Controlled Turbulence in a Heat Exchanger

A heat exchanger is in general any device for transferring thermal energy from one fluid to another. Commonly a piece of metal (steel, copper, aluminum, cast iron, etc) constitutes the exchanger. The effectiveness of the transfer depends on many variables including type and thickness of the metal, surface conditions on both sides, fluid velocities and temperatures on both sides.

In the specific example at hand we have hot gases (smoke), cool air to be heated, steel exchanger walls, with carefully planned air movements on both sides, all with the intention of being the most effective heat exchange geometry possible.

Following is a list of the important parameters that affect the results of heat exchange from the smoke side:

  1. Smoke air path area (at every point)
  2. Fire Intensity
  3. Fire Size and Location
  4. Amount of excess air
  5. Smoke Temperature

These all affect the average smoke velocity. If this velocity is too low then measures must be taken to increase the localized velocity near the exchanger surface. This is because low relative velocity causes "laminar" (layered) air flow along the exchanger surface. The layer right against the wall transfers heat satisfactorily but then tends to insulate the rest of the layers of air from the exchanger surface. The overall effect is poor heat transfer.

A simple aerodynamic analysis using the localized Reynold's number can determine the "critical" velocity below which laminar flow occurs. This must obviously be done for every portion of the exchange surface. Wherever laminar or critical flows are indicated design modifications should be contemplated. Above the critical velocity turbulence forms. Limited ( or controlled) turbulence should exist at all points on the heat exchange surface ON BOTH SIDES OF IT. Turbulence causes the thin layer of air near the surface that has already experienced considerable heat transfer, to be swept away from the surface to be replaced by other air. Most people are surprised that nearly all heat exchange occurs within a 1/8" thick layer of air next to the exchanger surface. Overall heat transfer efficiency is improved by moving the air against and then away from the exchange surface in a relatively brief time scale. Introducing turbulence is a very effective way of doing this.

There are some parameters which affect the heat transfer from the other side of the heat exchanger surface:

  1. House air path area (at all points)
  2. Blower size (capacity)
  3. Local forced air temperature

These affect the exchange on the room air side. Calculations for turbulence, etc., are similar but of course values are different.

Text Font Face
Text Size
(for printing)

We at JUCA selected a specific sequence of heat exchanger air flow to best take advantage of localized characteristics to pre-heat, heat, and boost the house air temp. In addition, the transitions from blower to pre-heater and between each internal area of heat exchange are "squarish" by design to initiate some extra turbulence in the succeeding exchanger passageways. Another purpose was to present a static pressure load when the optional blowers that we offer are used. Modern large furnace blowers MUST work against a load. If under-loaded they can over-rev changing electrical slip and phase angle parameters ultimately causing the motor to burn out.

Other parameters which affect performance of heat exchangers

  1. Metal type
  2. Metal thickness
  3. Surface condition
  4. Fluid characteristics (smoke density, etc)

Aluminum or copper would be ideal heat exchanger materials since they exhibit very high conductivity. Unfortunately they have low melting points and are quite susceptible to corrosion and oxidation, and so they are impractical for a furnace heat exchanger. Cast iron is durable but the necessary thickness of good castings reduces the overall available conductivity. Steel seems to show the best combination of durability and reasonably high conductivity for a warm air furnace application. Radiant stoves are not so dependent on rapid conductivity and so for them cast iron shows greater advantage.

As a unit is used a thin layer of soot or creosote coats the hot side of the exchanger surfaces. It would seem obvious that this will inhibit heat exchange by introducing another R factor into the energy balance equation. Computer simulations suggest a substantial (often 10%) reduction in net efficiency due to differing amounts of these deposits. This is in contrast to our experimental results. Our testing over the past 8 years has shown a reduction, but much less than would reasonably be expected, generally 1 to 3%. We don't yet have a good theoretical interpretation of this discrepancy, but it may involve the fact that the "lampblack" soot improves the surface emissivity to nearly that of a black body radiator, thereby improving the radiation capture (and release) probability. Further study in this area is needed.

For a geometrically complicated structure like the JUCA unit the determinations of all applicable aerodynamic and thermodynamic coefficients for all areas gets pretty involved. For the last several years we have been able to use a digital computer to "crunch" these numbers in analyzing product improvements. This approach will probably become necessary for woodburning products in general in the future as the sophistication of the industry improves.

The JUCA Home Page is at: juca