Introduction
A heat pipe is often used as if it were simply a copper rod with unusually good conduction, and that is the easiest way to misuse it. It works on a two phase cycle where the working fluid vaporises in the evaporator, condenses in the condenser, and is pumped back by capillary action in the wick. Its effective thermal conductivity can reach several hundred times that of a copper rod of the same size, but its transport capacity has a ceiling, and that ceiling depends strongly on mounting orientation. This article starts from the tension between capillary pressure and flow resistance in the wick, compares sintered, mesh and axial groove wicks at different tilt angles, explains the trigger conditions of the capillary, sonic, entrainment and boiling limits, shows why a vapour chamber wins on spreading resistance rather than total resistance, and closes with an actionable selection order and four common mistakes.
1. The two phase cycle: why it beats a copper bar by two orders of magnitude
A heat pipe is a sealed tube evacuated and backfilled with a small charge of working fluid, with a wick lining the inner wall. When one end is heated the fluid vaporises there, the vapour flows to the other end under a small pressure difference, condenses on the cooler wall and releases latent heat, and the condensate is drawn back to the evaporator by capillary action in the wick. There are no moving parts and no external power, and the driving forces are latent heat and capillary pressure.
Converting this into an effective thermal conductivity shows the scale of the difference. A heat pipe six millimetres in diameter and two hundred millimetres long, counting only conduction through the wick at both ends and through the wall, gives a thermal resistance of about zero point zero three four kelvin per watt, while a solid copper bar of the same size gives about thirteen point six kelvin per watt. Moving twenty watts needs a temperature difference of zero point seven kelvin for the pipe and two hundred and seventy two kelvin for the bar. Real products add contact resistance and vapour flow losses and are typically several times worse than the ideal figure, but the two still differ by about two orders of magnitude.
That gap defines where a heat pipe fits. It creates no cooling capacity and does not change the total heat transfer area, it only moves heat elsewhere with a very small temperature drop. It therefore has to work with enough fin area and enough airflow to be useful. Treating the heat pipe as if it were the heat sink itself is the first beginner mistake, and it shows up as a design that has a heat pipe fitted yet sees almost no drop in junction temperature.
2. Three wick types: capillary pressure and flow resistance are two sides of one trade
The job of the wick is to return condensate from the condenser to the evaporator. The pressure it can oppose is set by its capillary pressure, and a larger capillary pressure fights gravity and flow resistance better, but the wick also has to let liquid flow, and a finer channel means more resistance. The two requirements pull in opposite directions, and the three production wick types are simply three points on that line.
A sintered wick is a porous layer of metal powder sintered onto the wall, with the gaps between particles as the channels. It gives the highest capillary pressure, reaching six thousand three hundred pascals in one typical data set, so it resists gravity best, still moving forty five watts with the evaporator tilted thirty degrees upward and thirty seven watts at ninety degrees. The price is high flow resistance, which is why it loses to the groove type when horizontal. A mesh wick sits in between, with a capillary pressure around two thousand two hundred pascals and a capacity of one hundred and fifty eight watts with gravity assist in the vertical orientation.
An axial groove wick uses wide shallow channels cut into the wall. It has the lowest flow resistance and moves three hundred and thirty watts horizontally, the highest of the three, but its capillary pressure is only about six hundred and sixty two pascals, so as soon as the evaporator points upward gravity pulls the liquid back and at thirty degrees of tilt the capacity drops to zero. One comparison makes it vivid. A two hundred millimetre water column produces about one thousand nine hundred and twenty nine pascals, close to three times the capillary pressure of a groove wick, so with the evaporator tilted thirty degrees upward the gravity component alone is nine hundred and sixty five pascals and the groove type can no longer cope.
3. Orientation: a parameter that has to be stated on the drawing
Orientation is not a small correction for a heat pipe, it is the difference between working and not working. With the evaporator below and the condenser above, gravity assists return and capacity peaks. Reversed, with the evaporator above, the wick has to fight gravity alone and capacity can fall to zero. Horizontally there is no gravity burden, so the type with the lowest flow resistance wins, which is why the groove type is strongest horizontally and the sintered type strongest when tilted.
That gives one selection rule. Equipment whose mounting orientation is fixed, such as a cabinet or a telecom base station that always stands vertical, can use a groove wick and take the best horizontal or gravity assisted performance. Equipment whose orientation changes, such as a laptop, a tablet or a handheld instrument, must use a sintered wick, because there is always one worst case attitude and the design has to be checked against that worst case.
A real incident shows the consequence. An industrial control cabinet had its heat pipe cooler mounted upside down because of site constraints, and the equipment overheated and shut down in summer. The original design had the heat source at the bottom with gravity returning the liquid, but after inversion the heat source sat at the top and the weak capillary force of the groove wick could not return the liquid, so the pipe dried out, and restoring the mounting orientation made the problem disappear. The root cause was that the drawing did not state the mounting orientation, and the assembly floor had no reason to think it mattered.
4. Four transport limits: capillary, sonic, entrainment and boiling
A heat pipe is not more capable the more heat you push through it. It has four ceilings and the first one reached sets the real capacity. The capillary limit is the most common, meaning the capillary pressure is no longer enough to cover liquid flow resistance, vapour flow resistance and gravity together. It varies with temperature because the surface tension and latent heat of the working fluid vary with temperature, which is why the same pipe has very different capacity at different operating temperatures.
The sonic limit appears during low temperature start up. At low temperature the vapour density is small and the speed of sound is low, so the vapour velocity easily reaches the local sonic speed and chokes, which pins the transported heat and prevents any further increase. Its typical symptom is that heat cannot get away at the instant of a cold start and recovers by itself once the temperature rises. If the equipment has to start in a cold environment this term has to be checked separately rather than judging from steady state capacity.
The entrainment limit is the vapour flow blowing the liquid away. Vapour and liquid flow in opposite directions inside the pipe, and at high enough velocity the shear tears droplets off the wick surface and carries them back to the condenser, starving the evaporator. This term triggers more easily on thin pipes, high power and long pipes. The boiling limit occurs when the heat flux in the evaporator is too high, so bubbles form inside the wick, block the channels and break the capillary cycle. The usual order of assessment is to compute the capillary limit first and then confirm that the other three do not trigger earlier.
5. Vapour chambers: winning on spreading resistance rather than total resistance
A vapour chamber can be seen as a flattened two dimensional heat pipe. It runs on the same two phase cycle, but the vapour can spread across the whole plane instead of flowing only along the tube axis. The benefit concentrates on spreading resistance, which is the extra temperature drop paid on the conduction path when heat enters over a small area and has to fan out into a much larger base plate.
When does spreading resistance matter? When the heat source is much smaller than the base. A die may be a few millimetres square while the heat sink base is tens of millimetres square, and the process of spreading heat laterally through copper itself produces a noticeable temperature drop, with the point directly above the die measurably hotter than the average of the base. A vapour chamber moves heat laterally by two phase transport, which effectively replaces that lateral conduction with a near isothermal process and pushes this term down.
Conversely, if the heat source is already wide, such as the base plate of a large power module, the spreading resistance is small to begin with and the benefit of a vapour chamber is limited. Vapour chamber thickness is also structurally constrained, and flattening narrows the vapour channel and deforms the wick, lowering the limits, while bending and flattening likewise reduce the capacity of a heat pipe itself. The judging order in selection is therefore to compute the share of spreading resistance in the total first, and then decide whether the cost and thickness of a vapour chamber are justified by it.
6. A selection order and four common mistakes
Selection follows six steps. First set the heat flow and the temperature difference budget, stating how many watts have to be moved and how much temperature drop is allowed. Second fix the orientation and the worst case attitude, deriving the wick type from it. Third choose the working fluid by operating temperature, comparing figures of merit rather than relying on habit. Fourth compute the four limits, confirming margin on the capillary limit and that the other three do not trigger earlier. Fifth decide on a vapour chamber by the share of spreading resistance. Sixth confirm the forming method, because bending, flattening and pressing all require recomputing the limits with post forming values.
The first common mistake is ignoring orientation. That leads to dry out and overheating in the worst case, and the remedy is to include gravity in the limit calculation and to state the mounting orientation on the drawing. The second is modelling a heat pipe in system simulation as nothing more than a high conductivity bar. That produces a temperature distribution but cannot detect an exceeded limit, and the remedy is to separately confirm that the transported heat stays inside the limits. The third is ignoring the effect of bending and flattening, because forming deforms the wick and the vapour channel and lowers the limits, and the remedy is to use post forming values.
The fourth is not checking cold start. At low temperature the vapour density is low and the sonic limit is easily reached, which shows up as insufficient cooling on a cold machine that recovers by itself after running for a while, and the remedy is to confirm the sonic limit and the freezing point of the working fluid under start up conditions. Collapsing all of this into one sentence, a heat pipe is a device with an orientation, a ceiling and a start up condition, and treating it as a passive conducting bar is the common origin of most field failures.
Figure Notes
Figure 1｜Heat pipe versus a solid copper bar of the same sizeFigure 2｜Transport capacity of three wick types by orientationFigure 3｜Capacity of groove and sintered wicks against tilt angleFigure 4｜The two phase cycle inside a heat pipe and where the four limits actFigure 5｜Four common heat pipe selection mistakes and their remedies