Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Increasing Heat Transfer Area

  1. Aug 2, 2011 #1
    I have an assembly where I'm using a starwound cable heater to heat up a flow of air. But the air is not heated to the desired temperature. So I have decided to increase the heat transfer area by using metallic wool or some meshing around the heater. can anybody plz suggest me which one would be a better choice?
  2. jcsd
  3. Aug 2, 2011 #2


    User Avatar
    Science Advisor
    Gold Member

    Can you decrease the speed of the flow, or recirculate it a couple of times?
  4. Aug 2, 2011 #3
    At first I tried with a low flow rate but it didn't worked as much as it is working with a moderate flow rate. My initial option was to recirculate the hot air again, but the design limits the ability to recirculate the air without any recirculation pump. And I couldn't find any small pump able to recirculate hot air. :(
  5. Aug 2, 2011 #4


    User Avatar
    Science Advisor
    Homework Helper

    You could use the coil to heat up a conventional extruded metal heatsink, and design the system so the air is forced to flow between the heatsink fins. That way you should be able to predict if the system will do what you want, from the "degrees per watt" specification of the heatsink.

    Adding meshing or metal wool won't achieve much unless the coil can heat up the meshing etc efficiently. That may be difficult to achieve.
  6. Aug 2, 2011 #5
    Please take a look at my assembly and the heater. Also, is it possible to add fins to the heater because the clearance between the heater and the tubular casing is small.

    Attached Files:

  7. Aug 3, 2011 #6


    User Avatar
    Science Advisor
    Homework Helper

    You don't say what the physical size is, or what mass flow or heat transfer rate you want.

    I assume the air outlet is the side tube half way up the casing.

    I think one problem is that most of the air flow is not in contact with the heating coil which looks small compared with the size of the casing. I would try a spiral wound coil that covers the whole area so all the air has to flow through it. Make the gaps between the turns a similar size to the diameter of the wire. If the spiral is smaller than the casing, block off the sides so all the airflow has to go through the heating coil.

    You can probably buy spiral coils ready wound. There seems to be room to stack several above each other if you need to that.

    My first idea using a heatsink would work better in a rectangular casing. From your drawing there doesn't seem to be any reason you couldn't change that. Extruded heatsinks come in a wide range of sizes but I was thinking about the ones sold as BGA heatsinks, meant for cooling "ball grid array" chips like computer processors. Bond a heating pad onto the heatsink where the chip would be (you could use a heating wire and thermal cement). Again block off the sides as required so all the airflow has to go through the fins of the heatsink.

    Those heatsinks are designed for forced airflow and in computers this is created by a fan, but if you already have a forced flow through the system you won't need a fan as well.

    The heatsink idea would be a bigger design change, but it has the advantage that you can calculate reliably what heat transfer rate you will get, rather than having to experiment.

    Attached Files:

  8. Aug 3, 2011 #7
    The main problem is, I don't have enough time left to change the design completely right now. I could only modify the current design. My desired temperature is 350C but at 350C heater temp. I get hot air of around 180C. This is way too low for my purpose. The air is supplied by a compressed air cylinder and I'm supplying the air at 2 bar. I have attached a schematic of the system. The heater is around 10-15mm shorter than the casing length.

    Is there any calculation to get the optimum flow rate required to attain this temperature? i also think the air is not getting enough contact with the heater.

    Attached Files:

  9. Aug 4, 2011 #8


    User Avatar
    Science Advisor
    Homework Helper

    If you think about Newton's law of cooling, then the heater temperature must always be higher than the exit air temperature, otherwise there will be no heat transfer! In a system like this with a fairly short heating time, a temperature difference of the order of 100C doesn't seem unreasonable.

    From the mass flow rate of the air and the required temperature rise, you know the power that the heater needs to supply (Power = Cp x delta T x mass / second).

    If you ignore heat loss through the casing etc, then if you apply that much electrical power to the heating cable, the cable will heat up to whatever temperature it needs to be at, to reach to the steady state condition where all the heat transfers into the air.

    So long as you don't exceed its max temperature rating, the cable temperature can look after itself (and IIRC these type of heater cables can be rated up to 1000C or more).

    So, if you want a feedback loop to control the system, it would be better to use the exit air temperature to control the electrical power supplied, and just monitor the cable temperature for safety.

    If you change the air mass flow, you wll need to change the heater power in the same proportion, but (perhaps non-intuitively) the heater cable temperature will not change by much.
  10. Aug 10, 2011 #9
    I have tried with steel wire wool around the heater to increase the heat transfer. It has increased quite significantly but still not able to achieve the required temperature. Can preheating of the air be an effective solution to this problem?
  11. Aug 10, 2011 #10


    User Avatar
    Science Advisor
    Gold Member

    Your problem is that your heater has not been properly designed given you requirements. Resistive air heaters generally have a large surface area/flow area ratio, and will tend to be overpowered for the application so they don't have to run at 100% power all the time.

    I would recommend redesigning your heater to incorporate a series of 2 or 3 nichrome meshes instead of the current coil. It will increase your heat transfer area significantly and double as a turbulent mixer for mixing the heated air.
  12. Aug 11, 2011 #11
    thnx for the suggestions.

    Is there any general guideline (apart from the energy balance eqn) for calculating the heater capacity to achieve the desired temperature? My target is heating a stream of air to 350C which is @20C.
  13. Aug 11, 2011 #12


    User Avatar
    Science Advisor
    Gold Member

    The general guideline for power requirements would be:

    1- Given a mass flow rate and specific heat capacity of the flow, you can calculate the minimum require power needed to heat it. Given the total power requirement and a safety factor, you can specify the voltage and current the wire would have to operate at (and the temperature it would reach based in total length (see #2).

    2- Designing the heater hesh is more challenging because it would reaquire some heat transfer calculations for convection over a cylinder. Generally, if you know the speed and temperature of the flow, the temperature of the wire, the size heating wire you plan to use, and the total heat flow power, you can calculate the total length of wire you need in the flow to achieve the desired heating.
    Last edited: Aug 11, 2011
  14. Aug 15, 2011 #13
    The heater I'm currently using has wattage rating of 420W and watt density of 16W/in2.

    I am still frustrated with the heater performance. According to a energy balance heater is capable of heating air at much higher flow rate than what I am flowing.
  15. Aug 17, 2011 #14
    Where's that 420W going? It must be either entirely heating the air, or not being used at all. Is the heater running at continuous duty? If so then no tweaking heatsinks will help.

    But I suspect the heater isn't even operating half the time. That's why it's not heating properly. You could probably squeeze a little more power out of it just by running it at over-temperature (same wattage but higher duty cycle). Or simply put an obstruction in the flow so more of the air passes over the heater.
    Last edited: Aug 17, 2011
  16. Aug 21, 2011 #15
    The heater is running continuously. I have also tried putting obstruction, it helps to increase the heat transfer but still the temperature is not reaching that of the heater.
  17. Aug 21, 2011 #16
    The air temperature can never reach that of the heater, so that shouldn't be a goal.

    1) Energy balance says that 420W should heat the air from 20C to 350C.
    2) Real life says that 420W heats the air from 20C to 180C.

    One of those must be wrong.
    1) Are you sure? Maybe show your calculations.
    2) Heat loss through the walls? Measurement error caused by poor mixing? Is the input really 20C? The expansion from the tank would have cooled it somewhat, but surely not to -200C!!!

    I don't think you can win by modifying the geometry. Using fins/mesh is just a way to operate the heater at a lower temperature for the same power output. Your heater is already producing as much power as it can so there's nothing to gain from running it colder.
  18. Aug 22, 2011 #17
    I am going to try with an air preheater to see whether the temperature reaches the 350C mark. Do you think preheating the air upto a certain temperature could help?
  19. Aug 22, 2011 #18
    For sure. Or maybe increase the voltage to the heater if the insulation is rated for it.

    I'm suddenly wondering why the heater's surface(?) temperature is the same as the desired air temperature. Is that just a coincidence? Are you sure it has no thermostat?
  20. Aug 22, 2011 #19
    I have a temperature controller. I can select what temperature the heater should be operating at. Also, there is a thermocouple attached to the heaters surface to check it temperature is same as the set temperature.
  21. Aug 22, 2011 #20
    OH! That explains a lot of things! So really the heater isn't using a constant 420W? It may be much less. Are you measuring the electrical power?
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook