How Can Increasing the Heat Transfer Area Improve Air Heating Efficiency?

[ruzfactor]

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 please suggest me which one would be a better choice?

 

[Mech_Engineer]

Can you decrease the speed of the flow, or recirculate it a couple of times?

 

[ruzfactor]

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. :(

 

[AlephZero]

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.

 

[ruzfactor]

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.

 

[AlephZero]

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.

 

[ruzfactor]

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.

 

[AlephZero]

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.

 

[ruzfactor]

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?

 

[Mech_Engineer]

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?

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.

 

[ruzfactor]

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.

 

[Mech_Engineer]

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.

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 reacquire 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.

 

[ruzfactor]

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.

 

[Unrest]

The heater I'm currently using has wattage rating of 420W and watt density of 16W/in2.

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.

 

[ruzfactor]

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.

 

[Unrest]

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.

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.

 

[ruzfactor]

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.

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?

 

[Unrest]

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?

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?

 

[ruzfactor]

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.

 

[Unrest]

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.

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?

 

[ruzfactor]

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?

I am not measuring the electrical power. I just set what temperature I want the heater to be and check the heater temperature after that (both done in the temperature controller).

 

[Unrest]

I am not measuring the electrical power. I just set what temperature I want the heater to be and check the heater temperature after that (both done in the temperature controller).

I'm confused what you're trying to do. Make the exit air temperature 350C? In that case you (according to your calculations) just need to provide 420W, and it doesn't matter what the heater's temperature is.

Since it's thermostatically controlled, the heater won't be providing 420W on average, so it won't heat the air as much as you calculated.

Why not just turn the temperature up?

 

[ruzfactor]

I have tried at 400C, 500C but there is not much improvement in heating the air.

 

[ruzfactor]

well, preheating also did not worked as the temperature of the powder bed reached only 180C after 4 hrs of heating! :(

 

[Unrest]

How does the powder bed fit in with the air heater? Is that where the air goes to? Where's the preheater?I really think you should work out what, if anything, was wrong with your calculations. Then you can see the best way to solve it. You calculated the power needed to heat the air, and it didn't work. Why not? Check the power you're really adding with the preheater and the coil heater combined. Is it still less than what you calculated you need?

 

[ruzfactor]

How does the powder bed fit in with the air heater? Is that where the air goes to? Where's the preheater?I really think you should work out what, if anything, was wrong with your calculations. Then you can see the best way to solve it. You calculated the power needed to heat the air, and it didn't work. Why not? Check the power you're really adding with the preheater and the coil heater combined. Is it still less than what you calculated you need?

Powder bed is inside a reactor vessel and, the reactor is enclosed with another vessel ( with gap between them for air flow). Please see the earlier posts in this thread for a schematic. At a relatively moderate flow rate, the heat transfer is slow within the bed and at a high flow rate the heat transfer increase significantly. But, the preheater and heater temperature drops well below the set temperature at high flow rate.

My calculation is just an energy balance for the coil heater to see whether the heater is actually capable of heating the air.

How do I calculate the flow rate for the heater to actually heat the air to 350C? Energy balance shows it can, but practically it is not capable. :(

 

[Unrest]

How do I calculate the flow rate for the heater to actually heat the air to 350C? Energy balance shows it can, but practically it is not capable. :(

Air has specific heat of 1.0 kJ/(kg K)
330kJ/s should heat air with a mass flow rate of 1kg/s by 330C

You calculated 420W, so I suppose the mass flow rate is 1.3g/s (2.7 cfm at atmospheric pressure)

Maybe the mass flow rate is actually higher than that?
Maybe you're not really supplying 420W?
Maybe too much heat is being lost through the walls?
Maybe the flow rate is so low that the powder bed takes too much time to heat up?

 

[ruzfactor]

Air has specific heat of 1.0 kJ/(kg K)
330kJ/s should heat air with a mass flow rate of 1kg/s by 330C

You calculated 420W, so I suppose the mass flow rate is 1.3g/s (2.7 cfm at atmospheric pressure)

Maybe the mass flow rate is actually higher than that?
Maybe you're not really supplying 420W?
Maybe too much heat is being lost through the walls?
Maybe the flow rate is so low that the powder bed takes too much time to heat up?

I did not calculate 420W. It is the rating of the heater that is being used in the system. Last day I set the coil heater to 500C and the prheater to be 400C. After flowing the air at high flow rate, both the temperature dropped to 288C and 175C respectively. The powder bed has very low thermal conductivity (~1), but that is not my concern. If the heater is not heating to 350C, how can I achieve 350C in the reactor. As the coil heater is almost in contact with the wall, I have used some insulation around the outer wall to prevent heat loss to the surrounding. Is there any benefit of using moist air instead of dry air?? How can I calculate the mass of air required to constantly expose reactor outer shell to 350C?

 

[Unrest]

How can I calculate the mass of air required to constantly expose reactor outer shell to 350C?

Power = specific heat * mass flow rate * temperature difference
-> mass flow rate = 1.3g/s

If the air's flowing faster than that then it should not reach 350C without the preheater.

That assumes no heat loss. I think you should estimate the heat lost through the walls. Just assume the inside of the wall is at the air temperature (or 350C) and find the heat flux through conduction in the metal, convection outside, and radiation outside. Another way is to measure the temperature gradient (say inside and outside of wall).


Where's the preheater in the system? Before the coil heater or between the coil heater and powder bed? Or in the powder bed?

 

[ruzfactor]

Preheater is between the coil heater and compressed air line..

 

[Unrest]

It seems more and more like you're losing too much heat through the walls.

How about this:
- Measure the flow rate
- Measure the temperature increase of the air
- Calculate the power that would ideally be required to achieve that (like I did)
- Measure the electrical power used by the heaters
- The difference between the electrical power used, and the power needed to heat the air should be equal to the rate of heat loss.
- Assume heat loss is constant with temperature (terrible assumption, but it's easy!)
- Add another heater whose power is equal to the heat loss rate.Another idea:
Disconnect the compressed air and let it flow by natural convection. If that gives >350C air temperature then you can gradually increase the flow rate till it's just right. If it doesn't work then there's no hope without replacing the heaters or insulating it more.

I don't think water will help. It'll carry heat out the walls faster as well as to the reactor faster.

 

[ruzfactor]

I think you are right. Too much heat is lost through the walls. Most of it is in the wall surrounding the coil heater. So at moderate flow, air looses the heat through the walls. At high flow the air looses less heat but the heater is unable to cope with high flow, that's why the heater temperature decreases from the set temp. Please correct me if I am wrong.

I am now planning to use the inline heater to heat the air an directly send to the reactor vessel part by removing the coil heater part.

Btw, will it be possible to approximate the heat loss through the walls? Although I am using some insulation around the walls but still heat is lost.

 

[Unrest]

Sounds like it. But it sounds like you're just guessing. Why not calculate what should happen, then you won't have to spend ages modifying it just to find it still doesn't work.

Can you measure or estimate the mass flow rate?

I guess you can safely assume the heater is using full power because it can't reach it's set temperature.

 

[ruzfactor]

Sounds like it. But it sounds like you're just guessing. Why not calculate what should happen, then you won't have to spend ages modifying it just to find it still doesn't work.

Can you measure or estimate the mass flow rate?

I guess you can safely assume the heater is using full power because it can't reach it's set temperature.

Unfortunately I don't have any flow measuring device attached to my system. I really want to calculate what should happen, but don't know where to start from.

 

[Unrest]

Did you try my "other idea" of using a very low flow rate (eg natural convection) to see if it does work slowly.

If so then it's all fine, just increase the flow rate till it just barely works and you're done. Unless that's too slow.

Other than that, I think knowing the flow rate is kind of crucial, and you won't be able to design anything without that. Your air's coming from a tank, right? How about weigh the tank before and after to see how much mass was lost during the operating time.

 

[ruzfactor]

I have tried with low flow rate. The heat transfer is much less. My air is coming out from a compressed air cylinder. Could you please explain why the heater is not able to heat the air? could it be because of heat loss through the walls?

 

[Unrest]

I have tried with low flow rate. The heat transfer is much less. My air is coming out from a compressed air cylinder. Could you please explain why the heater is not able to heat the air? could it be because of heat loss through the walls?

Which heat transfer is much less? Heat flux into the reactor? Does that matter if it heats more slowly? Did you see the air temperature (or heater temperature) increase with a low flow rate?

I'd really like some idea of the flow rate to quantify it. However it's hard to imagine anything else besides loss through the walls.

 

[ruzfactor]

Heat transfer is much less in case of low flow rate. I need the reactor to be heated within a shorter period. But it all failed. A new design is the only solution I guess. Is the heat capacity of damp air higher than the dry air? How can I prove (theoretically) that damp air could improve heat transfer? Need some help on that.

 

[ruzfactor]

Is there any effect on temperature, when a flow experience sudden enlargement?

 

[Unrest]

It's a pity you can't know the mass flow rate. With that and your existing measurements you can confidently estimate the losses and predict what wattage heater is required.

I probably know no more than you about using wet air. My guess is it would have a higher heat capacity, but I don't think that's important, just using a higher flow rate should do the same job.

I really think you should work out what went wrong with this design before trying something fancy. Otherwise the fancy thing will fail too. You can probably do it just by replacing the heating element if you have some data to make predictions with.

Is there a reason you don't have the heater wound around the reactor? Why transfer it so far through air? And why is the air flowing?

When gas expands it cools. That will be happening to the gas coming out of the tank. It could easily be below 0C, but probably not enough to make a difference to the overall design.