# Which is a more efficient method of heat transfer?

I'm attempting to determine airflow direction. The two methods I propose are either:

1) Blowing air across a heating element, with that hot air then blown over a thermistor that sits directly opposite the air direction. The idea that the heat from the heating element is blown over the thermistor causing its temperature to rise.

2) Blowing air across a thermistor which is being heated by the element its sat upon. The idea that the thermistors is cooled by the air blowing over it.

With a ring of thermistors I can then determine the direction of the air flow.

Which of the two methods would be a more efficeint use of the air flow with regards to heating or cooling the thermistor i.e. which of the two methods would see a greater change in temperature of the thermistor for the same airflow?

• I see how in method 1, the air must be coming from opposite the thermistor that heats up the most, but can you expand on how method 2 can be used to determine the direction of airflow? It seems like whether the flow is from the left or right, all thermistors sitting on heating elements will be cooled. I'm sure I'm missing something simple. Perhaps you could upload a diagram with the dimensions labeled to help evaluate the heat transfer modes that are relevant. – Chemomechanics Mar 27 '18 at 16:28
• @Chemomechanics I cant upload anything at the moment. But essentially method 2 relies on a structure that is essentially multiple tubes; one for each thermistor, poining out in multiple directions and redirecting the air flow over the thermistor. as for dimensions in both cases the thermisotrs would occupy a ring stucture of diameter no bigger than 10cm. This may not even be feasible hence running it through this forum. Ill try and upload a diagram later tonight – Pop24 Mar 27 '18 at 16:56

Since you are referring to a ring, I assume it is a 2D problem, although, it could, at a cost, be extended to a 3D problem.

I also assume that, in method 1, you are talking about one heating element installed in the middle of the ring of thermistors. If this is not a correct assumption, I would consider this approach anyway.

Again, assuming a 2D scenario, you could use a PCB to place and wire all your devices. Method 1 structure could look like a clock:

For the best operation, the heating element in the middle should be thermally insulated from the sensors (i.e., no copper plane), so that most of the heat transfer is done by the moving air, not by the board. To further improve the insulation, you could remove some of the PCB material between the heater and the sensors (white ring on the drawing).

The sensors around the ring should also have minimum thermal contact with the PCB, to minimize their thermal mass and make them more responsive.

You can use thermistors or many other types of temperature sensors.

I believe that method 1 should be more efficient than method 2, because you can use just one heating element rather than multiple heating elements, i.e., one for each thermistor.

Method 1 should be more accurate, since you would not have to worry about all heating elements having the same initial temperature and, most likely, you would not need any calibration.

With method 1, you would not need any pipes to direct the air flow.

You can control the sensitivity and resolution of your device by changing the number of sensors, the radius of the ring and the temperature of the heating element.

In summary, method 1 seems to be better: more efficient, more accurate, easier to implement and operate.

• Apologies for the late response, Method 2 was what i had in mindm and this is what i've opted for. The tips on insulating between heating element and thermistors is a good one that i hadnt considered. – Pop24 Jul 2 '18 at 8:46
• On my intial results unintentionally i found that airflow over the sensors actually cooled down the thermisotrs closest to the direction of flow, as the heating element had heated the thermistors (method 2 instead of the intended method 1). This isnt an issue per se as the rate of change of temperature could be used to find the direction. However because the thermisotrs werent isolated properly from the heating element the temperatures of them all in still air varied by up to 3-4 Degrees. Isolating may fix the issues so i have more uniform temperatures in still air. – Pop24 Jul 2 '18 at 8:46