Multiple heatsinks in one enclosure: airflow and heat

Two heatsinks can share one fan, but they do not automatically receive the airflow used for their published thermal resistance. Each heatsink needs the correct local air velocity and a known inlet-air temperature. In series, the downstream heatsink receives warmer air. In parallel, the fan flow divides between branches. Calculate those conditions first, then apply each heatsink’s thermal resistance to its own heat load.

What you need

Gather these inputs before calculating temperatures:

• Heat dissipated into each heatsink, in watts, including the worst sustained operating condition.
• Each heatsink’s thermal-resistance curve versus air velocity, with its test arrangement and temperature reference.
• The fan’s airflow-versus-static-pressure curve at the intended supply voltage and speed.
• Dimensions of ducts, branch openings, fin passages and bypass gaps.
• Pressure-loss information for heatsinks, filters, guards, ducts and other restrictions.
• The highest expected temperature of the air entering each cooling path.
• Package temperature limits and the thermal resistance of each mounting interface.

A rating of 1.50°C/W at 500 linear feet per minute, or LFM, describes a particular airflow condition. The velocity is approximately 2.54 m/s. It is not a promise that any fan marked with that velocity produces the same result.

Use the heat sink thermal resistance article for resistance definitions and thermal-path fundamentals. The calculation below focuses on sharing airflow between separate heatsinks.

Calculate airflow and inlet temperature

1. Draw the complete air path. Mark the fan, heatsinks, restrictions, bypass gaps and return or exhaust opening. Identify whether air passes through both heatsinks sequentially or divides between them. A layout can contain both arrangements, so follow the actual path rather than the heatsinks’ positions alone.

2. Find the fan operating point. Actual flow occurs where the fan curve meets the installed system’s pressure-loss curve. Include every restriction in the path. Free-air flow describes operation without the installed resistance; it is not the flow through the assembled enclosure. For series heatsinks, their pressure losses add at the same flow. For parallel branches, flow divides according to each branch’s resistance.

3. Determine the flow through each heatsink. Subtract bypass flow before calculating useful cooling flow. Air travelling around a fin stack does not provide the same cooling as air travelling through it. Parallel branches receive equal flow only when their pressure losses are sufficiently similar. A larger opening or shorter path can take most of the available air.

4. Convert flow to the velocity used by the heatsink data. In customary units, velocity in ft/min equals flow in CFM divided by area in ft². Match the manufacturer’s area definition and measurement location. Approach velocity ahead of the heatsink is not necessarily the velocity inside its narrower fin passages. Do not compare one measurement with a curve based on the other.

5. Calculate the local inlet-air temperature. For a downstream heatsink, add the upstream air-temperature rise to the entering stream temperature. For parallel branches supplied from a common plenum, check whether both branches actually receive equally cool air. Exhaust recirculation can warm one branch even when the drawing shows separate paths.

6. Apply the appropriate thermal resistance. Calculate heatsink base temperature as T_sink = T_inlet + P_sink × R_sink-air. Use the resistance corresponding to the actual velocity and installation conditions. Confirm that the published resistance uses inlet air as its temperature reference before applying this equation directly.

A nearby fan can produce a strongly uneven velocity profile across a heatsink. Adequate average velocity does not prove that every fin passage receives adequate flow. Ducting, a plenum or a different fan position may improve distribution, but each change also affects pressure loss.

For whole-enclosure ventilation calculations, use the enclosure fan airflow calculation. That calculation establishes bulk heat removal, not the distribution between individual heatsinks.

Work through two heatsinks sharing one fan

Consider two identical heatsinks, each dissipating 40 W. Assume their documented sink-to-inlet-air resistance is 1.50°C/W at a verified approach velocity of 500 LFM.

For this example, each approach duct has a section measuring 4 × 4 in [102 × 102 mm]. Assume the duct arrangement matches the heatsink data and prevents bypass. These dimensions are calculation assumptions, not a catalog enclosure size.

The approach area is:

A = (4 × 4) / 144 = 0.111 ft²

The flow required through each approach section is:

Q = 500 × 0.111 = 55.6 CFM

Series airflow

Both heatsinks receive the same 55.6 CFM stream, assuming no leakage between them. However, the fan must overcome the combined resistance of both heatsinks and the remaining air path.

Estimate air heating from:

ΔT_air = P_to-air / (ρ × c_p × Q)

Here, Q must be in m³/s. At approximately 25°C, use air density of 1.18 kg/m³ and specific heat of 1,005 J/(kg·K). With flow expressed in CFM, the same approximation becomes:

ΔT_air [°C] ≈ P_to-air [W] / (0.56 × Q [CFM])

Assuming all 40 W from the first heatsink enters the air, its bulk temperature rise is:

ΔT_air ≈ 40 / (0.56 × 55.6) = 1.3°C

With 25°C entering the first heatsink, approximately 26.3°C enters the second. Assuming both installations still justify 1.50°C/W, their calculated base temperatures are 85°C and 86.3°C.

The second value assumes adequate mixing before the downstream heatsink. A concentrated hot wake can produce a higher local inlet temperature than the bulk calculation suggests.

Parallel airflow

Now place the heatsinks in identical parallel branches. If total useful fan flow remains 55.6 CFM, an equal split gives each branch only 27.8 CFM.

Through the same approach areas, each heatsink receives 250 LFM, not 500 LFM. Read the thermal resistance at 250 LFM from the applicable curve. Do not retain 1.50°C/W or assume that halving velocity simply doubles resistance.

To provide 500 LFM in both branches, total useful flow must reach approximately 111 CFM. The fan must deliver that flow at the installed operating pressure.

These are conditional calculations, not guaranteed operating temperatures. Fan selection, bypass, inlet uniformity and test conditions must support the assumptions.

Calculate package and junction temperatures

Heatsink base temperature is not package case temperature. For one device transferring its entire 40 W load through a mounting interface rated at 0.20°C/W, calculate:

T_case = T_sink + P_device × R_case-sink

The interface adds 8°C. In the series example, estimated case temperatures become 93°C and approximately 94.3°C.

Use interface data for the actual mounting pressure, contact area and installation method. A thicker insulating pad or poor contact can change the result substantially.

If the device manufacturer provides an applicable junction-to-case resistance, the simplified junction calculation is:

T_junction = T_case + P_device × R_junction-case

That series model assumes the stated heat follows the modeled path. Check the device documentation before applying it to packages with significant alternative heat paths.

When several devices share one heatsink, the heatsink carries their combined load. Each device still has its own interface drop, and local spreading resistance may require a more detailed model. Do not combine two separately loaded heatsinks as parallel thermal resistances merely because one fan serves both.

Checks before accepting the layout

Validate the assembled arrangement at the highest specified inlet temperature and representative maximum heat load. Follow applicable electrical safe-work procedures when installing sensors or making adjustments.

Measure air temperature immediately upstream of each heatsink, clear of direct contact with hot metal. Measure heatsink base and package temperatures at the locations required by their documentation. Use several airflow readings where the velocity profile is uneven.

Check these conditions during validation:

• Both heatsinks remain below their calculated and documented temperature limits, with an appropriate design margin.
• Downstream inlet temperature agrees reasonably with the predicted air heating.
• Branch airflow remains balanced with all covers, guards and filters installed.
• Bypass gaps and exhaust recirculation are controlled.
• Fan failure, reduced speed and filter loading produce the required alarm or protective response.
• Temperatures have stabilized before results are accepted.

For layout changes, the component layout article addresses clearance, hot zones and circulation paths.

If the fan only recirculates enclosed air, the example’s 25°C supply temperature must come from a separate enclosure thermal calculation. The fan does not establish that temperature. The internal fan heat-path article explains why enclosure walls or cooling equipment must still reject the total load.

If more space or different openings are required, send the airflow layout with your custom enclosure requirements. Include heatsink clearances, fan and duct footprints, cutout dimensions, mounting details and total heat load. Use the enclosure selector when choosing the enclosure becomes the next step; dimensional fit alone does not establish thermal suitability.

Mistakes that invalidate the calculation

Treating fan flow as local fin velocity. A fan’s total CFM does not establish velocity at either heatsink. Flow area, bypass and distribution determine the useful local value.

Ignoring the additional pressure loss. Adding a second heatsink can reduce total flow, including in a series arrangement. Recalculate the operating point rather than carrying over the original airflow.

Using outside ambient for every heatsink. A downstream sink sees upstream heating. An internally recirculating system sees an enclosure air temperature determined by its heat-rejection path.

Assuming downstream performance is unchanged. Warmer air is only one difference. The upstream heatsink can also change turbulence and velocity distribution, making the original resistance curve less representative.

Adding ventilation without reviewing environmental protection. A fan opening changes the enclosure design. Review the complete assembly rather than assuming the original enclosure rating remains applicable. Our enclosure ratings are manufacturer ratings; the enclosures are not UL listed.

Expecting internal circulation to solve inadequate heat rejection. Better airflow can reduce local temperatures while the enclosure still accumulates too much heat. Compare the available options in the enclosure cooling methods article before increasing fan speed alone.

Enclosures for this job

Catalog sizes that suit the job this article answers, picked from its subject; open one for its full specifications.

Questions