Help with Peltier modules 1

[Syncro5]

Hi all,

I'm somewhat of a novice with electronics but I am currently designing a system which needs to heat 2l of liquid up to 64°C for a short time (around 2 min) and then quickly cool it to below 7°C.

Is it possibe to use peltier modules for both aspects of the system, heating first and then reversing the current to immediately cool?

If so say we are opperating from an ambient temperature of 20°C. Would a Peltier module with a Delta T of 90°C mean the hot side increases to 65°C (ambient of 20 + 45) and the cold side decreases to -25°C (ambient of 20 - 45).

Any info on the matter would be greatly appreciated!

Thanks,

Joe.

 

[itsmoked]

2 liters?

No.

2 grams? Yes.

You should consider reservoirs of heat and cool.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[VEBill]

I agree with Keith's point.

It's a Physics 101 problem. Refer to Thermal Mass (a.k.a Heat Capacity)

Think how long it takes for an electric kettle (~1500 watts !!!) to raise a litre or two of water to ~100°C. A typical TEC will be many, many, many times less powerful (in either direction) and would thus take far longer than indicated by your choice of temporal domain adjectives.

 

[waross]

My first though was reservoirs or some type of thermal mass storage. Then I saw that itsmoked had beaten me to it.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[itsmoked]

Keith Cress
kcress -

http://www.flaminsystems.com

 

[MikeHalloran]

Sure, you can do it with Peltier cells. ... but you'll need an awful lot of them, and a huge high current DC power supply to run them. Other means of heating and cooling will be much cheaper to buy and to run.

If you want to change temperature fast, you need a lot of surface area. Think a heat exchanger with a lot of tubes, or a chamber, thin in one dimension and large in the other two, or maybe a large tall thin annulus. Selection depends on the nature of the process fluid you wish to heat and cool.

As mentioned, you would surround the process fluid with hot fluid from a hot reservoir, then quickly drain the hot fluid and pump in some cold fluid from a cold reservoir, then drain that to repeat the cycle. It's possible to use one heat transfer fluid and heat and cool it dynamically, as in a reversible heat pump, but it's simpler to use the separately heated and cooled reservoirs.





Mike Halloran
Pembroke Pines, FL, USA

 

[waross]

Another option is to use a reservoir of hot liquid and a heat exchanger. Circulate your sample through one section of the heat exchanger and circulate the hot fluid from the reservoir through the other section of the heat exchanger. With a reservoir much hotter than your target temperature you will get rapid heating of the sample. You could use a reservoir of cold fluid and the same heat exchanger and pump for cooling.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[IRstuff]

Your values result in 3kW of heating, but 4.26kW of cooling, which would require 80 of these puppies:

http://www.ferrotec.com/index.php?id=module_detail&mod_id=114

TTFN
faq731-376

 

[VEBill]

80 of those (at 5.5cm x 5.5cm each) would surround approximately 8L of volume; as opposed to 2L. I suspect that the inherent impracticality is the subtle point of IRstuff's link and the estimate of Qty 80.

...to heat 2l of liquid up to 64°C for a short time (around 2 min) and then quickly cool it to below 7°C. ...to immediately cool? ...

The OP didn't actually give any rate of change values. The only absolute time reference "around 2 min" could be interpreted to be the time spent at a steady +64°C. The only information about the rate of change are the undefined adjectives quickly and immediately.

 

[IRstuff]

Well, if 2min isn't the transition time, then my estimate of 80 is way off

TTFN
faq731-376

 

[btrueblood]

And the 8L coverage is probably more of an upper limit, one can easily conceive of configurations where the 80 units surround a working volume of less than 8L, down to zero.

In all seriousness, there are many polymeraze chain reaction (PCR) reactors (for genetic research) on the market using Pletier devices operating through temperatures similar to the OP's temperature range. One problem that develops is that thermal cycling develops thermal expansion stresses where the thermoelectric chips (typically BiTe) bond to the ceramic face plates. This set up fatigue stresses with each cycle. There is a Russian company that makes Peltier devices but with a compliant bonding layer between the BiTe chips and the ceramic faceplates; the compliant layer helps relieve the thermal expansion strains. In testing I did for a Peltier chiller, these devices withstood better than a factor of 100 more cycles than the more traditional solder/braze bonded peltier devices. A quick google reminded me of that company name, it's website is

http://sctbnord.com/article_157.html

 

[Syncro5]

Thank you everyone for your response! I have looked at having hot and cold reservoirs and heat exchangers. My thoughts were that the hot side would be relatively cheap and easy to do but the cold side would require a complete refrigeration system (vapour compression cycle being the most appropriate?) which seems expensive and bulky! I imagined I would need more than one Peltier module but had a maximum of maybe 10 in mind.

Just to clarify the 2 min is the time in which the liquid must remain at 64°C. Getting there shouldn't take any longer than 10 - 15 minutes though and likewise with the cooling would be around 15 minutes also. There will be a system in place keeping the mixture homogenised so it can be heated uniformly.

Another point I really should have mentioned is that the liquid being heated will change properties above 70°C so the heat source needs to remain below that temp to stop the properties of the liquid changing local to the heat source. So yes.. a big surface area to volume is key.

Is the general consensus here that using a separate heating and cooling system to generate hot and cool reservoirs would be more cost-effective than using the Peltier modules?

I am aware this is starting to stray a little from electrical to thermal dynamics so I'll repost in a separate forum when I progress from the initial "is it possible" stage.

 

[VEBill]

...64°C. Getting there shouldn't take any longer than 10 - 15 minutes though and likewise with the cooling would be around 15 minutes also...

Luxury. The project just became perfectly feasible in any number of ways.

 

[IRstuff]

A factor of 8 drops the TECs down to 10, which is more feasible. Reservoirs would still need some means of controlling temperature, so TECs might still be required somewhere.

Will, or can, the fluid be agitated for better thermal conduction?

TTFN
faq731-376

 

[Syncro5]

Will, or can, the fluid be agitated for better thermal conduction?

- Yes it will have. Something like a rotating mixer in the centre of the container and baffle plates around the outside (which could also be used for increasing the SA for heat transfer).

So this option may be feasible! To relate back to the original post, does the Delta T of the Peltier module split equally either side of the ambient operating temp? (i.e. when using a module with a Delta T of 90°C, operating in a 20°C environment, does the hot side raise to +65°C and the cold side drop to -25°C). I'm sure it won't be quite as easy as that but for a there or there about rule?

 

[IRstuff]

Yes and no. The pumping efficiency is radically different for the two cases.

TTFN
faq731-376

 

[Syncro5]

The refrigeration phase will have a significantly different "ambient" temperature to the previous heating phase. I can see how that case is considerably more complex as the "ambient" temp will be decreasing constanly.

I really appreciate all the help, I think I have enough support here to buy some peltier modules and start prototyping!

Thanks again,

Joe.

 

[DHambley]

To answer your question about both heating and cooling; Yes. They are bi-directional.

 

[itsmoked]

How about this?

Use a mechanical heat pump. Use the reservoir scheme. Use the mechanical to heat one reservoir and cool the other. Make the heat exchange fluid something that works at both temps. Glycol of some sort perhaps. Then just alternately flood the cooling chamber from the appropriate reservoir.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[Syncro5]

How about this?

Use a mechanical heat pump. Use the reservoir scheme. Use the mechanical to heat one reservoir and cool the other. Make the heat exchange fluid something that works at both temps. Glycol of some sort perhaps. Then just alternately flood the cooling chamber from the appropriate reservoir.

I wasn't sure what you meant by "mechanical heat pump" so did a quick google search and all I can find are vapour compression refrigeration systems.. is that what you had in mind? Heat one reservoir of exchange fluid using the condenser coils and cool another using the evapourator coils? That option still seems, to me, like a more expensive option but you are the experts! I'll get in touch with some refrigeration companies too.

Thanks again!

Joe.