downstream heat exchanger. The problem is, dehumidifiers take the air down to the dew point by stripping liquid condensate...but your next heat exchanger will take the air even colder, i.e. below the frost point of the remaining air. And frost buildup eventually chokes the heat transfer rate, unless you have a way to mechanically strip and remove the frost. There are heat transfer systems that can do this quite efficiently with mechanical removal of the ice...a commercial ice plant visit might give you some ideas. But...do your ice chunks then blow downstream into some other piece of equipment? How is the hypersonic vehicle research outlook these days anyway?

 

[jwseifed]

brtueblood I shouldn't comment on the application of this technology, but the industry is definitely "hyper" competitive.

It seeming more and more to me that frost is a problem we have to live with. Insulating or heating the surface would reduce frost but also reduce heat transfer, which is already being reduced due to the frost lol so pick your poison I guess. A dehumidifier is essentially just another heat exchanger, which would run into frosting problems of its own.

We are actually planning on using an ethanol based solution. One company supposedly managed to solve this problem by injecting methanol into the air stream. They inject it at the cold end and then re-inject it upstream and somehow this created dry air...? The article is below, maybe you can understand it better 'cause I'm pretty confused about it.

https://aviationweek.com/space/reaction-engines-reveals-secret-sabre-frost-control-technology

 

[hacksaw]

"the air has to be compressed before being injected into the engine’s combustion chambers. Without pre-cooling, the heat generated by compression would make the air hot enough to melt the engine.

The challenge for the engineers was to find a way to cool the air quickly without frost forming on the heat exchanger, which would clog it up and stop it working."

In counter-flow arrangements, you will see frost buildup progressing along the cold tube.

In cross-flow you'll see limited build up.

 

[jwseifed]

hacksaw What's the theory behind that? From my understanding, a counterflow system is more thermally efficient.