Fabrication and testing of heat pipes

From Thermal-FluidsPedia

Jump to: navigation, search



Heat is the name given to energy as it is transferred from one region to another via the thermal processes of convection (moving fluid), conduction (via a mechanical connection), and radiation (thermal emission). There are a great many processes where it is important to cool or heat materials. The most efficient way to do this is by using heat pipes. Since there is a great deal of information on this subject on the Web, only one reference will be cited here—the monograph by Faghri[1]. In his introduction he states (p. 1):

The advantage of using a heat pipe over other conventional methods is that large quantities of heat can be transported through a small cross-sectional area over a considerable distance with no additional power input to the system. Furthermore, design and manufacturing simplicity, small end-to-end temperature drops, and the ability to control and transfer high heat rates at various temperature levels are unique features of heat pipes.

Heat is transferred from the heat source by the evaporation of a working liquid in a sealed evacuated container to a heat sink where the condensation of the liquid gives up its heat. Heat pipes are manufactured to function from liquid nitrogen temperatures to the boiling points of metals. Table 1 gives the typical operating characteristics of a number of working liquids, along with the vessel materials used. Heat pipes with metals that evaporate like sodium are used in nuclear reactors.

Table 1. Operating Characteristics of Several Working Heat Pipe Liquids

Temperature Range/̊C Working Liquid Vessel Materials
-200 to -80 liquid nitrogen stainless steel
-70 to +60 liquid ammonia Ni, Al, stainless steel
-45 to +120 methanol Cu, Ni, stainless steel
+5 to +230 water Cu, Ni
+500 to +900 sodium Ni, stainless steel
+900 to +1500 lithium niobium + 1% zirconium
+1500 to +2000 silver tantalum + 5% tungsten

Figure 1 is a schematic representation of a heat pipe. As indicated in Table 1 various metals are used for making the container. For easy-to-make heat pipes the materials of choice are copper and borosilicate glass tubing. A small amount of working fluid is sealed into the tube under vacuum. Wicking material inside the tubing enhances fluid flow.

Heat Pipe Schematic
Figure 1. Schematic of a heat pipe.

The liquids discussed in this paper are ones that can be used at ambient temperatures: water, methanol, 95% ethanol, and acetone. Relevant physical properties of these liquids are shown in Table 2. Of special interest is the enthalpy of vaporization per mL of solvent. Notice that this value for water is significantly higher than those of the other liquids; water transfers the most heat per mL of evaporation of these four liquids. For ambient temperatures heat pipes containing water are the best (and safest) choice. However, it is usually easier to prepare a heat pipe with the more volatile organic liquids, although working with these flammable solvents is not recommended.

Table 2. Properties of Four Liquids.a

Liquid ΔHvap/(kJ/mol) NBP/̊C VP/torr ΔHvap/(kJ/mL)
water 40.65 100 23.8 2.25
methanol 35.3 64.7 125 0.866
ethanol 38.6 78.3 59.0 0.658
acetone 29.1 56.2 229 0.393

a NBP is the normal boiling point at 101325 Pa (1 atm). The ΔHvap values are at the NBP, and VP is at 25̊C.

Commercial heat pipes function well in both vertical and horizontal orientations. The heat pipes described in the next section work well in the vertical orientation since the condensed vapor simply runs down the tube to be re-evaporated. Commercial heat pipes have proprietary wicking materials inside which allow the condensed liquid to return efficiently to the hot section. The wicks are usually powdered metals which are sintered onto the inner surface of the metal pipes. Sometimes a wire mesh (like copper or aluminum window screening) is used for the wick. Heat pipes are also made as small rectangular blocks for use in electronic equipment like laptop computers—these may be as thin as a few millimeters. They require very efficient wicking. Also, for effective cooling the hot end needs to have an effective heat exchanger which can remove the heat. Simple cooling devices use fins and air movement, while other devices utilize liquid cooling coils.


A design criterion for the heat pipes to be described was that they would be able to be made without special tools or skills. Copper tubing in the 1/4, 3/8, and ½ inch sizes were used in making many heat pipes, with the 1/4 inch size preferred. This does require some skill in using a propane torch and soldering copper. Other heat pipes were made from borosilicate glass. The only rigid clear plastic tubing that met the heating requirements was Lexan. Many tests with this tubing showed it to not be practical because it is such a good thermal insulator.

A simple test was used to find out if a heat pipe was successfully constructed. A small inexpensive hot pot was used to heat water to about 90̊C. The heat pipe was immersed in this hot water to a depth of 3-5 cm. The top of the heat pipe was held and the time noted until the top got too hot to hold. Our commercial 1/4 inch copper heat pipe of 12 inch length took just 5-6 sec to get too hot to hold. Our better heat pipes took 10-20 sec to get too hot to hold. A wonderful demonstration of the effectiveness of heat pipes to transfer heat (Expt. 1) is to place into hot water one of the effective heat pipes and a similar length of 1/4 inch solid copper, brass, aluminum, or steel rod. For the solid metal rods heat is transferred up the rod via conduction (a slow process), while for the heat pipes the transfer is via vapor movement (a much faster process). This experiment can be quantified by measuring the length of time it takes for the rod to get too hot to touch near its top (or one-half way up its length). The temperature at various points along the length of a rod (or heat pipe) can be determined by using an inexpensive bead sensor digital thermometer such as #4233 available from www.control3.com. A gauze pad (such as those used for cosmetics removal) is used to hold the sensor against the rod safely. For this experiment a length of copper tubing (with and without its lower end sealed) can also be tested.

Copper Tubing heat Pipes – Although heat pipes may be fabricated from any diameter copper tubing, the least expensive and easiest to work with is 1/4 inch soft copper. We have made heat pipes out of 1/4, 3/8, and ½ inch copper tubing, and in lengths from12 to 36 inches. A 1/4 x 12 inch tube is recommended. The "bottom" end is sealed by soldering on an end cap, or by crimping it shut and filling the flattened end with solder. Figure 2 shows a variety of ways of sealing the bottom ends. The “top” end needs to have a way of sealing after a sufficient amount of fluid has been boiled off. Figure 3 shows several ways of sealing the top end. A short (ca. 2 inch) length of flexible plastic or rubber tubing is fastened to the top end by wiring it down, or with a hose clamp. For ca. $6.00 a 1/4 inch copper plumbing ball valve with a compression fitting also works well. The plastic or rubber tubing is most easily sealed with a thumb screw hose clamp (available from scientific suppliers). An alternative method is to fabricate out of a piece of 2.5 x 2.5 x 0.25 inch hard wood or pressed wood a sealing wedge where the spacing of the wedge compresses the tubing shut when the wedge is forced onto the tubing. A hemostat can also be used. Figure 4 shows several wedge seals.

For a 1/4 inch tube the initial charge is 1 mL of water. (Larger diameter tubes take up to 2 mL.) The water in the hot pot is heated to ca. 90̊C. Evaporation of the water is via an inexpensive plastic water aspirator (Nalgene # 6140-0010, ca. $20.00) that is connected to a cold water faucet. In our set-up this aspirator could attain pressures as low as 80 torr. Pressure tubing is fastened to the water aspirator. The connection from the aspirator tubing to the top of the heat pipe is via a two-ended tapered plastic adapter or, if appropriate, a short length of copper tubing. Adapters of different sizes may be fabricated by nesting several diameters of copper tubing and soldering them in place (see Fig. 4). For ease of handling a ring stand with clamps holds the heat pipe while it is being evacuated. The water aspirators we have used vary in the low pressure provided by the style of the aspirator and the temperature and flow rate of the water. Thus, it is not possible to recommend a length of time for evaporating a sufficient amount of water before sealing. After the heat pipe’s lower end is immersed in hot water and the water aspirator is turned on seal the heat pipe after two min and test it as described above (after cooling first). For a given set-up you will find a length of time that results in a heat pipe that is too hot to hold in 10-20 sec. For our 1/4 inch copper heat pipes the useful evaporation time ranged between 30 and 60 sec. Since you cannot see how much water is left in the opaque copper tubes, the evaporation time must be determined by trial and error.

Borosilicate Glass Heat Pipes – A propane torch (readily available in any hardware store) can heat up borosilicate glass sufficiently to make test-tube bottom seals in 6-10 mm OD tubing. Some skill at glass-blowing is helpful (RB is experienced at this). The top ends of the borosilicate tubes are also sealed in ways similar to those described above for copper tubes. If you are a skilled glass-blower, it is also possible to seal the top end with a torch when sufficient fluid has been evaporated and the tubing is still evacuated to make permanently sealed tubes. There are two alternative methods for sealing the bottom ends of glass tubes: (1) a glob of epoxy cement fills the end of the tube and is also shaped around the edge of the tubing; and (2) a silicone rubber stopper which is safe to use in boiling water. The epoxy cement seal also serves as a site for bubble formation for easier evaporation. You will have to experiment with sealing times to make functioning borosilicate heat tubes since it is difficult to see how much water is left in the tube when immersed. If too much liquid is left behind, it takes longer for the heat pipe to get too hot at the upper end.

Another way of making borosilicate glass heat pipes using only rubber stoppers was successful with a 12 inch length of 15 mm OD tubing. A propane torch can be used to fire-polish the end of this tubing, and for all such tubing. A rubber stopper seals the bottom end. This is then wrapped with several layers of aluminum foil to protect the stopper from over-heating. 2 mL of water are added. An inexpensive heat gun (e.g., Harbor Freight) was used to boil off the water using the lower heat setting. When steam was seen to be emerging for 2 min, a lightly greased rubber stopper was used to seal the top end, the heat turned off, and the lower end placed in cold water. This heat pipe took 30-45 sec to get too hot to touch.

Expt. #2. Determine how long it takes for each of the heat pipes you have made tp get to hot to touch at different points along their lengths, and at the top. The digital thermometer can yield the temperature at each of these locations.

Heat Pipes at Shallow Angles – The heat pipes described above do not have the efficient wicking that is in commercial heat pipes. They are best used and demonstrated in the vertical mode with hot water, or at shallow angles using a hot air gun on low heat. At shallow angles and using a hot air gun as the heat source they take significantly longer to get too hot to touch. This is because the heat exchange from hot air is much poorer than that from hot water. The commercial copper heat pipe got too hot to touch in 35 sec when at a 5̊ angle using the heat gun. A glass heat pipe filled with thin copper wire at a 10̊ angle got too hot to touch in 30 sec, and one of the 12 x 1/4 inch OD glass units did the same in 90 sec. A 30̊ angle resulted in too hot to touch glass and copper heat pipes in 90-180 sec. This was also the case for the glass unit containing 95% ethanol. Patience is needed to demonstrate these heat pipes at shallow angles.

Expt. #3. For a particular heat pipe determine how long it takes for the top end to get too hot to touch (or reach a particular temperature) at 5, 10, 30, and 45̊ angles.

Heat Pipes With Wicking Material – Copper and borosilicate glass heat pipes were tested with several possible wicking materials. The best two were aluminum screening in a tight roll and hemp fiber twine. An aluminum screening filled heat pipe (13" long x 10 mm OD) took 80 sec to heat up at a 5̊ angle, 40 sec at a 30̊ angle, and 17 sec vertically in hot water. The fiber-filled borosilicate heat pipe took 125 sec at a 30̊ angle, and 45 sec vertically in hot water. The 3/8" OD copper tube heat pipe filled with aluminum screening took 45 sec at a 5̊ angle, and 11 sec vertically in hot water.

Expt. #4. Experiment with several materials sealed within the heat pipe you have made to determine the efficiency at 0, 5, and 30̊ angles. Test a commercial heat pipe (if available) at these angles.

Cooling the Hot End – The efficiency of a heat pipe in transferring heat depends on the method of cooling the hot end (amongst other design features). For Expt. #5 test various ways of cooling vertical heat pipes. This can be quantified by filling a 125 mL Erlenmeyer flask with hot water (ca. 90̊C), and measuring the temperature of the water after 10 minutes. The flask needs to be insulated in some way, such as wrapping it completely in cotton or wool batting. A control measurement uses no cooling at the hot end. Ways of cooling are: (1) use a small pancake fan (used in electronic devices) on the hot end; (2) use the fan, but attach 5 cm square fins (punch a hole) made of copper or aluminum screening; (3) cool with cold water passing through a coil of 1/8 in refrigerator tubing or small diameter plastic tubing wrapped around the end of the heat pipe; and (4) fastening in place several layers of cloth (about 3 cm wide), wetting the cloth with water, and blowing on the cloth with the pancake fan.


Heat pipes are in common use in many applications as the most effective way to transfer heat from one region to another. The easy-to-make heat pipes described in this paper are an excellent illustration of a practical and dramatic application of the magnitude of the enthalpy of vaporization.