Cooling Fan Performance Curves: Free-Air CFM vs Installed Airflow

A cooling fan's free-air CFM rating is not the airflow it will necessarily deliver through a heavy-duty radiator package. Compare the fan's pressure-versus-flow curve with the resistance of the intended air path, using compatible test conditions. The useful result is the installed operating point, not the largest flow or pressure number printed beside a product photograph.

This guide is for truck, bus and off-highway parts teams reviewing cooling-fan performance information with their engineering counterparts. It explains how to read and compare evidence. It does not calculate a particular vehicle's cooling requirement, authorize a different fan or prescribe a workshop airflow test. Mechanical compatibility and safe installation remain separate requirements.

Start With the Question the Curve Must Answer

A replacement inquiry and a cooling-system redesign need different evidence. In an identified replacement, the first question is whether the proposed fan matches the approved configuration. A curve can help explain performance differences, but it does not replace the application reference. In a redesign, the engineering team must first define the air delivery required by the cooling package and the conditions under which it is needed.

For example, a distributor may receive a request for a larger fan because a machine runs hot during stationary work. That symptom does not establish that the original fan is undersized. The supplied curve may describe a sound fan operating at a speed the machine never reaches during the complaint. Alternatively, the original fan may be correct while the installed air path differs from the intended configuration. These are different questions with different purchasing consequences.

Write the question in terms of the decision being made. “Can this alternative deliver the specified flow against the specified resistance at the available fan speed?” is a performance comparison. “Is this the approved replacement for the installed fan?” is an application check. “Why does coolant temperature rise?” is a system investigation. Keeping those questions separate prevents a catalog graph from becoming an answer to a problem it was not designed to solve.

When the required duty point is unavailable, the proper output of a curve review is a request for missing engineering information. It is not a guessed airflow target based on engine displacement, horsepower or fan diameter. A purchasing team can identify an incomplete offer without attempting to perform the cooling-system designer's work.

Read the Axes Before Comparing the Curves

A basic fan curve plots volume flow on one axis and pressure on another. Additional curves may show input power, efficiency or sound. The presence of several lines does not mean that every line represents a different fan. First read the legend, units and scale attached to each line. A power trace using the right-hand axis must not be interpreted against the pressure scale on the left.

The graph's title and notes are part of the evidence. Retain the fan reference, diameter or configuration, speed, air density and test arrangement with the picture. A cropped screenshot containing only a colored curve can look useful while omitting the conditions that make it interpretable. Obtain the complete page or manufacturer-generated selection record before extracting values for a comparison.

The AMCA explanation of fan curves distinguishes airflow, pressure, power and efficiency and shows how to read them at the same operating condition. Its examples concern ventilation fans. Use the reading method, not its example duty values, when reviewing engine-cooling data.

Unit conversion is a presentation task, not a performance correction. Converting a flow value into another volume-per-time unit does not adjust for a different air density, rotational speed or test installation. If two quotations use different units and different test conditions, there are two separate issues to resolve. Fixing the units alone does not make the curves comparable.

Also inspect axis multipliers. A label that scales airflow by hundreds or thousands changes how the printed numbers must be read. Never reconstruct an unclear multiplier from what seems plausible for the fan. Ask for a legible original. An uncertainty in the axis can alter the whole interpretation, whereas another attractive product photograph adds no performance evidence.

Free-Air Flow and Maximum Pressure Are Different Endpoints

A free-air rating describes flow at little or no imposed pressure resistance under the stated test arrangement. The actual installation places a radiator, other heat exchangers, grilles, guards and surrounding structures in the air path. The fan therefore has to work against an air-path condition that the free-air endpoint does not represent.

At the other end of a pressure-versus-flow graph, a fan may develop pressure while moving very little air. That pressure figure is not the pressure available while simultaneously delivering the maximum free-air flow. Treating both endpoints as if they occur together creates a fictional operating condition that is not on the curve.

This is why replacing “choose the highest CFM” with “choose the highest static pressure” does not solve the selection problem. Either rule can select the wrong part. The required combination of flow and pressure must be supported at the relevant speed and conditions. A fan with a high pressure endpoint may still be unsuitable for a duty requiring substantial airflow farther along the curve.

Consider two hypothetical quotations. Offer A prominently states its free-air flow. Offer B gives a lower headline flow but includes data at the required resistance. Those descriptions do not establish that B is better, because A's pressure capability remains unknown. They establish that B has supplied more decision-relevant evidence. Request A's corresponding curve before comparing performance or rejecting it.

The same principle applies when a seller gives a single airflow value without stating pressure. Do not label the number false merely because its conditions are missing. Label the claim incomplete for the present decision. Ask what was measured or calculated, which configuration it describes, and where the operating condition appears in the supporting document.

Find the Installed Operating Point, Not a Convenient Point on the Fan Line

Figure — On compatible static-pressure and airflow axes, the fan and system-resistance curves meet at an installed operating point distinct from the free-air endpoint. These original schematic curves contain no measured values, exact mathematical fit or product ranking.

The system-resistance curve describes the pressure requirement of an air path at different flows. The fan and system operate together where their compatible curves intersect. That intersection expresses two conditions at once: the fan supplies the pressure needed by the air path, and the air path accepts the flow being delivered.

Reading only the fan line answers what the fan can provide under its stated conditions. Reading only the system line answers what the air path requires. Neither line independently establishes installed airflow. A comparison must bring the two together and identify the particular operating condition being evaluated.

For an illustrative comparison, imagine two fan curves crossing each other. At a low-resistance duty, the first fan intersects the system curve at a higher flow. At a different duty with greater resistance, the second fan could provide the higher intersection flow. There is no contradiction: the ranking changed because the system requirement changed. The illustration does not assign any measured values or performance ranking to actual products.

This also explains why a single demonstration on an unrelated radiator is weak evidence for another machine. The demonstration may be genuine while describing a different air path. Without the relationship between that fixture and the target package, the result cannot establish the target operating point.

Ask the responsible engineer whether the supplied system information applies to the complete package, a single heat exchanger or another defined boundary. A curve for one radiator core should not silently stand for the entire front-end assembly. Keep the scope visible when the evidence moves from engineering to purchasing so an incomplete resistance value does not become the accepted specification by repetition.

Keep Static and Total Pressure on the Same Basis

Pressure terminology matters because different quantities can appear on otherwise similar graphs. Fan static pressure and fan total pressure are not interchangeable labels. A comparison that uses one supplier's static-pressure line and another supplier's total-pressure line can make one option appear stronger without representing a real advantage.

AMCA's curve-reading article explains the distinction through the velocity-pressure contribution at a defined outlet. For a purchasing review, the immediate lesson is practical: retain the pressure definition and the measurement or calculation boundary. Do not change the column heading to the generic word “pressure” when combining source data into a comparison sheet.

Suppose an engineer provides a required duty using static pressure, while a quotation reports only total pressure. The buyer should request a compatible statement or an engineering conversion based on the proper configuration. Subtracting a guessed allowance is not an acceptable shortcut. The required information includes more than the two numbers visible in the offer.

The word resistance can also be used loosely in a conversation. Ask whether it refers to a pressure loss at a specified flow, a complete resistance curve or a general description such as “high-restriction core.” Only the first two provide a quantified relationship suitable for comparison. The third may describe a concern, but it does not define a duty point.

Precision here reduces unnecessary argument between suppliers. Two parties can both report accurate data using different conventions. The useful response is to align the definitions, not accuse one party of exaggeration before the test basis is understood. If the definitions cannot be aligned, keep the alternatives unranked on that attribute.

Speed and Air Density Belong Beside Every Airflow Claim

Figure — Keep the exact fan configuration, rotational basis, air density, pressure definition and test arrangement with the curve. Static and total pressure are different quantities; the illustrative fields contain no test values or completed selection.

A fan curve belongs to its stated rotational condition. Engine speed, fan-drive input speed, commanded speed and actual fan speed are not automatically the same value. A curve prepared for one fan speed does not demonstrate that the installed drive can reach or sustain it during the operating condition under review.

For an engine-driven installation, the drive relationship must come from the approved configuration. For an electrically driven unit, a nominal voltage or a command percentage alone does not establish the actual speed represented by a graph. Ask whether the document describes a particular control condition, measured rotational speed or a family of speed curves. This is identification of the test basis, not an instruction to alter controls.

Air density is equally important to interpreting aerodynamic data. It is not enough to label one curve “hot climate” and another “standard.” The documented condition must be explicit. The Horton eRev guide includes density settings derived from altitude, temperature and humidity, as well as a manual density input. That is a concrete example of a manufacturer treating the environment as an input rather than an afterthought.

Volume flow and mass flow also answer different questions. CFM expresses a volume passing per unit time; the mass associated with that volume depends on density. A cooling-system requirement must therefore retain the environmental basis used by the engineering team. A large volume figure cannot be translated into a heat-rejection promise without the remaining thermal and installation information.

Do not create a revised curve by changing a graph label to the destination country's summer temperature. Ask the manufacturer or qualified engineer for the applicable conditions and correction method. If an offer provides only standard-condition data, it may remain useful as a reference, but the environmental comparison is unfinished.

The Test Installation Can Change the Meaning of a Curve

Figure — Shroud, immersion, clearance and discharge-side obstruction belong to the conditions attached to fan data. The arrangements are illustrative and the arrows show direction only, not flow rate, measured advantage, a prescribed clearance or retrofit approval.

A fan does not operate independently of nearby structures. The test arrangement can include a shroud, an inlet or outlet configuration, a particular blade position and an obstruction representing surrounding equipment. The same fan reference can therefore be associated with several valid performance presentations for different arrangements.

Horton's eRev guide describes engine-silhouette settings for an obstruction immediately behind the fan discharge. It also identifies immersion as the proportion of the fan within the shroud. Its data notes describe different shroud configurations, clearances and test combinations. These details are not instructions to choose a universal immersion or clearance. They show why the setup attached to a curve must travel with the curve.

Imagine an offer whose data includes a representative discharge obstruction and another whose data does not. A direct overlay might attribute all the difference to blade design. Some of the difference could instead arise from the conditions represented. The engineer needs to determine whether either setup is suitable for the intended application before treating the overlay as a product comparison.

There is an important distinction between acknowledging installation effects and inventing a correction factor. Saying that the installed condition matters is supported. Reducing every catalog airflow number by a standard percentage is not a substitute for compatible data. Different arrangements need their own assessment; a convenient fixed deduction can conceal the missing information rather than account for it.

Keep mechanical fitment in its own record. The existing radiator cooling-fan matching guide addresses rotation, blade geometry, hub depth and shroud relationships. This curve review starts after the configuration has been identified. It does not authorize modifying a blade, moving a shroud or changing a mounting arrangement to imitate a test fixture.

Review the Whole Air Path Without Inventing a Universal Stack Loss

A heavy-duty cooling package may contain more than a radiator. Other heat exchangers and surrounding structures can affect the route taken by the air. Before comparing fan data, ask what components and boundaries were included in the resistance information supplied for the package.

A pressure loss stated for an individual component belongs to its stated flow and conditions. It is not a fixed penalty that can be added to any arbitrary fan rating. Likewise, placing several individual catalog pressure numbers in a spreadsheet does not automatically create a valid complete-system curve. The engineer must establish which components share the relevant air path and how the installed arrangement is represented.

Air that passes through the fan is not necessarily identical to useful airflow across every intended part of the cooling package. A gap, alternate path or recirculating route can make a total-flow observation insufficient to establish the desired distribution. A curve is valuable, but it does not show every local airflow detail around a radiator face or inside an engine compartment.

This is a reason to request an application assessment, not a reason to dismiss published data. A manufacturer curve answers a controlled question. An installation evaluation answers how the selected equipment behaves within the machine. Good technical purchasing keeps both forms of evidence and does not ask either one to certify something outside its scope.

When a package changes, identify the changed component rather than preserving an old system description by habit. A different cooler, guard or surrounding enclosure may require the existing airflow assessment to be revisited. The buyer need not calculate the change, but should not present the earlier package evidence as unchanged merely because the fan part number remains the same.

Read Power and Sound at the Same Duty Point

A lower-power claim is meaningful only when its task is clear. Two fans consuming different power while delivering different flow against different resistance are not performing the same job. Start with the required duty, then inspect the corresponding power information for each acceptable option.

Check what the power figure includes. A graph can describe aerodynamic shaft power while a complete electric assembly has additional motor and control losses. A mechanical system also has a drive boundary that must be understood. Do not put unlike quantities in one column labelled “consumption” and calculate a savings percentage from them.

The same caution applies to peak and operating values. A component's maximum rating is not its consumption at every point, and a favorable reading at one point is not its full-duty energy use. Estimating an operating-cost difference requires an agreed duty profile and consistent system boundaries. This article does not turn a single curve point into an annual fuel-saving or electrical-saving claim.

Sound data needs its own definition. Sound power and sound pressure at a location describe different information, and the test or calculation method matters. Horton's eRev guide expressly limits its calculated sound guidance to relative comparison within the software and warns that comparisons with other data or a specific installation can be inaccurate. Keep that limitation attached when a sound figure is copied into a quotation review.

A purchasing team can still compare evidence quality when the performance figures are not yet comparable. One offer may identify its power boundary and sound method; another may only use the word quiet. Record that difference as a documentation gap. Do not convert the adjective into a numerical advantage or reject a fully specified option because the other description sounds more persuasive.

Separate Stationary Cooling From Road-Air Assistance

The operating condition should explain why the fan is needed. A truck climbing slowly, a bus working through repeated stops and an off-highway machine performing stationary work do not present the same airflow situation as a vehicle moving steadily at road speed. A curve review should remain tied to the condition responsible for the engineering requirement.

The Horton HS6 product sheet discusses a heavy-duty on-road design in relation to both fan airflow and free ram-air passage. That supports considering the wider operating situation. It does not supply a universal vehicle speed above which any fan becomes unnecessary, or establish a retrofit choice for another truck.

Do not add a catalog free-air fan figure to an independently estimated road-air figure and call the result total installed airflow. Those quantities may represent different pressure conditions and air paths. The combined operating situation requires a system assessment rather than arithmetic on unrelated headline values.

A successful highway observation can therefore leave a stationary requirement unresolved. The reverse is also true: satisfying one stationary duty does not establish every operating condition. State which requirement the evidence addresses. That makes an approval useful without extending it to a broader claim than the supporting evaluation can carry.

This distinction also keeps the curve article separate from fan-clutch diagnosis. If the unresolved question is whether a commanded clutch engages and produces the expected speed response, use the fan-clutch engagement guide and the applicable service procedure. A stronger catalog curve cannot explain why an existing drive is not delivering its commanded operating state.

Work Through Two Offers Without Declaring a Winner Too Early

Consider a hypothetical replacement review for a defined cooling package. Both offers appear to match the application dimensions. Offer A includes a curve at a stated speed and standard air condition, without the installation notes. Offer B includes a complete selection report with a different speed and a defined discharge obstruction. Neither offer should immediately win the airflow comparison.

First separate the known differences from the unknowns. The rotational conditions differ, so the reported lines cannot be treated as a same-speed comparison. The installation definition is present for one offer and missing for the other. The application match also remains independent of the performance graph. Listing these facts is more useful than averaging the two flow values or selecting the larger one.

Next ask for a comparison at the approved duty and compatible conditions. If the suppliers provide it, the engineer can determine whether each option satisfies the requirement and identify any remaining installation limits. If one cannot provide it, that offer remains unsubstantiated for the present performance decision. Lack of evidence is not proof that the product is defective; it is a reason not to approve the claimed equivalence.

Now imagine that both options meet the airflow requirement, but one power figure covers the fan shaft and the other covers the electric assembly. The airflow question may be sufficiently answered while the efficiency ranking remains open. Technical approval is not always an all-or-nothing judgment. Clearly separating the settled attributes from the unresolved ones allows purchasing to move only as far as the evidence supports.

Finally, keep the accepted configuration linked to the evidence. A later offer using a different blade variant or a different speed basis is not automatically covered by the earlier comparison. The record should make that change visible without requiring a new reviewer to reconstruct the original discussion from emails.

Do Not Turn a Coarse Graph Into False Precision

Some documents provide a few tabulated points, while others show a small printed graph. Those formats support different levels of precision. A line read from a low-resolution image should not become a spreadsheet value with several decimal places. The extra digits come from the reader's presentation, not additional test evidence.

Suppose the required duty falls between two published points. A buyer may use an approximate reading to identify whether an offer deserves further review. If acceptance depends on a small margin, however, obtain the manufacturer's value or a suitable engineering assessment at that duty. An approximate visual reading should not decide a close technical pass when the source cannot support that precision.

A more serious problem arises when the required condition falls beyond the published range. Extending the last visible segment creates an extrapolation, not another manufacturer-supplied point. The shape might change outside the displayed region, or the manufacturer may have limited the presentation for reasons absent from the excerpt. Ask for the applicable data instead of making the graph long enough to reach the requirement.

Keep these distinctions visible in the review: a stated manufacturer value, an approximate graph reading and an engineering calculation are different evidence types. One can help locate the next question without replacing another. If a quotation changes from an approximate selection to a contractual performance claim, the supporting information must change with it.

Also retain any indicated recommended operating region. A curve that is drawn across a broad range does not necessarily designate every point as an acceptable continuous operating choice. Refer questions about the permitted region to the manufacturer rather than treating the existence of a plotted line as blanket approval.

Recognize What a Published Curve Cannot Prove

A performance curve is not a structural-speed approval, a blade-damage assessment or an interchange certificate. It does not establish that the hub, fasteners, material and guarding suit the machine. Those requirements remain necessary even when the aerodynamic comparison is persuasive.

Nor does a graph prove the condition of a delivered unit. A correctly identified fan can arrive damaged, and a visually sound fan can carry an unconfirmed reference. The cooling-fan receiving guide addresses physical release separately. Keeping the processes distinct avoids using an attractive curve to waive inspection or using a clean carton to validate aerodynamic performance.

A curve also cannot independently diagnose an overheating complaint. The relationship between coolant condition, thermal load and installed airflow requires the relevant system evidence. If a supplier presents a new fan as the remedy, ask which finding connects the proposed change to the actual problem. A catalog improvement that addresses no established deficiency may add cost without resolving the complaint.

The final review should identify the evidence that answers the original question, the conditions under which it applies and the remaining limits. It may conclude that the proposed fan is a documented match, that another operating condition needs engineering review, or that the available graph is insufficient. Each can be a useful result. The purpose is a defensible cooling-fan decision, not a forced ranking of incomplete quotations.

When discussing an alternative, ask for the configuration-specific curve and its conditions before accepting an airflow claim. Keep the complete manufacturer document with the approved application record. That small discipline preserves the difference between a fan that looks promising in free air and evidence that addresses the installed duty.

For an Elecdura cooling-fan enquiry by email, provide the installed reference, application and available performance documents with the quantity required. Identify any unresolved curve or installation question so the response can separate available product information from work needing the responsible engineer.