Hey guys,

I have no background whatsoever with fluid mechanics and have a basic question for my design.

I need to cool down a camera using a liquid coolant. It runs through tubes of a specific diameter ID and has to be set at a specific flow rate FR.

Now, the design is as follows. I will have a carboy filled with water which is gravity fed through the tube into the camera to cool it.

I understand that the height of the water inside the carboy needs to be kept constant in order to maintain a constant pressure (and avoid turbulent flow? I need to minimize vibrations).

My question is: How do I calculate the height of the water needed to get the right pressure/flow rate I need to cool the camera? I'm assuming the height of the carboy itself makes a difference...how do I take this all into account?

Thanks guys

 

Depending upon the budget for this project I would look at a new or old medical IV 'pump'.

These have the ability to also set and control the flow rate using gravity feed.

 

Do you know the mass flow rate you need to cool the camera? It sounds like you do. This translates into a flow velocity at the exit of the heat exchanger (mass flow is the fluid density multiplied by the cross sectional area and the flow velocity). This, in turn, will let you figure the fluid pressure at the exit of the heat exchanger.

Now, you know the length of tubing needed, the ID, the material and finish of the tubing inside diameter, and the flow velocity. You need to know the pressure drop (often called "head loss") of the heat exchanger's tubing. This accounts for frictional losses in the tubing. The usual method is to use the Darcy-Weisbach formula or, if you're using water, the Hazen-Williams formula. Once you derive the pressure drop, you'll get the pressure required at the entrance to the heat exchanger. If the tubing is not a straight chunk of tubing, there will be corrections needed; these effectively add to the length of tubing. However, I can't tell you how to calculated this correction; you'll probably have to make an engineering judgement by studying the losses for various fittings (especially elbows) or contact a fluids engineer.

Then you can use Bernoulli's equation to find the required water level in the carboy. There are some niceties with respect to the geometry of how the fluid is taken from the carboy to get to the heat exchanger, but those can be ignored for the first approximation.

There are lots of old books you can download from google books to help with this. You can also get a copy of Perry and Chilton's "Chemical Engineer's Handbook", which is a true work of art. I recommend getting a used copy from the 1970's or so -- I got mine for around $10 delivered via Amazon -- an absolute steal, as it was in new condition with no writing in it. The chapter on fluid flow will give you all the stuff you need, although it will probably be a bit confusing at first, as there seem to be so many technical details. But the basic problem solving strategy should probably be similar to what I outlined above.

It would also be wise to check the reasonableness of your answers using simple thermodynamic arguments as well as referring to a heat transfer text such as Holman (a 30 year old copy is fine).

Even after the paper modeling, the wise designer will be conservative and check things out when powering up. Try to power the load for a short time, then see if the coolant temperature differences are about what you expect. All of these design tasks are well established, but the devil is in the details and it's not hard to make a mistake in either understanding or numbers.

You don't describe the whole system -- do you have a pump returning the coolant to the reservoir? If so, you'll also have to factor in the temperature of the fluid in the reservoir increasing over time. Also, you don't say whether the thermal load is continuous or powered on and off. These things will also affect the design.

 

Would this be more of a job for a Peltier? We used Peltier-cooled photomultipliers for ionospheric research....worked very nicely...no muss, no fuss.

Eric

 

At your location, wouldn't just opening a window do as well?

What color photons were you counting?